Merge branch 'master' of http://www.denx.de/git/u-boot

CHANGELOG

@@ -2,6 +2,80 @@
 Changes since U-Boot 1.1.4:
 ======================================================================
 
+* Minor cleanup for PCS440EP board
+
+* Update PCS440EP port to fit into one flash device (incl. environment)
+  Patch by Stefan Roese, 06 Jun 2006
+
+* Add support for PCS440EP board
+  Patch by Stefan Roese, 02 Jun 2006
+
+* Fix examples/Makefile; some build targets were lost
+
+* Fix watchdog handling in CFI flash driver
+  Just use udelay() when waiting for status changes which will
+  implicitely trigger the watchdog.
+
+* Fix PCI to memory window size problems on PM82x boards
+  We use the "automatic" mode that was used for  the  MPC8266ADS  and
+  MPC8272 boards. Eventually this should be used on all boards?]
+  Patch by Wolfgang Grandegger, 17 Jan 2006
+
+* Correct GPIO setup (UART1/IRQ's) on yosemite & yellowstone
+  Patch by Stefan Roese, 29 May 2006
+
+* Update Intel IXP4xx support
+  - Add IXP4xx NPE ethernet MAC support
+  - Add support for Intel IXDPG425 board
+  - Add support for Prodrive PDNB3 board
+  - Add IRQ support
+  Patch by Stefan Roese, 23 May 2006
+
+* Fix problem in PVR detection for 440GR
+  Patch by Stefan Roese, 18 May 2006
+
+* Fix gcc 3.4.x AFLAGS setting for m68k platform.
+
+* Enable autoboot for M5271EVB board.
+
+* Changed default ramdisk addr in yosemite/yellowstone ports
+  Patch by Stefan Roese, 15 May 2006
+
+* Fix PCMCIA support on virtlab2
+
+* Add support for VirtLab2 board
+  (needed because of differences in the PCMCIA hardware).
+
+* Minor cleanup.
+
+* Update yosemite configuration to enable flash write buffer support
+  Patch by Stefan Roese, 10 May 2006
+
+* Fix compile warnings in common/xyzModem.c
+  Patch by Stefan Roese, 10 May 2006
+
+* Add support for AMCC 440EP Rev C and 440GR Rev B
+  Patch by John Otken, 08 May 2006
+
+* OMAP 5912/OSK: update EMIFS CS1 timings:
+  Problems have been seen in the linux kernel's smc91x network driver
+  due to improper bus timings.  The latest 2.6 OMAP kernels currently
+  have a workaround, but this fix belongs in u-boot.
+  Patch by Kevin Hilman, 13 Oct 2005
+
+* Fix REG_MPU_LOAD_TIMER definition in multiple OMAP ports
+  Patch by Hiroki Kaminaga, 11 Mar 2006
+
+* Update omap5912osk board support
+  - Fix OMAP support that omap5912osk compiles in current source tree
+  - Update with code from "http://omap.spectrumdigital.com/osk5912"
+    to fix problems with DDR initialization
+  - Fix timer setup
+  - Use CFI flash driver and support complete 32MB of onboard flash
+  - Add "print_cpuinfo()" and "checkboard()" functions to display
+    CPU (with frequency) and Board infos
+  Patch by Stefan Roese, 10 May 2006
+
 * Fix watchdog issues for ColdFire boards.
 
 * Add M5271EVB board support.

MAINTAINERS

@@ -283,6 +283,7 @@ Stefan Roese <sr@denx.de>
 	ebony			PPC440GP
 	ocotea			PPC440GX
 	p3p440			PPC440GP
+	pcs440ep		PPC440EP
 	sycamore		PPC405GPr
 	walnut			PPC405GP
 	yellowstone		PPC440GR
@@ -434,6 +435,11 @@ Dave Peverley <dpeverley@mpc-data.co.uk>
 
 	omap730p2		ARM926EJS
 
+Stefan Roese <sr@denx.de>
+
+	ixdpg425		xscale
+	pdnb3			xscale
+
 Robert Schwebel <r.schwebel@pengutronix.de>
 
 	csb226			xscale

MAKEALL

@@ -71,11 +71,11 @@ LIST_4xx="	\
 	HH405		HUB405		JSE		KAREF		\
 	luan		METROBOX	MIP405		MIP405T		\
 	ML2		ml300		ocotea		OCRTC		\
-	ORSG		p3p440		PCI405		PIP405		\
-	PLU405		PMC405	        PPChameleonEVB	sbc405		\
-	VOH405		VOM405          W7OLMC		W7OLMG		\
-	walnut		WUH405		XPEDITE1K	yellowstone	\
-	yosemite							\
+	ORSG		p3p440		PCI405		pcs440ep	\
+	PIP405		PLU405		PMC405	        PPChameleonEVB	\
+	sbc405		VOH405		VOM405          W7OLMC		\
+	W7OLMG		walnut		WUH405		XPEDITE1K	\
+	yellowstone	yosemite					\
 "
 
 #########################################################################
@@ -209,7 +209,7 @@ LIST_pxa="	\
 	zylonite							\
 "
 
-LIST_ixp="ixdp425"
+LIST_ixp="ixdp425	ixdpg425	pdnb3"
 
 
 LIST_arm="	\

Makefile

@@ -759,7 +759,8 @@ TQM850M_config		\
 TQM855M_config		\
 TQM860M_config		\
 TQM862M_config		\
-TQM866M_config:		unconfig
+TQM866M_config		\
+virtlab2_config:	unconfig
 	@ >include/config.h
 	@[ -z "$(findstring _LCD,$@)" ] || \
 		{ echo "#define CONFIG_LCD"		>>include/config.h ; \
@@ -921,6 +922,9 @@ p3p440_config:	unconfig
 PCI405_config:	unconfig
 	@./mkconfig $(@:_config=) ppc ppc4xx pci405 esd
 
+pcs440ep_config:	unconfig
+	@./mkconfig $(@:_config=) ppc ppc4xx pcs440ep
+
 PIP405_config:	unconfig
 	@./mkconfig $(@:_config=) ppc ppc4xx pip405 mpl
 
@@ -1718,12 +1722,18 @@ innokom_config	:	unconfig
 ixdp425_config	:	unconfig
 	@./mkconfig $(@:_config=) arm ixp ixdp425
 
+ixdpg425_config	:	unconfig
+	@./mkconfig $(@:_config=) arm ixp ixdp425
+
 lubbock_config	:	unconfig
 	@./mkconfig $(@:_config=) arm pxa lubbock
 
 logodl_config	:	unconfig
 	@./mkconfig $(@:_config=) arm pxa logodl
 
+pdnb3_config	:	unconfig
+	@./mkconfig $(@:_config=) arm ixp pdnb3 prodrive
+
 pxa255_idp_config:	unconfig
 	@./mkconfig $(@:_config=) arm pxa pxa255_idp
 

board/amcc/yellowstone/yellowstone.c

@@ -79,8 +79,8 @@ int board_early_init_f(void)
 	out32(GPIO1_ISR2L, in32(GPIO1_ISR2L) | 0x00010000);
 
 	/* external interrupts IRQ0...3 */
-	out32(GPIO1_TCR, in32(GPIO1_TCR) & ~0x0f000000);
-	out32(GPIO1_TSRL, in32(GPIO1_TSRL) & ~0x00005500);
+	out32(GPIO1_TCR, in32(GPIO1_TCR) & ~0x00f00000);
+	out32(GPIO1_TSRL, in32(GPIO1_TSRL) & ~0x0000ff00);
 	out32(GPIO1_ISR1L, in32(GPIO1_ISR1L) | 0x00005500);
 
 #if 0 /* test-only */

board/amcc/yosemite/yosemite.c

@@ -79,8 +79,8 @@ int board_early_init_f(void)
 	out32(GPIO1_ISR2L, in32(GPIO1_ISR2L) | 0x00010000);
 
 	/* external interrupts IRQ0...3 */
-	out32(GPIO1_TCR, in32(GPIO1_TCR) & ~0x0f000000);
-	out32(GPIO1_TSRL, in32(GPIO1_TSRL) & ~0x00005500);
+	out32(GPIO1_TCR, in32(GPIO1_TCR) & ~0x00f00000);
+	out32(GPIO1_TSRL, in32(GPIO1_TSRL) & ~0x0000ff00);
 	out32(GPIO1_ISR1L, in32(GPIO1_ISR1L) | 0x00005500);
 
 	/*setup USB 2.0 */

board/ixdp425/Makefile

@@ -25,7 +25,7 @@ include $(TOPDIR)/config.mk
 
 LIB	= lib$(BOARD).a
 
-OBJS	:= ixdp425.o flash.o
+OBJS	:= ixdp425.o
 
 $(LIB):	$(OBJS) $(SOBJS)
 	$(AR) crv $@ $^

board/ixdp425/config.mk

@@ -1,2 +1,4 @@
-#TEXT_BASE = 0x00100000
 TEXT_BASE = 0x00f80000
+
+# include NPE ethernet driver
+BOARDLIBS = cpu/ixp/npe/libnpe.a

board/ixdp425/ixdp425.c

@@ -1,4 +1,7 @@
 /*
+ * (C) Copyright 2006
+ * Stefan Roese, DENX Software Engineering, sr@denx.de.
+ *
  * (C) Copyright 2002
  * Kyle Harris, Nexus Technologies, Inc. kharris@nexus-tech.net
  *
@@ -25,24 +28,21 @@
  * MA 02111-1307 USA
  */
 
-#include <asm/arch/ixp425.h>
 #include <common.h>
+#include <command.h>
+#include <malloc.h>
+#include <asm/arch/ixp425.h>
 
 DECLARE_GLOBAL_DATA_PTR;
 
 /*
  * Miscelaneous platform dependent initialisations
  */
-
-/**********************************************************/
-
 int board_post_init (void)
 {
 	return (0);
 }
 
-/**********************************************************/
-
 int board_init (void)
 {
 	/* arch number of IXDP */
@@ -51,10 +51,58 @@ int board_init (void)
 	/* adress of boot parameters */
 	gd->bd->bi_boot_params = 0x00000100;
 
+#ifdef CONFIG_IXDPG425
+	/* arch number of IXDP */
+	gd->bd->bi_arch_number = MACH_TYPE_IXDPG425;
+
+	/*
+	 * Get realtek RTL8305 switch and SLIC out of reset
+	 */
+	GPIO_OUTPUT_SET(CFG_GPIO_SWITCH_RESET_N);
+	GPIO_OUTPUT_ENABLE(CFG_GPIO_SWITCH_RESET_N);
+	GPIO_OUTPUT_SET(CFG_GPIO_SLIC_RESET_N);
+	GPIO_OUTPUT_ENABLE(CFG_GPIO_SLIC_RESET_N);
+
+	/*
+	 * Setup GPIO's for PCI INTA & INTB
+	 */
+	GPIO_OUTPUT_DISABLE(CFG_GPIO_PCI_INTA_N);
+	GPIO_INT_ACT_LOW_SET(CFG_GPIO_PCI_INTA_N);
+	GPIO_OUTPUT_DISABLE(CFG_GPIO_PCI_INTB_N);
+	GPIO_INT_ACT_LOW_SET(CFG_GPIO_PCI_INTB_N);
+
+	/*
+	 * Setup GPIO's for 33MHz clock output
+	 */
+	*IXP425_GPIO_GPCLKR = 0x01FF01FF;
+	GPIO_OUTPUT_ENABLE(CFG_GPIO_PCI_CLK);
+	GPIO_OUTPUT_ENABLE(CFG_GPIO_EXTBUS_CLK);
+#endif
+
 	return 0;
 }
 
-/**********************************************************/
+/*
+ * Check Board Identity
+ */
+int checkboard(void)
+{
+	char *s = getenv("serial#");
+
+#ifdef CONFIG_IXDPG425
+	puts("Board: IXDPG425 - Intel Network Gateway Reference Platform");
+#else
+	puts("Board: IXDP425 - Intel Development Platform");
+#endif
+
+	if (s != NULL) {
+		puts(", serial# ");
+		puts(s);
+	}
+	putc('\n');
+
+	return (0);
+}
 
 int dram_init (void)
 {
@@ -64,8 +112,7 @@ int dram_init (void)
 	return (0);
 }
 
-/**********************************************************/
-
+#if (CONFIG_COMMANDS & CFG_CMD_PCI) || defined(CONFIG_PCI)
 extern struct pci_controller hose;
 extern void pci_ixp_init(struct pci_controller * hose);
 
@@ -75,3 +122,4 @@ void pci_init_board(void)
 
 	pci_ixp_init(&hose);
 }
+#endif

board/m5271evb/m5271evb.c

@@ -53,10 +53,9 @@ long int initdram (int board_type) {
 	 * Check to see if the SDRAM has already been initialized
 	 * by a run control tool
 	 */
-	if (!(mbar_readLong(MCF_SDRAMC_DACR0) & MCF_SDRAMC_DACRn_RE))
-	{
+	if (!(mbar_readLong(MCF_SDRAMC_DACR0) & MCF_SDRAMC_DACRn_RE)) {
 		/* Initialize DRAM Control Register: DCR */
-		mbar_writeShort(MCF_SDRAMC_DCR, 
+		mbar_writeShort(MCF_SDRAMC_DCR,
 				MCF_SDRAMC_DCR_RTIM(0x01)
 				| MCF_SDRAMC_DCR_RC(0x30));
 
@@ -74,7 +73,7 @@ long int initdram (int board_type) {
 				| MCF_SDRAMC_DACRn_PS(0));
 
 		/* Initialize DMR0 */
-		mbar_writeLong(MCF_SDRAMC_DMR0, 
+		mbar_writeLong(MCF_SDRAMC_DMR0,
 				MCF_SDRAMC_DMRn_BAM_16M
 				| MCF_SDRAMC_DMRn_V);
 

board/omap1610inn/lowlevel_init.S

@@ -382,7 +382,7 @@ REG_WATCHDOG:
 	.word 0xfffec808
 
 REG_MPU_LOAD_TIMER:
-	.word 0xfffec600
+	.word 0xfffec504
 REG_MPU_CNTL_TIMER:
 	.word 0xfffec500
 

board/omap5912osk/lowlevel_init.S

@@ -41,6 +41,13 @@ _TEXT_BASE:
 .globl lowlevel_init
 lowlevel_init:
 
+	/*------------------------------------------------------*
+	 * Ensure i-cache is enabled                            *
+	 * To configure TC regs without fetching instruction    *
+	 *------------------------------------------------------*/
+	mrc	p15, 0, r0, c1, c0
+	orr	r0, r0, #0x1000
+	mcr	p15, 0, r0, c1, c0
 
 	/*------------------------------------------------------*
 	 *mask all IRQs by setting all bits in the INTMR default*
@@ -59,33 +66,34 @@ lowlevel_init:
 	str	r1,	[r0]
 
 	/*------------------------------------------------------*
-	 * Set up ARM CLM registers (IDLECT2)  		        *
+	 * Set up ARM CLM registers (IDLECT2)                   *
 	 *------------------------------------------------------*/
 	ldr	r0,	REG_ARM_IDLECT2
 	ldr	r1,	VAL_ARM_IDLECT2
 	str	r1,	[r0]
 
 	/*------------------------------------------------------*
-	 * Set up ARM CLM registers (IDLECT3)		        *
+	 * Set up ARM CLM registers (IDLECT3)                   *
 	 *------------------------------------------------------*/
 	ldr	r0,	REG_ARM_IDLECT3
 	ldr	r1,	VAL_ARM_IDLECT3
 	str	r1,	[r0]
 
-
-	mov	r1,	#0x01		/* PER_EN bit */
+	mov	r1,	#0x01               /* PER_EN bit */
 	ldr	r0,	REG_ARM_RSTCT2
-	strh	r1,	[r0]		/* CLKM; Peripheral reset. */
+	strh	r1,	[r0]            /* CLKM; Peripheral reset. */
 
-	/* Set CLKM to Sync-Scalable	*/
-	/* I supposedly need to enable the dsp clock before switching */
-	mov	r1,	#0x0000
+	/* Set CLKM to Sync-Scalable */
+	mov	r1,	#0x1000
 	ldr	r0,	REG_ARM_SYSST
-	strh	r1,	[r0]
-	mov	r0,	#0x400
-1:
-	subs	r0,	r0,	#0x1	/* wait for any bubbles to finish */
+
+	mov	r2,	#0
+1:	cmp	r2,	#1
+	streqh	r1,	[r0]
+	add	r2,	r2,	#1
+	cmp	r2,	#0x100              /* wait for any bubbles to finish */
 	bne	1b
+
 	ldr	r1,	VAL_ARM_CKCTL
 	ldr	r0,	REG_ARM_CKCTL
 	strh	r1,	[r0]
@@ -107,17 +115,16 @@ lowlevel_init:
 	ldr	r1,	VAL_DPLL1_CTL
 	ldr	r0,	REG_DPLL1_CTL
 	strh	r1,	[r0]
-	ands	r1,	r1,	#0x10	/* Check if PLL is enabled. */
-	beq	lock_end	/* Do not look for lock if BYPASS selected */
+	ands	r1,	r1,	#0x10       /* Check if PLL is enabled. */
+	beq	lock_end                /* Do not look for lock if BYPASS selected */
 2:
 	ldrh	r1,	[r0]
-	ands	r1,	r1,	#0x01	/*	Check the LOCK bit.*/
-	beq 2b			/*	loop until bit goes hi. */
+	ands	r1,	r1,	#0x01       /* Check the LOCK bit.*/
+	beq 2b                      /* loop until bit goes hi. */
 lock_end:
 
-
 	/*------------------------------------------------------*
-	 * Turn off the watchdog during init...			*
+	 * Turn off the watchdog during init...                 *
  	 *------------------------------------------------------*/
 	ldr	r0,	REG_WATCHDOG
 	ldr	r1,	WATCHDOG_VAL1
@@ -143,30 +150,49 @@ watch2Wait:
 	tst	r1,	#0x10
 	bne	watch2Wait
 
-
 	/* Set memory timings corresponding to the new clock speed */
+	ldr	r3, VAL_SDRAM_CONFIG_SDF0
 
 	/* Check execution location to determine current execution location
 	 * and branch to appropriate initialization code.
 	 */
-	/* Load physical SDRAM base. */
-	mov 	r0, 	#0x10000000
-	/* Get current execution location. */
-	mov 	r1, 	pc
-	/* Compare. */
-	cmp 	r1, 	r0
-	/* Skip over EMIF-fast initialization if running from SDRAM. */
-	bge 	skip_sdram
+	mov	r0,	#0x10000000         /* Load physical SDRAM base. */
+	mov	r1,	pc                  /* Get current execution location. */
+	cmp	r1,	r0                  /* Compare. */
+	bge	skip_sdram              /* Skip over EMIF-fast initialization if running from SDRAM. */
+
+	/* identify the device revision, -- TMX or TMP(TMS) */
+	ldr	r0, REG_DEVICE_ID
+	ldr	r1, [r0]
+
+	ldr	r0, VAL_DEVICE_ID_TMP
+	mov	r1, r1, lsl #15
+	mov	r1, r1, lsr #16
+	cmp	r0, r1
+	bne	skip_TMP_Patch
+
+	/* Enable TMP/TMS device new features */
+	mov	r0, #1
+	ldr	r1, REG_TC_EMIFF_DOUBLER
+	str	r0, [r1]
+
+	/* Enable new ac parameters */
+	mov	r0, #0x0b
+	ldr	r1, REG_SDRAM_CONFIG2
+	str	r0, [r1]
+
+	ldr	r3, VAL_SDRAM_CONFIG_SDF1
+
+skip_TMP_Patch:
 
 	/*
 	* Delay for SDRAM initialization.
 	*/
-	mov	r3,	#0x1800		/* value should be checked */
+	mov	r0,	#0x1800		/* value should be checked */
 3:
-	subs	r3,	r3,	#0x1	/* Decrement count */
+	subs	r0,	r0,	#0x1	/* Decrement count */
 	bne 	3b
 
-
 	/*
 	 * Set SDRAM control values. Disable refresh before MRS command.
 	 */
@@ -178,14 +204,15 @@ watch2Wait:
 
 	/* config register */
 	ldr	r0,	REG_SDRAM_CONFIG
-	ldr	r1,	SDRAM_CONFIG_VAL
-	str	r1,	[r0]
+	str	r3,	[r0]
 
 	/* manual command register */
 	ldr	r0,	REG_SDRAM_MANUAL_CMD
+
 	/* issue set cke high */
 	mov	r1,	#CMD_SDRAM_CKE_SET_HIGH
 	str	r1,	[r0]
+
 	/* issue nop */
 	mov	r1,	#CMD_SDRAM_NOP
 	str	r1,	[r0]
@@ -228,25 +255,23 @@ waitMDDR1:
 	str	r1,	[r0]
 
 	/* delay loop */
-	mov	r2,	#0x0100
+	mov	r0,	#0x0100
 waitMDDR2:
-	subs	r2,	r2,	#1
+	subs	r0,	r0,	#1
 	bne	waitMDDR2
 
 	/*
 	 * Delay for SDRAM initialization.
 	 */
-	mov	r3,	#0x1800
+	mov	r0,	#0x1800
 4:
-	subs	r3,	r3,	#1	/* Decrement count. */
+	subs	r0,	r0,	#1	/* Decrement count. */
 	bne	4b
 	b	common_tc
 
 skip_sdram:
-
 	ldr	r0,	REG_SDRAM_CONFIG
-	ldr	r1,	SDRAM_CONFIG_VAL
-	str	r1,	[r0]
+	str	r3,	[r0]
 
 common_tc:
 	/* slow interface */
@@ -257,10 +282,15 @@ common_tc:
 	ldr	r1,	VAL_TC_EMIFS_CS1_CONFIG
 	ldr	r0,	REG_TC_EMIFS_CS1_CONFIG
 	str	r1,	[r0] /* Chip Select 1 */
+
 	ldr	r1,	VAL_TC_EMIFS_CS3_CONFIG
 	ldr	r0,	REG_TC_EMIFS_CS3_CONFIG
 	str	r1,	[r0] /* Chip Select 3 */
 
+	ldr	r1,	VAL_TC_EMIFS_DWS
+	ldr	r0,	REG_TC_EMIFS_DWS
+	str	r1,	[r0] /* Enable EMIFS.RDY for CS1 (ether) */
+
 #ifdef CONFIG_H2_OMAP1610
 	/* inserting additional 2 clock cycle hold time for LAN */
 	ldr     r0,     REG_TC_EMIFS_CS1_ADVANCED
@@ -282,8 +312,9 @@ common_tc:
 	/* the literal pools origin */
 	.ltorg
 
-
-REG_TC_EMIFS_CONFIG:		/* 32 bits */
+REG_DEVICE_ID:			/* 32 bits */
+	.word 0xfffe2004
+REG_TC_EMIFS_CONFIG:
 	.word 0xfffecc0c
 REG_TC_EMIFS_CS0_CONFIG:	/* 32 bits */
 	.word 0xfffecc10
@@ -293,7 +324,8 @@ REG_TC_EMIFS_CS2_CONFIG:	/* 32 bits */
 	.word 0xfffecc18
 REG_TC_EMIFS_CS3_CONFIG:	/* 32 bits */
 	.word 0xfffecc1c
-
+REG_TC_EMIFS_DWS:		/* 32 bits */
+	.word 0xfffecc40
 #ifdef CONFIG_H2_OMAP1610
 REG_TC_EMIFS_CS1_ADVANCED:	/* 32 bits */
 	.word 0xfffecc54
@@ -302,18 +334,17 @@ REG_TC_EMIFS_CS1_ADVANCED:	/* 32 bits */
 /* MPU clock/reset/power mode control registers */
 REG_ARM_CKCTL:			/* 16 bits */
 	.word 0xfffece00
-
 REG_ARM_IDLECT3:		/* 16 bits */
 	.word 0xfffece24
 REG_ARM_IDLECT2:		/* 16 bits */
 	.word 0xfffece08
 REG_ARM_IDLECT1:		/* 16 bits */
 	.word 0xfffece04
-
 REG_ARM_RSTCT2:			/* 16 bits */
 	.word 0xfffece14
 REG_ARM_SYSST:			/* 16 bits */
 	.word 0xfffece18
+
 /* DPLL control registers */
 REG_DPLL1_CTL:			/* 16 bits */
 	.word 0xfffecf00
@@ -335,6 +366,10 @@ WSPRDOG_VAL2:
  counter @8192 rows, 10 ns, 8 burst */
 REG_SDRAM_CONFIG:
 	.word 0xfffecc20
+REG_SDRAM_CONFIG2:
+	.word 0xfffecc3c
+REG_TC_EMIFF_DOUBLER:		/* 32 bits */
+	.word 0xfffecc60
 
 /* Operation register */
 REG_SDRAM_OPERATION:
@@ -356,35 +391,47 @@ REG_SDRAM_EMRS1:
 REG_DLL_WRT_CONTROL:
 	.word 0xfffecc68
 DLL_WRT_CONTROL_VAL:
-	.word 0x03f00002
+	.word 0x03f00002    /* Phase of 72deg, write offset +31  */
 
 /* URD DLL register */
 REG_DLL_URD_CONTROL:
 	.word 0xfffeccc0
 DLL_URD_CONTROL_VAL:
-	.word 0x00800002
+	.word 0x00800002    /* Phase of 72deg, read offset +31  */
 
 /* LRD DLL register */
 REG_DLL_LRD_CONTROL:
 	.word 0xfffecccc
+DLL_LRD_CONTROL_VAL:
+	.word 0x00800002    /* read offset +31  */
 
 REG_WATCHDOG:
 	.word 0xfffec808
+WATCHDOG_VAL1:
+	.word 0x000000f5
+WATCHDOG_VAL2:
+	.word 0x000000a0
 
 REG_MPU_LOAD_TIMER:
-	.word 0xfffec600
+	.word 0xfffec504
 REG_MPU_CNTL_TIMER:
 	.word 0xfffec500
+VAL_MPU_LOAD_TIMER:
+	.word 0xffffffff
+VAL_MPU_CNTL_TIMER:
+	.word 0xffffffa1
 
 /* 96 MHz Samsung Mobile DDR */
-SDRAM_CONFIG_VAL:
-	.word 0x001200f4
+/* Original setting for TMX device */
+VAL_SDRAM_CONFIG_SDF0:
+    .word 0x0014e6fe
 
-DLL_LRD_CONTROL_VAL:
-	.word 0x00800002
+/* NEW_SYS_FREQ mode (valid only TMP/TMS devices) */
+VAL_SDRAM_CONFIG_SDF1:
+    .word 0x0114e6fe
 
 VAL_ARM_CKCTL:
-	.word 0x3000
+	.word 0x2000		/* was:	0x3000, now use CLK_REF for timer input */
 VAL_DPLL1_CTL:
 	.word 0x2830
 
@@ -392,11 +439,15 @@ VAL_DPLL1_CTL:
 VAL_TC_EMIFS_CS0_CONFIG:
 	.word 0x002130b0
 VAL_TC_EMIFS_CS1_CONFIG:
-	.word 0x00001131
+	.word 0x00001133
 VAL_TC_EMIFS_CS2_CONFIG:
 	.word 0x000055f0
 VAL_TC_EMIFS_CS3_CONFIG:
-	.word 0x88011131
+	.word 0x88013141
+VAL_TC_EMIFS_DWS:	/* Enable EMIFS.RDY for CS1 access (ether) */
+	.word 0x000000c0
+VAL_DEVICE_ID_TMP:	/* TMP/TMS=0xb65f, TMX=0xb58c */
+	.word 0xb65f
 #endif
 
 #ifdef CONFIG_H2_OMAP1610
@@ -407,36 +458,20 @@ VAL_TC_EMIFS_CS1_CONFIG:
 VAL_TC_EMIFS_CS2_CONFIG:
 	.word 0xf800f22a
 VAL_TC_EMIFS_CS3_CONFIG:
-	.word 0x88011131
+	.word 0x88013141
 VAL_TC_EMIFS_CS1_ADVANCED:
 	.word 0x00000022
 #endif
 
-VAL_TC_EMIFF_SDRAM_CONFIG:
-	.word 0x010290fc
-VAL_TC_EMIFF_MRS:
-	.word 0x00000027
-
 VAL_ARM_IDLECT1:
 	.word 0x00000400
-
 VAL_ARM_IDLECT2:
 	.word 0x00000886
 VAL_ARM_IDLECT3:
 	.word 0x00000015
 
-WATCHDOG_VAL1:
-	.word 0x000000f5
-WATCHDOG_VAL2:
-	.word 0x000000a0
-
-VAL_MPU_LOAD_TIMER:
-	.word 0xffffffff
-VAL_MPU_CNTL_TIMER:
-	.word 0xffffffa1
-
 /* command values */
-.equ CMD_SDRAM_NOP,		0x00000000
-.equ CMD_SDRAM_PRECHARGE,	0x00000001
-.equ CMD_SDRAM_AUTOREFRESH,	0x00000002
-.equ CMD_SDRAM_CKE_SET_HIGH,	0x00000007
+.equ CMD_SDRAM_NOP,             0x00000000
+.equ CMD_SDRAM_PRECHARGE,       0x00000001
+.equ CMD_SDRAM_AUTOREFRESH,     0x00000002
+.equ CMD_SDRAM_CKE_SET_HIGH,    0x00000007

board/omap5912osk/omap5912osk.c

@@ -288,3 +288,21 @@ void peripheral_power_enable (void)
 
 	*SW_CLOCK_REQUEST |= UART1_48MHZ_ENABLE;
 }
+
+/*
+ * Check Board Identity
+ */
+int checkboard(void)
+{
+	char *s = getenv("serial#");
+
+	puts("Board: OSK5912");
+
+	if (s != NULL) {
+		puts(", serial# ");
+		puts(s);
+	}
+	putc('\n');
+
+	return (0);
+}

board/omap730p2/lowlevel_init.S

@@ -317,7 +317,7 @@ REG_WATCHDOG:
 	.word 0xfffec808
 
 REG_MPU_LOAD_TIMER:
-	.word 0xfffec600
+	.word 0xfffec504
 REG_MPU_CNTL_TIMER:
 	.word 0xfffec500
 

board/pcs440ep/Makefile

@@ -0,0 +1,47 @@
+#
+# (C) Copyright 2006
+# Wolfgang Denk, DENX Software Engineering, wd@denx.de.
+#
+# See file CREDITS for list of people who contributed to this
+# project.
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation; either version 2 of
+# the License, or (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+# MA 02111-1307 USA
+#
+
+include $(TOPDIR)/config.mk
+
+LIB	= lib$(BOARD).a
+
+OBJS	= $(BOARD).o flash.o
+SOBJS	= init.o
+
+$(LIB):	$(OBJS) $(SOBJS)
+	$(AR) crv $@ $(OBJS)
+
+clean:
+	rm -f $(SOBJS) $(OBJS)
+
+distclean:	clean
+	rm -f $(LIB) core *.bak .depend
+
+#########################################################################
+
+.depend:	Makefile $(SOBJS:.o=.S) $(OBJS:.o=.c)
+		$(CC) -M $(CFLAGS) $(SOBJS:.o=.S) $(OBJS:.o=.c) > $@
+
+sinclude .depend
+
+#########################################################################

board/pcs440ep/config.mk

@@ -0,0 +1,44 @@
+#
+# (C) Copyright 2006
+# Wolfgang Denk, DENX Software Engineering, wd@denx.de.
+#
+# See file CREDITS for list of people who contributed to this
+# project.
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation; either version 2 of
+# the License, or (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+# MA 02111-1307 USA
+#
+
+#
+# PCS440EP board
+#
+
+#TEXT_BASE = 0x00001000
+
+ifeq ($(ramsym),1)
+TEXT_BASE = 0xFBD00000
+else
+TEXT_BASE = 0xFFFA0000
+endif
+
+PLATFORM_CPPFLAGS += -DCONFIG_440=1
+
+ifeq ($(debug),1)
+PLATFORM_CPPFLAGS += -DDEBUG
+endif
+
+ifeq ($(dbcr),1)
+PLATFORM_CPPFLAGS += -DCFG_INIT_DBCR=0x8cff0000
+endif

board/pcs440ep/flash.c

@@ -0,0 +1,608 @@
+/*
+ * (C) Copyright 2006
+ * Stefan Roese, DENX Software Engineering, sr@denx.de.
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+#include <asm/processor.h>
+
+#ifndef CFG_FLASH_READ0
+#define CFG_FLASH_READ0		0x0000	/* 0 is standard			*/
+#define CFG_FLASH_READ1		0x0001	/* 1 is standard			*/
+#define CFG_FLASH_READ2		0x0002	/* 2 is standard			*/
+#endif
+
+flash_info_t	flash_info[CFG_MAX_FLASH_BANKS]; /* info for FLASH chips	*/
+
+/*
+ * Functions
+ */
+static int write_word(flash_info_t *info, ulong dest, ulong data);
+static ulong flash_get_size(vu_long *addr, flash_info_t *info);
+
+unsigned long flash_init(void)
+{
+	unsigned long size_b0, size_b1;
+	int i;
+	unsigned long base_b0, base_b1;
+
+	/* Init: no FLASHes known */
+	for (i = 0; i < CFG_MAX_FLASH_BANKS; ++i) {
+		flash_info[i].flash_id = FLASH_UNKNOWN;
+	}
+
+	/* Static FLASH Bank configuration here - FIXME XXX */
+
+	base_b0 = FLASH_BASE0_PRELIM;
+	size_b0 = flash_get_size ((vu_long *) base_b0, &flash_info[0]);
+
+	if (flash_info[0].flash_id == FLASH_UNKNOWN) {
+		printf ("## Unknown FLASH on Bank 0 - Size = 0x%08lx = %ld MB\n",
+				size_b0, size_b0 << 20);
+	}
+
+	base_b1 = FLASH_BASE1_PRELIM;
+	size_b1 = flash_get_size ((vu_long *) base_b1, &flash_info[1]);
+
+	return (size_b0 + size_b1);
+}
+
+void flash_print_info(flash_info_t *info)
+{
+	int i;
+	int k;
+	int size;
+	int erased;
+	volatile unsigned long *flash;
+
+	if (info->flash_id == FLASH_UNKNOWN) {
+		printf ("missing or unknown FLASH type\n");
+		return;
+	}
+
+	switch (info->flash_id & FLASH_VENDMASK) {
+	case FLASH_MAN_AMD:	printf ("AMD ");		break;
+	case FLASH_MAN_FUJ:	printf ("FUJITSU ");		break;
+	case FLASH_MAN_SST:	printf ("SST ");		break;
+	case FLASH_MAN_EXCEL:	printf ("Excel Semiconductor "); break;
+	default:		printf ("Unknown Vendor ");	break;
+	}
+
+	switch (info->flash_id & FLASH_TYPEMASK) {
+	case FLASH_AM400B:	printf ("AM29LV400B (4 Mbit, bottom boot sect)\n");
+		break;
+	case FLASH_AM400T:	printf ("AM29LV400T (4 Mbit, top boot sector)\n");
+		break;
+	case FLASH_AM040:	printf ("AM29LV040B (4 Mbit, uniform sector size)\n");
+		break;
+	case FLASH_AM800B:	printf ("AM29LV800B (8 Mbit, bottom boot sect)\n");
+		break;
+	case FLASH_AM800T:	printf ("AM29LV800T (8 Mbit, top boot sector)\n");
+		break;
+	case FLASH_AM160B:	printf ("AM29LV160B (16 Mbit, bottom boot sect)\n");
+		break;
+	case FLASH_AM160T:	printf ("AM29LV160T (16 Mbit, top boot sector)\n");
+		break;
+	case FLASH_AM320T:	printf ("AM29LV320T (32 M, top sector)\n");
+		break;
+	case FLASH_AM320B:	printf ("AM29LV320B (32 M, bottom sector)\n");
+		break;
+	case FLASH_AMDL322T:	printf ("AM29DL322T (32 M, top sector)\n");
+		break;
+	case FLASH_AMDL322B:	printf ("AM29DL322B (32 M, bottom sector)\n");
+		break;
+	case FLASH_AMDL323T:	printf ("AM29DL323T (32 M, top sector)\n");
+		break;
+	case FLASH_AMDL323B:	printf ("AM29DL323B (32 M, bottom sector)\n");
+		break;
+	case FLASH_SST020:	printf ("SST39LF/VF020 (2 Mbit, uniform sector size)\n");
+		break;
+	case FLASH_SST040:	printf ("SST39LF/VF040 (4 Mbit, uniform sector size)\n");
+		break;
+	default:		printf ("Unknown Chip Type\n");
+		break;
+	}
+
+	printf ("  Size: %ld MB in %d Sectors\n",
+		info->size >> 20, info->sector_count);
+
+	printf ("  Sector Start Addresses:");
+	for (i=0; i<info->sector_count; ++i) {
+#ifdef CFG_FLASH_EMPTY_INFO
+		/*
+		 * Check if whole sector is erased
+		 */
+		if (i != (info->sector_count-1))
+			size = info->start[i+1] - info->start[i];
+		else
+			size = info->start[0] + info->size - info->start[i];
+		erased = 1;
+		flash = (volatile unsigned long *)info->start[i];
+		size = size >> 2;	/* divide by 4 for longword access */
+		for (k=0; k<size; k++) {
+			if (*flash++ != 0xffffffff) {
+				erased = 0;
+				break;
+			}
+		}
+
+		if ((i % 5) == 0)
+			printf ("\n   ");
+		/* print empty and read-only info */
+		printf (" %08lX%s%s",
+			info->start[i],
+			erased ? " E" : "  ",
+			info->protect[i] ? "RO " : "   ");
+#else
+		if ((i % 5) == 0)
+			printf ("\n   ");
+		printf (" %08lX%s",
+			info->start[i],
+			info->protect[i] ? " (RO)" : "     ");
+#endif
+
+	}
+	printf ("\n");
+	return;
+}
+
+/*
+ * The following code cannot be run from FLASH!
+ */
+static ulong flash_get_size(vu_long *addr, flash_info_t *info)
+{
+	short i;
+	short n;
+	volatile CFG_FLASH_WORD_SIZE value;
+	ulong base = (ulong)addr;
+	volatile CFG_FLASH_WORD_SIZE *addr2 = (volatile CFG_FLASH_WORD_SIZE *)addr;
+
+	/* Write auto select command: read Manufacturer ID */
+	addr2[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+	addr2[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+	addr2[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00900090;
+
+	value = addr2[CFG_FLASH_READ0];
+
+	switch (value) {
+	case (CFG_FLASH_WORD_SIZE)AMD_MANUFACT:
+		info->flash_id = FLASH_MAN_AMD;
+		break;
+	case (CFG_FLASH_WORD_SIZE)FUJ_MANUFACT:
+		info->flash_id = FLASH_MAN_FUJ;
+		break;
+	case (CFG_FLASH_WORD_SIZE)SST_MANUFACT:
+		info->flash_id = FLASH_MAN_SST;
+		break;
+	case (CFG_FLASH_WORD_SIZE)EXCEL_MANUFACT:
+		info->flash_id = FLASH_MAN_EXCEL;
+		break;
+	default:
+		info->flash_id = FLASH_UNKNOWN;
+		info->sector_count = 0;
+		info->size = 0;
+		return (0);			/* no or unknown flash	*/
+	}
+
+	value = addr2[CFG_FLASH_READ1];		/* device ID	*/
+
+	switch (value) {
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV400T:
+		info->flash_id += FLASH_AM400T;
+		info->sector_count = 11;
+		info->size = 0x00080000;
+		break;				/* => 0.5 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV400B:
+		info->flash_id += FLASH_AM400B;
+		info->sector_count = 11;
+		info->size = 0x00080000;
+		break;				/* => 0.5 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV040B:
+		info->flash_id += FLASH_AM040;
+		info->sector_count = 8;
+		info->size = 0x0080000;		/* => 0.5 MB	*/
+		break;
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV800T:
+		info->flash_id += FLASH_AM800T;
+		info->sector_count = 19;
+		info->size = 0x00100000;
+		break;				/* => 1 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV800B:
+		info->flash_id += FLASH_AM800B;
+		info->sector_count = 19;
+		info->size = 0x00100000;
+		break;				/* => 1 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV160T:
+		info->flash_id += FLASH_AM160T;
+		info->sector_count = 35;
+		info->size = 0x00200000;
+		break;				/* => 2 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV160B:
+		info->flash_id += FLASH_AM160B;
+		info->sector_count = 35;
+		info->size = 0x00200000;
+		break;				/* => 2 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV320T:
+		info->flash_id += FLASH_AM320T;
+		info->sector_count = 71;
+		info->size = 0x00400000;
+		break;				/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV320B:
+		info->flash_id += FLASH_AM320B;
+		info->sector_count = 71;
+		info->size = 0x00400000;
+		break;				/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_DL322T:
+		info->flash_id += FLASH_AMDL322T;
+		info->sector_count = 71;
+		info->size = 0x00400000;
+		break;				/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_DL322B:
+		info->flash_id += FLASH_AMDL322B;
+		info->sector_count = 71;
+		info->size = 0x00400000;
+		break;				/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_DL323T:
+		info->flash_id += FLASH_AMDL323T;
+		info->sector_count = 71;
+		info->size = 0x00400000;
+		break;				/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_DL323B:
+		info->flash_id += FLASH_AMDL323B;
+		info->sector_count = 71;
+		info->size = 0x00400000;
+		break;				/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)SST_ID_xF020:
+		info->flash_id += FLASH_SST020;
+		info->sector_count = 64;
+		info->size = 0x00040000;
+		break;				/* => 256 kB	*/
+
+	case (CFG_FLASH_WORD_SIZE)SST_ID_xF040:
+		info->flash_id += FLASH_SST040;
+		info->sector_count = 128;
+		info->size = 0x00080000;
+		break;				/* => 512 kB	*/
+
+	default:
+		info->flash_id = FLASH_UNKNOWN;
+		return (0);			/* => no or unknown flash */
+
+	}
+
+	/* set up sector start address table */
+	if (((info->flash_id & FLASH_VENDMASK) == FLASH_MAN_SST) ||
+	    ((info->flash_id & FLASH_TYPEMASK) == FLASH_AM640U)) {
+		for (i = 0; i < info->sector_count; i++)
+			info->start[i] = base + (i * 0x00001000);
+	} else if ((info->flash_id & FLASH_TYPEMASK) == FLASH_AM040) {
+		for (i = 0; i < info->sector_count; i++)
+			info->start[i] = base + (i * 0x00010000);
+	} else if (((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL322B) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL323B) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AM320B) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL324B)) {
+		/* set sector offsets for bottom boot block type	*/
+		for (i=0; i<8; ++i) {		/*  8 x 8k boot sectors	*/
+			info->start[i] = base;
+			base += 8 << 10;
+		}
+		while (i < info->sector_count) {	/* 64k regular sectors	*/
+			info->start[i] = base;
+			base += 64 << 10;
+			++i;
+		}
+	} else if (((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL322T) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL323T) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AM320T) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL324T)) {
+		/* set sector offsets for top boot block type		*/
+		base += info->size;
+		i = info->sector_count;
+		for (n=0; n<8; ++n) {		/*  8 x 8k boot sectors	*/
+			base -= 8 << 10;
+			--i;
+			info->start[i] = base;
+		}
+		while (i > 0) {			/* 64k regular sectors	*/
+			base -= 64 << 10;
+			--i;
+			info->start[i] = base;
+		}
+	} else {
+		if (info->flash_id & FLASH_BTYPE) {
+			/* set sector offsets for bottom boot block type	*/
+			info->start[0] = base + 0x00000000;
+			info->start[1] = base + 0x00004000;
+			info->start[2] = base + 0x00006000;
+			info->start[3] = base + 0x00008000;
+			for (i = 4; i < info->sector_count; i++) {
+				info->start[i] = base + (i * 0x00010000) - 0x00030000;
+			}
+		} else {
+			/* set sector offsets for top boot block type		*/
+			i = info->sector_count - 1;
+			info->start[i--] = base + info->size - 0x00004000;
+			info->start[i--] = base + info->size - 0x00006000;
+			info->start[i--] = base + info->size - 0x00008000;
+			for (; i >= 0; i--) {
+				info->start[i] = base + i * 0x00010000;
+			}
+		}
+	}
+
+	/* check for protected sectors */
+	for (i = 0; i < info->sector_count; i++) {
+		/* read sector protection at sector address, (A7 .. A0) = 0x02 */
+		/* D0 = 1 if protected */
+		addr2 = (volatile CFG_FLASH_WORD_SIZE *)(info->start[i]);
+		if ((info->flash_id & FLASH_VENDMASK) != FLASH_MAN_AMD)
+			info->protect[i] = 0;
+		else
+			info->protect[i] = addr2[CFG_FLASH_READ2] & 1;
+	}
+
+	/*
+	 * Prevent writes to uninitialized FLASH.
+	 */
+	if (info->flash_id != FLASH_UNKNOWN) {
+		addr2 = (CFG_FLASH_WORD_SIZE *)info->start[0];
+		*addr2 = (CFG_FLASH_WORD_SIZE)0x00F000F0;	/* reset bank */
+	}
+
+	return (info->size);
+}
+
+
+int flash_erase(flash_info_t *info, int s_first, int s_last)
+{
+	volatile CFG_FLASH_WORD_SIZE *addr = (CFG_FLASH_WORD_SIZE *)(info->start[0]);
+	volatile CFG_FLASH_WORD_SIZE *addr2;
+	int flag, prot, sect, l_sect;
+	ulong start, now, last;
+
+	if ((s_first < 0) || (s_first > s_last)) {
+		if (info->flash_id == FLASH_UNKNOWN)
+			printf ("- missing\n");
+		else
+			printf ("- no sectors to erase\n");
+		return 1;
+	}
+
+	if (info->flash_id == FLASH_UNKNOWN) {
+		printf ("Can't erase unknown flash type - aborted\n");
+		return 1;
+	}
+
+	prot = 0;
+	for (sect=s_first; sect<=s_last; ++sect)
+		if (info->protect[sect])
+			prot++;
+
+	if (prot)
+		printf ("- Warning: %d protected sectors will not be erased!\n", prot);
+	else
+		printf ("\n");
+
+	l_sect = -1;
+
+	/* Disable interrupts which might cause a timeout here */
+	flag = disable_interrupts();
+
+	/* Start erase on unprotected sectors */
+	for (sect = s_first; sect<=s_last; sect++) {
+		if (info->protect[sect] == 0) {	/* not protected */
+			addr2 = (CFG_FLASH_WORD_SIZE *)(info->start[sect]);
+			if ((info->flash_id & FLASH_VENDMASK) == FLASH_MAN_SST) {
+				addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+				addr[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+				addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00800080;
+				addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+				addr[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+				addr2[0] = (CFG_FLASH_WORD_SIZE)0x00300030;  /* sector erase */
+
+				/* re-enable interrupts if necessary */
+				if (flag) {
+					enable_interrupts();
+					flag = 0;
+				}
+
+				/* data polling for D7 */
+				start = get_timer (0);
+				while ((addr2[0] & (CFG_FLASH_WORD_SIZE)0x00800080) !=
+				       (CFG_FLASH_WORD_SIZE)0x00800080) {
+					if (get_timer(start) > CFG_FLASH_WRITE_TOUT)
+						return (1);
+				}
+			} else {
+				if (sect == s_first) {
+					addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+					addr[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+					addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00800080;
+					addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+					addr[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+				}
+				addr2[0] = (CFG_FLASH_WORD_SIZE)0x00300030;  /* sector erase */
+			}
+			l_sect = sect;
+		}
+	}
+
+	/* re-enable interrupts if necessary */
+	if (flag)
+		enable_interrupts();
+
+	/* wait at least 80us - let's wait 1 ms */
+	udelay (1000);
+
+	/*
+	 * We wait for the last triggered sector
+	 */
+	if (l_sect < 0)
+		goto DONE;
+
+	start = get_timer (0);
+	last  = start;
+	addr = (CFG_FLASH_WORD_SIZE *)(info->start[l_sect]);
+	while ((addr[0] & (CFG_FLASH_WORD_SIZE)0x00800080) != (CFG_FLASH_WORD_SIZE)0x00800080) {
+		if ((now = get_timer(start)) > CFG_FLASH_ERASE_TOUT) {
+			printf ("Timeout\n");
+			return 1;
+		}
+		/* show that we're waiting */
+		if ((now - last) > 1000) {	/* every second */
+			putc ('.');
+			last = now;
+		}
+	}
+
+DONE:
+	/* reset to read mode */
+	addr = (CFG_FLASH_WORD_SIZE *)info->start[0];
+	addr[0] = (CFG_FLASH_WORD_SIZE)0x00F000F0;	/* reset bank */
+
+	printf (" done\n");
+	return 0;
+}
+
+/*
+ * Copy memory to flash, returns:
+ * 0 - OK
+ * 1 - write timeout
+ * 2 - Flash not erased
+ */
+int write_buff(flash_info_t *info, uchar *src, ulong addr, ulong cnt)
+{
+	ulong cp, wp, data;
+	int i, l, rc;
+
+	wp = (addr & ~3);	/* get lower word aligned address */
+
+	/*
+	 * handle unaligned start bytes
+	 */
+	if ((l = addr - wp) != 0) {
+		data = 0;
+		for (i=0, cp=wp; i<l; ++i, ++cp) {
+			data = (data << 8) | (*(uchar *)cp);
+		}
+		for (; i<4 && cnt>0; ++i) {
+			data = (data << 8) | *src++;
+			--cnt;
+			++cp;
+		}
+		for (; cnt==0 && i<4; ++i, ++cp) {
+			data = (data << 8) | (*(uchar *)cp);
+		}
+
+		if ((rc = write_word(info, wp, data)) != 0) {
+			return (rc);
+		}
+		wp += 4;
+	}
+
+	/*
+	 * handle word aligned part
+	 */
+	while (cnt >= 4) {
+		data = 0;
+		for (i=0; i<4; ++i)
+			data = (data << 8) | *src++;
+		if ((rc = write_word(info, wp, data)) != 0)
+			return (rc);
+		wp  += 4;
+		cnt -= 4;
+	}
+
+	if (cnt == 0)
+		return (0);
+
+	/*
+	 * handle unaligned tail bytes
+	 */
+	data = 0;
+	for (i=0, cp=wp; i<4 && cnt>0; ++i, ++cp) {
+		data = (data << 8) | *src++;
+		--cnt;
+	}
+	for (; i<4; ++i, ++cp)
+		data = (data << 8) | (*(uchar *)cp);
+
+	return (write_word(info, wp, data));
+}
+
+/*
+ * Write a word to Flash, returns:
+ * 0 - OK
+ * 1 - write timeout
+ * 2 - Flash not erased
+ */
+static int write_word(flash_info_t *info, ulong dest, ulong data)
+{
+	volatile CFG_FLASH_WORD_SIZE *addr2 = (CFG_FLASH_WORD_SIZE *)(info->start[0]);
+	volatile CFG_FLASH_WORD_SIZE *dest2 = (CFG_FLASH_WORD_SIZE *)dest;
+	volatile CFG_FLASH_WORD_SIZE *data2 = (CFG_FLASH_WORD_SIZE *)&data;
+	ulong start;
+	int flag;
+	int i;
+
+	/* Check if Flash is (sufficiently) erased */
+	if ((*((vu_long *)dest) & data) != data)
+		return (2);
+
+	/* Disable interrupts which might cause a timeout here */
+	flag = disable_interrupts();
+
+	for (i=0; i<4/sizeof(CFG_FLASH_WORD_SIZE); i++) {
+		addr2[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+		addr2[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+		addr2[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00A000A0;
+
+		dest2[i] = data2[i];
+
+		/* re-enable interrupts if necessary */
+		if (flag)
+			enable_interrupts();
+
+		/* data polling for D7 */
+		start = get_timer (0);
+		while ((dest2[i] & (CFG_FLASH_WORD_SIZE)0x00800080) !=
+		       (data2[i] & (CFG_FLASH_WORD_SIZE)0x00800080)) {
+			if (get_timer(start) > CFG_FLASH_WRITE_TOUT)
+				return (1);
+		}
+	}
+
+	return (0);
+}

board/pcs440ep/init.S

@@ -0,0 +1,113 @@
+/*
+ * (C) Copyright 2006
+ * Stefan Roese, DENX Software Engineering, sr@denx.de.
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.	 See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <ppc_asm.tmpl>
+#include <config.h>
+
+/* General */
+#define TLB_VALID   0x00000200
+
+/* Supported page sizes */
+
+#define SZ_1K	    0x00000000
+#define SZ_4K	    0x00000010
+#define SZ_16K	    0x00000020
+#define SZ_64K	    0x00000030
+#define SZ_256K	    0x00000040
+#define SZ_1M	    0x00000050
+#define SZ_8M       0x00000060
+#define SZ_16M	    0x00000070
+#define SZ_256M	    0x00000090
+
+/* Storage attributes */
+#define SA_W	    0x00000800	    /* Write-through */
+#define SA_I	    0x00000400	    /* Caching inhibited */
+#define SA_M	    0x00000200	    /* Memory coherence */
+#define SA_G	    0x00000100	    /* Guarded */
+#define SA_E	    0x00000080	    /* Endian */
+
+/* Access control */
+#define AC_X	    0x00000024	    /* Execute */
+#define AC_W	    0x00000012	    /* Write */
+#define AC_R	    0x00000009	    /* Read */
+
+/* Some handy macros */
+
+#define EPN(e)		((e) & 0xfffffc00)
+#define TLB0(epn,sz)	( (EPN((epn)) | (sz) | TLB_VALID ) )
+#define TLB1(rpn,erpn)	( ((rpn)&0xfffffc00) | (erpn) )
+#define TLB2(a)		( (a)&0x00000fbf )
+
+#define tlbtab_start\
+	mflr    r1  ;\
+	bl 0f	    ;
+
+#define tlbtab_end\
+	.long 0, 0, 0	;   \
+0:	mflr    r0	;   \
+	mtlr    r1	;   \
+	blr		;
+
+#define tlbentry(epn,sz,rpn,erpn,attr)\
+	.long TLB0(epn,sz),TLB1(rpn,erpn),TLB2(attr)
+
+
+/**************************************************************************
+ * TLB TABLE
+ *
+ * This table is used by the cpu boot code to setup the initial tlb
+ * entries. Rather than make broad assumptions in the cpu source tree,
+ * this table lets each board set things up however they like.
+ *
+ *  Pointer to the table is returned in r1
+ *
+ *************************************************************************/
+
+    .section .bootpg,"ax"
+    .globl tlbtab
+
+tlbtab:
+    tlbtab_start
+
+    /*
+     * BOOT_CS (FLASH) must be first. Before relocation SA_I can be off to use the
+     * speed up boot process. It is patched after relocation to enable SA_I
+     */
+    tlbentry( CFG_BOOT_BASE_ADDR, SZ_256M, CFG_BOOT_BASE_ADDR, 0, AC_R|AC_W|AC_X|SA_G/*|SA_I*/)
+
+    /* TLB-entry for init-ram in dcache (SA_I must be turned off!) */
+    tlbentry( CFG_INIT_RAM_ADDR, SZ_64K, CFG_INIT_RAM_ADDR, 0, AC_R|AC_W|AC_X|SA_G )
+
+    tlbentry( CFG_SDRAM_BASE, SZ_256M, CFG_SDRAM_BASE, 0, AC_R|AC_W|AC_X|SA_G|SA_I )
+    tlbentry( CFG_PCI_BASE, SZ_256M, CFG_PCI_BASE, 0, AC_R|AC_W|SA_G|SA_I )
+
+    /* PCI */
+    tlbentry( CFG_PCI_MEMBASE, SZ_256M, CFG_PCI_MEMBASE, 0, AC_R|AC_W|SA_G|SA_I )
+    tlbentry( CFG_PCI_MEMBASE1, SZ_256M, CFG_PCI_MEMBASE1, 0, AC_R|AC_W|SA_G|SA_I )
+    tlbentry( CFG_PCI_MEMBASE2, SZ_256M, CFG_PCI_MEMBASE2, 0, AC_R|AC_W|SA_G|SA_I )
+    tlbentry( CFG_PCI_MEMBASE3, SZ_256M, CFG_PCI_MEMBASE3, 0, AC_R|AC_W|SA_G|SA_I )
+
+    /* USB 2.0 Device */
+    tlbentry( CFG_USB_DEVICE, SZ_1K, 0x50000000, 0, AC_R|AC_W|SA_G|SA_I )
+
+    tlbtab_end

board/pcs440ep/pcs440ep.c

@@ -0,0 +1,379 @@
+/*
+ * (C) Copyright 2006
+ * Stefan Roese, DENX Software Engineering, sr@denx.de.
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.	 See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+#include <ppc4xx.h>
+#include <asm/processor.h>
+#include <spd_sdram.h>
+
+DECLARE_GLOBAL_DATA_PTR;
+
+extern flash_info_t flash_info[CFG_MAX_FLASH_BANKS]; /* info for FLASH chips	*/
+
+static void set_leds(int val)
+{
+	unsigned char led[16] = {0x0, 0x8, 0x4, 0xc, 0x2, 0xa, 0x6, 0xe,
+				 0x1, 0x9, 0x5, 0xd, 0x3, 0xb, 0x7, 0xf};
+	out32(GPIO0_OR, (in32(GPIO0_OR) & ~0x78000000) | (led[val] << 27));
+}
+
+int board_early_init_f(void)
+{
+	register uint reg;
+
+	set_leds(0);			/* display boot info counter */
+
+	/*--------------------------------------------------------------------
+	 * Setup the external bus controller/chip selects
+	 *-------------------------------------------------------------------*/
+	mtdcr(ebccfga, xbcfg);
+	reg = mfdcr(ebccfgd);
+	mtdcr(ebccfgd, reg | 0x04000000);	/* Set ATC */
+
+	/*--------------------------------------------------------------------
+	 * GPIO's are alreay setup in cpu/ppc4xx/cpu_init.c
+	 * via define from board config file.
+	 *-------------------------------------------------------------------*/
+
+	/*--------------------------------------------------------------------
+	 * Setup the interrupt controller polarities, triggers, etc.
+	 *-------------------------------------------------------------------*/
+	mtdcr(uic0sr, 0xffffffff);	/* clear all */
+	mtdcr(uic0er, 0x00000000);	/* disable all */
+	mtdcr(uic0cr, 0x00000001);	/* UIC1 crit is critical */
+	mtdcr(uic0pr, 0xfffffe1f);	/* per ref-board manual */
+	mtdcr(uic0tr, 0x01c00000);	/* per ref-board manual */
+	mtdcr(uic0vr, 0x00000001);	/* int31 highest, base=0x000 */
+	mtdcr(uic0sr, 0xffffffff);	/* clear all */
+
+	mtdcr(uic1sr, 0xffffffff);	/* clear all */
+	mtdcr(uic1er, 0x00000000);	/* disable all */
+	mtdcr(uic1cr, 0x00000000);	/* all non-critical */
+	mtdcr(uic1pr, 0xffffe0ff);	/* per ref-board manual */
+	mtdcr(uic1tr, 0x00ffc000);	/* per ref-board manual */
+	mtdcr(uic1vr, 0x00000001);	/* int31 highest, base=0x000 */
+	mtdcr(uic1sr, 0xffffffff);	/* clear all */
+
+	/*--------------------------------------------------------------------
+	 * Setup other serial configuration
+	 *-------------------------------------------------------------------*/
+	mfsdr(sdr_pci0, reg);
+	mtsdr(sdr_pci0, 0x80000000 | reg);	/* PCI arbiter enabled */
+	mtsdr(sdr_pfc0, 0x00000100);	/* Pin function: enable GPIO49-63 */
+	mtsdr(sdr_pfc1, 0x00048000);	/* Pin function: UART0 has 4 pins, select IRQ5 */
+
+	return 0;
+}
+
+int misc_init_r (void)
+{
+	uint pbcr;
+	int size_val = 0;
+
+	/* Re-do sizing to get full correct info */
+	mtdcr(ebccfga, pb0cr);
+	pbcr = mfdcr(ebccfgd);
+	switch (gd->bd->bi_flashsize) {
+	case 1 << 20:
+		size_val = 0;
+		break;
+	case 2 << 20:
+		size_val = 1;
+		break;
+	case 4 << 20:
+		size_val = 2;
+		break;
+	case 8 << 20:
+		size_val = 3;
+		break;
+	case 16 << 20:
+		size_val = 4;
+		break;
+	case 32 << 20:
+		size_val = 5;
+		break;
+	case 64 << 20:
+		size_val = 6;
+		break;
+	case 128 << 20:
+		size_val = 7;
+		break;
+	}
+	pbcr = (pbcr & 0x0001ffff) | gd->bd->bi_flashstart | (size_val << 17);
+	mtdcr(ebccfga, pb0cr);
+	mtdcr(ebccfgd, pbcr);
+
+	/* adjust flash start and offset */
+	gd->bd->bi_flashstart = 0 - gd->bd->bi_flashsize;
+	gd->bd->bi_flashoffset = 0;
+
+	/* Monitor protection ON by default */
+	(void)flash_protect(FLAG_PROTECT_SET,
+			    -CFG_MONITOR_LEN,
+			    0xffffffff,
+			    &flash_info[1]);
+
+	/* Env protection ON by default */
+	(void)flash_protect(FLAG_PROTECT_SET,
+			    CFG_ENV_ADDR_REDUND,
+			    CFG_ENV_ADDR_REDUND + 2*CFG_ENV_SECT_SIZE - 1,
+			    &flash_info[1]);
+
+	return 0;
+}
+
+int checkboard(void)
+{
+	char *s = getenv("serial#");
+
+	printf("Board: PCS440EP");
+	if (s != NULL) {
+		puts(", serial# ");
+		puts(s);
+	}
+	putc('\n');
+
+	return (0);
+}
+
+long int initdram (int board_type)
+{
+	long dram_size = 0;
+
+	set_leds(1);			/* display boot info counter */
+	dram_size = spd_sdram();
+	set_leds(2);			/* display boot info counter */
+
+	return dram_size;
+}
+
+#if defined(CFG_DRAM_TEST)
+int testdram(void)
+{
+	unsigned long *mem = (unsigned long *)0;
+	const unsigned long kend = (1024 / sizeof(unsigned long));
+	unsigned long k, n;
+
+	mtmsr(0);
+
+	for (k = 0; k < CFG_KBYTES_SDRAM;
+	     ++k, mem += (1024 / sizeof(unsigned long))) {
+		if ((k & 1023) == 0) {
+			printf("%3d MB\r", k / 1024);
+		}
+
+		memset(mem, 0xaaaaaaaa, 1024);
+		for (n = 0; n < kend; ++n) {
+			if (mem[n] != 0xaaaaaaaa) {
+				printf("SDRAM test fails at: %08x\n",
+				       (uint) & mem[n]);
+				return 1;
+			}
+		}
+
+		memset(mem, 0x55555555, 1024);
+		for (n = 0; n < kend; ++n) {
+			if (mem[n] != 0x55555555) {
+				printf("SDRAM test fails at: %08x\n",
+				       (uint) & mem[n]);
+				return 1;
+			}
+		}
+	}
+	printf("SDRAM test passes\n");
+	return 0;
+}
+#endif
+
+/*************************************************************************
+ *  pci_pre_init
+ *
+ *  This routine is called just prior to registering the hose and gives
+ *  the board the opportunity to check things. Returning a value of zero
+ *  indicates that things are bad & PCI initialization should be aborted.
+ *
+ *	Different boards may wish to customize the pci controller structure
+ *	(add regions, override default access routines, etc) or perform
+ *	certain pre-initialization actions.
+ *
+ ************************************************************************/
+#if defined(CONFIG_PCI) && defined(CFG_PCI_PRE_INIT)
+int pci_pre_init(struct pci_controller *hose)
+{
+	unsigned long addr;
+
+	/*-------------------------------------------------------------------------+
+	  | Set priority for all PLB3 devices to 0.
+	  | Set PLB3 arbiter to fair mode.
+	  +-------------------------------------------------------------------------*/
+	mfsdr(sdr_amp1, addr);
+	mtsdr(sdr_amp1, (addr & 0x000000FF) | 0x0000FF00);
+	addr = mfdcr(plb3_acr);
+	mtdcr(plb3_acr, addr | 0x80000000);
+
+	/*-------------------------------------------------------------------------+
+	  | Set priority for all PLB4 devices to 0.
+	  +-------------------------------------------------------------------------*/
+	mfsdr(sdr_amp0, addr);
+	mtsdr(sdr_amp0, (addr & 0x000000FF) | 0x0000FF00);
+	addr = mfdcr(plb4_acr) | 0xa0000000;	/* Was 0x8---- */
+	mtdcr(plb4_acr, addr);
+
+	/*-------------------------------------------------------------------------+
+	  | Set Nebula PLB4 arbiter to fair mode.
+	  +-------------------------------------------------------------------------*/
+	/* Segment0 */
+	addr = (mfdcr(plb0_acr) & ~plb0_acr_ppm_mask) | plb0_acr_ppm_fair;
+	addr = (addr & ~plb0_acr_hbu_mask) | plb0_acr_hbu_enabled;
+	addr = (addr & ~plb0_acr_rdp_mask) | plb0_acr_rdp_4deep;
+	addr = (addr & ~plb0_acr_wrp_mask) | plb0_acr_wrp_2deep;
+	mtdcr(plb0_acr, addr);
+
+	/* Segment1 */
+	addr = (mfdcr(plb1_acr) & ~plb1_acr_ppm_mask) | plb1_acr_ppm_fair;
+	addr = (addr & ~plb1_acr_hbu_mask) | plb1_acr_hbu_enabled;
+	addr = (addr & ~plb1_acr_rdp_mask) | plb1_acr_rdp_4deep;
+	addr = (addr & ~plb1_acr_wrp_mask) | plb1_acr_wrp_2deep;
+	mtdcr(plb1_acr, addr);
+
+	return 1;
+}
+#endif				/* defined(CONFIG_PCI) && defined(CFG_PCI_PRE_INIT) */
+
+/*************************************************************************
+ *  pci_target_init
+ *
+ *	The bootstrap configuration provides default settings for the pci
+ *	inbound map (PIM). But the bootstrap config choices are limited and
+ *	may not be sufficient for a given board.
+ *
+ ************************************************************************/
+#if defined(CONFIG_PCI) && defined(CFG_PCI_TARGET_INIT)
+void pci_target_init(struct pci_controller *hose)
+{
+	/*--------------------------------------------------------------------------+
+	 * Set up Direct MMIO registers
+	 *--------------------------------------------------------------------------*/
+	/*--------------------------------------------------------------------------+
+	  | PowerPC440 EP PCI Master configuration.
+	  | Map one 1Gig range of PLB/processor addresses to PCI memory space.
+	  |   PLB address 0xA0000000-0xDFFFFFFF ==> PCI address 0xA0000000-0xDFFFFFFF
+	  |   Use byte reversed out routines to handle endianess.
+	  | Make this region non-prefetchable.
+	  +--------------------------------------------------------------------------*/
+	out32r(PCIX0_PMM0MA, 0x00000000);	/* PMM0 Mask/Attribute - disabled b4 setting */
+	out32r(PCIX0_PMM0LA, CFG_PCI_MEMBASE);	/* PMM0 Local Address */
+	out32r(PCIX0_PMM0PCILA, CFG_PCI_MEMBASE);	/* PMM0 PCI Low Address */
+	out32r(PCIX0_PMM0PCIHA, 0x00000000);	/* PMM0 PCI High Address */
+	out32r(PCIX0_PMM0MA, 0xE0000001);	/* 512M + No prefetching, and enable region */
+
+	out32r(PCIX0_PMM1MA, 0x00000000);	/* PMM0 Mask/Attribute - disabled b4 setting */
+	out32r(PCIX0_PMM1LA, CFG_PCI_MEMBASE2);	/* PMM0 Local Address */
+	out32r(PCIX0_PMM1PCILA, CFG_PCI_MEMBASE2);	/* PMM0 PCI Low Address */
+	out32r(PCIX0_PMM1PCIHA, 0x00000000);	/* PMM0 PCI High Address */
+	out32r(PCIX0_PMM1MA, 0xE0000001);	/* 512M + No prefetching, and enable region */
+
+	out32r(PCIX0_PTM1MS, 0x00000001);	/* Memory Size/Attribute */
+	out32r(PCIX0_PTM1LA, 0);	/* Local Addr. Reg */
+	out32r(PCIX0_PTM2MS, 0);	/* Memory Size/Attribute */
+	out32r(PCIX0_PTM2LA, 0);	/* Local Addr. Reg */
+
+	/*--------------------------------------------------------------------------+
+	 * Set up Configuration registers
+	 *--------------------------------------------------------------------------*/
+
+	/* Program the board's subsystem id/vendor id */
+	pci_write_config_word(0, PCI_SUBSYSTEM_VENDOR_ID,
+			      CFG_PCI_SUBSYS_VENDORID);
+	pci_write_config_word(0, PCI_SUBSYSTEM_ID, CFG_PCI_SUBSYS_ID);
+
+	/* Configure command register as bus master */
+	pci_write_config_word(0, PCI_COMMAND, PCI_COMMAND_MASTER);
+
+	/* 240nS PCI clock */
+	pci_write_config_word(0, PCI_LATENCY_TIMER, 1);
+
+	/* No error reporting */
+	pci_write_config_word(0, PCI_ERREN, 0);
+
+	pci_write_config_dword(0, PCI_BRDGOPT2, 0x00000101);
+
+}
+#endif				/* defined(CONFIG_PCI) && defined(CFG_PCI_TARGET_INIT) */
+
+/*************************************************************************
+ *  pci_master_init
+ *
+ ************************************************************************/
+#if defined(CONFIG_PCI) && defined(CFG_PCI_MASTER_INIT)
+void pci_master_init(struct pci_controller *hose)
+{
+	unsigned short temp_short;
+
+	/*--------------------------------------------------------------------------+
+	  | Write the PowerPC440 EP PCI Configuration regs.
+	  |   Enable PowerPC440 EP to be a master on the PCI bus (PMM).
+	  |   Enable PowerPC440 EP to act as a PCI memory target (PTM).
+	  +--------------------------------------------------------------------------*/
+	pci_read_config_word(0, PCI_COMMAND, &temp_short);
+	pci_write_config_word(0, PCI_COMMAND,
+			      temp_short | PCI_COMMAND_MASTER |
+			      PCI_COMMAND_MEMORY);
+}
+#endif				/* defined(CONFIG_PCI) && defined(CFG_PCI_MASTER_INIT) */
+
+/*************************************************************************
+ *  is_pci_host
+ *
+ *	This routine is called to determine if a pci scan should be
+ *	performed. With various hardware environments (especially cPCI and
+ *	PPMC) it's insufficient to depend on the state of the arbiter enable
+ *	bit in the strap register, or generic host/adapter assumptions.
+ *
+ *	Rather than hard-code a bad assumption in the general 440 code, the
+ *	440 pci code requires the board to decide at runtime.
+ *
+ *	Return 0 for adapter mode, non-zero for host (monarch) mode.
+ *
+ *
+ ************************************************************************/
+#if defined(CONFIG_PCI)
+int is_pci_host(struct pci_controller *hose)
+{
+	/* PCS440EP is always configured as host. */
+	return (1);
+}
+#endif				/* defined(CONFIG_PCI) */
+
+/*************************************************************************
+ *  hw_watchdog_reset
+ *
+ *	This routine is called to reset (keep alive) the watchdog timer
+ *
+ ************************************************************************/
+#if defined(CONFIG_HW_WATCHDOG)
+void hw_watchdog_reset(void)
+{
+
+}
+#endif

board/pcs440ep/u-boot.lds

@@ -0,0 +1,141 @@
+/*
+ * (C) Copyright 2006
+ * Wolfgang Denk, DENX Software Engineering, wd@denx.de.
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+OUTPUT_ARCH(powerpc)
+SEARCH_DIR(/lib); SEARCH_DIR(/usr/lib); SEARCH_DIR(/usr/local/lib); SEARCH_DIR(/usr/local/powerpc-any-elf/lib);
+/* Do we need any of these for elf?
+   __DYNAMIC = 0;    */
+SECTIONS
+{
+  .resetvec 0xFFFFFFFC :
+  {
+    *(.resetvec)
+  } = 0xffff
+
+  .bootpg 0xFFFFF000 :
+  {
+    cpu/ppc4xx/start.o	(.bootpg)
+  } = 0xffff
+
+  /* Read-only sections, merged into text segment: */
+  . = + SIZEOF_HEADERS;
+  .interp : { *(.interp) }
+  .hash          : { *(.hash)		}
+  .dynsym        : { *(.dynsym)		}
+  .dynstr        : { *(.dynstr)		}
+  .rel.text      : { *(.rel.text)		}
+  .rela.text     : { *(.rela.text) 	}
+  .rel.data      : { *(.rel.data)		}
+  .rela.data     : { *(.rela.data) 	}
+  .rel.rodata    : { *(.rel.rodata) 	}
+  .rela.rodata   : { *(.rela.rodata) 	}
+  .rel.got       : { *(.rel.got)		}
+  .rela.got      : { *(.rela.got)		}
+  .rel.ctors     : { *(.rel.ctors)	}
+  .rela.ctors    : { *(.rela.ctors)	}
+  .rel.dtors     : { *(.rel.dtors)	}
+  .rela.dtors    : { *(.rela.dtors)	}
+  .rel.bss       : { *(.rel.bss)		}
+  .rela.bss      : { *(.rela.bss)		}
+  .rel.plt       : { *(.rel.plt)		}
+  .rela.plt      : { *(.rela.plt)		}
+  .init          : { *(.init)	}
+  .plt : { *(.plt) }
+  .text      :
+  {
+    cpu/ppc4xx/start.o	(.text)
+    board/pcs440ep/init.o	(.text)
+
+    *(.text)
+    *(.fixup)
+    *(.got1)
+  }
+  _etext = .;
+  PROVIDE (etext = .);
+  .rodata    :
+  {
+    *(.rodata)
+    *(.rodata1)
+    *(.rodata.str1.4)
+    *(.eh_frame)
+  }
+  .fini      : { *(.fini)    } =0
+  .ctors     : { *(.ctors)   }
+  .dtors     : { *(.dtors)   }
+
+  /* Read-write section, merged into data segment: */
+  . = (. + 0x00FF) & 0xFFFFFF00;
+  _erotext = .;
+  PROVIDE (erotext = .);
+  .reloc   :
+  {
+    *(.got)
+    _GOT2_TABLE_ = .;
+    *(.got2)
+    _FIXUP_TABLE_ = .;
+    *(.fixup)
+  }
+  __got2_entries = (_FIXUP_TABLE_ - _GOT2_TABLE_) >> 2;
+  __fixup_entries = (. - _FIXUP_TABLE_) >> 2;
+
+  .data    :
+  {
+    *(.data)
+    *(.data1)
+    *(.sdata)
+    *(.sdata2)
+    *(.dynamic)
+    CONSTRUCTORS
+  }
+  _edata  =  .;
+  PROVIDE (edata = .);
+
+  . = .;
+  __u_boot_cmd_start = .;
+  .u_boot_cmd : { *(.u_boot_cmd) }
+  __u_boot_cmd_end = .;
+
+  . = .;
+  __start___ex_table = .;
+  __ex_table : { *(__ex_table) }
+  __stop___ex_table = .;
+
+  . = ALIGN(256);
+  __init_begin = .;
+  .text.init : { *(.text.init) }
+  .data.init : { *(.data.init) }
+  . = ALIGN(256);
+  __init_end = .;
+
+  __bss_start = .;
+  .bss       :
+  {
+   *(.sbss) *(.scommon)
+   *(.dynbss)
+   *(.bss)
+   *(COMMON)
+  }
+
+  _end = . ;
+  PROVIDE (end = .);
+}

board/prodrive/common/flash.c

@@ -0,0 +1,556 @@
+/*
+ * (C) Copyright 2006
+ * Stefan Roese, DENX Software Engineering, sr@denx.de.
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+#include <asm/processor.h>
+
+flash_info_t	flash_info[CFG_MAX_FLASH_BANKS]; /* info for FLASH chips	*/
+
+/*
+ * Functions
+ */
+static int write_word(flash_info_t *info, ulong dest, ulong data);
+
+void flash_print_info(flash_info_t *info)
+{
+	int i;
+	int k;
+	int size;
+	int erased;
+	volatile unsigned long *flash;
+
+	if (info->flash_id == FLASH_UNKNOWN) {
+		printf ("missing or unknown FLASH type\n");
+		return;
+	}
+
+	switch (info->flash_id & FLASH_VENDMASK) {
+	case FLASH_MAN_AMD:	printf ("AMD ");		break;
+	case FLASH_MAN_FUJ:	printf ("FUJITSU ");		break;
+	case FLASH_MAN_SST:	printf ("SST ");		break;
+	case FLASH_MAN_EXCEL:	printf ("Excel Semiconductor "); break;
+	default:		printf ("Unknown Vendor ");	break;
+	}
+
+	switch (info->flash_id & FLASH_TYPEMASK) {
+	case FLASH_AM400B:	printf ("AM29LV400B (4 Mbit, bottom boot sect)\n");
+		break;
+	case FLASH_AM400T:	printf ("AM29LV400T (4 Mbit, top boot sector)\n");
+		break;
+	case FLASH_AM800B:	printf ("AM29LV800B (8 Mbit, bottom boot sect)\n");
+		break;
+	case FLASH_AM800T:	printf ("AM29LV800T (8 Mbit, top boot sector)\n");
+		break;
+	case FLASH_AM160B:	printf ("AM29LV160B (16 Mbit, bottom boot sect)\n");
+		break;
+	case FLASH_AM160T:	printf ("AM29LV160T (16 Mbit, top boot sector)\n");
+		break;
+	case FLASH_AM320T:	printf ("AM29LV320T (32 M, top sector)\n");
+		break;
+	case FLASH_AM320B:	printf ("AM29LV320B (32 M, bottom sector)\n");
+		break;
+	case FLASH_AMDL322T:	printf ("AM29DL322T (32 M, top sector)\n");
+		break;
+	case FLASH_AMDL322B:	printf ("AM29DL322B (32 M, bottom sector)\n");
+		break;
+	case FLASH_AMDL323T:	printf ("AM29DL323T (32 M, top sector)\n");
+		break;
+	case FLASH_AMDL323B:	printf ("AM29DL323B (32 M, bottom sector)\n");
+		break;
+	case FLASH_SST020:	printf ("SST39LF/VF020 (2 Mbit, uniform sector size)\n");
+		break;
+	case FLASH_SST040:	printf ("SST39LF/VF040 (4 Mbit, uniform sector size)\n");
+		break;
+	default:		printf ("Unknown Chip Type\n");
+		break;
+	}
+
+	printf ("  Size: %ld MB in %d Sectors\n",
+		info->size >> 20, info->sector_count);
+
+	printf ("  Sector Start Addresses:");
+	for (i=0; i<info->sector_count; ++i) {
+#ifdef CFG_FLASH_EMPTY_INFO
+		/*
+		 * Check if whole sector is erased
+		 */
+		if (i != (info->sector_count-1))
+			size = info->start[i+1] - info->start[i];
+		else
+			size = info->start[0] + info->size - info->start[i];
+		erased = 1;
+		flash = (volatile unsigned long *)info->start[i];
+		size = size >> 2;        /* divide by 4 for longword access */
+		for (k=0; k<size; k++) {
+			if (*flash++ != 0xffffffff) {
+				erased = 0;
+				break;
+			}
+		}
+
+		if ((i % 5) == 0)
+			printf ("\n   ");
+		/* print empty and read-only info */
+		printf (" %08lX%s%s",
+			info->start[i],
+			erased ? " E" : "  ",
+			info->protect[i] ? "RO " : "   ");
+#else
+		if ((i % 5) == 0)
+			printf ("\n   ");
+		printf (" %08lX%s",
+			info->start[i],
+			info->protect[i] ? " (RO)" : "     ");
+#endif
+
+	}
+	printf ("\n");
+	return;
+}
+
+/*
+ * The following code cannot be run from FLASH!
+ */
+static ulong flash_get_size(vu_long *addr, flash_info_t *info)
+{
+	short i;
+	short n;
+	CFG_FLASH_WORD_SIZE value;
+	ulong base = (ulong)addr;
+	volatile CFG_FLASH_WORD_SIZE *addr2 = (CFG_FLASH_WORD_SIZE *)addr;
+
+	/* Write auto select command: read Manufacturer ID */
+	addr2[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+	addr2[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+	addr2[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00900090;
+
+	value = addr2[CFG_FLASH_READ0];
+
+	switch (value) {
+	case (CFG_FLASH_WORD_SIZE)AMD_MANUFACT:
+		info->flash_id = FLASH_MAN_AMD;
+		break;
+	case (CFG_FLASH_WORD_SIZE)FUJ_MANUFACT:
+		info->flash_id = FLASH_MAN_FUJ;
+		break;
+	case (CFG_FLASH_WORD_SIZE)SST_MANUFACT:
+		info->flash_id = FLASH_MAN_SST;
+		break;
+	case (CFG_FLASH_WORD_SIZE)EXCEL_MANUFACT:
+		info->flash_id = FLASH_MAN_EXCEL;
+		break;
+	default:
+		info->flash_id = FLASH_UNKNOWN;
+		info->sector_count = 0;
+		info->size = 0;
+		return (0);			/* no or unknown flash	*/
+	}
+
+	value = addr2[CFG_FLASH_READ1];		/* device ID		*/
+
+	switch (value) {
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV400T:
+		info->flash_id += FLASH_AM400T;
+		info->sector_count = 11;
+		info->size = 0x00080000;
+		break;				/* => 0.5 MB		*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV400B:
+		info->flash_id += FLASH_AM400B;
+		info->sector_count = 11;
+		info->size = 0x00080000;
+		break;				/* => 0.5 MB		*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV800T:
+		info->flash_id += FLASH_AM800T;
+		info->sector_count = 19;
+		info->size = 0x00100000;
+		break;				/* => 1 MB		*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV800B:
+		info->flash_id += FLASH_AM800B;
+		info->sector_count = 19;
+		info->size = 0x00100000;
+		break;				/* => 1 MB		*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV160T:
+		info->flash_id += FLASH_AM160T;
+		info->sector_count = 35;
+		info->size = 0x00200000;
+		break;				/* => 2 MB		*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV160B:
+		info->flash_id += FLASH_AM160B;
+		info->sector_count = 35;
+		info->size = 0x00200000;
+		break;				/* => 2 MB		*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV320T:
+		info->flash_id += FLASH_AM320T;
+		info->sector_count = 71;
+		info->size = 0x00400000;  break;	/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_LV320B:
+		info->flash_id += FLASH_AM320B;
+		info->sector_count = 71;
+		info->size = 0x00400000;  break;	/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_DL322T:
+		info->flash_id += FLASH_AMDL322T;
+		info->sector_count = 71;
+		info->size = 0x00400000;  break;	/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_DL322B:
+		info->flash_id += FLASH_AMDL322B;
+		info->sector_count = 71;
+		info->size = 0x00400000;  break;	/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_DL323T:
+		info->flash_id += FLASH_AMDL323T;
+		info->sector_count = 71;
+		info->size = 0x00400000;  break;	/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)AMD_ID_DL323B:
+		info->flash_id += FLASH_AMDL323B;
+		info->sector_count = 71;
+		info->size = 0x00400000;  break;	/* => 4 MB	*/
+
+	case (CFG_FLASH_WORD_SIZE)SST_ID_xF020:
+		info->flash_id += FLASH_SST020;
+		info->sector_count = 64;
+		info->size = 0x00040000;
+		break;				/* => 256 kB		*/
+
+	case (CFG_FLASH_WORD_SIZE)SST_ID_xF040:
+		info->flash_id += FLASH_SST040;
+		info->sector_count = 128;
+		info->size = 0x00080000;
+		break;				/* => 512 kB		*/
+
+	default:
+		info->flash_id = FLASH_UNKNOWN;
+		return (0);			/* => no or unknown flash */
+
+	}
+
+	/* set up sector start address table */
+	if ((info->flash_id & FLASH_VENDMASK) == FLASH_MAN_SST) {
+		for (i = 0; i < info->sector_count; i++)
+			info->start[i] = base + (i * 0x00001000);
+	} else if (((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL322B) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL323B) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AM320B) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL324B)) {
+		/* set sector offsets for bottom boot block type	*/
+		for (i=0; i<8; ++i) {		/*  8 x 8k boot sectors	*/
+			info->start[i] = base;
+			base += 8 << 10;
+		}
+		while (i < info->sector_count) {	/* 64k regular sectors	*/
+			info->start[i] = base;
+			base += 64 << 10;
+			++i;
+		}
+	} else if (((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL322T) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL323T) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AM320T) ||
+		   ((info->flash_id & FLASH_TYPEMASK) == FLASH_AMDL324T)) {
+		/* set sector offsets for top boot block type		*/
+		base += info->size;
+		i = info->sector_count;
+		for (n=0; n<8; ++n) {		/*  8 x 8k boot sectors	*/
+			base -= 8 << 10;
+			--i;
+			info->start[i] = base;
+		}
+		while (i > 0) {			/* 64k regular sectors	*/
+			base -= 64 << 10;
+			--i;
+			info->start[i] = base;
+		}
+	} else {
+		if (info->flash_id & FLASH_BTYPE) {
+			/* set sector offsets for bottom boot block type	*/
+			info->start[0] = base + 0x00000000;
+			info->start[1] = base + 0x00004000;
+			info->start[2] = base + 0x00006000;
+			info->start[3] = base + 0x00008000;
+			for (i = 4; i < info->sector_count; i++) {
+				info->start[i] = base + (i * 0x00010000) - 0x00030000;
+			}
+		} else {
+			/* set sector offsets for top boot block type		*/
+			i = info->sector_count - 1;
+			info->start[i--] = base + info->size - 0x00004000;
+			info->start[i--] = base + info->size - 0x00006000;
+			info->start[i--] = base + info->size - 0x00008000;
+			for (; i >= 0; i--) {
+				info->start[i] = base + i * 0x00010000;
+			}
+		}
+	}
+
+	/* check for protected sectors */
+	for (i = 0; i < info->sector_count; i++) {
+		/* read sector protection at sector address, (A7 .. A0) = 0x02 */
+		/* D0 = 1 if protected */
+		addr2 = (volatile CFG_FLASH_WORD_SIZE *)(info->start[i]);
+		if ((info->flash_id & FLASH_VENDMASK) != FLASH_MAN_AMD)
+			info->protect[i] = 0;
+		else
+			info->protect[i] = addr2[CFG_FLASH_READ2] & 1;
+	}
+
+	/*
+	 * Prevent writes to uninitialized FLASH.
+	 */
+	if (info->flash_id != FLASH_UNKNOWN) {
+		addr2 = (CFG_FLASH_WORD_SIZE *)info->start[0];
+		*addr2 = (CFG_FLASH_WORD_SIZE)0x00F000F0;	/* reset bank */
+	}
+
+	return (info->size);
+}
+
+
+int flash_erase(flash_info_t *info, int s_first, int s_last)
+{
+	volatile CFG_FLASH_WORD_SIZE *addr = (CFG_FLASH_WORD_SIZE *)(info->start[0]);
+	volatile CFG_FLASH_WORD_SIZE *addr2;
+	int flag, prot, sect, l_sect;
+	ulong start, now, last;
+
+	if ((s_first < 0) || (s_first > s_last)) {
+		if (info->flash_id == FLASH_UNKNOWN)
+			printf ("- missing\n");
+		else
+			printf ("- no sectors to erase\n");
+		return 1;
+	}
+
+	if (info->flash_id == FLASH_UNKNOWN) {
+		printf ("Can't erase unknown flash type - aborted\n");
+		return 1;
+	}
+
+	prot = 0;
+	for (sect=s_first; sect<=s_last; ++sect)
+		if (info->protect[sect])
+			prot++;
+
+	if (prot)
+		printf ("- Warning: %d protected sectors will not be erased!\n", prot);
+	else
+		printf ("\n");
+
+	l_sect = -1;
+
+	/* Disable interrupts which might cause a timeout here */
+	flag = disable_interrupts();
+
+	/* Start erase on unprotected sectors */
+	for (sect = s_first; sect<=s_last; sect++) {
+		if (info->protect[sect] == 0) {	/* not protected */
+			addr2 = (CFG_FLASH_WORD_SIZE *)(info->start[sect]);
+			if ((info->flash_id & FLASH_VENDMASK) == FLASH_MAN_SST) {
+				addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+				addr[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+				addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00800080;
+				addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+				addr[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+				addr2[0] = (CFG_FLASH_WORD_SIZE)0x00300030;  /* sector erase */
+
+				/* re-enable interrupts if necessary */
+				if (flag) {
+					enable_interrupts();
+					flag = 0;
+				}
+
+				/* data polling for D7 */
+				start = get_timer (0);
+				while ((addr2[0] & (CFG_FLASH_WORD_SIZE)0x00800080) !=
+				       (CFG_FLASH_WORD_SIZE)0x00800080) {
+					if (get_timer(start) > CFG_FLASH_WRITE_TOUT)
+						return (1);
+				}
+			} else {
+				if (sect == s_first) {
+					addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+					addr[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+					addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00800080;
+					addr[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+					addr[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+				}
+				addr2[0] = (CFG_FLASH_WORD_SIZE)0x00300030;  /* sector erase */
+			}
+			l_sect = sect;
+		}
+	}
+
+	/* re-enable interrupts if necessary */
+	if (flag)
+		enable_interrupts();
+
+	/* wait at least 80us - let's wait 1 ms */
+	udelay (1000);
+
+	/*
+	 * We wait for the last triggered sector
+	 */
+	if (l_sect < 0)
+		goto DONE;
+
+	start = get_timer (0);
+	last  = start;
+	addr = (CFG_FLASH_WORD_SIZE *)(info->start[l_sect]);
+	while ((addr[0] & (CFG_FLASH_WORD_SIZE)0x00800080) != (CFG_FLASH_WORD_SIZE)0x00800080) {
+		if ((now = get_timer(start)) > CFG_FLASH_ERASE_TOUT) {
+			printf ("Timeout\n");
+			return 1;
+		}
+		/* show that we're waiting */
+		if ((now - last) > 1000) {	/* every second */
+			putc ('.');
+			last = now;
+		}
+	}
+
+DONE:
+	/* reset to read mode */
+	addr = (CFG_FLASH_WORD_SIZE *)info->start[0];
+	addr[0] = (CFG_FLASH_WORD_SIZE)0x00F000F0;	/* reset bank */
+
+	printf (" done\n");
+	return 0;
+}
+
+/*
+ * Copy memory to flash, returns:
+ * 0 - OK
+ * 1 - write timeout
+ * 2 - Flash not erased
+ */
+int write_buff(flash_info_t *info, uchar *src, ulong addr, ulong cnt)
+{
+	ulong cp, wp, data;
+	int i, l, rc;
+
+	wp = (addr & ~3);	/* get lower word aligned address */
+
+	/*
+	 * handle unaligned start bytes
+	 */
+	if ((l = addr - wp) != 0) {
+		data = 0;
+		for (i=0, cp=wp; i<l; ++i, ++cp) {
+			data = (data << 8) | (*(uchar *)cp);
+		}
+		for (; i<4 && cnt>0; ++i) {
+			data = (data << 8) | *src++;
+			--cnt;
+			++cp;
+		}
+		for (; cnt==0 && i<4; ++i, ++cp) {
+			data = (data << 8) | (*(uchar *)cp);
+		}
+
+		if ((rc = write_word(info, wp, data)) != 0) {
+			return (rc);
+		}
+		wp += 4;
+	}
+
+	/*
+	 * handle word aligned part
+	 */
+	while (cnt >= 4) {
+		data = 0;
+		for (i=0; i<4; ++i)
+			data = (data << 8) | *src++;
+		if ((rc = write_word(info, wp, data)) != 0)
+			return (rc);
+		wp  += 4;
+		cnt -= 4;
+	}
+
+	if (cnt == 0)
+		return (0);
+
+	/*
+	 * handle unaligned tail bytes
+	 */
+	data = 0;
+	for (i=0, cp=wp; i<4 && cnt>0; ++i, ++cp) {
+		data = (data << 8) | *src++;
+		--cnt;
+	}
+	for (; i<4; ++i, ++cp)
+		data = (data << 8) | (*(uchar *)cp);
+
+	return (write_word(info, wp, data));
+}
+
+/*
+ * Write a word to Flash, returns:
+ * 0 - OK
+ * 1 - write timeout
+ * 2 - Flash not erased
+ */
+static int write_word(flash_info_t *info, ulong dest, ulong data)
+{
+	volatile CFG_FLASH_WORD_SIZE *addr2 = (CFG_FLASH_WORD_SIZE *)(info->start[0]);
+	volatile CFG_FLASH_WORD_SIZE *dest2 = (CFG_FLASH_WORD_SIZE *)dest;
+	volatile CFG_FLASH_WORD_SIZE *data2 = (CFG_FLASH_WORD_SIZE *)&data;
+	ulong start;
+	int flag;
+	int i;
+
+	/* Check if Flash is (sufficiently) erased */
+	if ((*((vu_long *)dest) & data) != data)
+		return (2);
+
+	/* Disable interrupts which might cause a timeout here */
+	flag = disable_interrupts();
+
+	for (i=0; i<4/sizeof(CFG_FLASH_WORD_SIZE); i++) {
+		addr2[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00AA00AA;
+		addr2[CFG_FLASH_ADDR1] = (CFG_FLASH_WORD_SIZE)0x00550055;
+		addr2[CFG_FLASH_ADDR0] = (CFG_FLASH_WORD_SIZE)0x00A000A0;
+
+		dest2[i] = data2[i];
+
+		/* re-enable interrupts if necessary */
+		if (flag)
+			enable_interrupts();
+
+		/* data polling for D7 */
+		start = get_timer (0);
+		while ((dest2[i] & (CFG_FLASH_WORD_SIZE)0x00800080) !=
+		       (data2[i] & (CFG_FLASH_WORD_SIZE)0x00800080)) {
+			if (get_timer(start) > CFG_FLASH_WRITE_TOUT)
+				return (1);
+		}
+	}
+
+	return (0);
+}

board/prodrive/common/fpga.c

@@ -0,0 +1,183 @@
+/*
+ * (C) Copyright 2006
+ * Stefan Roese, DENX Software Engineering, sr@denx.de.
+ *
+ * (C) Copyright 2001-2004
+ * Matthias Fuchs, esd gmbh germany, matthias.fuchs@esd-electronics.com
+ * Stefan Roese, esd gmbh germany, stefan.roese@esd-electronics.com
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+#include <asm/processor.h>
+#include <command.h>
+
+/* ------------------------------------------------------------------------- */
+
+#ifdef FPGA_DEBUG
+#define DBG(x...) printf(x)
+#else
+#define DBG(x...)
+#endif /* DEBUG */
+
+#define FPGA_PRG		CFG_FPGA_PRG /* FPGA program pin (cpu output)*/
+#define FPGA_CLK		CFG_FPGA_CLK /* FPGA clk pin (cpu output)    */
+#define FPGA_DATA		CFG_FPGA_DATA /* FPGA data pin (cpu output)  */
+#define FPGA_DONE		CFG_FPGA_DONE /* FPGA done pin (cpu input)   */
+#define FPGA_INIT		CFG_FPGA_INIT /* FPGA init pin (cpu input)   */
+
+#define ERROR_FPGA_PRG_INIT_LOW  -1        /* Timeout after PRG* asserted   */
+#define ERROR_FPGA_PRG_INIT_HIGH -2        /* Timeout after PRG* deasserted */
+#define ERROR_FPGA_PRG_DONE      -3        /* Timeout after programming     */
+
+#ifndef OLD_VAL
+# define OLD_VAL		0
+#endif
+
+#if 0 /* test-only */
+#define FPGA_WRITE_1 { \
+		SET_FPGA(OLD_VAL | FPGA_PRG | 0        | FPGA_DATA);  /* set clock to 0 */ \
+		SET_FPGA(OLD_VAL | FPGA_PRG | 0        | FPGA_DATA);  /* set data to 1  */	\
+		SET_FPGA(OLD_VAL | FPGA_PRG | FPGA_CLK | FPGA_DATA);  /* set clock to 1 */	\
+		SET_FPGA(OLD_VAL | FPGA_PRG | FPGA_CLK | FPGA_DATA);} /* set data to 1  */
+
+#define FPGA_WRITE_0 { \
+		SET_FPGA(OLD_VAL | FPGA_PRG | 0        | FPGA_DATA);  /* set clock to 0 */	\
+		SET_FPGA(OLD_VAL | FPGA_PRG | 0        | 0        );  /* set data to 0  */	\
+		SET_FPGA(OLD_VAL | FPGA_PRG | FPGA_CLK | 0        );  /* set clock to 1 */	\
+		SET_FPGA(OLD_VAL | FPGA_PRG | FPGA_CLK | FPGA_DATA);} /* set data to 1  */
+#else
+#define FPGA_WRITE_1 { \
+		SET_FPGA(OLD_VAL | FPGA_PRG | 0        | FPGA_DATA);  /* set data to 1  */	\
+		SET_FPGA(OLD_VAL | FPGA_PRG | FPGA_CLK | FPGA_DATA);} /* set data to 1  */
+
+#define FPGA_WRITE_0 { \
+		SET_FPGA(OLD_VAL | FPGA_PRG | 0        | 0        );   /* set data to 0  */	\
+		SET_FPGA(OLD_VAL | FPGA_PRG | FPGA_CLK | 0        );}  /* set data to 1  */
+#endif
+
+static int fpga_boot(unsigned char *fpgadata, int size)
+{
+	int i,index,len;
+	int count;
+	int j;
+
+	/* display infos on fpgaimage */
+	index = 15;
+	for (i=0; i<4; i++) {
+		len = fpgadata[index];
+		DBG("FPGA: %s\n", &(fpgadata[index+1]));
+		index += len+3;
+	}
+
+	/* search for preamble 0xFFFFFFFF */
+	while (1) {
+		if ((fpgadata[index] == 0xff) && (fpgadata[index+1] == 0xff) &&
+		    (fpgadata[index+2] == 0xff) && (fpgadata[index+3] == 0xff))
+			break; /* preamble found */
+		else
+			index++;
+	}
+
+	DBG("FPGA: configdata starts at position 0x%x\n",index);
+	DBG("FPGA: length of fpga-data %d\n", size-index);
+
+	/*
+	 * Setup port pins for fpga programming
+	 */
+	SET_FPGA(FPGA_PRG | FPGA_CLK | FPGA_DATA);            /* set pins to high */
+
+	DBG("%s, ",(FPGA_DONE_STATE == 0) ? "NOT DONE" : "DONE" );
+	DBG("%s\n",(FPGA_INIT_STATE == 0) ? "NOT INIT" : "INIT" );
+
+	/*
+	 * Init fpga by asserting and deasserting PROGRAM*
+	 */
+	SET_FPGA(0 | FPGA_CLK | FPGA_DATA);             /* set prog active */
+
+	/* Wait for FPGA init line low */
+	count = 0;
+	while (FPGA_INIT_STATE) {
+		udelay(1000); /* wait 1ms */
+		/* Check for timeout - 100us max, so use 3ms */
+		if (count++ > 3) {
+			DBG("FPGA: Booting failed!\n");
+			return ERROR_FPGA_PRG_INIT_LOW;
+		}
+	}
+
+	DBG("%s, ",(FPGA_DONE_STATE == 0) ? "NOT DONE" : "DONE" );
+	DBG("%s\n",(FPGA_INIT_STATE == 0) ? "NOT INIT" : "INIT" );
+
+	/* deassert PROGRAM* */
+	SET_FPGA(FPGA_PRG | FPGA_CLK | FPGA_DATA);           /* set prog inactive */
+
+	/* Wait for FPGA end of init period .  */
+	count = 0;
+	while (!(FPGA_INIT_STATE)) {
+		udelay(1000); /* wait 1ms */
+		/* Check for timeout */
+		if (count++ > 3) {
+			DBG("FPGA: Booting failed!\n");
+			return ERROR_FPGA_PRG_INIT_HIGH;
+		}
+	}
+
+	DBG("%s, ",(FPGA_DONE_STATE == 0) ? "NOT DONE" : "DONE" );
+	DBG("%s\n",(FPGA_INIT_STATE == 0) ? "NOT INIT" : "INIT" );
+
+	DBG("write configuration data into fpga\n");
+	/* write configuration-data into fpga... */
+
+	/*
+	 * Load uncompressed image into fpga
+	 */
+	for (i=index; i<size; i++) {
+		for (j=0; j<8; j++) {
+			if ((fpgadata[i] & 0x80) == 0x80) {
+				FPGA_WRITE_1;
+			} else {
+				FPGA_WRITE_0;
+			}
+			fpgadata[i] <<= 1;
+		}
+	}
+
+	DBG("%s, ",(FPGA_DONE_STATE == 0) ? "NOT DONE" : "DONE" );
+	DBG("%s\n",(FPGA_INIT_STATE == 0) ? "NOT INIT" : "INIT" );
+
+	/*
+	 * Check if fpga's DONE signal - correctly booted ?
+	 */
+
+	/* Wait for FPGA end of programming period .  */
+	count = 0;
+	while (!(FPGA_DONE_STATE)) {
+		udelay(1000); /* wait 1ms */
+		/* Check for timeout */
+		if (count++ > 3) {
+			DBG("FPGA: Booting failed!\n");
+			return ERROR_FPGA_PRG_DONE;
+		}
+	}
+
+	DBG("FPGA: Booting successful!\n");
+	return 0;
+}

board/prodrive/pdnb3/Makefile

@@ -0,0 +1,46 @@
+#
+# (C) Copyright 2006
+# Wolfgang Denk, DENX Software Engineering, wd@denx.de.
+#
+# See file CREDITS for list of people who contributed to this
+# project.
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation; either version 2 of
+# the License, or (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+# MA 02111-1307 USA
+#
+
+include $(TOPDIR)/config.mk
+
+LIB	= lib$(BOARD).a
+
+OBJS	:= flash.o pdnb3.o nand.o
+
+$(LIB):	$(OBJS) $(SOBJS)
+	$(AR) crv $@ $^
+
+clean:
+	rm -f $(SOBJS) $(OBJS)
+
+distclean:	clean
+	rm -f $(LIB) core *.bak .depend
+
+#########################################################################
+
+.depend:	Makefile $(SOBJS:.o=.S) $(OBJS:.o=.c)
+		$(CC) -M $(CPPFLAGS) $(SOBJS:.o=.S) $(OBJS:.o=.c) > $@
+
+-include .depend
+
+#########################################################################

board/prodrive/pdnb3/config.mk

@@ -0,0 +1,4 @@
+TEXT_BASE = 0x01f00000
+
+# include NPE ethernet driver
+BOARDLIBS = cpu/ixp/npe/libnpe.a

board/prodrive/pdnb3/flash.c

@@ -0,0 +1,85 @@
+/*
+ * (C) Copyright 2006
+ * Stefan Roese, DENX Software Engineering, sr@denx.de.
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+#include <asm/arch/ixp425.h>
+
+/*
+ * include common flash code (for esd boards)
+ */
+#include "../common/flash.c"
+
+/*
+ * Prototypes
+ */
+static ulong flash_get_size (vu_long * addr, flash_info_t * info);
+
+static inline ulong ld(ulong x)
+{
+	ulong k = 0;
+
+	while (x >>= 1)
+		++k;
+
+	return k;
+}
+
+unsigned long flash_init(void)
+{
+	unsigned long size;
+	int i;
+
+	/* Init: no FLASHes known */
+	for (i=0; i<CFG_MAX_FLASH_BANKS; i++)
+		flash_info[i].flash_id = FLASH_UNKNOWN;
+
+	size = flash_get_size((vu_long *)FLASH_BASE0_PRELIM, &flash_info[0]);
+
+	if (flash_info[0].flash_id == FLASH_UNKNOWN)
+		printf ("## Unknown FLASH on Bank 0 - Size = 0x%08lx = %ld MB\n",
+			size, size<<20);
+
+	/* Reconfigure CS0 to actual FLASH size */
+	*IXP425_EXP_CS0 = (*IXP425_EXP_CS0 & ~0x00003C00) | ((ld(size) - 9) << 10);
+
+	/* Monitor protection ON by default */
+	flash_protect(FLAG_PROTECT_SET,
+		      CFG_MONITOR_BASE, CFG_MONITOR_BASE + monitor_flash_len - 1,
+		      &flash_info[CFG_MAX_FLASH_BANKS - 1]);
+
+	/* Environment protection ON by default */
+	flash_protect(FLAG_PROTECT_SET,
+		      CFG_ENV_ADDR,
+		      CFG_ENV_ADDR + CFG_ENV_SECT_SIZE - 1,
+		      &flash_info[CFG_MAX_FLASH_BANKS - 1]);
+
+	/* Redundant environment protection ON by default */
+	flash_protect(FLAG_PROTECT_SET,
+		      CFG_ENV_ADDR_REDUND,
+		      CFG_ENV_ADDR_REDUND + CFG_ENV_SIZE_REDUND - 1,
+		      &flash_info[CFG_MAX_FLASH_BANKS - 1]);
+
+	flash_info[0].size = size;
+
+	return size;
+}

board/prodrive/pdnb3/nand.c

@@ -0,0 +1,171 @@
+/*
+ * (C) Copyright 2006
+ * Stefan Roese, DENX Software Engineering, sr@denx.de.
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+
+#if (CONFIG_COMMANDS & CFG_CMD_NAND)
+
+#include <nand.h>
+
+struct pdnb3_ndfc_regs {
+	uchar cmd;
+	uchar wait;
+	uchar addr;
+	uchar term;
+	uchar data;
+};
+
+static u8 hwctl;
+static struct pdnb3_ndfc_regs *pdnb3_ndfc;
+
+#define readb(addr)	*(volatile u_char *)(addr)
+#define readl(addr)	*(volatile u_long *)(addr)
+#define writeb(d,addr)	*(volatile u_char *)(addr) = (d)
+
+/*
+ * The PDNB3 has a NAND Flash Controller (NDFC) that handles all accesses to
+ * the NAND devices.  The NDFC has command, address and data registers that
+ * when accessed will set up the NAND flash pins appropriately.  We'll use the
+ * hwcontrol function to save the configuration in a global variable.
+ * We can then use this information in the read and write functions to
+ * determine which NDFC register to access.
+ *
+ * There is one NAND devices on the board, a Hynix HY27US08561A (32 MByte).
+ */
+static void pdnb3_nand_hwcontrol(struct mtd_info *mtd, int cmd)
+{
+	switch (cmd) {
+	case NAND_CTL_SETCLE:
+		hwctl |= 0x1;
+		break;
+	case NAND_CTL_CLRCLE:
+		hwctl &= ~0x1;
+		break;
+
+	case NAND_CTL_SETALE:
+		hwctl |= 0x2;
+		break;
+	case NAND_CTL_CLRALE:
+		hwctl &= ~0x2;
+		break;
+
+	case NAND_CTL_SETNCE:
+		break;
+	case NAND_CTL_CLRNCE:
+		writeb(0x00, &(pdnb3_ndfc->term));
+		break;
+	}
+}
+
+static void pdnb3_nand_write_byte(struct mtd_info *mtd, u_char byte)
+{
+	if (hwctl & 0x1)
+		writeb(byte, &(pdnb3_ndfc->cmd));
+	else if (hwctl & 0x2)
+		writeb(byte, &(pdnb3_ndfc->addr));
+	else
+		writeb(byte, &(pdnb3_ndfc->data));
+}
+
+static u_char pdnb3_nand_read_byte(struct mtd_info *mtd)
+{
+	return readb(&(pdnb3_ndfc->data));
+}
+
+static void pdnb3_nand_write_buf(struct mtd_info *mtd, const u_char *buf, int len)
+{
+	int i;
+
+	for (i = 0; i < len; i++) {
+		if (hwctl & 0x1)
+			writeb(buf[i], &(pdnb3_ndfc->cmd));
+		else if (hwctl & 0x2)
+			writeb(buf[i], &(pdnb3_ndfc->addr));
+		else
+			writeb(buf[i], &(pdnb3_ndfc->data));
+	}
+}
+
+static void pdnb3_nand_read_buf(struct mtd_info *mtd, u_char *buf, int len)
+{
+	int i;
+
+	if (len % 4) {
+		for (i = 0; i < len; i++)
+			buf[i] = readb(&(pdnb3_ndfc->data));
+	} else {
+		ulong *ptr = (ulong *)buf;
+		int count = len >> 2;
+
+		for (i = 0; i < count; i++)
+			*ptr++ = readl(&(pdnb3_ndfc->data));
+	}
+}
+
+static int pdnb3_nand_verify_buf(struct mtd_info *mtd, const u_char *buf, int len)
+{
+	int i;
+
+	for (i = 0; i < len; i++)
+		if (buf[i] != readb(&(pdnb3_ndfc->data)))
+			return i;
+
+	return 0;
+}
+
+static int pdnb3_nand_dev_ready(struct mtd_info *mtd)
+{
+	volatile u_char val;
+
+	/*
+	 * Blocking read to wait for NAND to be ready
+	 */
+	val = readb(&(pdnb3_ndfc->wait));
+
+	/*
+	 * Return always true
+	 */
+	return 1;
+}
+
+void board_nand_init(struct nand_chip *nand)
+{
+	pdnb3_ndfc = (struct pdnb3_ndfc_regs *)CFG_NAND_BASE;
+
+	nand->eccmode = NAND_ECC_SOFT;
+
+	/* Set address of NAND IO lines (Using Linear Data Access Region) */
+	nand->IO_ADDR_R = (void __iomem *) ((ulong) pdnb3_ndfc + 0x4);
+	nand->IO_ADDR_W = (void __iomem *) ((ulong) pdnb3_ndfc + 0x4);
+	/* Reference hardware control function */
+	nand->hwcontrol  = pdnb3_nand_hwcontrol;
+	/* Set command delay time */
+	nand->hwcontrol  = pdnb3_nand_hwcontrol;
+	nand->write_byte = pdnb3_nand_write_byte;
+	nand->read_byte  = pdnb3_nand_read_byte;
+	nand->write_buf  = pdnb3_nand_write_buf;
+	nand->read_buf   = pdnb3_nand_read_buf;
+	nand->verify_buf = pdnb3_nand_verify_buf;
+	nand->dev_ready  = pdnb3_nand_dev_ready;
+}
+#endif

board/prodrive/pdnb3/pdnb3.c

@@ -0,0 +1,249 @@
+/*
+ * (C) Copyright 2006
+ * Stefan Roese, DENX Software Engineering, sr@denx.de.
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+#include <command.h>
+#include <malloc.h>
+#include <asm/arch/ixp425.h>
+
+DECLARE_GLOBAL_DATA_PTR;
+
+/* Prototypes */
+int gunzip(void *, int, unsigned char *, unsigned long *);
+int do_reset(cmd_tbl_t *cmdtp, int flag, int argc, char *argv[]);
+
+/* predefine these here for FPGA programming (before including fpga.c) */
+#define SET_FPGA(data)	*IXP425_GPIO_GPOUTR = (data)
+#define FPGA_DONE_STATE (*IXP425_GPIO_GPINR & CFG_FPGA_DONE)
+#define FPGA_INIT_STATE (*IXP425_GPIO_GPINR & CFG_FPGA_INIT)
+#define OLD_VAL		old_val
+
+static unsigned long old_val = 0;
+
+/*
+ * include common fpga code (for prodrive boards)
+ */
+#include "../common/fpga.c"
+
+/*
+ * Miscelaneous platform dependent initialisations
+ */
+int board_post_init(void)
+{
+	return (0);
+}
+
+int board_init(void)
+{
+	/* arch number of PDNB3 */
+	gd->bd->bi_arch_number = MACH_TYPE_PDNB3;
+
+	/* adress of boot parameters */
+	gd->bd->bi_boot_params = 0x00000100;
+
+	GPIO_OUTPUT_SET(CFG_GPIO_FPGA_RESET);
+	GPIO_OUTPUT_ENABLE(CFG_GPIO_FPGA_RESET);
+
+	GPIO_OUTPUT_SET(CFG_GPIO_SYS_RUNNING);
+	GPIO_OUTPUT_ENABLE(CFG_GPIO_SYS_RUNNING);
+
+	/*
+	 * Setup GPIO's for FPGA programming
+	 */
+	GPIO_OUTPUT_CLEAR(CFG_GPIO_PRG);
+	GPIO_OUTPUT_CLEAR(CFG_GPIO_CLK);
+	GPIO_OUTPUT_CLEAR(CFG_GPIO_DATA);
+	GPIO_OUTPUT_ENABLE(CFG_GPIO_PRG);
+	GPIO_OUTPUT_ENABLE(CFG_GPIO_CLK);
+	GPIO_OUTPUT_ENABLE(CFG_GPIO_DATA);
+	GPIO_OUTPUT_DISABLE(CFG_GPIO_INIT);
+	GPIO_OUTPUT_DISABLE(CFG_GPIO_DONE);
+
+	/*
+	 * Setup GPIO's for interrupts
+	 */
+	GPIO_OUTPUT_DISABLE(CFG_GPIO_PCI_INTA);
+	GPIO_INT_ACT_LOW_SET(CFG_GPIO_PCI_INTA);
+	GPIO_OUTPUT_DISABLE(CFG_GPIO_PCI_INTB);
+	GPIO_INT_ACT_LOW_SET(CFG_GPIO_PCI_INTB);
+	GPIO_OUTPUT_DISABLE(CFG_GPIO_RESTORE_INT);
+	GPIO_INT_ACT_LOW_SET(CFG_GPIO_RESTORE_INT);
+	GPIO_OUTPUT_DISABLE(CFG_GPIO_RESTART_INT);
+	GPIO_INT_ACT_LOW_SET(CFG_GPIO_RESTART_INT);
+
+	/*
+	 * Setup GPIO's for 33MHz clock output
+	 */
+	*IXP425_GPIO_GPCLKR = 0x01FF0000;
+	GPIO_OUTPUT_ENABLE(CFG_GPIO_CLK_33M);
+
+	/*
+	 * Setup other chip select's
+	 */
+	*IXP425_EXP_CS1 = CFG_EXP_CS1;
+
+	return 0;
+}
+
+/*
+ * Check Board Identity
+ */
+int checkboard(void)
+{
+	char *s = getenv("serial#");
+
+	puts("Board: PDNB3");
+
+	if (s != NULL) {
+		puts(", serial# ");
+		puts(s);
+	}
+	putc('\n');
+
+	return (0);
+}
+
+int dram_init(void)
+{
+	gd->bd->bi_dram[0].start = PHYS_SDRAM_1;
+	gd->bd->bi_dram[0].size  = PHYS_SDRAM_1_SIZE;
+
+	return (0);
+}
+
+int do_fpga_boot(unsigned char *fpgadata)
+{
+	unsigned char *dst;
+	int status;
+	int index;
+	int i;
+	ulong len = CFG_MALLOC_LEN;
+
+	/*
+	 * Setup GPIO's for FPGA programming
+	 */
+	GPIO_OUTPUT_CLEAR(CFG_GPIO_PRG);
+	GPIO_OUTPUT_CLEAR(CFG_GPIO_CLK);
+	GPIO_OUTPUT_CLEAR(CFG_GPIO_DATA);
+
+	/*
+	 * Save value so no readback is required upon programming
+	 */
+	old_val = *IXP425_GPIO_GPOUTR;
+
+	/*
+	 * First try to decompress fpga image (gzip compressed?)
+	 */
+	dst = malloc(CFG_FPGA_MAX_SIZE);
+	if (gunzip(dst, CFG_FPGA_MAX_SIZE, (uchar *)fpgadata, &len) != 0) {
+		printf("Error: Image has to be gzipp'ed!\n");
+		return -1;
+	}
+
+	status = fpga_boot(dst, len);
+	if (status != 0) {
+		printf("\nFPGA: Booting failed ");
+		switch (status) {
+		case ERROR_FPGA_PRG_INIT_LOW:
+			printf("(Timeout: INIT not low after asserting PROGRAM*)\n ");
+			break;
+		case ERROR_FPGA_PRG_INIT_HIGH:
+			printf("(Timeout: INIT not high after deasserting PROGRAM*)\n ");
+			break;
+		case ERROR_FPGA_PRG_DONE:
+			printf("(Timeout: DONE not high after programming FPGA)\n ");
+			break;
+		}
+
+		/* display infos on fpgaimage */
+		index = 15;
+		for (i=0; i<4; i++) {
+			len = dst[index];
+			printf("FPGA: %s\n", &(dst[index+1]));
+			index += len+3;
+		}
+		putc ('\n');
+		/* delayed reboot */
+		for (i=5; i>0; i--) {
+			printf("Rebooting in %2d seconds \r",i);
+			for (index=0;index<1000;index++)
+				udelay(1000);
+		}
+		putc('\n');
+		do_reset(NULL, 0, 0, NULL);
+	}
+
+	puts("FPGA:  ");
+
+	/* display infos on fpgaimage */
+	index = 15;
+	for (i=0; i<4; i++) {
+		len = dst[index];
+		printf("%s ", &(dst[index+1]));
+		index += len+3;
+	}
+	putc('\n');
+
+	free(dst);
+
+	/*
+	 * Reset FPGA
+	 */
+	GPIO_OUTPUT_CLEAR(CFG_GPIO_FPGA_RESET);
+	udelay(10);
+	GPIO_OUTPUT_SET(CFG_GPIO_FPGA_RESET);
+
+	return (0);
+}
+
+int do_fpga(cmd_tbl_t *cmdtp, int flag, int argc, char *argv[])
+{
+	ulong addr;
+
+	if (argc < 2) {
+		printf ("Usage:\n%s\n", cmdtp->usage);
+		return 1;
+	}
+
+	addr = simple_strtoul(argv[1], NULL, 16);
+
+	return do_fpga_boot((unsigned char *)addr);
+}
+
+U_BOOT_CMD(
+	fpga,     2,     0,      do_fpga,
+	"fpga    - boot FPGA\n",
+	"address size\n    - boot FPGA with gzipped image at <address>\n"
+);
+
+#if (CONFIG_COMMANDS & CFG_CMD_PCI) || defined(CONFIG_PCI)
+extern struct pci_controller hose;
+extern void pci_ixp_init(struct pci_controller * hose);
+
+void pci_init_board(void)
+{
+	extern void pci_ixp_init (struct pci_controller *hose);
+
+	pci_ixp_init(&hose);
+}
+#endif

board/prodrive/pdnb3/u-boot.lds

@@ -0,0 +1,56 @@
+/*
+ * (C) Copyright 2006
+ * Wolfgang Denk, DENX Software Engineering, wd@denx.de.
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+OUTPUT_FORMAT("elf32-bigarm", "elf32-bigarm", "elf32-bigarm")
+OUTPUT_ARCH(arm)
+ENTRY(_start)
+SECTIONS
+{
+	. = 0x00000000;
+
+	. = ALIGN(4);
+	.text      :
+	{
+	  cpu/ixp/start.o	(.text)
+	  *(.text)
+	}
+
+	. = ALIGN(4);
+	.rodata : { *(.rodata) }
+
+	. = ALIGN(4);
+	.data : { *(.data) }
+
+	. = ALIGN(4);
+	.got : { *(.got) }
+
+	. = .;
+	__u_boot_cmd_start = .;
+	.u_boot_cmd : { *(.u_boot_cmd) }
+	__u_boot_cmd_end = .;
+
+	. = ALIGN(4);
+	__bss_start = .;
+	.bss : { *(.bss) }
+	_end = .;
+}

board/r5200/r5200.c

@@ -47,8 +47,7 @@ long int initdram (int board_type) {
 	 * Check to see if the SDRAM has already been initialized
 	 * by a run control tool
 	 */
-	if (!(mbar_readLong(MCF_SDRAMC_DACR0) & MCF_SDRAMC_DACRn_RE))
-	{
+	if (!(mbar_readLong(MCF_SDRAMC_DACR0) & MCF_SDRAMC_DACRn_RE)) {
 		/*
 		 * Initialize DRAM Control Register: DCR
 		 */
@@ -67,7 +66,7 @@ long int initdram (int board_type) {
 		/*
 		 * Initialize DMR0
 		 */
-		mbar_writeLong(MCF_SDRAMC_DMR0, 
+		mbar_writeLong(MCF_SDRAMC_DMR0,
 				MCF_SDRAMC_DMRn_BAM_8M
 				| MCF_SDRAMC_DMRn_V);
 

board/tqm8xx/tqm8xx.c

@@ -1,5 +1,5 @@
 /*
- * (C) Copyright 2000-2004
+ * (C) Copyright 2000-2006
  * Wolfgang Denk, DENX Software Engineering, wd@denx.de.
  *
  * See file CREDITS for list of people who contributed to this
@@ -124,6 +124,9 @@ int checkboard (void)
 			break;
 		putc (*s);
 	}
+#ifdef CONFIG_VIRTLAB2
+	puts (" (Virtlab2)");
+#endif
 	putc ('\n');
 
 	return (0);

common/cmd_pcmcia.c

@@ -1,5 +1,5 @@
 /*
- * (C) Copyright 2000-2004
+ * (C) Copyright 2000-2006
  * Wolfgang Denk, DENX Software Engineering, wd@denx.de.
  *
  * See file CREDITS for list of people who contributed to this
@@ -718,7 +718,8 @@ static int hardware_disable(int slot)
 /* SC8xx   Boards by SinoVee Microsystems				*/
 /* -------------------------------------------------------------------- */
 
-#if defined(CONFIG_TQM8xxL) || defined(CONFIG_SVM_SC8xx)
+#if (defined(CONFIG_TQM8xxL) || defined(CONFIG_SVM_SC8xx)) \
+	&& !defined(CONFIG_VIRTLAB2)
 
 #if defined(CONFIG_TQM8xxL)
 #define PCMCIA_BOARD_MSG "TQM8xxL"
@@ -1021,6 +1022,183 @@ done:
 
 #endif	/* TQM8xxL */
 
+/* -------------------------------------------------------------------- */
+/* Virtlab2 Board by TQ Components					*/
+/* -------------------------------------------------------------------- */
+
+#if defined(CONFIG_VIRTLAB2)
+#define PCMCIA_BOARD_MSG "Virtlab2"
+
+static int hardware_enable(int slot)
+{
+	volatile pcmconf8xx_t	*pcmp =
+		(pcmconf8xx_t *)&(((immap_t *)CFG_IMMR)->im_pcmcia);
+	volatile unsigned char	*powerctl =
+		(volatile unsigned char *)PCMCIA_CTRL;
+	volatile sysconf8xx_t	*sysp =
+		(sysconf8xx_t *)(&(((immap_t *)CFG_IMMR)->im_siu_conf));
+	unsigned int		reg, mask;
+
+	debug ("hardware_enable: " PCMCIA_BOARD_MSG " Slot %c\n", 'A'+slot);
+
+	udelay(10000);
+
+	/*
+	 * Configure SIUMCR to enable PCMCIA port B
+	 */
+	sysp->sc_siumcr &= ~SIUMCR_DBGC11;	/* set DBGC to 00 */
+
+	/* clear interrupt state, and disable interrupts */
+	pcmp->pcmc_pscr =  PCMCIA_MASK(slot);
+	pcmp->pcmc_per &= ~PCMCIA_MASK(slot);
+
+	/*
+	 * Disable interrupts, DMA, and PCMCIA buffers
+	 * (isolate the interface) and assert RESET signal
+	 */
+	debug ("Disable PCMCIA buffers and assert RESET\n");
+	reg = __MY_PCMCIA_GCRX_CXRESET;		/* active high */
+	reg |= __MY_PCMCIA_GCRX_CXOE;		/* active low  */
+
+	PCMCIA_PGCRX(slot) = reg;
+	udelay(500);
+
+	/* remove all power */
+	*powerctl = 0;
+
+	/*
+	 * Make sure there is a card in the slot, then configure the interface.
+	 */
+	udelay(10000);
+	debug ("[%d] %s: PIPR(%p)=0x%x\n", __LINE__,__FUNCTION__,
+		&(pcmp->pcmc_pipr),pcmp->pcmc_pipr);
+
+	if (pcmp->pcmc_pipr & (0x18000000 >> (slot << 4))) {
+		printf ("   No Card found\n");
+		return (1);
+	}
+
+	/*
+	 * Power On.
+	 */
+	mask = PCMCIA_VS1(slot) | PCMCIA_VS2(slot);
+	reg  = pcmp->pcmc_pipr;
+	debug ("PIPR: 0x%x ==> VS1=o%s, VS2=o%s\n", reg,
+		(reg&PCMCIA_VS1(slot))?"n":"ff",
+		(reg&PCMCIA_VS2(slot))?"n":"ff");
+
+	if ((reg & mask) == mask) {
+		*powerctl = 2;	/* Enable 5V Vccout */
+		puts (" 5.0V card found: ");
+	} else {
+		*powerctl = 1;	/* Enable 3.3 V Vccout */
+		puts (" 3.3V card found: ");
+	}
+
+	udelay(1000);
+	debug ("Enable PCMCIA buffers and stop RESET\n");
+	reg  =  PCMCIA_PGCRX(slot);
+	reg &= ~__MY_PCMCIA_GCRX_CXRESET;	/* active high */
+	reg &= ~__MY_PCMCIA_GCRX_CXOE;		/* active low  */
+
+	PCMCIA_PGCRX(slot) = reg;
+
+	udelay(250000);	/* some cards need >150 ms to come up :-( */
+
+	debug ("# hardware_enable done\n");
+
+	return (0);
+}
+
+#if (CONFIG_COMMANDS & CFG_CMD_PCMCIA)
+static int hardware_disable(int slot)
+{
+	volatile unsigned char	*powerctl =
+		(volatile unsigned char *)PCMCIA_CTRL;
+	unsigned long		 reg;
+
+	debug ("hardware_disable: " PCMCIA_BOARD_MSG " Slot %c\n", 'A'+slot);
+
+	/* remove all power */
+	*powerctl = 0;
+
+	debug ("Disable PCMCIA buffers and assert RESET\n");
+	reg = __MY_PCMCIA_GCRX_CXRESET;		/* active high */
+	reg |= __MY_PCMCIA_GCRX_CXOE;		/* active low  */
+
+	PCMCIA_PGCRX(slot) = reg;
+
+	udelay(10000);
+
+	return (0);
+}
+#endif
+
+static int voltage_set(int slot, int vcc, int vpp)
+{
+#ifdef DEBUG
+	volatile pcmconf8xx_t	*pcmp =
+		(pcmconf8xx_t *)(&(((immap_t *)CFG_IMMR)->im_pcmcia));
+#endif
+	volatile unsigned char	*powerctl =
+		(volatile unsigned char *)PCMCIA_CTRL;
+	unsigned long		reg;
+
+	debug ("voltage_set: " PCMCIA_BOARD_MSG
+		" Slot %c, Vcc=%d.%d, Vpp=%d.%d\n",
+		'A'+slot, vcc/10, vcc%10, vpp/10, vcc%10);
+
+	/*
+	 * Disable PCMCIA buffers (isolate the interface)
+	 * and assert RESET signal
+	 */
+	debug ("Disable PCMCIA buffers and assert RESET\n");
+	reg  = PCMCIA_PGCRX(slot);
+	reg |= __MY_PCMCIA_GCRX_CXRESET;	/* active high */
+	reg |= __MY_PCMCIA_GCRX_CXOE;		/* active low  */
+
+	PCMCIA_PGCRX(slot) = reg;
+	udelay(500);
+
+	/*
+	 * Configure pins for 5 Volts Enable and 3 Volts enable,
+	 * Turn off all power.
+	 */
+	debug ("PCMCIA power OFF\n");
+	reg = 0;
+	switch(vcc) {
+	case  0: 		break;
+	case 33: reg = 0x0001;	break;
+	case 50: reg = 0x0002;	break;
+	default: 		goto done;
+	}
+
+	/* Checking supported voltages */
+
+	debug ("PIPR: 0x%x --> %s\n", pcmp->pcmc_pipr,
+		(pcmp->pcmc_pipr & 0x00008000) ? "only 5 V" : "can do 3.3V");
+
+	*powerctl = reg;
+
+	if (reg) {
+		debug ("PCMCIA powered at %sV\n", (reg&0x0004) ? "5.0" : "3.3");
+	} else {
+		debug ("PCMCIA powered down\n");
+	}
+
+done:
+	debug ("Enable PCMCIA buffers and stop RESET\n");
+	reg  =  PCMCIA_PGCRX(slot);
+	reg &= ~__MY_PCMCIA_GCRX_CXRESET;	/* active high */
+	reg &= ~__MY_PCMCIA_GCRX_CXOE;		/* active low  */
+
+	PCMCIA_PGCRX(slot) = reg;
+	udelay(500);
+
+	debug ("voltage_set: " PCMCIA_BOARD_MSG " Slot %c, DONE\n", slot+'A');
+	return (0);
+}
+#endif	/* CONFIG_VIRTLAB2 */
 
 /* -------------------------------------------------------------------- */
 /* LWMON Board								*/

common/soft_i2c.c

@@ -33,6 +33,9 @@
 #include <asm/io.h>
 #include <asm/arch/hardware.h>
 #endif
+#ifdef	CONFIG_IXP425			/* only valid for IXP425 */
+#include <asm/arch/ixp425.h>
+#endif
 #include <i2c.h>
 
 #if defined(CONFIG_SOFT_I2C)

common/xyzModem.c

@@ -379,13 +379,13 @@ xyzModem_get_hdr(void)
     }
 
     /* Header found, now read the data */
-    res = CYGACC_COMM_IF_GETC_TIMEOUT(*xyz.__chan, &xyz.blk);
+    res = CYGACC_COMM_IF_GETC_TIMEOUT(*xyz.__chan, (char *)&xyz.blk);
     ZM_DEBUG(zm_save(xyz.blk));
     if (!res) {
         ZM_DEBUG(zm_dump(__LINE__));
         return xyzModem_timeout;
     }
-    res = CYGACC_COMM_IF_GETC_TIMEOUT(*xyz.__chan, &xyz.cblk);
+    res = CYGACC_COMM_IF_GETC_TIMEOUT(*xyz.__chan, (char *)&xyz.cblk);
     ZM_DEBUG(zm_save(xyz.cblk));
     if (!res) {
         ZM_DEBUG(zm_dump(__LINE__));
@@ -403,14 +403,14 @@ xyzModem_get_hdr(void)
             return xyzModem_timeout;
         }
     }
-    res = CYGACC_COMM_IF_GETC_TIMEOUT(*xyz.__chan, &xyz.crc1);
+    res = CYGACC_COMM_IF_GETC_TIMEOUT(*xyz.__chan, (char *)&xyz.crc1);
     ZM_DEBUG(zm_save(xyz.crc1));
     if (!res) {
         ZM_DEBUG(zm_dump(__LINE__));
         return xyzModem_timeout;
     }
     if (xyz.crc_mode) {
-        res = CYGACC_COMM_IF_GETC_TIMEOUT(*xyz.__chan, &xyz.crc2);
+        res = CYGACC_COMM_IF_GETC_TIMEOUT(*xyz.__chan, (char *)&xyz.crc2);
         ZM_DEBUG(zm_save(xyz.crc2));
         if (!res) {
             ZM_DEBUG(zm_dump(__LINE__));
@@ -450,7 +450,10 @@ xyzModem_get_hdr(void)
 int
 xyzModem_stream_open(connection_info_t *info, int *err)
 {
-    int console_chan, stat=0;
+#ifdef REDBOOT
+    int console_chan;
+#endif
+    int stat = 0;
     int retries = xyzModem_MAX_RETRIES;
     int crc_retries = xyzModem_MAX_RETRIES_WITH_CRC;
 
@@ -510,7 +513,7 @@ xyzModem_stream_open(connection_info_t *info, int *err)
                 /* skip filename */
                 while (*xyz.bufp++);
                 /* get the length */
-                parse_num(xyz.bufp, &xyz.file_length, NULL, " ");
+                parse_num((char *)xyz.bufp, &xyz.file_length, NULL, " ");
 #endif
                 /* The rest of the file name data block quietly discarded */
                 xyz.tx_ack = true;

config.mk

@@ -143,14 +143,15 @@ CFLAGS := $(CPPFLAGS) -Wall -Wno-trigraphs
 endif
 endif
 
+AFLAGS_DEBUG := -Wa,-gstabs
+
 # turn jbsr into jsr for m68k
 ifeq ($(ARCH),m68k)
 ifeq ($(findstring 3.4,$(shell $(CC) --version)),3.4)
 AFLAGS_DEBUG := -Wa,-gstabs,-S
 endif
-else
-AFLAGS_DEBUG := -Wa,-gstabs
 endif
+
 AFLAGS := $(AFLAGS_DEBUG) -D__ASSEMBLY__ $(CPPFLAGS)
 
 LDFLAGS += -Bstatic -T $(LDSCRIPT) -Ttext $(TEXT_BASE) $(PLATFORM_LDFLAGS)

cpu/arm926ejs/Makefile

@@ -26,7 +26,7 @@ include $(TOPDIR)/config.mk
 LIB	= lib$(CPU).a
 
 START	= start.o
-OBJS	= interrupts.o cpu.o
+OBJS	= interrupts.o cpu.o cpuinfo.o
 
 all:	.depend $(START) $(LIB)
 

cpu/arm926ejs/cpuinfo.c

@@ -0,0 +1,244 @@
+/*
+ * OMAP1 CPU identification code
+ *
+ * Copyright (C) 2004 Nokia Corporation
+ * Written by Tony Lindgren <tony@atomide.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <common.h>
+#include <command.h>
+#include <arm926ejs.h>
+
+#if defined(CONFIG_DISPLAY_CPUINFO) && defined(CONFIG_OMAP)
+
+#define omap_readw(x)		*(volatile unsigned short *)(x)
+#define omap_readl(x)		*(volatile unsigned long *)(x)
+
+#define ARRAY_SIZE(x)		(sizeof(x) / sizeof((x)[0]))
+
+#define OMAP_DIE_ID_0		0xfffe1800
+#define OMAP_DIE_ID_1		0xfffe1804
+#define OMAP_PRODUCTION_ID_0	0xfffe2000
+#define OMAP_PRODUCTION_ID_1	0xfffe2004
+#define OMAP32_ID_0		0xfffed400
+#define OMAP32_ID_1		0xfffed404
+
+struct omap_id {
+	u16	jtag_id;	/* Used to determine OMAP type */
+	u8	die_rev;	/* Processor revision */
+	u32	omap_id;	/* OMAP revision */
+	u32	type;		/* Cpu id bits [31:08], cpu class bits [07:00] */
+};
+
+/* Register values to detect the OMAP version */
+static struct omap_id omap_ids[] = {
+	{ .jtag_id = 0xb574, .die_rev = 0x2, .omap_id = 0x03310315, .type = 0x03100000},
+	{ .jtag_id = 0x355f, .die_rev = 0x0, .omap_id = 0x03320000, .type = 0x07300100},
+	{ .jtag_id = 0xb55f, .die_rev = 0x0, .omap_id = 0x03320000, .type = 0x07300300},
+	{ .jtag_id = 0xb470, .die_rev = 0x0, .omap_id = 0x03310100, .type = 0x15100000},
+	{ .jtag_id = 0xb576, .die_rev = 0x0, .omap_id = 0x03320000, .type = 0x16100000},
+	{ .jtag_id = 0xb576, .die_rev = 0x2, .omap_id = 0x03320100, .type = 0x16110000},
+	{ .jtag_id = 0xb576, .die_rev = 0x3, .omap_id = 0x03320100, .type = 0x16100c00},
+	{ .jtag_id = 0xb576, .die_rev = 0x0, .omap_id = 0x03320200, .type = 0x16100d00},
+	{ .jtag_id = 0xb613, .die_rev = 0x0, .omap_id = 0x03320300, .type = 0x1610ef00},
+	{ .jtag_id = 0xb613, .die_rev = 0x0, .omap_id = 0x03320300, .type = 0x1610ef00},
+	{ .jtag_id = 0xb576, .die_rev = 0x1, .omap_id = 0x03320100, .type = 0x16110000},
+	{ .jtag_id = 0xb58c, .die_rev = 0x2, .omap_id = 0x03320200, .type = 0x16110b00},
+	{ .jtag_id = 0xb58c, .die_rev = 0x3, .omap_id = 0x03320200, .type = 0x16110c00},
+	{ .jtag_id = 0xb65f, .die_rev = 0x0, .omap_id = 0x03320400, .type = 0x16212300},
+	{ .jtag_id = 0xb65f, .die_rev = 0x1, .omap_id = 0x03320400, .type = 0x16212300},
+	{ .jtag_id = 0xb65f, .die_rev = 0x1, .omap_id = 0x03320500, .type = 0x16212300},
+	{ .jtag_id = 0xb5f7, .die_rev = 0x0, .omap_id = 0x03330000, .type = 0x17100000},
+	{ .jtag_id = 0xb5f7, .die_rev = 0x1, .omap_id = 0x03330100, .type = 0x17100000},
+	{ .jtag_id = 0xb5f7, .die_rev = 0x2, .omap_id = 0x03330100, .type = 0x17100000},
+};
+
+/*
+ * Get OMAP type from PROD_ID.
+ * 1710 has the PROD_ID in bits 15:00, not in 16:01 as documented in TRM.
+ * 1510 PROD_ID is empty, and 1610 PROD_ID does not make sense.
+ * Undocumented register in TEST BLOCK is used as fallback; This seems to
+ * work on 1510, 1610 & 1710. The official way hopefully will work in future
+ * processors.
+ */
+static u16 omap_get_jtag_id(void)
+{
+	u32 prod_id, omap_id;
+
+	prod_id = omap_readl(OMAP_PRODUCTION_ID_1);
+	omap_id = omap_readl(OMAP32_ID_1);
+
+	/* Check for unusable OMAP_PRODUCTION_ID_1 on 1611B/5912 and 730 */
+	if (((prod_id >> 20) == 0) || (prod_id == omap_id))
+		prod_id = 0;
+	else
+		prod_id &= 0xffff;
+
+	if (prod_id)
+		return prod_id;
+
+	/* Use OMAP32_ID_1 as fallback */
+	prod_id = ((omap_id >> 12) & 0xffff);
+
+	return prod_id;
+}
+
+/*
+ * Get OMAP revision from DIE_REV.
+ * Early 1710 processors may have broken OMAP_DIE_ID, it contains PROD_ID.
+ * Undocumented register in the TEST BLOCK is used as fallback.
+ * REVISIT: This does not seem to work on 1510
+ */
+static u8 omap_get_die_rev(void)
+{
+	u32 die_rev;
+
+	die_rev = omap_readl(OMAP_DIE_ID_1);
+
+	/* Check for broken OMAP_DIE_ID on early 1710 */
+	if (((die_rev >> 12) & 0xffff) == omap_get_jtag_id())
+		die_rev = 0;
+
+	die_rev = (die_rev >> 17) & 0xf;
+	if (die_rev)
+		return die_rev;
+
+	die_rev = (omap_readl(OMAP32_ID_1) >> 28) & 0xf;
+
+	return die_rev;
+}
+
+static unsigned long dpll1(void)
+{
+	unsigned short pll_ctl_val = omap_readw(DPLL_CTL_REG);
+	unsigned long rate;
+
+	rate = CONFIG_SYS_CLK_FREQ; /* Base xtal rate */
+	if (pll_ctl_val & 0x10) {
+		/* PLL enabled, apply multiplier and divisor */
+		if (pll_ctl_val & 0xf80)
+			rate *= (pll_ctl_val & 0xf80) >> 7;
+		rate /= ((pll_ctl_val & 0x60) >> 5) + 1;
+	} else {
+		/* PLL disabled, apply bypass divisor */
+		switch (pll_ctl_val & 0xc) {
+		case 0:
+			break;
+		case 0x4:
+			rate /= 2;
+			break;
+		default:
+			rate /= 4;
+			break;
+		}
+	}
+
+	return rate;
+}
+
+static unsigned long armcore(void)
+{
+	unsigned short arm_ckctl = omap_readw(ARM_CKCTL);
+
+	return (dpll1() >> ((arm_ckctl & 0x0030) >> 4));
+}
+
+int print_cpuinfo (void)
+{
+	int i;
+	u16 jtag_id;
+	u8 die_rev;
+	u32 omap_id;
+	u8 cpu_type;
+	u32 system_serial_high;
+	u32 system_serial_low;
+	u32 system_rev = 0;
+
+	jtag_id = omap_get_jtag_id();
+	die_rev = omap_get_die_rev();
+	omap_id = omap_readl(OMAP32_ID_0);
+
+#ifdef DEBUG
+	printf("OMAP_DIE_ID_0: 0x%08x\n", omap_readl(OMAP_DIE_ID_0));
+	printf("OMAP_DIE_ID_1: 0x%08x DIE_REV: %i\n",
+	       omap_readl(OMAP_DIE_ID_1),
+	       (omap_readl(OMAP_DIE_ID_1) >> 17) & 0xf);
+	printf("OMAP_PRODUCTION_ID_0: 0x%08x\n", omap_readl(OMAP_PRODUCTION_ID_0));
+	printf("OMAP_PRODUCTION_ID_1: 0x%08x JTAG_ID: 0x%04x\n",
+	       omap_readl(OMAP_PRODUCTION_ID_1),
+	       omap_readl(OMAP_PRODUCTION_ID_1) & 0xffff);
+	printf("OMAP32_ID_0: 0x%08x\n", omap_readl(OMAP32_ID_0));
+	printf("OMAP32_ID_1: 0x%08x\n", omap_readl(OMAP32_ID_1));
+	printf("JTAG_ID: 0x%04x DIE_REV: %i\n", jtag_id, die_rev);
+#endif
+
+	system_serial_high = omap_readl(OMAP_DIE_ID_0);
+	system_serial_low = omap_readl(OMAP_DIE_ID_1);
+
+	/* First check only the major version in a safe way */
+	for (i = 0; i < ARRAY_SIZE(omap_ids); i++) {
+		if (jtag_id == (omap_ids[i].jtag_id)) {
+			system_rev = omap_ids[i].type;
+			break;
+		}
+	}
+
+	/* Check if we can find the die revision */
+	for (i = 0; i < ARRAY_SIZE(omap_ids); i++) {
+		if (jtag_id == omap_ids[i].jtag_id && die_rev == omap_ids[i].die_rev) {
+			system_rev = omap_ids[i].type;
+			break;
+		}
+	}
+
+	/* Finally check also the omap_id */
+	for (i = 0; i < ARRAY_SIZE(omap_ids); i++) {
+		if (jtag_id == omap_ids[i].jtag_id
+		    && die_rev == omap_ids[i].die_rev
+		    && omap_id == omap_ids[i].omap_id) {
+			system_rev = omap_ids[i].type;
+			break;
+		}
+	}
+
+	/* Add the cpu class info (7xx, 15xx, 16xx, 24xx) */
+	cpu_type = system_rev >> 24;
+
+	switch (cpu_type) {
+	case 0x07:
+		system_rev |= 0x07;
+		break;
+	case 0x03:
+	case 0x15:
+		system_rev |= 0x15;
+		break;
+	case 0x16:
+	case 0x17:
+		system_rev |= 0x16;
+		break;
+	case 0x24:
+		system_rev |= 0x24;
+		break;
+	default:
+		printf("Unknown OMAP cpu type: 0x%02x\n", cpu_type);
+	}
+
+	printf("CPU:   OMAP%04x", system_rev >> 16);
+	if ((system_rev >> 8) & 0xff)
+		printf("%x", (system_rev >> 8) & 0xff);
+#ifdef DEBUG
+	printf(" revision %i handled as %02xxx id: %08x%08x",
+	       die_rev, system_rev & 0xff, system_serial_low, system_serial_high);
+#endif
+	printf(" at %ld.%01ld MHz (DPLL1=%ld.%01ld MHz)\n",
+	       armcore() / 1000000, (armcore() / 100000) % 10,
+	       dpll1() / 1000000, (dpll1() / 100000) % 10);
+
+	return 0;
+}
+
+#endif /* #if defined(CONFIG_DISPLAY_CPUINFO) && defined(CONFIG_OMAP) */

cpu/ixp/config.mk

@@ -27,7 +27,7 @@ BIG_ENDIAN = y
 PLATFORM_RELFLAGS += -fno-strict-aliasing  -fno-common -ffixed-r8 \
 	-msoft-float -mbig-endian
 
-PLATFORM_CPPFLAGS += -mbig-endian -march=armv4 -mtune=strongarm1100
+PLATFORM_CPPFLAGS += -mbig-endian -march=armv5te -mtune=strongarm1100
 # =========================================================================
 #
 # Supply options according to compiler version

cpu/ixp/cpu.c

@@ -34,10 +34,47 @@
 #include <command.h>
 #include <asm/arch/ixp425.h>
 
+ulong loops_per_jiffy;
+
 #ifdef CONFIG_USE_IRQ
 DECLARE_GLOBAL_DATA_PTR;
 #endif
 
+#if defined(CONFIG_DISPLAY_CPUINFO)
+int print_cpuinfo (void)
+{
+	unsigned long id;
+	int speed = 0;
+
+	asm ("mrc p15, 0, %0, c0, c0, 0":"=r" (id));
+
+	puts("CPU:   Intel IXP425 at ");
+	switch ((id & 0x000003f0) >> 4) {
+	case 0x1c:
+		loops_per_jiffy = 887467;
+		speed = 533;
+		break;
+
+	case 0x1d:
+		loops_per_jiffy = 666016;
+		speed = 400;
+		break;
+
+	case 0x1f:
+		loops_per_jiffy = 442901;
+		speed = 266;
+		break;
+	}
+
+	if (speed)
+		printf("%d MHz\n", speed);
+	else
+		puts("unknown revision\n");
+
+	return 0;
+}
+#endif /* CONFIG_DISPLAY_CPUINFO */
+
 int cpu_init (void)
 {
 	/*
@@ -48,7 +85,9 @@ int cpu_init (void)
 	FIQ_STACK_START = IRQ_STACK_START - CONFIG_STACKSIZE_IRQ;
 #endif
 
+#if (CONFIG_COMMANDS & CFG_CMD_PCI) || defined (CONFIG_PCI)
 	pci_init();
+#endif
 	return 0;
 }
 
@@ -154,3 +193,25 @@ void pci_init(void)
 	return;
 }
 */
+
+#ifdef CONFIG_BOOTCOUNT_LIMIT
+
+void bootcount_store (ulong a)
+{
+	volatile ulong *save_addr = (volatile ulong *)(CFG_BOOTCOUNT_ADDR);
+
+	save_addr[0] = a;
+	save_addr[1] = BOOTCOUNT_MAGIC;
+}
+
+ulong bootcount_load (void)
+{
+	volatile ulong *save_addr = (volatile ulong *)(CFG_BOOTCOUNT_ADDR);
+
+	if (save_addr[1] != BOOTCOUNT_MAGIC)
+		return 0;
+	else
+		return save_addr[0];
+}
+
+#endif /* CONFIG_BOOTCOUNT_LIMIT */

cpu/ixp/interrupts.c

@@ -1,5 +1,7 @@
-/* vi: set ts=8 sw=8 noet: */
 /*
+ * (C) Copyright 2006
+ * Stefan Roese, DENX Software Engineering, sr@denx.de.
+ *
  * (C) Copyright 2002
  * Sysgo Real-Time Solutions, GmbH <www.elinos.com>
  * Marius Groeger <mgroeger@sysgo.de>
@@ -31,22 +33,85 @@
 #include <asm/arch/ixp425.h>
 
 #ifdef CONFIG_USE_IRQ
+/*
+ * When interrupts are enabled, use timer 2 for time/delay generation...
+ */
+
+#define FREQ		66666666
+#define CLOCK_TICK_RATE	(((FREQ / CFG_HZ & ~IXP425_OST_RELOAD_MASK) + 1) * CFG_HZ)
+#define LATCH		((CLOCK_TICK_RATE + CFG_HZ/2) / CFG_HZ)	/* For divider */
+
+struct _irq_handler {
+	void                *m_data;
+	void (*m_func)( void *data);
+};
+
+static struct _irq_handler IRQ_HANDLER[N_IRQS];
+
+static volatile ulong timestamp;
+
 /* enable IRQ/FIQ interrupts */
-void enable_interrupts (void)
+void enable_interrupts(void)
 {
-#error: interrupts not implemented yet
+	unsigned long temp;
+	__asm__ __volatile__("mrs %0, cpsr\n"
+			     "bic %0, %0, #0x80\n"
+			     "msr cpsr_c, %0"
+			     : "=r" (temp)
+			     :
+			     : "memory");
 }
 
-
 /*
  * disable IRQ/FIQ interrupts
  * returns true if interrupts had been enabled before we disabled them
  */
-int disable_interrupts (void)
+int disable_interrupts(void)
 {
-#error: interrupts not implemented yet
+	unsigned long old,temp;
+	__asm__ __volatile__("mrs %0, cpsr\n"
+			     "orr %1, %0, #0x80\n"
+			     "msr cpsr_c, %1"
+			     : "=r" (old), "=r" (temp)
+			     :
+			     : "memory");
+	return (old & 0x80) == 0;
 }
-#else
+
+static void default_isr(void *data)
+{
+	printf("default_isr():  called for IRQ %d, Interrupt Status=%x PR=%x\n",
+	       (int)data, *IXP425_ICIP, *IXP425_ICIH);
+}
+
+static int next_irq(void)
+{
+	return (((*IXP425_ICIH & 0x000000fc) >> 2) - 1);
+}
+
+static void timer_isr(void *data)
+{
+	unsigned int *pTime = (unsigned int *)data;
+
+	(*pTime)++;
+
+	/*
+	 * Reset IRQ source
+	 */
+	*IXP425_OSST = IXP425_OSST_TIMER_2_PEND;
+}
+
+ulong get_timer (ulong base)
+{
+	return timestamp - base;
+}
+
+void reset_timer (void)
+{
+	timestamp = 0;
+}
+
+#else /* #ifdef CONFIG_USE_IRQ */
 void enable_interrupts (void)
 {
 	return;
@@ -55,8 +120,7 @@ int disable_interrupts (void)
 {
 	return 0;
 }
-#endif
-
+#endif /* #ifdef CONFIG_USE_IRQ */
 
 void bad_mode (void)
 {
@@ -140,19 +204,46 @@ void do_fiq (struct pt_regs *pt_regs)
 {
 	printf ("fast interrupt request\n");
 	show_regs (pt_regs);
-	bad_mode ();
+	printf("IRQ=%08lx FIQ=%08lx\n", *IXP425_ICIH, *IXP425_ICFH);
 }
 
 void do_irq (struct pt_regs *pt_regs)
 {
+#ifdef CONFIG_USE_IRQ
+	int irq = next_irq();
+
+	IRQ_HANDLER[irq].m_func(IRQ_HANDLER[irq].m_data);
+#else
 	printf ("interrupt request\n");
 	show_regs (pt_regs);
 	bad_mode ();
+#endif
 }
 
-
 int interrupt_init (void)
 {
-	/* nothing happens here - we don't setup any IRQs */
+#ifdef CONFIG_USE_IRQ
+	int i;
+
+	/* install default interrupt handlers */
+	for (i = 0; i < N_IRQS; i++) {
+		IRQ_HANDLER[i].m_data = (void *)i;
+		IRQ_HANDLER[i].m_func = default_isr;
+	}
+
+	/* install interrupt handler for timer */
+	IRQ_HANDLER[IXP425_TIMER_2_IRQ].m_data = (void *)&timestamp;
+	IRQ_HANDLER[IXP425_TIMER_2_IRQ].m_func = timer_isr;
+
+	/* setup the Timer counter value */
+	*IXP425_OSRT2 = (LATCH & ~IXP425_OST_RELOAD_MASK) | IXP425_OST_ENABLE;
+
+	/* configure interrupts for IRQ mode */
+	*IXP425_ICLR = 0x00000000;
+
+	/* enable timer irq */
+	*IXP425_ICMR = (1 << IXP425_TIMER_2_IRQ);
+#endif
+
 	return (0);
 }

cpu/ixp/npe/IxEthAcc.c

@@ -0,0 +1,261 @@
+/**
+ * @file IxEthAcc.c
+ *
+ * @author Intel Corporation
+ * @date 20-Feb-2001
+ *
+ * @brief This file contains the implementation of the IXP425 Ethernet Access Component
+ *
+ * Design Notes:
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+
+
+#include "IxEthAcc.h"
+#ifdef CONFIG_IXP425_COMPONENT_ETHDB
+#include "IxEthDB.h"
+#endif
+#include "IxFeatureCtrl.h"
+
+#include "IxEthAcc_p.h"
+#include "IxEthAccMac_p.h"
+#include "IxEthAccMii_p.h"
+
+/**
+ * @addtogroup IxEthAcc
+ *@{
+ */
+
+
+/**
+ * @brief System-wide information data strucure.
+ *
+ * @ingroup IxEthAccPri
+ *
+ */
+
+IxEthAccInfo ixEthAccDataInfo;
+extern PUBLIC IxEthAccMacState ixEthAccMacState[];
+extern PUBLIC IxOsalMutex ixEthAccControlInterfaceMutex;
+
+/**
+ * @brief System-wide information
+ *
+ * @ingroup IxEthAccPri
+ *
+ */
+BOOL ixEthAccServiceInit = FALSE;
+
+/* global filtering bit mask */
+PUBLIC UINT32 ixEthAccNewSrcMask;
+
+/**
+ * @brief Per port information data strucure.
+ *
+ * @ingroup IxEthAccPri
+ *
+ */
+
+IxEthAccPortDataInfo ixEthAccPortData[IX_ETH_ACC_NUMBER_OF_PORTS];
+
+PUBLIC IxEthAccStatus ixEthAccInit()
+{
+#ifdef CONFIG_IXP425_COMPONENT_ETHDB
+  /*
+   * Initialize Control plane
+   */
+  if (ixEthDBInit() != IX_ETH_ACC_SUCCESS)
+  {
+      IX_ETH_ACC_WARNING_LOG("ixEthAccInit: EthDB init failed\n", 0, 0, 0, 0, 0, 0);
+
+      return IX_ETH_ACC_FAIL;
+  }
+#endif
+
+  if (IX_FEATURE_CTRL_SWCONFIG_ENABLED == ixFeatureCtrlSwConfigurationCheck (IX_FEATURECTRL_ETH_LEARNING))
+  {
+      ixEthAccNewSrcMask = (~0); /* want all the bits */
+  }
+  else
+  {
+      ixEthAccNewSrcMask = (~IX_ETHACC_NE_NEWSRCMASK); /* want all but the NewSrc bit */
+  }
+
+  /*
+   * Initialize Data plane
+   */
+   if ( ixEthAccInitDataPlane()  != IX_ETH_ACC_SUCCESS )
+   {
+      IX_ETH_ACC_WARNING_LOG("ixEthAccInit: data plane init failed\n", 0, 0, 0, 0, 0, 0);
+
+       return IX_ETH_ACC_FAIL;
+   }
+
+
+   if ( ixEthAccQMgrQueuesConfig() != IX_ETH_ACC_SUCCESS )
+   {
+      IX_ETH_ACC_WARNING_LOG("ixEthAccInit: queue config failed\n", 0, 0, 0, 0, 0, 0);
+
+       return IX_ETH_ACC_FAIL;
+   }
+
+   /*
+    * Initialize MII
+    */
+   if ( ixEthAccMiiInit() != IX_ETH_ACC_SUCCESS )
+   {
+      IX_ETH_ACC_WARNING_LOG("ixEthAccInit: Mii init failed\n", 0, 0, 0, 0, 0, 0);
+
+       return IX_ETH_ACC_FAIL;
+   }
+
+   /*
+    * Initialize MAC I/O memory
+    */
+   if (ixEthAccMacMemInit() != IX_ETH_ACC_SUCCESS)
+   {
+      IX_ETH_ACC_WARNING_LOG("ixEthAccInit: Mac init failed\n", 0, 0, 0, 0, 0, 0);
+
+     return IX_ETH_ACC_FAIL;
+   }
+
+   /*
+    * Initialize control plane interface lock
+    */
+   if (ixOsalMutexInit(&ixEthAccControlInterfaceMutex) != IX_SUCCESS)
+   {
+       IX_ETH_ACC_WARNING_LOG("ixEthAccInit: Control plane interface lock initialization failed\n", 0, 0, 0, 0, 0, 0);
+
+       return IX_ETH_ACC_FAIL;
+   }
+
+   /* initialiasation is complete */
+   ixEthAccServiceInit = TRUE;
+
+   return IX_ETH_ACC_SUCCESS;
+
+}
+
+PUBLIC void ixEthAccUnload(void)
+{
+    IxEthAccPortId portId;
+
+    if ( IX_ETH_ACC_IS_SERVICE_INITIALIZED() )
+    {
+       /* check none of the port is still active */
+       for (portId = 0; portId < IX_ETH_ACC_NUMBER_OF_PORTS; portId++)
+       {
+	   if ( IX_ETH_IS_PORT_INITIALIZED(portId) )
+	   {
+	       if (ixEthAccMacState[portId].portDisableState == ACTIVE)
+	       {
+		   IX_ETH_ACC_WARNING_LOG("ixEthAccUnload: port %u still active, bail out\n", portId, 0, 0, 0, 0, 0);
+		   return;
+	       }
+	   }
+       }
+
+       /* unmap the memory areas */
+       ixEthAccMiiUnload();
+       ixEthAccMacUnload();
+
+       /* set all ports as uninitialized */
+       for (portId = 0; portId < IX_ETH_ACC_NUMBER_OF_PORTS; portId++)
+       {
+	       ixEthAccPortData[portId].portInitialized = FALSE;
+       }
+
+       /* uninitialize the service */
+       ixEthAccServiceInit = FALSE;
+   }
+}
+
+PUBLIC IxEthAccStatus ixEthAccPortInit( IxEthAccPortId portId)
+{
+
+  IxEthAccStatus ret=IX_ETH_ACC_SUCCESS;
+
+   if ( ! IX_ETH_ACC_IS_SERVICE_INITIALIZED() )
+   {
+   	return(IX_ETH_ACC_FAIL);
+   }
+
+   /*
+    * Check for valid port
+    */
+
+   if ( ! IX_ETH_ACC_IS_PORT_VALID(portId) )
+   {
+       return (IX_ETH_ACC_INVALID_PORT);
+   }
+
+   if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+   {
+       IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot initialize Eth port.\n",(INT32) portId,0,0,0,0,0);
+       return IX_ETH_ACC_SUCCESS ;
+   }
+
+   if ( IX_ETH_IS_PORT_INITIALIZED(portId) )
+   {
+   	/* Already initialized */
+   	return(IX_ETH_ACC_FAIL);
+   }
+
+   if(ixEthAccMacInit(portId)!=IX_ETH_ACC_SUCCESS)
+   {
+       return IX_ETH_ACC_FAIL;
+   }
+
+   /*
+     * Set the port init flag.
+     */
+
+    ixEthAccPortData[portId].portInitialized = TRUE;
+
+#ifdef CONFIG_IXP425_COMPONENT_ETHDB
+    /* init learning/filtering database structures for this port */
+    ixEthDBPortInit(portId);
+#endif
+
+    return(ret);
+}
+
+

cpu/ixp/npe/IxEthAccCommon.c

@@ -0,0 +1,1049 @@
+/**
+ * @file IxEthAccCommon.c
+ *
+ * @author Intel Corporation
+ * @date 12-Feb-2002
+ *
+ * @brief This file contains the implementation common support routines for the component
+ *
+ * Design Notes:
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+/*
+ * Component header files
+ */
+
+#include "IxOsal.h"
+#include "IxEthAcc.h"
+#include "IxEthDB.h"
+#include "IxNpeMh.h"
+#include "IxEthDBPortDefs.h"
+#include "IxFeatureCtrl.h"
+#include "IxEthAcc_p.h"
+#include "IxEthAccQueueAssign_p.h"
+
+#include "IxEthAccDataPlane_p.h"
+#include "IxEthAccMii_p.h"
+
+/**
+ * @addtogroup IxEthAccPri
+ *@{
+ */
+
+extern IxEthAccInfo   ixEthAccDataInfo;
+
+/**
+ *
+ * @brief Maximum number of RX queues set to be the maximum number
+ * of traffic calsses.
+ *
+ */
+#define IX_ETHACC_MAX_RX_QUEUES \
+      (IX_ETH_DB_QOS_TRAFFIC_CLASS_7_RX_QUEUE_PROPERTY \
+      - IX_ETH_DB_QOS_TRAFFIC_CLASS_0_RX_QUEUE_PROPERTY \
+      + 1)
+
+/**
+ *
+ * @brief Maximum number of 128 entry RX queues
+ *
+ */
+#define IX_ETHACC_MAX_LARGE_RX_QUEUES 4
+
+/**
+ *
+ * @brief Data structure template for Default RX Queues
+ *
+ */
+IX_ETH_ACC_PRIVATE
+IxEthAccQregInfo ixEthAccQmgrRxDefaultTemplate =
+  {
+    IX_ETH_ACC_RX_FRAME_ETH_Q, 	     /**< Queue ID */
+    "Eth Rx Q",
+    ixEthRxFrameQMCallback,          /**< Functional callback */
+    (IxQMgrCallbackId) 0,	     /**< Callback tag	      */
+    IX_QMGR_Q_SIZE128,		     /**< Allocate Max Size Q */
+    IX_QMGR_Q_ENTRY_SIZE1,	     /**< Queue Entry Sizes - all Q entries are single word entries   */
+    TRUE,			     /**< Enable Q notification at startup */
+    IX_ETH_ACC_RX_FRAME_ETH_Q_SOURCE,/**< Q Condition to drive callback   */
+    IX_QMGR_Q_WM_LEVEL0, 	     /**< Q Low water mark */
+    IX_QMGR_Q_WM_LEVEL1,	     /**< Q High water mark - needed by NPE */
+  };
+
+/**
+ *
+ * @brief Data structure template for Small RX Queues
+ *
+ */
+IX_ETH_ACC_PRIVATE
+IxEthAccQregInfo ixEthAccQmgrRxSmallTemplate =
+  {
+    IX_ETH_ACC_RX_FRAME_ETH_Q, 	     /**< Queue ID */
+    "Eth Rx Q",
+    ixEthRxFrameQMCallback,          /**< Functional callback */
+    (IxQMgrCallbackId) 0,	     /**< Callback tag	      */
+    IX_QMGR_Q_SIZE64,		     /**< Allocate Smaller Q */
+    IX_QMGR_Q_ENTRY_SIZE1,	     /**< Queue Entry Sizes - all Q entries are single word entries   */
+    TRUE,			     /**< Enable Q notification at startup */
+    IX_ETH_ACC_RX_FRAME_ETH_Q_SOURCE,/**< Q Condition to drive callback   */
+    IX_QMGR_Q_WM_LEVEL0, 	     /**< Q Low water mark */
+    IX_QMGR_Q_WM_LEVEL1,	     /**< Q High water mark - needed by NPE */
+  };
+
+
+/**
+ *
+ * @brief Data structure used to register & initialize the Queues
+ *
+ */
+IX_ETH_ACC_PRIVATE
+IxEthAccQregInfo ixEthAccQmgrStaticInfo[]=
+{
+  {
+    IX_ETH_ACC_RX_FREE_BUFF_ENET0_Q,
+    "Eth Rx Fr Q 1",
+    ixEthRxFreeQMCallback,
+    (IxQMgrCallbackId) IX_ETH_PORT_1,
+    IX_QMGR_Q_SIZE128,		     /**< Allocate Max Size Q */
+    IX_QMGR_Q_ENTRY_SIZE1,	     /**< Queue Entry Sizes - all Q entries are single word entries   */
+    FALSE,			     /**< Disable Q notification at startup */
+    IX_ETH_ACC_RX_FREE_BUFF_ENET0_Q_SOURCE, /**< Q Condition to drive callback  */
+    IX_QMGR_Q_WM_LEVEL0, 	     /***< Q Low water mark */
+    IX_QMGR_Q_WM_LEVEL64,	     /**< Q High water mark */
+  },
+
+  {
+    IX_ETH_ACC_RX_FREE_BUFF_ENET1_Q,
+    "Eth Rx Fr Q 2",
+    ixEthRxFreeQMCallback,
+    (IxQMgrCallbackId) IX_ETH_PORT_2,
+    IX_QMGR_Q_SIZE128,		     /**< Allocate Max Size Q */
+    IX_QMGR_Q_ENTRY_SIZE1,	     /**< Queue Entry Sizes - all Q entries are single word entries   */
+    FALSE,			     /**< Disable Q notification at startup */
+    IX_ETH_ACC_RX_FREE_BUFF_ENET1_Q_SOURCE,  /**< Q Condition to drive callback  */
+    IX_QMGR_Q_WM_LEVEL0, 	     /**< Q Low water mark */
+    IX_QMGR_Q_WM_LEVEL64,	     /**< Q High water mark */
+  },
+#ifdef __ixp46X
+  {
+    IX_ETH_ACC_RX_FREE_BUFF_ENET2_Q,
+    "Eth Rx Fr Q 3",
+    ixEthRxFreeQMCallback,
+    (IxQMgrCallbackId) IX_ETH_PORT_3,
+    IX_QMGR_Q_SIZE128,		     /**< Allocate Max Size Q */
+    IX_QMGR_Q_ENTRY_SIZE1,	     /**< Queue Entry Sizes - all Q entries are single word entries   */
+    FALSE,			     /**< Disable Q notification at startup */
+    IX_ETH_ACC_RX_FREE_BUFF_ENET2_Q_SOURCE,  /**< Q Condition to drive callback  */
+    IX_QMGR_Q_WM_LEVEL0, 	     /**< Q Low water mark */
+    IX_QMGR_Q_WM_LEVEL64,	     /**< Q High water mark */
+  },
+#endif
+  {
+     IX_ETH_ACC_TX_FRAME_ENET0_Q,
+    "Eth Tx Q 1",
+     ixEthTxFrameQMCallback,
+     (IxQMgrCallbackId) IX_ETH_PORT_1,
+    IX_QMGR_Q_SIZE128,		     /**< Allocate Max Size Q */
+    IX_QMGR_Q_ENTRY_SIZE1,	     /**< Queue Entry Sizes - all Q entries are single word entries   */
+    FALSE,			     /**< Disable Q notification at startup */
+    IX_ETH_ACC_TX_FRAME_ENET0_Q_SOURCE,	 /**< Q Condition to drive callback  */
+    IX_QMGR_Q_WM_LEVEL0, 	     /**< Q Low water mark */
+    IX_QMGR_Q_WM_LEVEL64,	     /**< Q High water mark */
+  },
+
+  {
+     IX_ETH_ACC_TX_FRAME_ENET1_Q,
+    "Eth Tx Q 2",
+     ixEthTxFrameQMCallback,
+     (IxQMgrCallbackId) IX_ETH_PORT_2,
+    IX_QMGR_Q_SIZE128,		     /**< Allocate Max Size Q */
+    IX_QMGR_Q_ENTRY_SIZE1,	     /**< Queue Entry Sizes - all Q entries are single word entries   */
+    FALSE,			     /**< Disable Q notification at startup */
+    IX_ETH_ACC_TX_FRAME_ENET1_Q_SOURCE,	     /**< Q Condition to drive callback  */
+    IX_QMGR_Q_WM_LEVEL0, 	     /**< Q Low water mark */
+    IX_QMGR_Q_WM_LEVEL64,	     /**< Q High water mark */
+  },
+#ifdef __ixp46X
+  {
+     IX_ETH_ACC_TX_FRAME_ENET2_Q,
+    "Eth Tx Q 3",
+     ixEthTxFrameQMCallback,
+     (IxQMgrCallbackId) IX_ETH_PORT_3,
+    IX_QMGR_Q_SIZE128,		     /**< Allocate Max Size Q */
+    IX_QMGR_Q_ENTRY_SIZE1,	     /** Queue Entry Sizes - all Q entries are single ord entries   */
+    FALSE,			     /** Disable Q notification at startup */
+    IX_ETH_ACC_TX_FRAME_ENET2_Q_SOURCE,	     /** Q Condition to drive callback  */
+    IX_QMGR_Q_WM_LEVEL0, 	     /* No queues use almost empty */
+    IX_QMGR_Q_WM_LEVEL64,	      /** Q High water mark - needed used  */
+  },
+#endif
+  {
+     IX_ETH_ACC_TX_FRAME_DONE_ETH_Q,
+    "Eth Tx Done Q",
+     ixEthTxFrameDoneQMCallback,
+     (IxQMgrCallbackId) 0,
+    IX_QMGR_Q_SIZE128,		     /**< Allocate Max Size Q */
+    IX_QMGR_Q_ENTRY_SIZE1,	     /**< Queue Entry Sizes - all Q entries are single word entries   */
+    TRUE,			     /**< Enable Q notification at startup */
+    IX_ETH_ACC_TX_FRAME_DONE_ETH_Q_SOURCE, /**< Q Condition to drive callback  */
+    IX_QMGR_Q_WM_LEVEL0, 	     /**< Q Low water mark */
+    IX_QMGR_Q_WM_LEVEL2,	     /**< Q High water mark - needed by NPE */
+  },
+
+  {  /* Null Termination entry
+      */
+     (IxQMgrQId)0,
+     (char *) NULL,
+     (IxQMgrCallback) NULL,
+     (IxQMgrCallbackId) 0,
+     0,
+     0,
+     0,
+     0,
+     0,
+     0
+  }
+
+};
+
+/**
+ *
+ * @brief Data structure used to register & initialize the Queues
+ *
+ * The structure will be filled at run time depending on the NPE
+ * image already loaded and the QoS configured in ethDB.
+ *
+ */
+IX_ETH_ACC_PRIVATE
+IxEthAccQregInfo ixEthAccQmgrRxQueuesInfo[IX_ETHACC_MAX_RX_QUEUES+1]=
+{
+  {  /* PlaceHolder for rx queues
+      * depending on the QoS configured
+      */
+     (IxQMgrQId)0,
+     (char *) NULL,
+     (IxQMgrCallback) NULL,
+     (IxQMgrCallbackId) 0,
+     0,
+     0,
+     0,
+     0,
+     0,
+     0
+  },
+
+  {  /* PlaceHolder for rx queues
+      * depending on the QoS configured
+      */
+     (IxQMgrQId)0,
+     (char *) NULL,
+     (IxQMgrCallback) NULL,
+     (IxQMgrCallbackId) 0,
+     0,
+     0,
+     0,
+     0,
+     0,
+     0
+  },
+
+  {  /* PlaceHolder for rx queues
+      * depending on the QoS configured
+      */
+     (IxQMgrQId)0,
+     (char *) NULL,
+     (IxQMgrCallback) NULL,
+     (IxQMgrCallbackId) 0,
+     0,
+     0,
+     0,
+     0,
+     0,
+     0
+  },
+
+  {  /* PlaceHolder for rx queues
+      * depending on the QoS configured
+      */
+     (IxQMgrQId)0,
+     (char *) NULL,
+     (IxQMgrCallback) NULL,
+     (IxQMgrCallbackId) 0,
+     0,
+     0,
+     0,
+     0,
+     0,
+     0
+  },
+
+  {  /* PlaceHolder for rx queues
+      * depending on the QoS configured
+      */
+     (IxQMgrQId)0,
+     (char *) NULL,
+     (IxQMgrCallback) NULL,
+     (IxQMgrCallbackId) 0,
+     0,
+     0,
+     0,
+     0,
+     0,
+     0
+  },
+
+  {  /* PlaceHolder for rx queues
+      * depending on the QoS configured
+      */
+      (IxQMgrQId)0,
+     (char *) NULL,
+     (IxQMgrCallback) NULL,
+     (IxQMgrCallbackId) 0,
+     0,
+     0,
+     0,
+     0,
+     0,
+     0
+  },
+
+  {  /* PlaceHolder for rx queues
+      * depending on the QoS configured
+      */
+     (IxQMgrQId)0,
+     (char *) NULL,
+     (IxQMgrCallback) NULL,
+     (IxQMgrCallbackId) 0,
+     0,
+     0,
+     0,
+     0,
+     0,
+     0
+  },
+
+  {  /* PlaceHolder for rx queues
+      * depending on the QoS configured
+      */
+     (IxQMgrQId)0,
+     (char *) NULL,
+     (IxQMgrCallback) NULL,
+     (IxQMgrCallbackId) 0,
+     0,
+     0,
+     0,
+     0,
+     0,
+     0
+  },
+
+  {  /* Null Termination entry
+      */
+     (IxQMgrQId)0,
+     (char *) NULL,
+     (IxQMgrCallback) NULL,
+     (IxQMgrCallbackId) 0,
+     0,
+     0,
+     0,
+     0,
+     0,
+     0
+  }
+
+};
+
+/* forward declarations */
+IX_ETH_ACC_PRIVATE IxEthAccStatus
+ixEthAccQMgrQueueSetup(IxEthAccQregInfo *qInfoDes);
+
+/**
+ * @fn ixEthAccQMgrQueueSetup(void)
+ *
+ * @brief Setup one queue and its event, and register the callback required
+ * by this component to the QMgr
+ *
+ * @internal
+ */
+IX_ETH_ACC_PRIVATE IxEthAccStatus
+ixEthAccQMgrQueueSetup(IxEthAccQregInfo *qInfoDes)
+{
+    /*
+     * Configure each Q.
+     */
+    if ( ixQMgrQConfig( qInfoDes->qName,
+			qInfoDes->qId,
+			qInfoDes->qSize,
+			qInfoDes->qWords) != IX_SUCCESS)
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+    if ( ixQMgrWatermarkSet( qInfoDes->qId,
+			     qInfoDes->AlmostEmptyThreshold,
+			     qInfoDes->AlmostFullThreshold
+			     ) != IX_SUCCESS)
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+    /*
+     * Set dispatcher priority.
+     */
+    if ( ixQMgrDispatcherPrioritySet( qInfoDes->qId,
+				      IX_ETH_ACC_QM_QUEUE_DISPATCH_PRIORITY)
+	 != IX_SUCCESS)
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+    /*
+     * Register callbacks for each Q.
+     */
+    if ( ixQMgrNotificationCallbackSet(qInfoDes->qId,
+				       qInfoDes->qCallback,
+				       qInfoDes->callbackTag)
+	 != IX_SUCCESS )
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+    /*
+     * Set notification condition for Q
+     */
+    if ( qInfoDes->qNotificationEnableAtStartup == TRUE )
+    {
+	if (   ixQMgrNotificationEnable(qInfoDes->qId,
+					qInfoDes->qConditionSource)
+	       != IX_SUCCESS )
+	{
+	    return IX_ETH_ACC_FAIL;
+	}
+    }
+
+    return(IX_ETH_ACC_SUCCESS);
+}
+
+/**
+ * @fn ixEthAccQMgrQueuesConfig(void)
+ *
+ * @brief Setup all the queues and register all callbacks required
+ * by this component to the QMgr
+ *
+ * The RxFree queues, tx queues, rx queues are configured statically
+ *
+ * Rx queues configuration is driven by QoS setup.
+ * Many Rx queues may be required when QoS is enabled (this depends
+ * on IxEthDB setup and the images being downloaded). The configuration
+ * of the rxQueues is done in many steps as follows:
+ *
+ * @li select all Rx queues as configured by ethDB for all ports
+ * @li sort the queues by traffic class
+ * @li build the priority dependency for all queues
+ * @li fill the configuration for all rx queues
+ * @li configure all statically configured queues
+ * @li configure all dynamically configured queues
+ *
+ * @param none
+ *
+ * @return IxEthAccStatus
+ *
+ * @internal
+ */
+IX_ETH_ACC_PUBLIC
+IxEthAccStatus ixEthAccQMgrQueuesConfig(void)
+{
+    struct
+    {
+	int npeCount;
+	UINT32 npeId;
+	IxQMgrQId qId;
+	IxEthDBProperty trafficClass;
+    } rxQueues[IX_ETHACC_MAX_RX_QUEUES];
+
+    UINT32 rxQueue = 0;
+    UINT32 rxQueueCount = 0;
+    IxQMgrQId ixQId =IX_QMGR_MAX_NUM_QUEUES;
+    IxEthDBStatus ixEthDBStatus = IX_ETH_DB_SUCCESS;
+    IxEthDBPortId ixEthDbPortId = 0;
+    IxEthAccPortId ixEthAccPortId = 0;
+    UINT32 ixNpeId = 0;
+    UINT32 ixHighestNpeId = 0;
+    UINT32 sortIterations = 0;
+    IxEthAccStatus ret = IX_ETH_ACC_SUCCESS;
+    IxEthAccQregInfo *qInfoDes = NULL;
+    IxEthDBProperty ixEthDBTrafficClass = IX_ETH_DB_QOS_TRAFFIC_CLASS_0_RX_QUEUE_PROPERTY;
+    IxEthDBPropertyType ixEthDBPropertyType = IX_ETH_DB_INTEGER_PROPERTY;
+    UINT32 ixEthDBParameter = 0;
+    BOOL completelySorted = FALSE;
+
+    /* Fill the corspondance between ports and queues
+     * This defines the mapping from port to queue Ids.
+     */
+
+    ixEthAccPortData[IX_ETH_PORT_1].ixEthAccRxData.rxFreeQueue
+	= IX_ETH_ACC_RX_FREE_BUFF_ENET0_Q;
+    ixEthAccPortData[IX_ETH_PORT_2].ixEthAccRxData.rxFreeQueue
+	= IX_ETH_ACC_RX_FREE_BUFF_ENET1_Q;
+#ifdef __ixp46X
+    ixEthAccPortData[IX_ETH_PORT_3].ixEthAccRxData.rxFreeQueue
+	= IX_ETH_ACC_RX_FREE_BUFF_ENET2_Q;
+#endif
+    ixEthAccPortData[IX_ETH_PORT_1].ixEthAccTxData.txQueue
+	= IX_ETH_ACC_TX_FRAME_ENET0_Q;
+    ixEthAccPortData[IX_ETH_PORT_2].ixEthAccTxData.txQueue
+	= IX_ETH_ACC_TX_FRAME_ENET1_Q;
+#ifdef __ixp46X
+    ixEthAccPortData[IX_ETH_PORT_3].ixEthAccTxData.txQueue
+	= IX_ETH_ACC_TX_FRAME_ENET2_Q;
+#endif
+    /* Fill the corspondance between ports and NPEs
+     * This defines the mapping from port to npeIds.
+     */
+
+    ixEthAccPortData[IX_ETH_PORT_1].npeId = IX_NPEMH_NPEID_NPEB;
+    ixEthAccPortData[IX_ETH_PORT_2].npeId = IX_NPEMH_NPEID_NPEC;
+#ifdef __ixp46X
+    ixEthAccPortData[IX_ETH_PORT_3].npeId = IX_NPEMH_NPEID_NPEA;
+#endif
+    /* set the default rx scheduling discipline */
+    ixEthAccDataInfo.schDiscipline = FIFO_NO_PRIORITY;
+
+    /*
+     * Queue Selection step:
+     *
+     * The following code selects all the queues and build
+     * a temporary array which contains for each queue
+     * - the queue Id,
+     * - the highest traffic class (in case of many
+     * priorities configured for the same queue on different
+     * ports)
+     * - the number of different Npes which are
+     * configured to write to this queue.
+     *
+     * The output of this loop is a temporary array of RX queues
+     * in any order.
+     *
+     */
+#ifdef CONFIG_IXP425_COMPONENT_ETHDB
+    for (ixEthAccPortId = 0;
+	 (ixEthAccPortId < IX_ETH_ACC_NUMBER_OF_PORTS)
+	     && (ret == IX_ETH_ACC_SUCCESS);
+	 ixEthAccPortId++)
+    {
+	/* map between ethDb and ethAcc port Ids */
+	ixEthDbPortId = (IxEthDBPortId)ixEthAccPortId;
+
+	/* map between npeId and ethAcc port Ids */
+	ixNpeId = IX_ETH_ACC_PORT_TO_NPE_ID(ixEthAccPortId);
+
+	/* Iterate thru the different priorities */
+	for (ixEthDBTrafficClass = IX_ETH_DB_QOS_TRAFFIC_CLASS_0_RX_QUEUE_PROPERTY;
+	     ixEthDBTrafficClass <= IX_ETH_DB_QOS_TRAFFIC_CLASS_7_RX_QUEUE_PROPERTY;
+	     ixEthDBTrafficClass++)
+	{
+	    ixEthDBStatus = ixEthDBFeaturePropertyGet(
+	      ixEthDbPortId,
+	      IX_ETH_DB_VLAN_QOS,
+	      ixEthDBTrafficClass,
+	      &ixEthDBPropertyType,
+	      (void *)&ixEthDBParameter);
+
+	    if (ixEthDBStatus == IX_ETH_DB_SUCCESS)
+	    {
+		/* This port and QoS class are mapped to
+		 * a RX queue.
+		 */
+		if (ixEthDBPropertyType == IX_ETH_DB_INTEGER_PROPERTY)
+		{
+		    /* remember the highest npe Id supporting ethernet */
+		    if (ixNpeId > ixHighestNpeId)
+		    {
+			ixHighestNpeId = ixNpeId;
+		    }
+
+		    /* search the queue in the list of queues
+		     * already used by an other port or QoS
+		     */
+		    for (rxQueue = 0;
+			 rxQueue < rxQueueCount;
+			 rxQueue++)
+		    {
+			if (rxQueues[rxQueue].qId == (IxQMgrQId)ixEthDBParameter)
+			{
+			    /* found an existing setup, update the number of ports
+			     * for this queue if the port maps to
+			     * a different NPE.
+			     */
+			    if (rxQueues[rxQueue].npeId != ixNpeId)
+			    {
+				rxQueues[rxQueue].npeCount++;
+				rxQueues[rxQueue].npeId = ixNpeId;
+			    }
+			    /* get the highest traffic class for this queue */
+			    if (rxQueues[rxQueue].trafficClass > ixEthDBTrafficClass)
+			    {
+				rxQueues[rxQueue].trafficClass = ixEthDBTrafficClass;
+			    }
+			    break;
+			}
+		    }
+		    if (rxQueue == rxQueueCount)
+		    {
+			/* new queue not found in the current list,
+			 * add a new entry.
+			 */
+			IX_OSAL_ASSERT(rxQueueCount < IX_ETHACC_MAX_RX_QUEUES);
+			rxQueues[rxQueueCount].qId = ixEthDBParameter;
+			rxQueues[rxQueueCount].npeCount = 1;
+			rxQueues[rxQueueCount].npeId = ixNpeId;
+			rxQueues[rxQueueCount].trafficClass = ixEthDBTrafficClass;
+			rxQueueCount++;
+		    }
+		}
+		else
+		{
+		    /* unexpected property type (not Integer) */
+		    ret = IX_ETH_ACC_FAIL;
+
+                    IX_ETH_ACC_WARNING_LOG("ixEthAccQMgrQueuesConfig: unexpected property type returned by EthDB\n", 0, 0, 0, 0, 0, 0);
+
+		    /* no point to continue to iterate */
+		    break;
+		}
+	    }
+	    else
+	    {
+		/* No Rx queue configured for this port
+		 * and this traffic class. Do nothing.
+		 */
+	    }
+	}
+
+        /* notify EthDB that queue initialization is complete and traffic class allocation is frozen */
+        ixEthDBFeaturePropertySet(ixEthDbPortId,
+            IX_ETH_DB_VLAN_QOS,
+            IX_ETH_DB_QOS_QUEUE_CONFIGURATION_COMPLETE,
+            NULL /* ignored */);
+    }
+
+#else
+
+    ixNpeId = IX_ETH_ACC_PORT_TO_NPE_ID(ixEthAccPortId);
+    rxQueues[0].qId = 4;
+    rxQueues[0].npeCount = 1;
+    rxQueues[0].npeId = ixNpeId;
+    rxQueues[0].trafficClass = IX_ETH_DB_QOS_TRAFFIC_CLASS_0_RX_QUEUE_PROPERTY;
+    rxQueueCount++;
+
+#endif
+
+    /* check there is at least 1 rx queue : there is no point
+     * to continue if there is no rx queue configured
+     */
+    if ((rxQueueCount == 0) || (ret == IX_ETH_ACC_FAIL))
+    {
+        IX_ETH_ACC_WARNING_LOG("ixEthAccQMgrQueuesConfig: no queues configured, bailing out\n", 0, 0, 0, 0, 0, 0);
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    /* Queue sort step:
+     *
+     * Re-order the array of queues by decreasing traffic class
+     * using a bubble sort. (trafficClass 0 is the lowest
+     * priority traffic, trafficClass 7 is the highest priority traffic)
+     *
+     * Primary sort order is traffic class
+     * Secondary sort order is npeId
+     *
+     * Note that a bubble sort algorithm is not very efficient when
+     * the number of queues grows . However, this is not a very bad choice
+     * considering the very small number of entries to sort. Also, bubble
+     * sort is extremely fast when the list is already sorted.
+     *
+     * The output of this loop is a sorted array of queues.
+     *
+     */
+    sortIterations = 0;
+    do
+    {
+	sortIterations++;
+	completelySorted = TRUE;
+	for (rxQueue = 0;
+	     rxQueue < rxQueueCount - sortIterations;
+	     rxQueue++)
+	{
+	    /* compare adjacent elements */
+	    if ((rxQueues[rxQueue].trafficClass <
+		rxQueues[rxQueue+1].trafficClass)
+		|| ((rxQueues[rxQueue].trafficClass ==
+		     rxQueues[rxQueue+1].trafficClass)
+		    &&(rxQueues[rxQueue].npeId <
+		       rxQueues[rxQueue+1].npeId)))
+	    {
+		/* swap adjacent elements */
+		int npeCount = rxQueues[rxQueue].npeCount;
+		UINT32 npeId = rxQueues[rxQueue].npeId;
+		IxQMgrQId qId = rxQueues[rxQueue].qId;
+		IxEthDBProperty trafficClass = rxQueues[rxQueue].trafficClass;
+		rxQueues[rxQueue].npeCount = rxQueues[rxQueue+1].npeCount;
+		rxQueues[rxQueue].npeId = rxQueues[rxQueue+1].npeId;
+		rxQueues[rxQueue].qId = rxQueues[rxQueue+1].qId;
+		rxQueues[rxQueue].trafficClass = rxQueues[rxQueue+1].trafficClass;
+		rxQueues[rxQueue+1].npeCount = npeCount;
+		rxQueues[rxQueue+1].npeId = npeId;
+		rxQueues[rxQueue+1].qId = qId;
+		rxQueues[rxQueue+1].trafficClass = trafficClass;
+		completelySorted = FALSE;
+	    }
+	}
+    }
+    while (!completelySorted);
+
+    /* Queue traffic class list:
+     *
+     * Fill an array of rx queues linked by ascending traffic classes.
+     *
+     * If the queues are configured as follows
+     *   qId 6 -> traffic class 0 (lowest)
+     *   qId 7 -> traffic class 0
+     *   qId 8 -> traffic class 6
+     *   qId 12 -> traffic class 7 (highest)
+     *
+     * Then the output of this loop will be
+     *
+     * higherPriorityQueue[6] = 8
+     * higherPriorityQueue[7] = 8
+     * higherPriorityQueue[8] = 12
+     * higherPriorityQueue[12] = Invalid queueId
+     * higherPriorityQueue[...] = Invalid queueId
+     *
+     * Note that this queue ordering does not handle all possibilities
+     * that could result from different rules associated with different
+     * ports, and inconsistencies in the rules. In all cases, the
+     * output of this  algorithm is a simple linked list of queues,
+     * without closed circuit.
+
+     * This list is implemented as an array with invalid values initialized
+     * with an "invalid" queue id which is the maximum number of queues.
+     *
+     */
+
+    /*
+     * Initialise the rx queue list.
+     */
+    for (rxQueue = 0; rxQueue < IX_QMGR_MAX_NUM_QUEUES; rxQueue++)
+    {
+	ixEthAccDataInfo.higherPriorityQueue[rxQueue] = IX_QMGR_MAX_NUM_QUEUES;
+    }
+
+    /* build the linked list for this NPE.
+     */
+    for (ixNpeId = 0;
+	 ixNpeId <= ixHighestNpeId;
+	 ixNpeId++)
+    {
+	/* iterate thru the sorted list of queues
+	 */
+	ixQId = IX_QMGR_MAX_NUM_QUEUES;
+	for (rxQueue = 0;
+	     rxQueue < rxQueueCount;
+	     rxQueue++)
+	{
+	    if (rxQueues[rxQueue].npeId == ixNpeId)
+	    {
+		ixEthAccDataInfo.higherPriorityQueue[rxQueues[rxQueue].qId] = ixQId;
+		/* iterate thru queues with the same traffic class
+		 * than the current queue. (queues are ordered by descending
+		 * traffic classes and npeIds).
+		 */
+		while ((rxQueue < rxQueueCount - 1)
+		       && (rxQueues[rxQueue].trafficClass
+			   == rxQueues[rxQueue+1].trafficClass)
+		       && (ixNpeId == rxQueues[rxQueue].npeId))
+		{
+		    rxQueue++;
+		    ixEthAccDataInfo.higherPriorityQueue[rxQueues[rxQueue].qId] = ixQId;
+		}
+		ixQId = rxQueues[rxQueue].qId;
+	    }
+	}
+    }
+
+    /* point on the first dynamic queue description */
+    qInfoDes = ixEthAccQmgrRxQueuesInfo;
+
+    /* update the list of queues with the rx queues */
+    for (rxQueue = 0;
+	 (rxQueue < rxQueueCount) && (ret == IX_ETH_ACC_SUCCESS);
+	 rxQueue++)
+    {
+	/* Don't utilize more than IX_ETHACC_MAX_LARGE_RX_QUEUES queues
+	 * with the full 128 entries.  For the lower priority queues, use
+	 * a smaller number of entries.  This ensures queue resources
+	 * remain available for other components.
+	 */
+	if( (rxQueueCount > IX_ETHACC_MAX_LARGE_RX_QUEUES) &&
+	    (rxQueue < rxQueueCount - IX_ETHACC_MAX_LARGE_RX_QUEUES) )
+	{
+	    /* add the small RX Queue setup template to the list of queues */
+	    memcpy(qInfoDes, &ixEthAccQmgrRxSmallTemplate, sizeof(*qInfoDes));
+	} else {
+	    /* add the default RX Queue setup template to the list of queues */
+	    memcpy(qInfoDes, &ixEthAccQmgrRxDefaultTemplate, sizeof(*qInfoDes));
+	}
+
+	/* setup the RxQueue ID */
+	qInfoDes->qId = rxQueues[rxQueue].qId;
+
+	/* setup the RxQueue watermark level
+	 *
+	 * Each queue can be filled by many NPEs. To avoid the
+	 * NPEs to write to a full queue, need to set the
+	 * high watermark level for nearly full condition.
+	 * (the high watermark level are a power of 2
+	 * starting from the top of the queue)
+	 *
+	 * Number of     watermark
+         *   ports        level
+         *    1             0
+	 *    2             1
+	 *    3             2
+	 *    4             4
+	 *    5             4
+	 *    6             8
+	 *    n          approx. 2**ceil(log2(n))
+	 */
+	if (rxQueues[rxQueue].npeCount == 1)
+	{
+	    qInfoDes->AlmostFullThreshold = IX_QMGR_Q_WM_LEVEL0;
+	}
+	else if (rxQueues[rxQueue].npeCount == 2)
+	{
+	    qInfoDes->AlmostFullThreshold = IX_QMGR_Q_WM_LEVEL1;
+	}
+	else if (rxQueues[rxQueue].npeCount == 3)
+	{
+	    qInfoDes->AlmostFullThreshold = IX_QMGR_Q_WM_LEVEL2;
+	}
+	else
+	{
+	    /* reach the maximum number for CSR 2.0 */
+            IX_ETH_ACC_WARNING_LOG("ixEthAccQMgrQueuesConfig: maximum number of NPEs per queue reached, bailing out\n", 0, 0, 0, 0, 0, 0);
+	    ret = IX_ETH_ACC_FAIL;
+	    break;
+	}
+
+	/* move to next queue entry */
+	++qInfoDes;
+    }
+
+    /* configure the static list (RxFree, Tx and TxDone queues) */
+    for (qInfoDes = ixEthAccQmgrStaticInfo;
+	 (qInfoDes->qCallback != (IxQMgrCallback) NULL )
+	     && (ret == IX_ETH_ACC_SUCCESS);
+	 ++qInfoDes)
+    {
+	ret = ixEthAccQMgrQueueSetup(qInfoDes);
+    }
+
+    /* configure the dynamic list (Rx queues) */
+    for (qInfoDes = ixEthAccQmgrRxQueuesInfo;
+	 (qInfoDes->qCallback != (IxQMgrCallback) NULL )
+	     && (ret == IX_ETH_ACC_SUCCESS);
+	 ++qInfoDes)
+    {
+	ret = ixEthAccQMgrQueueSetup(qInfoDes);
+    }
+
+    return(ret);
+}
+
+/**
+ * @fn ixEthAccQMgrRxQEntryGet(UINT32 *rxQueueEntries)
+ *
+ * @brief Add and return the total number of entries in all Rx queues
+ *
+ * @param UINT32 rxQueueEntries[in] number of entries in all queues
+ *
+ * @return void
+ *
+ * @note Rx queues configuration is driven by Qos Setup. There is a
+ * variable number of rx queues which are set at initialisation.
+ *
+ * @internal
+ */
+IX_ETH_ACC_PUBLIC
+void ixEthAccQMgrRxQEntryGet(UINT32 *numRxQueueEntries)
+{
+    UINT32 rxQueueLevel;
+    IxEthAccQregInfo *qInfoDes;;
+
+    *numRxQueueEntries = 0;
+
+    /* iterate thru rx queues */
+    for (qInfoDes = ixEthAccQmgrRxQueuesInfo;
+	 qInfoDes->qCallback != (IxQMgrCallback)NULL;
+	 ++qInfoDes)
+    {
+	/* retrieve the rx queue level */
+	rxQueueLevel = 0;
+	ixQMgrQNumEntriesGet(qInfoDes->qId, &rxQueueLevel);
+	(*numRxQueueEntries) += rxQueueLevel;
+    }
+}
+
+/**
+ * @fn ixEthAccQMgrRxCallbacksRegister(IxQMgrCallback ixQMgrCallback)
+ *
+ * @brief Change the callback registered to all rx queues.
+ *
+ * @param IxQMgrCallback ixQMgrCallback[in] QMgr callback to register
+ *
+ * @return IxEthAccStatus
+ *
+ * @note The user may decide to use different Rx mechanisms
+ * (e.g. receive many frames at the same time , or receive
+ *  one frame at a time, depending on the overall application
+ *  performances). A different QMgr callback is registered. This
+ *  way, there is no excessive pointer checks in the datapath.
+ *
+ * @internal
+ */
+IX_ETH_ACC_PUBLIC
+IxEthAccStatus ixEthAccQMgrRxCallbacksRegister(IxQMgrCallback ixQMgrCallback)
+{
+    IxEthAccQregInfo *qInfoDes;
+    IxEthAccStatus ret = IX_ETH_ACC_SUCCESS;
+
+    /* parameter check */
+    if (NULL == ixQMgrCallback)
+    {
+	ret = IX_ETH_ACC_FAIL;
+    }
+
+    /* iterate thru rx queues */
+    for (qInfoDes = ixEthAccQmgrRxQueuesInfo;
+	 (qInfoDes->qCallback != (IxQMgrCallback) NULL )
+	     && (ret == IX_ETH_ACC_SUCCESS);
+	 ++qInfoDes)
+    {
+	/* register the rx callback for all queues */
+	if (ixQMgrNotificationCallbackSet(qInfoDes->qId,
+					     ixQMgrCallback,
+					     qInfoDes->callbackTag
+					     ) != IX_SUCCESS)
+	{
+	    ret = IX_ETH_ACC_FAIL;
+	}
+    }
+    return(ret);
+}
+
+/**
+ * @fn ixEthAccSingleEthNpeCheck(IxEthAccPortId portId)
+ *
+ * @brief Check the npe exists for this port
+ *
+ * @param IxEthAccPortId portId[in] port
+ *
+ * @return IxEthAccStatus
+ *
+ * @internal
+ */
+IX_ETH_ACC_PUBLIC
+IxEthAccStatus ixEthAccSingleEthNpeCheck(IxEthAccPortId portId)
+{
+
+    /* If not IXP42X A0 stepping, proceed to check for existence of coprocessors */
+    if ((IX_FEATURE_CTRL_SILICON_TYPE_A0 !=
+        (ixFeatureCtrlProductIdRead() & IX_FEATURE_CTRL_SILICON_STEPPING_MASK))
+        || (IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X != ixFeatureCtrlDeviceRead ()))
+      {
+            if ((IX_ETH_PORT_1 == portId) &&
+                (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_ETH0) ==
+                 IX_FEATURE_CTRL_COMPONENT_ENABLED))
+            {
+                return IX_ETH_ACC_SUCCESS;
+            }
+
+            if ((IX_ETH_PORT_2 == portId) &&
+                (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_ETH1) ==
+                 IX_FEATURE_CTRL_COMPONENT_ENABLED))
+            {
+                return IX_ETH_ACC_SUCCESS;
+            }
+
+            if ((IX_ETH_PORT_3 == portId) &&
+                (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEA_ETH) ==
+                 IX_FEATURE_CTRL_COMPONENT_ENABLED))
+            {
+                return IX_ETH_ACC_SUCCESS;
+            }
+
+            return IX_ETH_ACC_FAIL;
+      }
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+/**
+ * @fn ixEthAccStatsShow(void)
+ *
+ * @brief Displays all EthAcc stats
+ *
+ * @return void
+ *
+ */
+void ixEthAccStatsShow(IxEthAccPortId portId)
+{
+    ixEthAccMdioShow();
+
+    printf("\nPort %u\nUnicast MAC : ", portId);
+    ixEthAccPortUnicastAddressShow(portId);
+    ixEthAccPortMulticastAddressShow(portId);
+    printf("\n");
+
+    ixEthAccDataPlaneShow();
+}
+
+
+

cpu/ixp/npe/IxEthAccControlInterface.c

@@ -0,0 +1,533 @@
+/**
+ * @file IxEthAccControlInterface.c
+ *
+ * @author Intel Corporation
+ * @date 
+ *
+ * @brief IX_ETH_ACC_PUBLIC wrappers for control plane functions
+ *
+ * Design Notes:
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxOsal.h"
+#include "IxEthAcc.h"
+#include "IxEthAcc_p.h"
+
+PUBLIC IxOsalMutex ixEthAccControlInterfaceMutex;
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortEnable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+        printf("EthAcc: (Mac) cannot enable port %d, service not initialized\n", portId);
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortEnablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortDisable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    /* check the context is iinitialized */
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortDisablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortEnabledQuery(IxEthAccPortId portId, BOOL *enabled)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortEnabledQueryPriv(portId, enabled);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortPromiscuousModeClear(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortPromiscuousModeClearPriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortPromiscuousModeSet(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortPromiscuousModeSetPriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortUnicastMacAddressSet(IxEthAccPortId portId, IxEthAccMacAddr *macAddr)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortUnicastMacAddressSetPriv(portId, macAddr);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortUnicastMacAddressGet(IxEthAccPortId portId, IxEthAccMacAddr *macAddr)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortUnicastMacAddressGetPriv(portId, macAddr);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortMulticastAddressJoin(IxEthAccPortId portId, IxEthAccMacAddr *macAddr)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortMulticastAddressJoinPriv(portId, macAddr);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortMulticastAddressJoinAll(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortMulticastAddressJoinAllPriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortMulticastAddressLeave(IxEthAccPortId portId, IxEthAccMacAddr *macAddr)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortMulticastAddressLeavePriv(portId, macAddr);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortMulticastAddressLeaveAll(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortMulticastAddressLeaveAllPriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortUnicastAddressShow(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+ 
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortUnicastAddressShowPriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC void 
+ixEthAccPortMulticastAddressShow(IxEthAccPortId portId)
+{
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return;
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    ixEthAccPortMulticastAddressShowPriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortDuplexModeSet(IxEthAccPortId portId, IxEthAccDuplexMode mode)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortDuplexModeSetPriv(portId, mode);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortDuplexModeGet(IxEthAccPortId portId, IxEthAccDuplexMode *mode)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortDuplexModeGetPriv(portId, mode);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortTxFrameAppendPaddingEnable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+     
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortTxFrameAppendPaddingEnablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortTxFrameAppendPaddingDisable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortTxFrameAppendPaddingDisablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortTxFrameAppendFCSEnable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortTxFrameAppendFCSEnablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortTxFrameAppendFCSDisable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortTxFrameAppendFCSDisablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortRxFrameAppendFCSEnable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortRxFrameAppendFCSEnablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccPortRxFrameAppendFCSDisable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortRxFrameAppendFCSDisablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccTxSchedulingDisciplineSet(IxEthAccPortId portId, IxEthAccSchedulerDiscipline sched)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccTxSchedulingDisciplineSetPriv(portId, sched);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus
+ixEthAccRxSchedulingDisciplineSet(IxEthAccSchedulerDiscipline sched)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccRxSchedulingDisciplineSetPriv(sched);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortNpeLoopbackEnable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccNpeLoopbackEnablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortTxEnable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortTxEnablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortRxEnable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortRxEnablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortNpeLoopbackDisable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccNpeLoopbackDisablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortTxDisable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortTxDisablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortRxDisable(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortRxDisablePriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus 
+ixEthAccPortMacReset(IxEthAccPortId portId)
+{
+    IxEthAccStatus result;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&ixEthAccControlInterfaceMutex, IX_OSAL_WAIT_FOREVER);
+    result = ixEthAccPortMacResetPriv(portId);
+    ixOsalMutexUnlock(&ixEthAccControlInterfaceMutex);
+    return result;
+}

cpu/ixp/npe/IxEthAccDataPlane.c

@@ -0,0 +1,2483 @@
+/**
+ * @file IxEthDataPlane.c
+ *
+ * @author Intel Corporation
+ * @date 12-Feb-2002
+ *
+ * @brief This file contains the implementation of the IXPxxx
+ * Ethernet Access Data plane component
+ *
+ * Design Notes:
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxNpeMh.h"
+#include "IxEthAcc.h"
+#include "IxEthDB.h"
+#include "IxOsal.h"
+#include "IxEthDBPortDefs.h"
+#include "IxFeatureCtrl.h"
+#include "IxEthAcc_p.h"
+#include "IxEthAccQueueAssign_p.h"
+
+extern PUBLIC IxEthAccMacState ixEthAccMacState[];
+extern PUBLIC UINT32 ixEthAccNewSrcMask;
+
+/**
+ * private functions prototype
+ */
+PRIVATE IX_OSAL_MBUF *
+ixEthAccEntryFromQConvert(UINT32 qEntry, UINT32 mask);
+
+PRIVATE UINT32
+ixEthAccMbufRxQPrepare(IX_OSAL_MBUF *mbuf);
+
+PRIVATE UINT32
+ixEthAccMbufTxQPrepare(IX_OSAL_MBUF *mbuf);
+
+PRIVATE IxEthAccStatus
+ixEthAccTxSwQHighestPriorityGet(IxEthAccPortId portId,
+				IxEthAccTxPriority *priorityPtr);
+
+PRIVATE IxEthAccStatus
+ixEthAccTxFromSwQ(IxEthAccPortId portId,
+		  IxEthAccTxPriority priority);
+
+PRIVATE IxEthAccStatus
+ixEthAccRxFreeFromSwQ(IxEthAccPortId portId);
+
+PRIVATE void
+ixEthAccMbufFromTxQ(IX_OSAL_MBUF *mbuf);
+
+PRIVATE void
+ixEthAccMbufFromRxQ(IX_OSAL_MBUF *mbuf);
+
+PRIVATE IX_STATUS
+ixEthAccQmgrLockTxWrite(IxEthAccPortId portId,
+			UINT32 qBuffer);
+
+PRIVATE IX_STATUS
+ixEthAccQmgrLockRxWrite(IxEthAccPortId portId,
+			UINT32 qBuffer);
+
+PRIVATE IX_STATUS
+ixEthAccQmgrTxWrite(IxEthAccPortId portId,
+		    UINT32 qBuffer,
+		    UINT32 priority);
+
+/**
+ * @addtogroup IxEthAccPri
+ *@{
+ */
+
+/* increment a counter only when stats are enabled */
+#define TX_STATS_INC(port,field) \
+        IX_ETH_ACC_STATS_INC(ixEthAccPortData[port].ixEthAccTxData.stats.field)
+#define RX_STATS_INC(port,field) \
+        IX_ETH_ACC_STATS_INC(ixEthAccPortData[port].ixEthAccRxData.stats.field)
+
+/* always increment the counter (mainly used for unexpected errors) */
+#define TX_INC(port,field) \
+        ixEthAccPortData[port].ixEthAccTxData.stats.field++
+#define RX_INC(port,field) \
+        ixEthAccPortData[port].ixEthAccRxData.stats.field++
+
+PRIVATE IxEthAccDataPlaneStats     ixEthAccDataStats;
+
+extern IxEthAccPortDataInfo   ixEthAccPortData[];
+extern IxEthAccInfo   ixEthAccDataInfo;
+
+PRIVATE IxOsalFastMutex txWriteMutex[IX_ETH_ACC_NUMBER_OF_PORTS];
+PRIVATE IxOsalFastMutex rxWriteMutex[IX_ETH_ACC_NUMBER_OF_PORTS];
+
+/**
+ *
+ * @brief Mbuf header conversion macros : they implement the
+ *  different conversions using a temporary value. They also double-check
+ *  that the parameters can be converted to/from NPE format.
+ *
+ */
+#if defined(__wince) && !defined(IN_KERNEL)
+#define PTR_VIRT2NPE(ptrSrc,dst) \
+  do { UINT32 temp; \
+      IX_OSAL_ENSURE(sizeof(ptrSrc) == sizeof(UINT32), "Wrong parameter type"); \
+      IX_OSAL_ENSURE(sizeof(dst) == sizeof(UINT32), "Wrong parameter type"); \
+      temp = (UINT32)IX_OSAL_MBUF_MBUF_VIRTUAL_TO_PHYSICAL_TRANSLATION((IX_OSAL_MBUF*)ptrSrc); \
+      (dst) = IX_OSAL_SWAP_BE_SHARED_LONG(temp); } \
+  while(0)
+
+#define PTR_NPE2VIRT(type,src,ptrDst) \
+  do { void *temp; \
+      IX_OSAL_ENSURE(sizeof(type) == sizeof(UINT32), "Wrong parameter type"); \
+      IX_OSAL_ENSURE(sizeof(src) == sizeof(UINT32), "Wrong parameter type"); \
+      IX_OSAL_ENSURE(sizeof(ptrDst) == sizeof(UINT32), "Wrong parameter type"); \
+      temp = (void *)IX_OSAL_SWAP_BE_SHARED_LONG(src); \
+      (ptrDst) = (type)IX_OSAL_MBUF_MBUF_PHYSICAL_TO_VIRTUAL_TRANSLATION(temp); } \
+  while(0)
+#else
+#define PTR_VIRT2NPE(ptrSrc,dst) \
+  do { UINT32 temp; \
+      IX_OSAL_ENSURE(sizeof(ptrSrc) == sizeof(UINT32), "Wrong parameter type"); \
+      IX_OSAL_ENSURE(sizeof(dst) == sizeof(UINT32), "Wrong parameter type"); \
+      temp = (UINT32)IX_OSAL_MMU_VIRT_TO_PHYS(ptrSrc); \
+      (dst) = IX_OSAL_SWAP_BE_SHARED_LONG(temp); } \
+  while(0)
+
+#define PTR_NPE2VIRT(type,src,ptrDst) \
+  do { void *temp; \
+      IX_OSAL_ENSURE(sizeof(type) == sizeof(UINT32), "Wrong parameter type"); \
+      IX_OSAL_ENSURE(sizeof(src) == sizeof(UINT32), "Wrong parameter type"); \
+      IX_OSAL_ENSURE(sizeof(ptrDst) == sizeof(UINT32), "Wrong parameter type"); \
+      temp = (void *)IX_OSAL_SWAP_BE_SHARED_LONG(src); \
+      (ptrDst) = (type)IX_OSAL_MMU_PHYS_TO_VIRT(temp); } \
+  while(0)
+#endif
+
+/**
+ *
+ * @brief Mbuf payload pointer conversion macros : Wince has its own
+ *  method to convert the buffer pointers
+ */
+#if defined(__wince) && !defined(IN_KERNEL)
+#define DATAPTR_VIRT2NPE(ptrSrc,dst) \
+  do { UINT32 temp; \
+      temp = (UINT32)IX_OSAL_MBUF_DATA_VIRTUAL_TO_PHYSICAL_TRANSLATION(ptrSrc); \
+      (dst) = IX_OSAL_SWAP_BE_SHARED_LONG(temp); } \
+  while(0)
+
+#else
+#define DATAPTR_VIRT2NPE(ptrSrc,dst) PTR_VIRT2NPE(IX_OSAL_MBUF_MDATA(ptrSrc),dst)
+#endif
+
+
+/* Flush the shared part of the mbuf header */
+#define IX_ETHACC_NE_CACHE_FLUSH(mbufPtr) \
+  do { \
+      IX_OSAL_CACHE_FLUSH(IX_ETHACC_NE_SHARED(mbufPtr), \
+			      sizeof(IxEthAccNe)); \
+    } \
+  while(0)
+
+/* Invalidate the shared part of the mbuf header */
+#define IX_ETHACC_NE_CACHE_INVALIDATE(mbufPtr) \
+  do { \
+      IX_OSAL_CACHE_INVALIDATE(IX_ETHACC_NE_SHARED(mbufPtr), \
+				   sizeof(IxEthAccNe)); \
+    } \
+  while(0)
+
+/* Preload one cache line (shared mbuf headers are aligned
+ * and their size is 1 cache line)
+ *
+ * IX_OSAL_CACHED  is defined when the mbuf headers are
+ * allocated from cached memory.
+ *
+ * Other processor on emulation environment may not implement
+ * preload function
+ */
+#ifdef IX_OSAL_CACHED
+	#if (CPU!=SIMSPARCSOLARIS) && !defined (__wince)
+		#define IX_ACC_DATA_CACHE_PRELOAD(ptr) \
+		do { /* preload a cache line (Xscale Processor) */ \
+			__asm__ (" pld [%0]\n": : "r" (ptr)); \
+		} \
+		while(0)
+	#else
+		/* preload not implemented on different processor */
+		#define IX_ACC_DATA_CACHE_PRELOAD(mbufPtr) \
+		do { /* nothing */ } while (0)
+	#endif
+#else
+	/* preload not needed if cache is not enabled */
+	#define IX_ACC_DATA_CACHE_PRELOAD(mbufPtr) \
+	do { /* nothing */ } while (0)
+#endif
+
+/**
+ *
+ * @brief function to retrieve the correct pointer from
+ * a queue entry posted by the NPE
+ *
+ * @param qEntry : entry from qmgr queue
+ *        mask : applicable mask for this queue
+ *        (4 most significant bits are used for additional informations)
+ *
+ * @return IX_OSAL_MBUF * pointer to mbuf header
+ *
+ * @internal
+ */
+PRIVATE IX_OSAL_MBUF *
+ixEthAccEntryFromQConvert(UINT32 qEntry, UINT32 mask)
+{
+    IX_OSAL_MBUF *mbufPtr;
+
+    if (qEntry != 0)
+    {
+        /* mask NPE bits (e.g. priority, port ...) */
+        qEntry &= mask;
+
+#if IX_ACC_DRAM_PHYS_OFFSET != 0
+        /* restore the original address pointer (if PHYS_OFFSET is not 0) */
+        qEntry |= (IX_ACC_DRAM_PHYS_OFFSET & ~IX_ETHNPE_QM_Q_RXENET_ADDR_MASK);
+#endif
+        /* get the mbuf pointer address from the npe-shared address */
+        qEntry -= offsetof(IX_OSAL_MBUF,ix_ne);
+
+        /* phys2virt mbuf */
+        mbufPtr = (IX_OSAL_MBUF *)IX_OSAL_MMU_PHYS_TO_VIRT(qEntry);
+
+        /* preload the cacheline shared with NPE */
+        IX_ACC_DATA_CACHE_PRELOAD(IX_ETHACC_NE_SHARED(mbufPtr));
+
+        /* preload the cacheline used by xscale */
+        IX_ACC_DATA_CACHE_PRELOAD(mbufPtr);
+    }
+    else
+    {
+	mbufPtr = NULL;
+    }
+
+    return mbufPtr;
+}
+
+/* Convert the mbuf header for NPE transmission */
+PRIVATE UINT32
+ixEthAccMbufTxQPrepare(IX_OSAL_MBUF *mbuf)
+{
+    UINT32 qbuf;
+    UINT32 len;
+
+    /* endianess swap for tci and flags
+       note: this is done only once, even for chained buffers */
+    IX_ETHACC_NE_FLAGS(mbuf)   = IX_OSAL_SWAP_BE_SHARED_SHORT(IX_ETHACC_NE_FLAGS(mbuf));
+    IX_ETHACC_NE_VLANTCI(mbuf) = IX_OSAL_SWAP_BE_SHARED_SHORT(IX_ETHACC_NE_VLANTCI(mbuf));
+
+    /* test for unchained mbufs */
+    if (IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(mbuf) == NULL)
+    {
+	/* "best case" scenario : unchained mbufs */
+	IX_ETH_ACC_STATS_INC(ixEthAccDataStats.unchainedTxMBufs);
+
+	/* payload pointer conversion */
+	DATAPTR_VIRT2NPE(mbuf, IX_ETHACC_NE_DATA(mbuf));
+
+	/* unchained mbufs : the frame length is the mbuf length
+	 * and the 2 identical lengths are stored in the same
+	 * word.
+	 */
+	len = IX_OSAL_MBUF_MLEN(mbuf);
+
+	/* set the length in both length and pktLen 16-bits fields */
+	len |= (len << IX_ETHNPE_ACC_LENGTH_OFFSET);
+	IX_ETHACC_NE_LEN(mbuf) = IX_OSAL_SWAP_BE_SHARED_LONG(len);
+
+	/* unchained mbufs : next contains 0 */
+	IX_ETHACC_NE_NEXT(mbuf) = 0;
+
+	/* flush shared header after all address conversions */
+	IX_ETHACC_NE_CACHE_FLUSH(mbuf);
+    }
+    else
+    {
+	/* chained mbufs */
+	IX_OSAL_MBUF *ptr = mbuf;
+	IX_OSAL_MBUF *nextPtr;
+	UINT32 frmLen;
+
+	/* get the frame length from the header of the first buffer */
+	frmLen = IX_OSAL_MBUF_PKT_LEN(mbuf);
+
+	do
+	{
+	    IX_ETH_ACC_STATS_INC(ixEthAccDataStats.chainedTxMBufs);
+
+	    /* payload pointer */
+	    DATAPTR_VIRT2NPE(ptr,IX_ETHACC_NE_DATA(ptr));
+	    /* Buffer length and frame length are stored in the same word */
+	    len = IX_OSAL_MBUF_MLEN(ptr);
+	    len = frmLen | (len << IX_ETHNPE_ACC_LENGTH_OFFSET);
+	    IX_ETHACC_NE_LEN(ptr) = IX_OSAL_SWAP_BE_SHARED_LONG(len);
+
+	    /* get the virtual next chain pointer */
+	    nextPtr = IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(ptr);
+	    if (nextPtr != NULL)
+	    {
+		/* shared pointer of the next buffer is chained */
+		PTR_VIRT2NPE(IX_ETHACC_NE_SHARED(nextPtr),
+			     IX_ETHACC_NE_NEXT(ptr));
+	    }
+	    else
+	    {
+		IX_ETHACC_NE_NEXT(ptr) = 0;
+	    }
+
+	    /* flush shared header after all address conversions */
+	    IX_ETHACC_NE_CACHE_FLUSH(ptr);
+
+	    /* move to next buffer */
+	    ptr = nextPtr;
+
+	    /* the frame length field is set only in the first buffer
+	     * and is zeroed in the next buffers
+	     */
+	    frmLen = 0;
+	}
+	while(ptr != NULL);
+
+    }
+
+    /* virt2phys mbuf itself */
+    qbuf = (UINT32)IX_OSAL_MMU_VIRT_TO_PHYS(
+		  IX_ETHACC_NE_SHARED(mbuf));
+
+    /* Ensure the bits which are reserved to exchange information with
+     * the NPE are cleared
+     *
+     * If the mbuf address is not correctly aligned, or from an
+     * incompatible memory range, there is no point to continue
+     */
+    IX_OSAL_ENSURE(((qbuf & ~IX_ETHNPE_QM_Q_TXENET_ADDR_MASK) == 0),
+	      "Invalid address range");
+
+    return qbuf;
+}
+
+/* Convert the mbuf header for NPE reception */
+PRIVATE UINT32
+ixEthAccMbufRxQPrepare(IX_OSAL_MBUF *mbuf)
+{
+    UINT32 len;
+    UINT32 qbuf;
+
+    if (IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(mbuf) == NULL)
+    {
+	/* "best case" scenario : unchained mbufs */
+	IX_ETH_ACC_STATS_INC(ixEthAccDataStats.unchainedRxFreeMBufs);
+
+	/* unchained mbufs : payload pointer */
+	DATAPTR_VIRT2NPE(mbuf, IX_ETHACC_NE_DATA(mbuf));
+
+	/* unchained mbufs : set the buffer length
+	* and the frame length field is zeroed
+	*/
+	len = (IX_OSAL_MBUF_MLEN(mbuf) << IX_ETHNPE_ACC_LENGTH_OFFSET);
+	IX_ETHACC_NE_LEN(mbuf) = IX_OSAL_SWAP_BE_SHARED_LONG(len);
+
+	/* unchained mbufs : next pointer is null */
+	IX_ETHACC_NE_NEXT(mbuf) = 0;
+
+	/* flush shared header after all address conversions */
+	IX_ETHACC_NE_CACHE_FLUSH(mbuf);
+
+	/* remove shared header cache line */
+	IX_ETHACC_NE_CACHE_INVALIDATE(mbuf);
+    }
+    else
+    {
+	/* chained mbufs */
+	IX_OSAL_MBUF *ptr = mbuf;
+	IX_OSAL_MBUF *nextPtr;
+
+	do
+	{
+	    /* chained mbufs */
+	    IX_ETH_ACC_STATS_INC(ixEthAccDataStats.chainedRxFreeMBufs);
+
+	    /* we must save virtual next chain pointer */
+	    nextPtr = IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(ptr);
+
+	    if (nextPtr != NULL)
+	    {
+		/* chaining pointer for NPE */
+		PTR_VIRT2NPE(IX_ETHACC_NE_SHARED(nextPtr),
+			     IX_ETHACC_NE_NEXT(ptr));
+	    }
+	    else
+	    {
+		IX_ETHACC_NE_NEXT(ptr) = 0;
+	    }
+
+	    /* payload pointer */
+	    DATAPTR_VIRT2NPE(ptr,IX_ETHACC_NE_DATA(ptr));
+
+	    /* buffer length */
+	    len = (IX_OSAL_MBUF_MLEN(ptr) << IX_ETHNPE_ACC_LENGTH_OFFSET);
+	    IX_ETHACC_NE_LEN(ptr) = IX_OSAL_SWAP_BE_SHARED_LONG(len);
+
+	    /* flush shared header after all address conversions */
+	    IX_ETHACC_NE_CACHE_FLUSH(ptr);
+
+	    /* remove shared header cache line */
+	    IX_ETHACC_NE_CACHE_INVALIDATE(ptr);
+
+	    /* next mbuf in the chain */
+	    ptr = nextPtr;
+	}
+	while(ptr != NULL);
+    }
+
+    /* virt2phys mbuf itself */
+    qbuf = (UINT32)IX_OSAL_MMU_VIRT_TO_PHYS(
+		  IX_ETHACC_NE_SHARED(mbuf));
+
+    /* Ensure the bits which are reserved to exchange information with
+     * the NPE are cleared
+     *
+     * If the mbuf address is not correctly aligned, or from an
+     * incompatible memory range, there is no point to continue
+     */
+    IX_OSAL_ENSURE(((qbuf & ~IX_ETHNPE_QM_Q_RXENET_ADDR_MASK) == 0),
+	      "Invalid address range");
+
+    return qbuf;
+}
+
+/* Convert the mbuf header after NPE transmission
+ * Since there is nothing changed by the NPE, there is no need
+ * to process anything but the update of internal stats
+ * when they are enabled
+*/
+PRIVATE void
+ixEthAccMbufFromTxQ(IX_OSAL_MBUF *mbuf)
+{
+#ifndef NDEBUG
+    /* test for unchained mbufs */
+    if (IX_ETHACC_NE_NEXT(mbuf) == 0)
+    {
+	/* unchained mbufs : update the stats */
+	IX_ETH_ACC_STATS_INC(ixEthAccDataStats.unchainedTxDoneMBufs);
+    }
+    else
+    {
+	/* chained mbufs : walk the chain and update the stats */
+	IX_OSAL_MBUF *ptr = mbuf;
+
+	do
+	{
+	    IX_ETH_ACC_STATS_INC(ixEthAccDataStats.chainedTxDoneMBufs);
+	    ptr = IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(ptr);
+	}
+	while (ptr != NULL);
+    }
+#endif
+}
+
+/* Convert the mbuf header after NPE reception */
+PRIVATE void
+ixEthAccMbufFromRxQ(IX_OSAL_MBUF *mbuf)
+{
+    UINT32 len;
+
+    /* endianess swap for tci and flags
+       note: this is done only once, even for chained buffers */
+    IX_ETHACC_NE_FLAGS(mbuf)   = IX_OSAL_SWAP_BE_SHARED_SHORT(IX_ETHACC_NE_FLAGS(mbuf));
+    IX_ETHACC_NE_VLANTCI(mbuf) = IX_OSAL_SWAP_BE_SHARED_SHORT(IX_ETHACC_NE_VLANTCI(mbuf));
+
+    /* test for unchained mbufs */
+    if (IX_ETHACC_NE_NEXT(mbuf) == 0)
+    {
+	/* unchained mbufs */
+	IX_ETH_ACC_STATS_INC(ixEthAccDataStats.unchainedRxMBufs);
+
+	/* get the frame length. it is the same than the buffer length */
+	len = IX_OSAL_SWAP_BE_SHARED_LONG(IX_ETHACC_NE_LEN(mbuf));
+	len &= IX_ETHNPE_ACC_PKTLENGTH_MASK;
+	IX_OSAL_MBUF_PKT_LEN(mbuf) = IX_OSAL_MBUF_MLEN(mbuf) = len;
+
+        /* clears the next packet field */
+	IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(mbuf) = NULL;
+    }
+    else
+    {
+	IX_OSAL_MBUF *ptr = mbuf;
+	IX_OSAL_MBUF *nextPtr;
+	UINT32 frmLen;
+
+	/* convert the frame length */
+	frmLen = IX_OSAL_SWAP_BE_SHARED_LONG(IX_ETHACC_NE_LEN(mbuf));
+	IX_OSAL_MBUF_PKT_LEN(mbuf) = (frmLen & IX_ETHNPE_ACC_PKTLENGTH_MASK);
+
+        /* chained mbufs */
+	do
+	{
+	    IX_ETH_ACC_STATS_INC(ixEthAccDataStats.chainedRxMBufs);
+
+	    /* convert the length */
+	    len = IX_OSAL_SWAP_BE_SHARED_LONG(IX_ETHACC_NE_LEN(ptr));
+	    IX_OSAL_MBUF_MLEN(ptr) = (len >> IX_ETHNPE_ACC_LENGTH_OFFSET);
+
+            /* get the next pointer */
+ 	    PTR_NPE2VIRT(IX_OSAL_MBUF *,IX_ETHACC_NE_NEXT(ptr), nextPtr);
+	    if (nextPtr != NULL)
+	    {
+		nextPtr = (IX_OSAL_MBUF *)((UINT8 *)nextPtr - offsetof(IX_OSAL_MBUF,ix_ne));
+	    }
+	    /* set the next pointer */
+	    IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(ptr) = nextPtr;
+
+	    /* move to the next buffer */
+	    ptr = nextPtr;
+	}
+	while (ptr != NULL);
+    }
+}
+
+/* write to qmgr if possible and report an overflow if not possible
+ * Use a fast lock to protect the queue write.
+ * This way, the tx feature is reentrant.
+ */
+PRIVATE IX_STATUS
+ixEthAccQmgrLockTxWrite(IxEthAccPortId portId, UINT32 qBuffer)
+{
+    IX_STATUS qStatus;
+    if (ixOsalFastMutexTryLock(&txWriteMutex[portId]) == IX_SUCCESS)
+    {
+	qStatus = ixQMgrQWrite(
+	       IX_ETH_ACC_PORT_TO_TX_Q_ID(portId),
+	       &qBuffer);
+#ifndef NDEBUG
+	if (qStatus != IX_SUCCESS)
+	{
+	    TX_STATS_INC(portId, txOverflow);
+	}
+#endif
+	ixOsalFastMutexUnlock(&txWriteMutex[portId]);
+    }
+    else
+    {
+	TX_STATS_INC(portId, txLock);
+	qStatus = IX_QMGR_Q_OVERFLOW;
+    }
+    return qStatus;
+}
+
+/* write to qmgr if possible and report an overflow if not possible
+ * Use a fast lock to protect the queue write.
+ * This way, the Rx feature is reentrant.
+ */
+PRIVATE IX_STATUS
+ixEthAccQmgrLockRxWrite(IxEthAccPortId portId, UINT32 qBuffer)
+{
+    IX_STATUS qStatus;
+    if (ixOsalFastMutexTryLock(&rxWriteMutex[portId]) == IX_SUCCESS)
+    {
+	qStatus = ixQMgrQWrite(
+	       IX_ETH_ACC_PORT_TO_RX_FREE_Q_ID(portId),
+	       &qBuffer);
+#ifndef NDEBUG
+	if (qStatus != IX_SUCCESS)
+	{
+	    RX_STATS_INC(portId, rxFreeOverflow);
+	}
+#endif
+	ixOsalFastMutexUnlock(&rxWriteMutex[portId]);
+    }
+    else
+    {
+	RX_STATS_INC(portId, rxFreeLock);
+	qStatus = IX_QMGR_Q_OVERFLOW;
+    }
+    return qStatus;
+}
+
+/*
+ * Set the priority and write to a qmgr queue.
+ */
+PRIVATE IX_STATUS
+ixEthAccQmgrTxWrite(IxEthAccPortId portId, UINT32 qBuffer, UINT32 priority)
+{
+    /* fill the priority field */
+    qBuffer |= (priority << IX_ETHNPE_QM_Q_FIELD_PRIOR_R);
+
+    return ixEthAccQmgrLockTxWrite(portId, qBuffer);
+}
+
+/**
+ *
+ * @brief This function will discover the highest priority S/W Tx Q that
+ *        has entries in it
+ *
+ * @param portId - (in) the id of the port whose S/W Tx queues are to be searched
+ *        priorityPtr - (out) the priority of the highest priority occupied q will be written
+ *                      here
+ *
+ * @return IX_ETH_ACC_SUCCESS if an occupied Q is found
+ *         IX_ETH_ACC_FAIL if no Q has entries
+ *
+ * @internal
+ */
+PRIVATE IxEthAccStatus
+ixEthAccTxSwQHighestPriorityGet(IxEthAccPortId portId,
+				IxEthAccTxPriority *priorityPtr)
+{
+    if (ixEthAccPortData[portId].ixEthAccTxData.schDiscipline
+	== FIFO_NO_PRIORITY)
+    {
+	if(IX_ETH_ACC_DATAPLANE_IS_Q_EMPTY(ixEthAccPortData[portId].
+	       ixEthAccTxData.txQ[IX_ETH_ACC_TX_DEFAULT_PRIORITY]))
+	{
+	    return IX_ETH_ACC_FAIL;
+	}
+	else
+	{
+	    *priorityPtr = IX_ETH_ACC_TX_DEFAULT_PRIORITY;
+	    TX_STATS_INC(portId,txPriority[*priorityPtr]);
+	    return IX_ETH_ACC_SUCCESS;
+	}
+    }
+    else
+    {
+	IxEthAccTxPriority highestPriority = IX_ETH_ACC_TX_PRIORITY_7;
+	while(1)
+	{
+	    if(!IX_ETH_ACC_DATAPLANE_IS_Q_EMPTY(ixEthAccPortData[portId].
+	       ixEthAccTxData.txQ[highestPriority]))
+	    {
+
+		*priorityPtr = highestPriority;
+		TX_STATS_INC(portId,txPriority[highestPriority]);
+		return IX_ETH_ACC_SUCCESS;
+
+	    }
+	    if (highestPriority == IX_ETH_ACC_TX_PRIORITY_0)
+	    {
+		return IX_ETH_ACC_FAIL;
+	    }
+	    highestPriority--;
+	}
+    }
+}
+
+/**
+ *
+ * @brief This function will take a buffer from a TX S/W Q and attempt
+ *        to add it to the relevant TX H/W Q
+ *
+ * @param portId - the port whose TX queue is to be written to
+ *        priority - identifies the queue from which the entry is to be read
+ *
+ * @internal
+ */
+PRIVATE IxEthAccStatus
+ixEthAccTxFromSwQ(IxEthAccPortId portId,
+		  IxEthAccTxPriority priority)
+{
+    IX_OSAL_MBUF        *mbuf;
+    IX_STATUS	   qStatus;
+
+    IX_OSAL_ENSURE((UINT32)priority <= (UINT32)7, "Invalid priority");
+
+    IX_ETH_ACC_DATAPLANE_REMOVE_MBUF_FROM_Q_HEAD(
+	ixEthAccPortData[portId].ixEthAccTxData.txQ[priority],
+	mbuf);
+
+    if (mbuf != NULL)
+    {
+	/*
+	 * Add the Tx buffer to the H/W Tx Q
+	 * We do not need to flush here as it is already done
+	 * in TxFrameSubmit().
+	 */
+	qStatus = ixEthAccQmgrTxWrite(
+	      portId,
+	      IX_OSAL_MMU_VIRT_TO_PHYS((UINT32)IX_ETHACC_NE_SHARED(mbuf)),
+	      priority);
+
+	if (qStatus == IX_SUCCESS)
+	{
+	    TX_STATS_INC(portId,txFromSwQOK);
+	    return IX_SUCCESS;
+	}
+	else if (qStatus == IX_QMGR_Q_OVERFLOW)
+	{
+	    /*
+	     * H/W Q overflow, need to save the buffer
+	     * back on the s/w Q.
+	     * we must put it back on the head of the q to avoid
+	     * reordering packet tx
+	     */
+	    TX_STATS_INC(portId,txFromSwQDelayed);
+	    IX_ETH_ACC_DATAPLANE_ADD_MBUF_TO_Q_HEAD(
+		ixEthAccPortData[portId].ixEthAccTxData.txQ[priority],
+		mbuf);
+
+	    /*enable Q notification*/
+	    qStatus = ixQMgrNotificationEnable(
+		IX_ETH_ACC_PORT_TO_TX_Q_ID(portId),
+		IX_ETH_ACC_PORT_TO_TX_Q_SOURCE(portId));
+
+            if (qStatus != IX_SUCCESS && qStatus != IX_QMGR_WARNING)
+            {
+		TX_INC(portId,txUnexpectedError);
+		IX_ETH_ACC_FATAL_LOG(
+	            "ixEthAccTxFromSwQ:Unexpected Error: %u\n",
+	            qStatus, 0, 0, 0, 0, 0);
+            }
+	}
+	else
+	{
+	    TX_INC(portId,txUnexpectedError);
+
+	    /* recovery attempt */
+	    IX_ETH_ACC_DATAPLANE_ADD_MBUF_TO_Q_HEAD(
+		ixEthAccPortData[portId].ixEthAccTxData.txQ[priority],
+		mbuf);
+
+	    IX_ETH_ACC_FATAL_LOG(
+		"ixEthAccTxFromSwQ:Error: unexpected QM status 0x%08X\n",
+		qStatus, 0, 0, 0, 0, 0);
+	}
+    }
+    else
+    {
+	/* sw queue is empty */
+    }
+    return IX_ETH_ACC_FAIL;
+}
+
+/**
+ *
+ * @brief This function will take a buffer from a RXfree S/W Q and attempt
+ *        to add it to the relevant RxFree H/W Q
+ *
+ * @param portId - the port whose RXFree queue is to be written to
+ *
+ * @internal
+ */
+PRIVATE IxEthAccStatus
+ixEthAccRxFreeFromSwQ(IxEthAccPortId portId)
+{
+    IX_OSAL_MBUF        *mbuf;
+    IX_STATUS	   qStatus = IX_SUCCESS;
+
+    IX_ETH_ACC_DATAPLANE_REMOVE_MBUF_FROM_Q_HEAD(
+	  ixEthAccPortData[portId].ixEthAccRxData.freeBufferList,
+	  mbuf);
+    if (mbuf != NULL)
+    {
+	/*
+	 * Add The Rx Buffer to the H/W Free buffer Q if possible
+	 */
+	qStatus = ixEthAccQmgrLockRxWrite(portId,
+		  IX_OSAL_MMU_VIRT_TO_PHYS(
+			 (UINT32)IX_ETHACC_NE_SHARED(mbuf)));
+
+	if (qStatus == IX_SUCCESS)
+	{
+	    RX_STATS_INC(portId,rxFreeRepFromSwQOK);
+	    /*
+	     * Buffer added to h/w Q.
+	     */
+	    return IX_SUCCESS;
+	}
+	else if (qStatus == IX_QMGR_Q_OVERFLOW)
+	{
+	    /*
+	     * H/W Q overflow, need to save the buffer back on the s/w Q.
+	     */
+	    RX_STATS_INC(portId,rxFreeRepFromSwQDelayed);
+
+	    IX_ETH_ACC_DATAPLANE_ADD_MBUF_TO_Q_HEAD(
+		   ixEthAccPortData[portId].ixEthAccRxData.freeBufferList,
+		   mbuf);
+	}
+	else
+	{
+	    /* unexpected qmgr error */
+	    RX_INC(portId,rxUnexpectedError);
+
+	    IX_ETH_ACC_DATAPLANE_ADD_MBUF_TO_Q_HEAD(
+		    ixEthAccPortData[portId].ixEthAccRxData.freeBufferList,
+		    mbuf);
+
+	    IX_ETH_ACC_FATAL_LOG("IxEthAccRxFreeFromSwQ:Error: unexpected QM status 0x%08X\n",
+				 qStatus, 0, 0, 0, 0, 0);
+	}
+    }
+    else
+    {
+	/* sw queue is empty */
+    }
+    return IX_ETH_ACC_FAIL;
+}
+
+
+IX_ETH_ACC_PUBLIC
+IxEthAccStatus ixEthAccInitDataPlane()
+{
+    UINT32 portId;
+
+    /*
+     * Initialize the service and register callback to other services.
+     */
+
+    IX_ETH_ACC_MEMSET(&ixEthAccDataStats,
+		      0,
+		      sizeof(ixEthAccDataStats));
+
+    for(portId=0; portId < IX_ETH_ACC_NUMBER_OF_PORTS; portId++)
+    {
+	ixOsalFastMutexInit(&txWriteMutex[portId]);
+	ixOsalFastMutexInit(&rxWriteMutex[portId]);
+
+	IX_ETH_ACC_MEMSET(&ixEthAccPortData[portId],
+			  0,
+			  sizeof(ixEthAccPortData[portId]));
+
+	ixEthAccPortData[portId].ixEthAccTxData.schDiscipline = FIFO_NO_PRIORITY;
+    }
+
+    return (IX_ETH_ACC_SUCCESS);
+}
+
+
+IX_ETH_ACC_PUBLIC
+IxEthAccStatus ixEthAccPortTxDoneCallbackRegister(IxEthAccPortId portId,
+						  IxEthAccPortTxDoneCallback
+						  txCallbackFn,
+						  UINT32 callbackTag)
+{
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+    if (!IX_ETH_ACC_IS_PORT_VALID(portId))
+    {
+	return (IX_ETH_ACC_INVALID_PORT);
+    }
+
+/* HACK: removing this code to enable NPE-A preliminary testing
+ *    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+ *    {
+ *        IX_ETH_ACC_WARNING_LOG("ixEthAccPortTxDoneCallbackRegister: Unavailable Eth %d: Cannot register TxDone Callback.\n",(INT32)portId,0,0,0,0,0);
+ *        return IX_ETH_ACC_SUCCESS ;
+ *    }
+ */
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+    if (txCallbackFn == 0)
+	/* Check for null function pointer here. */
+    {
+	return (IX_ETH_ACC_INVALID_ARG);
+    }
+    ixEthAccPortData[portId].ixEthAccTxData.txBufferDoneCallbackFn = txCallbackFn;
+    ixEthAccPortData[portId].ixEthAccTxData.txCallbackTag = callbackTag;
+    return (IX_ETH_ACC_SUCCESS);
+}
+
+
+IX_ETH_ACC_PUBLIC
+IxEthAccStatus ixEthAccPortRxCallbackRegister(IxEthAccPortId portId,
+					      IxEthAccPortRxCallback
+					      rxCallbackFn,
+					      UINT32 callbackTag)
+{
+    IxEthAccPortId port;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+    if (!IX_ETH_ACC_IS_PORT_VALID(portId))
+    {
+	return (IX_ETH_ACC_INVALID_PORT);
+    }
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("ixEthAccPortRxCallbackRegister: Unavailable Eth %d: Cannot register Rx Callback.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /* Check for null function pointer here. */
+    if (rxCallbackFn == NULL)
+    {
+	return (IX_ETH_ACC_INVALID_ARG);
+    }
+
+    /* Check the user is not changing the callback type
+     * when the port is enabled.
+    */
+    if (ixEthAccMacState[portId].portDisableState == ACTIVE)
+    {
+	for (port = 0; port < IX_ETH_ACC_NUMBER_OF_PORTS; port++)
+	{
+	    if ((ixEthAccMacState[port].portDisableState == ACTIVE)
+		&& (ixEthAccPortData[port].ixEthAccRxData.rxMultiBufferCallbackInUse == TRUE))
+	    {
+		/* one of the active ports has a different rx callback type.
+		 * Changing the callback type when the port is enabled
+		 * is not safe
+		 */
+		return (IX_ETH_ACC_INVALID_ARG);
+	    }
+	}
+    }
+
+    /* update the callback pointer : this is done before
+     * registering the new qmgr callback
+     */
+    ixEthAccPortData[portId].ixEthAccRxData.rxCallbackFn = rxCallbackFn;
+    ixEthAccPortData[portId].ixEthAccRxData.rxCallbackTag = callbackTag;
+
+    /* update the qmgr callback for rx queues */
+    if (ixEthAccQMgrRxCallbacksRegister(ixEthRxFrameQMCallback)
+	!= IX_ETH_ACC_SUCCESS)
+    {
+	/* unexpected qmgr error */
+        IX_ETH_ACC_FATAL_LOG("ixEthAccPortRxCallbackRegister: unexpected QMgr error, " \
+            "could not register Rx single-buffer callback\n", 0, 0, 0, 0, 0, 0);
+
+	RX_INC(portId,rxUnexpectedError);
+	return (IX_ETH_ACC_INVALID_ARG);
+    }
+
+    ixEthAccPortData[portId].ixEthAccRxData.rxMultiBufferCallbackInUse = FALSE;
+
+    return (IX_ETH_ACC_SUCCESS);
+}
+
+IX_ETH_ACC_PUBLIC
+IxEthAccStatus ixEthAccPortMultiBufferRxCallbackRegister(
+			 IxEthAccPortId portId,
+			 IxEthAccPortMultiBufferRxCallback
+			 rxCallbackFn,
+			 UINT32 callbackTag)
+{
+    IxEthAccPortId port;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+    if (!IX_ETH_ACC_IS_PORT_VALID(portId))
+    {
+	return (IX_ETH_ACC_INVALID_PORT);
+    }
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("ixEthAccPortMultiBufferRxCallbackRegister: Unavailable Eth %d: Cannot register Rx Callback.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /* Check for null function pointer here. */
+    if (rxCallbackFn == NULL)
+    {
+	return (IX_ETH_ACC_INVALID_ARG);
+    }
+
+    /* Check the user is not changing the callback type
+     * when the port is enabled.
+    */
+    if (ixEthAccMacState[portId].portDisableState == ACTIVE)
+    {
+	for (port = 0; port < IX_ETH_ACC_NUMBER_OF_PORTS; port++)
+	{
+	    if ((ixEthAccMacState[port].portDisableState == ACTIVE)
+		&& (ixEthAccPortData[port].ixEthAccRxData.rxMultiBufferCallbackInUse == FALSE))
+	    {
+		/* one of the active ports has a different rx callback type.
+		 * Changing the callback type when the port is enabled
+		 * is not safe
+		 */
+		return (IX_ETH_ACC_INVALID_ARG);
+	    }
+	}
+    }
+
+    /* update the callback pointer : this is done before
+     * registering the new qmgr callback
+     */
+    ixEthAccPortData[portId].ixEthAccRxData.rxMultiBufferCallbackFn = rxCallbackFn;
+    ixEthAccPortData[portId].ixEthAccRxData.rxMultiBufferCallbackTag = callbackTag;
+
+    /* update the qmgr callback for rx queues */
+    if (ixEthAccQMgrRxCallbacksRegister(ixEthRxMultiBufferQMCallback)
+	!= IX_ETH_ACC_SUCCESS)
+    {
+	/* unexpected qmgr error */
+	RX_INC(portId,rxUnexpectedError);
+
+        IX_ETH_ACC_FATAL_LOG("ixEthAccPortMultiBufferRxCallbackRegister: unexpected QMgr error, " \
+            "could not register Rx multi-buffer callback\n", 0, 0, 0, 0, 0, 0);
+
+	return (IX_ETH_ACC_INVALID_ARG);
+    }
+
+    ixEthAccPortData[portId].ixEthAccRxData.rxMultiBufferCallbackInUse = TRUE;
+
+    return (IX_ETH_ACC_SUCCESS);
+}
+
+IX_ETH_ACC_PUBLIC
+IxEthAccStatus ixEthAccPortTxFrameSubmit(IxEthAccPortId portId,
+					 IX_OSAL_MBUF *buffer,
+					 IxEthAccTxPriority priority)
+{
+    IX_STATUS	qStatus = IX_SUCCESS;
+    UINT32      qBuffer;
+    IxEthAccTxPriority highestPriority;
+    IxQMgrQStatus txQStatus;
+
+#ifndef NDEBUG
+    if (buffer == NULL)
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+    if (!IX_ETH_ACC_IS_PORT_VALID(portId))
+    {
+	return (IX_ETH_ACC_INVALID_PORT);
+    }
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_FATAL_LOG("ixEthAccPortTxFrameSubmit: Unavailable Eth %d: Cannot submit Tx Frame.\n",
+			     (INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_PORT_UNINITIALIZED ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+    if ((UINT32)priority > (UINT32)IX_ETH_ACC_TX_PRIORITY_7)
+    {
+	return (IX_ETH_ACC_INVALID_ARG);
+    }
+#endif
+
+    /*
+     * Need to Flush the MBUF and its contents (data) as it may be
+     * read from the NPE. Convert virtual addresses to physical addresses also.
+     */
+    qBuffer = ixEthAccMbufTxQPrepare(buffer);
+
+    /*
+     * If no fifo priority set on Xscale ...
+     */
+    if (ixEthAccPortData[portId].ixEthAccTxData.schDiscipline ==
+	FIFO_NO_PRIORITY)
+    {
+	/*
+	 * Add The Tx Buffer to the H/W Tx Q if possible
+	 * (the priority is passed to the NPE, because
+	 * the NPE is able to reorder the frames
+	 * before transmission to the underlying hardware)
+	 */
+	qStatus = ixEthAccQmgrTxWrite(portId,
+				      qBuffer,
+				      IX_ETH_ACC_TX_DEFAULT_PRIORITY);
+
+	if (qStatus == IX_SUCCESS)
+	{
+	    TX_STATS_INC(portId,txQOK);
+
+	    /*
+	     * "best case" scenario : Buffer added to h/w Q.
+	     */
+	    return (IX_SUCCESS);
+	}
+	else if (qStatus == IX_QMGR_Q_OVERFLOW)
+	{
+	    /*
+	     * We were unable to write the buffer to the
+	     * appropriate H/W Q,  Save it in the sw Q.
+	     * (use the default priority queue regardless of
+	     * input parameter)
+	     */
+	    priority = IX_ETH_ACC_TX_DEFAULT_PRIORITY;
+	}
+	else
+	{
+	    /* unexpected qmgr error */
+	    TX_INC(portId,txUnexpectedError);
+	    IX_ETH_ACC_FATAL_LOG(
+		"ixEthAccPortTxFrameSubmit:Error: qStatus = %u\n",
+		(UINT32)qStatus, 0, 0, 0, 0, 0);
+	    return (IX_ETH_ACC_FAIL);
+	}
+    }
+    else if (ixEthAccPortData[portId].ixEthAccTxData.schDiscipline ==
+	     FIFO_PRIORITY)
+    {
+
+	/*
+	 * For priority transmission, put the frame directly on the H/W queue
+	 * if the H/W queue is empty, otherwise, put it in a S/W Q
+	 */
+	ixQMgrQStatusGet(IX_ETH_ACC_PORT_TO_TX_Q_ID(portId), &txQStatus);
+	if((txQStatus & IX_QMGR_Q_STATUS_E_BIT_MASK) != 0)
+	{
+	    /*The tx queue is empty, check whether there are buffers on the s/w queues*/
+	    if(ixEthAccTxSwQHighestPriorityGet(portId,  &highestPriority)
+	       !=IX_ETH_ACC_FAIL)
+	    {
+		/*there are buffers on the s/w queues, submit them*/
+		ixEthAccTxFromSwQ(portId, highestPriority);
+
+		/* the queue was empty, 1 buffer is already supplied
+		 * but is likely to be immediately transmitted and the
+		 * hw queue is likely to be empty again, so submit
+		 * more from the sw queues
+		 */
+		if(ixEthAccTxSwQHighestPriorityGet(portId,  &highestPriority)
+		   !=IX_ETH_ACC_FAIL)
+		{
+		    ixEthAccTxFromSwQ(portId, highestPriority);
+		    /*
+		     * and force the buffer supplied to be placed
+		     * on a priority queue
+		     */
+		    qStatus = IX_QMGR_Q_OVERFLOW;
+		}
+		else
+		{
+		    /*there are no buffers in the s/w queues, submit directly*/
+		    qStatus = ixEthAccQmgrTxWrite(portId, qBuffer, priority);
+		}
+	    }
+	    else
+	    {
+		/*there are no buffers in the s/w queues, submit directly*/
+		qStatus = ixEthAccQmgrTxWrite(portId, qBuffer, priority);
+	    }
+	}
+	else
+	{
+	    qStatus = IX_QMGR_Q_OVERFLOW;
+	}
+    }
+    else
+    {
+	TX_INC(portId,txUnexpectedError);
+	IX_ETH_ACC_FATAL_LOG(
+	    "ixEthAccPortTxFrameSubmit:Error: wrong schedule discipline setup\n",
+	    0, 0, 0, 0, 0, 0);
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    if(qStatus == IX_SUCCESS )
+    {
+	TX_STATS_INC(portId,txQOK);
+	return IX_ETH_ACC_SUCCESS;
+    }
+    else if(qStatus == IX_QMGR_Q_OVERFLOW)
+    {
+	TX_STATS_INC(portId,txQDelayed);
+	/*
+	 * We were unable to write the buffer to the
+	 * appropriate H/W Q,  Save it in a s/w Q.
+	 */
+	IX_ETH_ACC_DATAPLANE_ADD_MBUF_TO_Q_TAIL(
+		ixEthAccPortData[portId].
+		ixEthAccTxData.txQ[priority],
+		buffer);
+
+	qStatus = ixQMgrNotificationEnable(
+		IX_ETH_ACC_PORT_TO_TX_Q_ID(portId),
+		IX_ETH_ACC_PORT_TO_TX_Q_SOURCE(portId));
+
+        if (qStatus != IX_SUCCESS)
+	{
+	    if (qStatus == IX_QMGR_WARNING)
+	    {
+		/* notification is enabled for a queue
+		 * which is already empty (the condition is already met)
+		 * and there will be no more queue event to drain the sw queue
+		 */
+		TX_STATS_INC(portId,txLateNotificationEnabled);
+
+		/* pull a buffer from the sw queue */
+		if(ixEthAccTxSwQHighestPriorityGet(portId,  &highestPriority)
+		   !=IX_ETH_ACC_FAIL)
+		{
+		    /*there are buffers on the s/w queues, submit from them*/
+		    ixEthAccTxFromSwQ(portId, highestPriority);
+		}
+	    }
+	    else
+	    {
+		TX_INC(portId,txUnexpectedError);
+		IX_ETH_ACC_FATAL_LOG(
+		     "ixEthAccPortTxFrameSubmit: unexpected Error: %u\n",
+		     qStatus, 0, 0, 0, 0, 0);
+	    }
+        }
+    }
+    else
+    {
+	TX_INC(portId,txUnexpectedError);
+	IX_ETH_ACC_FATAL_LOG(
+	     "ixEthAccPortTxFrameSubmit: unexpected Error: %u\n",
+	     qStatus, 0, 0, 0, 0, 0);
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    return (IX_ETH_ACC_SUCCESS);
+}
+
+
+/**
+ *
+ * @brief replenish: convert a chain of mbufs to the format
+ *        expected by the NPE
+ *
+  */
+
+IX_ETH_ACC_PUBLIC
+IxEthAccStatus ixEthAccPortRxFreeReplenish(IxEthAccPortId portId,
+					   IX_OSAL_MBUF *buffer)
+{
+    IX_STATUS	qStatus = IX_SUCCESS;
+    UINT32      qBuffer;
+
+    /*
+     * Check buffer is valid.
+     */
+
+#ifndef NDEBUG
+    /* check parameter value */
+    if (buffer == 0)
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+    if (!IX_ETH_ACC_IS_PORT_VALID(portId))
+    {
+	return (IX_ETH_ACC_INVALID_PORT);
+    }
+
+    /* check initialisation is done */
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_FATAL_LOG(" ixEthAccPortRxFreeReplenish: Unavailable Eth %d: Cannot replenish Rx Free Q.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_PORT_UNINITIALIZED ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+    /* check boundaries and constraints */
+    if (IX_OSAL_MBUF_MLEN(buffer) < IX_ETHNPE_ACC_RXFREE_BUFFER_LENGTH_MIN)
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+#endif
+
+    qBuffer = ixEthAccMbufRxQPrepare(buffer);
+
+    /*
+     * Add The Rx Buffer to the H/W Free buffer Q if possible
+     */
+    qStatus = ixEthAccQmgrLockRxWrite(portId, qBuffer);
+
+    if (qStatus == IX_SUCCESS)
+    {
+	RX_STATS_INC(portId,rxFreeRepOK);
+	/*
+	 * Buffer added to h/w Q.
+	 */
+	return (IX_SUCCESS);
+    }
+    else if (qStatus == IX_QMGR_Q_OVERFLOW)
+    {
+	RX_STATS_INC(portId,rxFreeRepDelayed);
+	/*
+	 * We were unable to write the buffer to the approprate H/W Q,
+	 * Save it in a s/w Q.
+	 */
+	IX_ETH_ACC_DATAPLANE_ADD_MBUF_TO_Q_TAIL(
+	    ixEthAccPortData[portId].ixEthAccRxData.freeBufferList,
+	    buffer);
+
+	qStatus = ixQMgrNotificationEnable(
+	    IX_ETH_ACC_PORT_TO_RX_FREE_Q_ID(portId),
+	    IX_ETH_ACC_PORT_TO_RX_FREE_Q_SOURCE(portId));
+
+        if (qStatus != IX_SUCCESS)
+	{
+	    if (qStatus == IX_QMGR_WARNING)
+	    {
+		/* notification is enabled for a queue
+		 * which is already empty (the condition is already met)
+		 * and there will be no more queue event to drain the sw queue
+		 * move an entry from the sw queue to the hw queue */
+		RX_STATS_INC(portId,rxFreeLateNotificationEnabled);
+		ixEthAccRxFreeFromSwQ(portId);
+	    }
+	    else
+	    {
+		RX_INC(portId,rxUnexpectedError);
+		IX_ETH_ACC_FATAL_LOG(
+		     "ixEthAccRxPortFreeReplenish:Error: %u\n",
+		     qStatus, 0, 0, 0, 0, 0);
+	    }
+        }
+    }
+    else
+    {
+	RX_INC(portId,rxUnexpectedError);
+	IX_ETH_ACC_FATAL_LOG(
+	    "ixEthAccRxPortFreeReplenish:Error: qStatus = %u\n",
+	    (UINT32)qStatus, 0, 0, 0, 0, 0);
+        return(IX_ETH_ACC_FAIL);
+    }
+    return (IX_ETH_ACC_SUCCESS);
+}
+
+
+IX_ETH_ACC_PUBLIC
+IxEthAccStatus ixEthAccTxSchedulingDisciplineSetPriv(IxEthAccPortId portId,
+						 IxEthAccSchedulerDiscipline
+						 sched)
+{
+    if (!IX_ETH_ACC_IS_PORT_VALID(portId))
+    {
+	return (IX_ETH_ACC_INVALID_PORT);
+    }
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("ixEthAccTxSchedulingDisciplineSet: Unavailable Eth %d: Cannot set Tx Scheduling Discipline.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    if (sched != FIFO_PRIORITY && sched != FIFO_NO_PRIORITY)
+    {
+	return (IX_ETH_ACC_INVALID_ARG);
+    }
+
+    ixEthAccPortData[portId].ixEthAccTxData.schDiscipline = sched;
+    return (IX_ETH_ACC_SUCCESS);
+}
+
+IX_ETH_ACC_PUBLIC
+IxEthAccStatus ixEthAccRxSchedulingDisciplineSetPriv(IxEthAccSchedulerDiscipline
+						 sched)
+{
+    if (sched != FIFO_PRIORITY && sched != FIFO_NO_PRIORITY)
+    {
+	return (IX_ETH_ACC_INVALID_ARG);
+    }
+
+    ixEthAccDataInfo.schDiscipline = sched;
+
+    return (IX_ETH_ACC_SUCCESS);
+}
+
+
+/**
+ * @fn ixEthRxFrameProcess(IxEthAccPortId portId, IX_OSAL_MBUF *mbufPtr)
+ *
+ * @brief process incoming frame :
+ *
+ * @param @ref IxQMgrCallback IxQMgrMultiBufferCallback
+ *
+ * @return none
+ *
+ * @internal
+ *
+ */
+IX_ETH_ACC_PRIVATE BOOL
+ixEthRxFrameProcess(IxEthAccPortId portId, IX_OSAL_MBUF *mbufPtr)
+{
+    UINT32 flags;
+    IxEthDBStatus result;
+
+#ifndef NDEBUG
+    /* Prudent to at least check the port is within range */
+    if (portId >= IX_ETH_ACC_NUMBER_OF_PORTS)
+    {
+	ixEthAccDataStats.unexpectedError++;
+	IX_ETH_ACC_FATAL_LOG(
+	     "ixEthRxFrameProcess: Illegal port: %u\n",
+	     (UINT32)portId, 0, 0, 0, 0, 0);
+	return FALSE;
+    }
+#endif
+
+    /* convert fields from mbuf header */
+    ixEthAccMbufFromRxQ(mbufPtr);
+
+    /* check about any special processing for this frame */
+    flags = IX_ETHACC_NE_FLAGS(mbufPtr);
+    if ((flags & (IX_ETHACC_NE_FILTERMASK | IX_ETHACC_NE_NEWSRCMASK)) == 0)
+    {
+	/* "best case" scenario : nothing special to do for this frame */
+	return TRUE;
+    }
+
+#ifdef CONFIG_IXP425_COMPONENT_ETHDB
+    /* if a new source MAC address is detected by the NPE,
+     * update IxEthDB with the portId and the MAC address.
+     */
+    if ((flags & IX_ETHACC_NE_NEWSRCMASK & ixEthAccNewSrcMask) != 0)
+    {
+        result = ixEthDBFilteringDynamicEntryProvision(portId,
+			  (IxEthDBMacAddr *) IX_ETHACC_NE_SOURCEMAC(mbufPtr));
+
+	if (result != IX_ETH_DB_SUCCESS && result != IX_ETH_DB_FEATURE_UNAVAILABLE)
+	{
+            if ((ixEthAccMacState[portId].portDisableState == ACTIVE) && (result != IX_ETH_DB_BUSY))
+            {
+	        RX_STATS_INC(portId, rxUnexpectedError);
+                IX_ETH_ACC_FATAL_LOG("ixEthRxFrameProcess: Failed to add source MAC \
+                                    to the Learning/Filtering database\n", 0, 0, 0, 0, 0, 0);
+            }
+            else
+            {
+                /* we expect this to fail during PortDisable, as EthDB is disabled for
+                 * that port and will refuse to learn new addresses
+		 */
+            }
+	}
+	else
+	{
+	    RX_STATS_INC(portId, rxUnlearnedMacAddress);
+	}
+    }
+#endif
+
+    /* check if this frame should have been filtered
+     * by the NPE and take the appropriate action
+     */
+    if (((flags & IX_ETHACC_NE_FILTERMASK) != 0)
+        && (ixEthAccMacState[portId].portDisableState == ACTIVE))
+    {
+        /* If the mbuf was allocated with a small data size, or the current data pointer is not
+         * within the allocated data area, then the buffer is non-standard and has to be
+         * replenished with the minimum size only
+         */
+        if( (IX_OSAL_MBUF_ALLOCATED_BUFF_LEN(mbufPtr) < IX_ETHNPE_ACC_RXFREE_BUFFER_LENGTH_MIN)
+           || ((UINT8 *)IX_OSAL_MBUF_ALLOCATED_BUFF_DATA(mbufPtr) > IX_OSAL_MBUF_MDATA(mbufPtr))
+           || ((UINT8 *)(IX_OSAL_MBUF_ALLOCATED_BUFF_DATA(mbufPtr) +
+              IX_OSAL_MBUF_ALLOCATED_BUFF_LEN(mbufPtr))
+               < IX_OSAL_MBUF_MDATA(mbufPtr)) )
+        {
+            /* set to minimum length */
+            IX_OSAL_MBUF_MLEN(mbufPtr) = IX_OSAL_MBUF_PKT_LEN(mbufPtr) =
+                IX_ETHNPE_ACC_RXFREE_BUFFER_LENGTH_MIN;
+        }
+        else
+        {
+            /* restore original length */
+            IX_OSAL_MBUF_MLEN(mbufPtr) = IX_OSAL_MBUF_PKT_LEN(mbufPtr) =
+                ( IX_OSAL_MBUF_ALLOCATED_BUFF_LEN(mbufPtr) -
+                 (IX_OSAL_MBUF_MDATA(mbufPtr) - (UINT8 *)IX_OSAL_MBUF_ALLOCATED_BUFF_DATA(mbufPtr)) );
+        }
+
+        /* replenish from here */
+        if (ixEthAccPortRxFreeReplenish(portId, mbufPtr) != IX_ETH_ACC_SUCCESS)
+        {
+                IX_ETH_ACC_FATAL_LOG("ixEthRxFrameProcess: Failed to replenish with filtered frame\
+                                      on port %d\n", portId, 0, 0, 0, 0, 0);
+        }
+
+        RX_STATS_INC(portId, rxFiltered);
+
+        /* indicate that frame should not be subjected to further processing */
+        return FALSE;
+    }
+
+    return TRUE;
+}
+
+
+/**
+ * @fn ixEthRxFrameQMCallback
+ *
+ * @brief receive callback for Frame receive Q from NPE
+ *
+ * Frames are passed one-at-a-time to the user
+ *
+ * @param @ref IxQMgrCallback
+ *
+ * @return none
+ *
+ * @internal
+ *
+ * Design note : while processing the entry X, entry X+1 is preloaded
+ * into memory to reduce the number of stall cycles
+ *
+ */
+void ixEthRxFrameQMCallback(IxQMgrQId qId, IxQMgrCallbackId callbackId)
+{
+    IX_OSAL_MBUF    *mbufPtr;
+    IX_OSAL_MBUF    *nextMbufPtr;
+    UINT32     qEntry;
+    UINT32     nextQEntry;
+    UINT32     *qEntryPtr;
+    UINT32     portId;
+    UINT32     destPortId;
+    UINT32     npeId;
+    UINT32     rxQReadStatus;
+
+    /*
+     * Design note : entries are read in a buffer, This buffer contains
+     * an extra zeroed entry so the loop will
+     * always terminate on a null entry, whatever the result of Burst read is.
+     */
+    UINT32 rxQEntry[IX_ETH_ACC_MAX_RX_FRAME_CONSUME_PER_CALLBACK + 1];
+
+    /*
+     * Indication of the number of times the callback is used.
+     */
+    IX_ETH_ACC_STATS_INC(ixEthAccDataStats.rxCallbackCounter);
+
+    do
+    {
+	/*
+	 * Indication of the number of times the queue is drained
+	 */
+	IX_ETH_ACC_STATS_INC(ixEthAccDataStats.rxCallbackBurstRead);
+
+	/* ensure the last entry of the array contains a zeroed value */
+	qEntryPtr = rxQEntry;
+	qEntryPtr[IX_ETH_ACC_MAX_RX_FRAME_CONSUME_PER_CALLBACK] = 0;
+
+	rxQReadStatus = ixQMgrQBurstRead(qId,
+		 IX_ETH_ACC_MAX_RX_FRAME_CONSUME_PER_CALLBACK,
+		 qEntryPtr);
+
+#ifndef NDEBUG
+	if ((rxQReadStatus != IX_QMGR_Q_UNDERFLOW)
+	    && (rxQReadStatus != IX_SUCCESS))
+	{
+	    ixEthAccDataStats.unexpectedError++;
+	    /*major error*/
+	    IX_ETH_ACC_FATAL_LOG(
+		"ixEthRxFrameQMCallback:Error: %u\n",
+		(UINT32)rxQReadStatus, 0, 0, 0, 0, 0);
+	    return;
+	}
+#endif
+
+	/* convert and preload the next entry
+	 * (the conversion function takes care about null pointers which
+	 * are used to mark the end of the loop)
+	 */
+	nextQEntry = *qEntryPtr;
+	nextMbufPtr = ixEthAccEntryFromQConvert(nextQEntry,
+			  IX_ETHNPE_QM_Q_RXENET_ADDR_MASK);
+
+	while(nextQEntry != 0)
+	{
+	    /* get the next entry */
+	    qEntry = nextQEntry;
+	    mbufPtr = nextMbufPtr;
+
+#ifndef NDEBUG
+	    if (mbufPtr == NULL)
+	    {
+		ixEthAccDataStats.unexpectedError++;
+		IX_ETH_ACC_FATAL_LOG(
+		    "ixEthRxFrameQMCallback: Null Mbuf Ptr\n",
+		    0, 0, 0, 0, 0, 0);
+		return;
+	    }
+#endif
+
+	    /* convert the next entry
+	     * (the conversion function takes care about null pointers which
+	     * are used to mark the end of the loop)
+	     */
+	    nextQEntry = *(++qEntryPtr);
+	    nextMbufPtr = ixEthAccEntryFromQConvert(nextQEntry,
+			      IX_ETHNPE_QM_Q_RXENET_ADDR_MASK);
+
+	    /*
+	     * Get Port and Npe ID from message.
+	     */
+	    npeId = ((IX_ETHNPE_QM_Q_RXENET_NPEID_MASK &
+		      qEntry) >> IX_ETHNPE_QM_Q_FIELD_NPEID_R);
+	    portId = IX_ETH_ACC_NPE_TO_PORT_ID(npeId);
+
+	    /* process frame, check the return code and skip the remaining of
+	     * the loop if the frame is to be filtered out
+	     */
+            if (ixEthRxFrameProcess(portId, mbufPtr))
+            {
+	        /* destination portId for this packet */
+	        destPortId = IX_ETHACC_NE_DESTPORTID(mbufPtr);
+
+                if (destPortId != IX_ETH_DB_UNKNOWN_PORT)
+                {
+                    destPortId = IX_ETH_DB_NPE_LOGICAL_ID_TO_PORT_ID(destPortId);
+                }
+
+	        /* test if QoS is enabled in ethAcc
+	        */
+	        if (ixEthAccDataInfo.schDiscipline == FIFO_PRIORITY)
+	        {
+		    /* check if there is a higher priority queue
+		    * which may require processing and then process it.
+		    */
+		    if (ixEthAccDataInfo.higherPriorityQueue[qId] < IX_QMGR_MAX_NUM_QUEUES)
+		    {
+		        ixEthRxFrameQMCallback(ixEthAccDataInfo.higherPriorityQueue[qId],
+					    callbackId);
+		    }
+	        }
+
+	        /*
+	        * increment priority stats
+	        */
+	        RX_STATS_INC(portId,rxPriority[IX_ETHACC_NE_QOS(mbufPtr)]);
+
+	        /*
+	        * increment callback count stats
+	        */
+	        RX_STATS_INC(portId,rxFrameClientCallback);
+
+	        /*
+	        * Call user level callback.
+	        */
+	        ixEthAccPortData[portId].ixEthAccRxData.rxCallbackFn(
+		    ixEthAccPortData[portId].ixEthAccRxData.rxCallbackTag,
+		    mbufPtr,
+		    destPortId);
+            }
+	}
+    } while (rxQReadStatus == IX_SUCCESS);
+}
+
+/**
+ * @fn ixEthRxMultiBufferQMCallback
+ *
+ * @brief receive callback for Frame receive Q from NPE
+ *
+ * Frames are passed as an array to the user
+ *
+ * @param @ref IxQMgrCallback
+ *
+ * @return none
+ *
+ * @internal
+ *
+ * Design note : while processing the entry X, entry X+1 is preloaded
+ * into memory to reduce the number of stall cycles
+ *
+ */
+void ixEthRxMultiBufferQMCallback(IxQMgrQId qId, IxQMgrCallbackId callbackId)
+{
+    IX_OSAL_MBUF    *mbufPtr;
+    IX_OSAL_MBUF    *nextMbufPtr;
+    UINT32     qEntry;
+    UINT32     nextQEntry;
+    UINT32     *qEntryPtr;
+    UINT32     portId;
+    UINT32     npeId;
+    UINT32     rxQReadStatus;
+    /*
+     * Design note : entries are read in a static buffer, This buffer contains
+     * an extra zeroed entry so the loop will
+     * always terminate on a null entry, whatever the result of Burst read is.
+     */
+    static UINT32 rxQEntry[IX_ETH_ACC_MAX_RX_FRAME_CONSUME_PER_CALLBACK + 1];
+    static IX_OSAL_MBUF *rxMbufPortArray[IX_ETH_ACC_NUMBER_OF_PORTS][IX_ETH_ACC_MAX_RX_FRAME_CONSUME_PER_CALLBACK + 1];
+    IX_OSAL_MBUF **rxMbufPtr[IX_ETH_ACC_NUMBER_OF_PORTS];
+
+    for (portId = 0; portId < IX_ETH_ACC_NUMBER_OF_PORTS; portId++)
+    {
+	rxMbufPtr[portId] = rxMbufPortArray[portId];
+    }
+
+    /*
+     * Indication of the number of times the callback is used.
+     */
+    IX_ETH_ACC_STATS_INC(ixEthAccDataStats.rxCallbackCounter);
+
+    do
+    {
+	/*
+	 * Indication of the number of times the queue is drained
+	 */
+	IX_ETH_ACC_STATS_INC(ixEthAccDataStats.rxCallbackBurstRead);
+
+	/* ensure the last entry of the array contains a zeroed value */
+	qEntryPtr = rxQEntry;
+	qEntryPtr[IX_ETH_ACC_MAX_RX_FRAME_CONSUME_PER_CALLBACK] = 0;
+
+	rxQReadStatus = ixQMgrQBurstRead(qId,
+		 IX_ETH_ACC_MAX_RX_FRAME_CONSUME_PER_CALLBACK,
+		 qEntryPtr);
+
+#ifndef NDEBUG
+	if ((rxQReadStatus != IX_QMGR_Q_UNDERFLOW)
+	    && (rxQReadStatus != IX_SUCCESS))
+	{
+	    ixEthAccDataStats.unexpectedError++;
+	    /*major error*/
+	    IX_ETH_ACC_FATAL_LOG(
+		"ixEthRxFrameMultiBufferQMCallback:Error: %u\n",
+		(UINT32)rxQReadStatus, 0, 0, 0, 0, 0);
+	    return;
+	}
+#endif
+
+	/* convert and preload the next entry
+	 * (the conversion function takes care about null pointers which
+	 * are used to mark the end of the loop)
+	 */
+	nextQEntry = *qEntryPtr;
+	nextMbufPtr = ixEthAccEntryFromQConvert(nextQEntry,
+			  IX_ETHNPE_QM_Q_RXENET_ADDR_MASK);
+
+	while(nextQEntry != 0)
+	{
+	    /* get the next entry */
+	    qEntry = nextQEntry;
+	    mbufPtr = nextMbufPtr;
+
+#ifndef NDEBUG
+	    if (mbufPtr == NULL)
+	    {
+		ixEthAccDataStats.unexpectedError++;
+		IX_ETH_ACC_FATAL_LOG(
+		    "ixEthRxFrameMultiBufferQMCallback:Error: Null Mbuf Ptr\n",
+		    0, 0, 0, 0, 0, 0);
+		return;
+	    }
+#endif
+
+	    /* convert the next entry
+	     * (the conversion function takes care about null pointers which
+	     * are used to mark the end of the loop)
+	     */
+	    nextQEntry = *(++qEntryPtr);
+	    nextMbufPtr = ixEthAccEntryFromQConvert(nextQEntry,
+			      IX_ETHNPE_QM_Q_RXENET_ADDR_MASK);
+
+	    /*
+	     * Get Port and Npe ID from message.
+	     */
+	    npeId = ((IX_ETHNPE_QM_Q_RXENET_NPEID_MASK &
+		      qEntry) >>
+		     IX_ETHNPE_QM_Q_FIELD_NPEID_R);
+	    portId = IX_ETH_ACC_NPE_TO_PORT_ID(npeId);
+
+	    /* skip the remaining of the loop if the frame is
+	     * to be filtered out
+	     */
+	    if (ixEthRxFrameProcess(portId, mbufPtr))
+	    {
+		/* store a mbuf pointer in an array */
+		*rxMbufPtr[portId]++ = mbufPtr;
+
+		/*
+		 * increment priority stats
+		 */
+		RX_STATS_INC(portId,rxPriority[IX_ETHACC_NE_QOS(mbufPtr)]);
+	    }
+
+	    /* test for QoS enabled in ethAcc */
+	    if (ixEthAccDataInfo.schDiscipline == FIFO_PRIORITY)
+	    {
+		/* check if there is a higher priority queue
+		 * which may require processing and then process it.
+		 */
+		if (ixEthAccDataInfo.higherPriorityQueue[qId] < IX_QMGR_MAX_NUM_QUEUES)
+		{
+		    ixEthRxMultiBufferQMCallback(ixEthAccDataInfo.higherPriorityQueue[qId],
+						 callbackId);
+		}
+	    }
+	}
+
+	/* check if any of the the arrays contains any entry */
+	for (portId = 0; portId < IX_ETH_ACC_NUMBER_OF_PORTS; portId++)
+	{
+	    if (rxMbufPtr[portId] != rxMbufPortArray[portId])
+	    {
+		/* add a last NULL pointer at the end of the
+		 * array of mbuf pointers
+		 */
+		*rxMbufPtr[portId] = NULL;
+
+		/*
+		 * increment callback count stats
+		 */
+		RX_STATS_INC(portId,rxFrameClientCallback);
+
+		/*
+		 * Call user level callback with an array of
+		 * buffers (NULL terminated)
+		 */
+		ixEthAccPortData[portId].ixEthAccRxData.
+		    rxMultiBufferCallbackFn(
+			    ixEthAccPortData[portId].ixEthAccRxData.
+			           rxMultiBufferCallbackTag,
+			    rxMbufPortArray[portId]);
+
+		/* reset the buffer pointer to the beginning of
+		 * the array
+		 */
+		rxMbufPtr[portId] = rxMbufPortArray[portId];
+	    }
+	}
+
+    } while (rxQReadStatus == IX_SUCCESS);
+}
+
+
+/**
+ * @brief  rxFree low event handler
+ *
+ */
+void ixEthRxFreeQMCallback(IxQMgrQId qId, IxQMgrCallbackId callbackId)
+{
+    IxEthAccPortId	portId = (IxEthAccPortId) callbackId;
+    int		        lockVal;
+    UINT32		maxQWritesToPerform = IX_ETH_ACC_MAX_RX_FREE_BUFFERS_LOAD;
+    IX_STATUS	        qStatus = IX_SUCCESS;
+
+    /*
+     * We have reached a low threshold on one of the Rx Free Qs
+     */
+
+    /*note that due to the fact that we are working off an Empty threshold, this callback
+      need only write a single entry to the Rx Free queue in order to re-arm the notification
+    */
+
+    RX_STATS_INC(portId,rxFreeLowCallback);
+
+    /*
+     * Get buffers from approprite S/W Rx freeBufferList Q.
+     */
+
+#ifndef NDEBUG
+    if (!IX_ETH_ACC_IS_PORT_VALID(portId))
+    {
+	ixEthAccDataStats.unexpectedError++;
+	IX_ETH_ACC_FATAL_LOG(
+	    "ixEthRxFreeQMCallback:Error: Invalid Port 0x%08X\n",
+	    portId, 0, 0, 0, 0, 0);
+	return;
+    }
+#endif
+    IX_ETH_ACC_DATA_PLANE_LOCK(lockVal);
+    if (IX_ETH_ACC_DATAPLANE_IS_Q_EMPTY(ixEthAccPortData[portId].
+					ixEthAccRxData.freeBufferList))
+    {
+	/*
+	 * Turn off Q callback notification for Q in Question.
+	 */
+	qStatus = ixQMgrNotificationDisable(
+	    IX_ETH_ACC_PORT_TO_RX_FREE_Q_ID(portId));
+
+
+	IX_ETH_ACC_DATA_PLANE_UNLOCK(lockVal);
+
+	if (qStatus != IX_SUCCESS)
+	{
+	    RX_INC(portId,rxUnexpectedError);
+	    IX_ETH_ACC_FATAL_LOG(
+		"ixEthRxFreeQMCallback:Error: unexpected QM status 0x%08X\n",
+		qStatus, 0, 0, 0, 0, 0);
+	    return;
+	}
+    }
+    else
+    {
+	IX_ETH_ACC_DATA_PLANE_UNLOCK(lockVal);
+	/*
+	 * Load the H/W Q with buffers from the s/w Q.
+	 */
+
+	do
+	{
+	    /*
+	     * Consume Q entries. - Note Q contains Physical addresss,
+	     * and have already been flushed to memory,
+	     * And endianess converted if required.
+	     */
+	    if (ixEthAccRxFreeFromSwQ(portId) != IX_SUCCESS)
+	    {
+		/*
+		 * No more entries in s/w Q.
+		 * Turn off Q callback indication
+		 */
+
+		IX_ETH_ACC_DATA_PLANE_LOCK(lockVal);
+		if (IX_ETH_ACC_DATAPLANE_IS_Q_EMPTY(ixEthAccPortData[portId].
+		    ixEthAccRxData.freeBufferList))
+		{
+		    qStatus = ixQMgrNotificationDisable(
+			IX_ETH_ACC_PORT_TO_RX_FREE_Q_ID(portId));
+		}
+		IX_ETH_ACC_DATA_PLANE_UNLOCK(lockVal);
+		break;
+	    }
+	}
+	while (--maxQWritesToPerform);
+    }
+}
+/**
+ * @fn Tx queue low event handler
+ *
+ */
+void
+ixEthTxFrameQMCallback(IxQMgrQId qId, IxQMgrCallbackId callbackId)
+{
+    IxEthAccPortId portId = (IxEthAccPortId) callbackId;
+    int		   lockVal;
+    UINT32	   maxQWritesToPerform = IX_ETH_ACC_MAX_TX_FRAME_TX_CONSUME_PER_CALLBACK;
+    IX_STATUS	   qStatus = IX_SUCCESS;
+    IxEthAccTxPriority highestPriority;
+
+
+    /*
+     * We have reached a low threshold on the Tx Q, and are being asked to
+     * supply a buffer for transmission from our S/W TX queues
+     */
+    TX_STATS_INC(portId,txLowThreshCallback);
+
+    /*
+     * Get buffers from approprite Q.
+     */
+
+#ifndef NDEBUG
+    if (!IX_ETH_ACC_IS_PORT_VALID(portId))
+    {
+	ixEthAccDataStats.unexpectedError++;
+	IX_ETH_ACC_FATAL_LOG(
+	    "ixEthTxFrameQMCallback:Error: Invalid Port 0x%08X\n",
+	    portId, 0, 0, 0, 0, 0);
+	return;
+    }
+#endif
+
+    do
+    {
+	/*
+	 * Consume Q entries. - Note Q contains Physical addresss,
+	 * and have already been flushed to memory,
+	 * and endianess already sone if required.
+	 */
+
+	IX_ETH_ACC_DATA_PLANE_LOCK(lockVal);
+
+	if(ixEthAccTxSwQHighestPriorityGet(portId, &highestPriority) ==
+	   IX_ETH_ACC_FAIL)
+	{
+	    /*
+	     * No more entries in s/w Q.
+	     * Turn off Q callback indication
+	     */
+	    qStatus = ixQMgrNotificationDisable(
+		IX_ETH_ACC_PORT_TO_TX_Q_ID(portId));
+
+	    IX_ETH_ACC_DATA_PLANE_UNLOCK(lockVal);
+
+	    if (qStatus != IX_SUCCESS)
+	    {
+		ixEthAccDataStats.unexpectedError++;
+		IX_ETH_ACC_FATAL_LOG(
+		    "ixEthTxFrameQMCallback:Error: unexpected QM status 0x%08X\n",
+		    qStatus, 0, 0, 0, 0, 0);
+	    }
+
+	    return;
+	}
+	else
+	{
+	    IX_ETH_ACC_DATA_PLANE_UNLOCK(lockVal);
+	    if (ixEthAccTxFromSwQ(portId,highestPriority)!=IX_SUCCESS)
+	    {
+                /* nothing left in the sw queue or the hw queues are
+                * full. There is no point to continue to drain the
+                * sw queues
+                */
+		return;
+	    }
+	}
+    }
+    while (--maxQWritesToPerform);
+}
+
+/**
+ * @brief TxDone event handler
+ *
+ * Design note : while processing the entry X, entry X+1 is preloaded
+ * into memory to reduce the number of stall cycles
+ *
+ */
+
+void
+ixEthTxFrameDoneQMCallback(IxQMgrQId qId, IxQMgrCallbackId callbackId)
+{
+    IX_OSAL_MBUF    *mbufPtr;
+    UINT32     qEntry;
+    UINT32     *qEntryPtr;
+    UINT32     txDoneQReadStatus;
+    UINT32     portId;
+    UINT32     npeId;
+
+    /*
+     * Design note : entries are read in a static buffer, This buffer contains
+     * an extra entyry (which is zeroed by the compiler), so the loop will
+     * always terminate on a null entry, whatever the result of Burst read is.
+     */
+    static UINT32 txDoneQEntry[IX_ETH_ACC_MAX_TX_FRAME_DONE_CONSUME_PER_CALLBACK + 1];
+
+    /*
+     * Indication that Tx frames have been transmitted from the NPE.
+     */
+
+    IX_ETH_ACC_STATS_INC(ixEthAccDataStats.txDoneCallbackCounter);
+
+    do{
+	qEntryPtr = txDoneQEntry;
+	txDoneQReadStatus = ixQMgrQBurstRead(IX_ETH_ACC_TX_FRAME_DONE_ETH_Q,
+		     IX_ETH_ACC_MAX_TX_FRAME_DONE_CONSUME_PER_CALLBACK,
+		     qEntryPtr);
+
+#ifndef NDEBUG
+	if (txDoneQReadStatus != IX_QMGR_Q_UNDERFLOW
+	    && (txDoneQReadStatus != IX_SUCCESS))
+	{
+	    /*major error*/
+	    ixEthAccDataStats.unexpectedError++;
+	    IX_ETH_ACC_FATAL_LOG(
+		"ixEthTxFrameDoneQMCallback:Error: %u\n",
+		(UINT32)txDoneQReadStatus, 0, 0, 0, 0, 0);
+	    return;
+	}
+#endif
+
+	qEntry = *qEntryPtr;
+
+	while(qEntry != 0)
+	{
+	    mbufPtr = ixEthAccEntryFromQConvert(qEntry,
+		      IX_ETHNPE_QM_Q_TXENET_ADDR_MASK);
+
+#ifndef NDEBUG
+	    if (mbufPtr == NULL)
+	    {
+		ixEthAccDataStats.unexpectedError++;
+		IX_ETH_ACC_FATAL_LOG(
+		    "ixEthTxFrameDoneQMCallback:Error: Null Mbuf Ptr\n",
+		    0, 0, 0, 0, 0, 0);
+		return;
+	    }
+#endif
+
+	    /* endianness conversions and stats updates */
+	    ixEthAccMbufFromTxQ(mbufPtr);
+
+	    /*
+	     * Get NPE id from message, then convert to portId.
+	     */
+	    npeId = ((IX_ETHNPE_QM_Q_TXENETDONE_NPEID_MASK &
+		       qEntry) >>
+		      IX_ETHNPE_QM_Q_FIELD_NPEID_R);
+	    portId = IX_ETH_ACC_NPE_TO_PORT_ID(npeId);
+
+#ifndef NDEBUG
+	    /* Prudent to at least check the port is within range */
+	    if (portId >= IX_ETH_ACC_NUMBER_OF_PORTS)
+	    {
+		ixEthAccDataStats.unexpectedError++;
+		IX_ETH_ACC_FATAL_LOG(
+		    "ixEthTxFrameDoneQMCallback: Illegal port: %u\n",
+		    (UINT32)portId, 0, 0, 0, 0, 0);
+		return;
+	    }
+#endif
+
+	    TX_STATS_INC(portId,txDoneClientCallback);
+
+	    /*
+	     * Call user level callback.
+	     */
+	    ixEthAccPortData[portId].ixEthAccTxData.txBufferDoneCallbackFn(
+		ixEthAccPortData[portId].ixEthAccTxData.txCallbackTag,
+		mbufPtr);
+
+	    /* move to next queue entry */
+	    qEntry = *(++qEntryPtr);
+
+	}
+    } while( txDoneQReadStatus == IX_SUCCESS );
+}
+
+IX_ETH_ACC_PUBLIC
+void ixEthAccDataPlaneShow(void)
+{
+    UINT32 numTx0Entries;
+    UINT32 numTx1Entries;
+    UINT32 numTxDoneEntries;
+    UINT32 numRxEntries;
+    UINT32 numRxFree0Entries;
+    UINT32 numRxFree1Entries;
+    UINT32 portId;
+#ifdef __ixp46X
+    UINT32 numTx2Entries;
+    UINT32 numRxFree2Entries;
+#endif
+#ifndef NDEBUG
+    UINT32 priority;
+    UINT32 numBuffersInRx=0;
+    UINT32 numBuffersInTx=0;
+    UINT32 numBuffersInSwQ=0;
+    UINT32 totalBuffers=0;
+    UINT32 rxFreeCallbackCounter = 0;
+    UINT32 txCallbackCounter = 0;
+#endif
+    UINT32 key;
+
+    /* snapshot of stats */
+    IxEthAccTxDataStats tx[IX_ETH_ACC_NUMBER_OF_PORTS];
+    IxEthAccRxDataStats rx[IX_ETH_ACC_NUMBER_OF_PORTS];
+    IxEthAccDataPlaneStats stats;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return;
+    }
+
+    /* get a reliable snapshot */
+    key = ixOsalIrqLock();
+
+    numTx0Entries = 0;
+    ixQMgrQNumEntriesGet(IX_ETH_ACC_TX_FRAME_ENET0_Q, &numTx0Entries);
+    numTx1Entries = 0;
+    ixQMgrQNumEntriesGet(IX_ETH_ACC_TX_FRAME_ENET1_Q, &numTx1Entries);
+    numTxDoneEntries = 0;
+    ixQMgrQNumEntriesGet( IX_ETH_ACC_TX_FRAME_DONE_ETH_Q, &numTxDoneEntries);
+    numRxEntries = 0;
+    ixEthAccQMgrRxQEntryGet(&numRxEntries);
+    numRxFree0Entries = 0;
+    ixQMgrQNumEntriesGet(IX_ETH_ACC_RX_FREE_BUFF_ENET0_Q, &numRxFree0Entries);
+    numRxFree1Entries = 0;
+    ixQMgrQNumEntriesGet(IX_ETH_ACC_RX_FREE_BUFF_ENET1_Q, &numRxFree1Entries);
+
+#ifdef __ixp46X
+    numTx2Entries = 0;
+    ixQMgrQNumEntriesGet(IX_ETH_ACC_TX_FRAME_ENET2_Q, &numTx2Entries);
+    numRxFree2Entries = 0;
+    ixQMgrQNumEntriesGet(IX_ETH_ACC_RX_FREE_BUFF_ENET2_Q, &numRxFree2Entries);
+#endif
+
+    for(portId=IX_ETH_PORT_1; portId < IX_ETH_ACC_NUMBER_OF_PORTS; portId++)
+    {
+	memcpy(&tx[portId],
+	       &ixEthAccPortData[portId].ixEthAccTxData.stats,
+	       sizeof(tx[portId]));
+	memcpy(&rx[portId],
+	       &ixEthAccPortData[portId].ixEthAccRxData.stats,
+	       sizeof(rx[portId]));
+    }
+    memcpy(&stats, &ixEthAccDataStats, sizeof(stats));
+
+    ixOsalIrqUnlock(key);
+
+#ifdef NDEBUG
+    printf("Detailed statistics collection not supported in this load\n");
+#endif
+
+    /* print snapshot */
+    for(portId=0; portId < IX_ETH_ACC_NUMBER_OF_PORTS; portId++)
+    {
+        /* If not IXP42X A0 stepping, proceed to check for existence of coprocessors */
+        if ((IX_FEATURE_CTRL_SILICON_TYPE_A0 !=
+	     (ixFeatureCtrlProductIdRead() & IX_FEATURE_CTRL_SILICON_STEPPING_MASK))
+	    || (IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X != ixFeatureCtrlDeviceRead ()))
+        {
+                if ((IX_ETH_PORT_1 == portId) &&
+                    (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_ETH0) ==
+                     IX_FEATURE_CTRL_COMPONENT_DISABLED))
+                {
+                   continue ;
+                }
+                if ((IX_ETH_PORT_2 == portId) &&
+                    (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_ETH1) ==
+                     IX_FEATURE_CTRL_COMPONENT_DISABLED))
+                {
+                    continue ;
+                }
+                if ((IX_ETH_PORT_3 == portId) &&
+                    (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEA_ETH) ==
+                     IX_FEATURE_CTRL_COMPONENT_DISABLED))
+                {
+                    continue ;
+                }
+        }
+
+	printf("PORT %u --------------------------------\n",
+	       portId);
+#ifndef NDEBUG
+	printf("Tx Done Frames                : %u\n",
+	       tx[portId].txDoneClientCallback +
+	       tx[portId].txDoneSwQDuringDisable +
+	       tx[portId].txDoneDuringDisable);
+	printf("Tx Frames                     : %u\n",
+	       tx[portId].txQOK + tx[portId].txQDelayed);
+	printf("Tx H/W Q Added OK             : %u\n",
+	       tx[portId].txQOK);
+	printf("Tx H/W Q Delayed              : %u\n",
+	       tx[portId].txQDelayed);
+	printf("Tx From S/W Q Added OK        : %u\n",
+	       tx[portId].txFromSwQOK);
+	printf("Tx From S/W Q Delayed         : %u\n",
+	       tx[portId].txFromSwQDelayed);
+	printf("Tx Overflow                   : %u\n",
+	       tx[portId].txOverflow);
+	printf("Tx Mutual Lock                : %u\n",
+	       tx[portId].txLock);
+	printf("Tx Late Ntf Enabled           : %u\n",
+	       tx[portId].txLateNotificationEnabled);
+	printf("Tx Low Thresh CB              : %u\n",
+	       tx[portId].txLowThreshCallback);
+	printf("Tx Done from H/W Q (Disable)  : %u\n",
+	       tx[portId].txDoneDuringDisable);
+	printf("Tx Done from S/W Q (Disable)  : %u\n",
+	       tx[portId].txDoneSwQDuringDisable);
+	for (priority = IX_ETH_ACC_TX_PRIORITY_0;
+	     priority <= IX_ETH_ACC_TX_PRIORITY_7;
+	     priority++)
+	{
+	    if (tx[portId].txPriority[priority])
+	    {
+		printf("Tx Priority %u                 : %u\n",
+		       priority,
+		       tx[portId].txPriority[priority]);
+	    }
+	}
+#endif
+	printf("Tx unexpected errors          : %u (should be 0)\n",
+	       tx[portId].txUnexpectedError);
+
+#ifndef NDEBUG
+	printf("Rx Frames                     : %u\n",
+	       rx[portId].rxFrameClientCallback +
+	       rx[portId].rxSwQDuringDisable+
+	       rx[portId].rxDuringDisable);
+	printf("Rx Free Replenish             : %u\n",
+	       rx[portId].rxFreeRepOK + rx[portId].rxFreeRepDelayed);
+	printf("Rx Free H/W Q Added OK        : %u\n",
+	       rx[portId].rxFreeRepOK);
+	printf("Rx Free H/W Q Delayed         : %u\n",
+	       rx[portId].rxFreeRepDelayed);
+	printf("Rx Free From S/W Q Added OK   : %u\n",
+	       rx[portId].rxFreeRepFromSwQOK);
+	printf("Rx Free From S/W Q Delayed    : %u\n",
+	       rx[portId].rxFreeRepFromSwQDelayed);
+	printf("Rx Free Overflow              : %u\n",
+	       rx[portId].rxFreeOverflow);
+	printf("Rx Free Mutual Lock           : %u\n",
+	       rx[portId].rxFreeLock);
+	printf("Rx Free Late Ntf Enabled      : %u\n",
+	       rx[portId].rxFreeLateNotificationEnabled);
+	printf("Rx Free Low CB                : %u\n",
+	       rx[portId].rxFreeLowCallback);
+	printf("Rx From H/W Q (Disable)       : %u\n",
+	       rx[portId].rxDuringDisable);
+	printf("Rx From S/W Q (Disable)       : %u\n",
+	       rx[portId].rxSwQDuringDisable);
+	printf("Rx unlearned Mac Address      : %u\n",
+	       rx[portId].rxUnlearnedMacAddress);
+        printf("Rx Filtered (Rx => RxFree)    : %u\n",
+            rx[portId].rxFiltered);
+
+	for (priority = IX_ETH_ACC_TX_PRIORITY_0;
+	     priority <= IX_ETH_ACC_TX_PRIORITY_7;
+	     priority++)
+	{
+	    if (rx[portId].rxPriority[priority])
+	    {
+		printf("Rx Priority %u                 : %u\n",
+		       priority,
+		       rx[portId].rxPriority[priority]);
+	    }
+	}
+#endif
+	printf("Rx unexpected errors          : %u (should be 0)\n",
+	       rx[portId].rxUnexpectedError);
+
+#ifndef NDEBUG
+	numBuffersInTx = tx[portId].txQOK +
+	    tx[portId].txQDelayed -
+	    tx[portId].txDoneClientCallback -
+	    tx[portId].txDoneSwQDuringDisable -
+	    tx[portId].txDoneDuringDisable;
+
+	printf("# Tx Buffers currently for transmission : %u\n",
+	       numBuffersInTx);
+
+	numBuffersInRx = rx[portId].rxFreeRepOK +
+	    rx[portId].rxFreeRepDelayed -
+	    rx[portId].rxFrameClientCallback -
+	    rx[portId].rxSwQDuringDisable -
+	    rx[portId].rxDuringDisable;
+
+	printf("# Rx Buffers currently for reception    : %u\n",
+	       numBuffersInRx);
+
+	totalBuffers += numBuffersInRx + numBuffersInTx;
+#endif
+    }
+
+    printf("---------------------------------------\n");
+
+#ifndef NDEBUG
+    printf("\n");
+    printf("Mbufs :\n");
+    printf("Tx Unchained mbufs            : %u\n",
+	   stats.unchainedTxMBufs);
+    printf("Tx Chained bufs               : %u\n",
+	   stats.chainedTxMBufs);
+    printf("TxDone Unchained mbufs        : %u\n",
+	   stats.unchainedTxDoneMBufs);
+    printf("TxDone Chained bufs           : %u\n",
+	   stats.chainedTxDoneMBufs);
+    printf("RxFree Unchained mbufs        : %u\n",
+	   stats.unchainedRxFreeMBufs);
+    printf("RxFree Chained bufs           : %u\n",
+	   stats.chainedRxFreeMBufs);
+    printf("Rx Unchained mbufs            : %u\n",
+	   stats.unchainedRxMBufs);
+    printf("Rx Chained bufs               : %u\n",
+	   stats.chainedRxMBufs);
+
+    printf("\n");
+    printf("Software queue usage :\n");
+    printf("Buffers added to S/W Q        : %u\n",
+	   stats.addToSwQ);
+    printf("Buffers removed from S/W Q    : %u\n",
+	   stats.removeFromSwQ);
+
+    printf("\n");
+    printf("Hardware queues callbacks :\n");
+
+    for(portId=0; portId < IX_ETH_ACC_NUMBER_OF_PORTS; portId++)
+    {
+	rxFreeCallbackCounter += rx[portId].rxFreeLowCallback;
+	txCallbackCounter += tx[portId].txLowThreshCallback;
+    }
+    printf("Tx Done QM Callback invoked   : %u\n",
+	   stats.txDoneCallbackCounter);
+    printf("Tx QM Callback invoked        : %u\n",
+	   txCallbackCounter);
+    printf("Rx QM Callback invoked        : %u\n",
+	   stats.rxCallbackCounter);
+    printf("Rx QM Callback burst read     : %u\n",
+	   stats.rxCallbackBurstRead);
+    printf("Rx Free QM Callback invoked   : %u\n",
+	   rxFreeCallbackCounter);
+#endif
+    printf("Unexpected errors in CB       : %u (should be 0)\n",
+	   stats.unexpectedError);
+    printf("\n");
+
+    printf("Hardware queues levels :\n");
+    printf("Transmit Port 1 Q             : %u \n",numTx0Entries);
+    printf("Transmit Port 2 Q             : %u \n",numTx1Entries);
+#ifdef __ixp46X
+    printf("Transmit Port 3 Q             : %u \n",numTx2Entries);
+#endif
+    printf("Transmit Done Q               : %u \n",numTxDoneEntries);
+    printf("Receive Q                     : %u \n",numRxEntries);
+    printf("Receive Free Port 1 Q         : %u \n",numRxFree0Entries);
+    printf("Receive Free Port 2 Q         : %u \n",numRxFree1Entries);
+#ifdef __ixp46X
+    printf("Receive Free Port 3 Q         : %u \n",numRxFree2Entries);
+#endif
+
+#ifndef NDEBUG
+    printf("\n");
+    printf("# Total Buffers accounted for : %u\n",
+	   totalBuffers);
+
+    numBuffersInSwQ = ixEthAccDataStats.addToSwQ -
+	ixEthAccDataStats.removeFromSwQ;
+
+    printf("    Buffers in S/W Qs         : %u\n",
+	   numBuffersInSwQ);
+    printf("    Buffers in H/W Qs or NPEs : %u\n",
+	   totalBuffers - numBuffersInSwQ);
+#endif
+
+    printf("Rx QoS Discipline             : %s\n",
+	   (ixEthAccDataInfo.schDiscipline ==
+	    FIFO_PRIORITY ) ? "Enabled" : "Disabled");
+
+    for(portId=0; portId < IX_ETH_ACC_NUMBER_OF_PORTS; portId++)
+    {
+	printf("Tx QoS Discipline port %u      : %s\n",
+	       portId,
+	       (ixEthAccPortData[portId].ixEthAccTxData.schDiscipline ==
+		FIFO_PRIORITY ) ? "Enabled" : "Disabled");
+    }
+    printf("\n");
+}
+
+
+
+
+

cpu/ixp/npe/IxEthAccMac.c

@@ -0,0 +1,2641 @@
+/**
+ * @file IxEthAccMac.c
+ *
+ * @author Intel Corporation
+ * @date
+ *
+ * @brief  MAC control functions
+ *
+ * Design Notes:
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxOsal.h"
+#include "IxNpeMh.h"
+#ifdef CONFIG_IXP425_COMPONENT_ETHDB
+#include "IxEthDB.h"
+#endif
+#include "IxEthDBPortDefs.h"
+#include "IxEthNpe.h"
+#include "IxEthAcc.h"
+#include "IxEthAccDataPlane_p.h"
+#include "IxEthAcc_p.h"
+#include "IxEthAccMac_p.h"
+
+/* Maximum number of retries during ixEthAccPortDisable, which
+ * is approximately 10 seconds
+*/
+#define IX_ETH_ACC_MAX_RETRY 500
+
+/* Maximum number of retries during ixEthAccPortDisable when expecting
+ * timeout
+ */
+#define IX_ETH_ACC_MAX_RETRY_TIMEOUT 5
+
+#define IX_ETH_ACC_VALIDATE_PORT_ID(portId) \
+    do                                                           \
+    {                                                            \
+        if(!IX_ETH_ACC_IS_PORT_VALID(portId))   \
+        {                                                        \
+	    return IX_ETH_ACC_INVALID_PORT;                      \
+        }                                                        \
+    } while(0)
+
+PUBLIC IxEthAccMacState ixEthAccMacState[IX_ETH_ACC_NUMBER_OF_PORTS];
+
+PRIVATE UINT32 ixEthAccMacBase[IX_ETH_ACC_NUMBER_OF_PORTS];
+
+/*Forward function declarations*/
+PRIVATE void
+ixEthAccPortDisableRx (IxEthAccPortId portId,
+		       IX_OSAL_MBUF * mBufPtr,
+		       BOOL useMultiBufferCallback);
+
+PRIVATE void
+ixEthAccPortDisableRxAndReplenish (IxEthAccPortId portId,
+				   IX_OSAL_MBUF * mBufPtr,
+				   BOOL useMultiBufferCallback);
+
+PRIVATE void
+ixEthAccPortDisableTxDone (UINT32 cbTag,
+			   IX_OSAL_MBUF *mbuf);
+
+PRIVATE void
+ixEthAccPortDisableTxDoneAndSubmit (UINT32 cbTag,
+				    IX_OSAL_MBUF *mbuf);
+
+PRIVATE void
+ixEthAccPortDisableRxCallback (UINT32 cbTag,
+			       IX_OSAL_MBUF * mBufPtr,
+			       UINT32 learnedPortId);
+
+PRIVATE void
+ixEthAccPortDisableMultiBufferRxCallback (UINT32 cbTag,
+					  IX_OSAL_MBUF **mBufPtr);
+
+PRIVATE IxEthAccStatus
+ixEthAccPortDisableTryTransmit(UINT32 portId);
+
+PRIVATE IxEthAccStatus
+ixEthAccPortDisableTryReplenish(UINT32 portId);
+
+PRIVATE IxEthAccStatus
+ixEthAccPortMulticastMacAddressGet (IxEthAccPortId portId,
+				    IxEthAccMacAddr *macAddr);
+
+PRIVATE IxEthAccStatus
+ixEthAccPortMulticastMacFilterGet (IxEthAccPortId portId,
+				   IxEthAccMacAddr *macAddr);
+
+PRIVATE void
+ixEthAccMacNpeStatsMessageCallback (IxNpeMhNpeId npeId,
+				    IxNpeMhMessage msg);
+
+PRIVATE void
+ixEthAccMacNpeStatsResetMessageCallback (IxNpeMhNpeId npeId,
+					 IxNpeMhMessage msg);
+
+PRIVATE void
+ixEthAccNpeLoopbackMessageCallback (IxNpeMhNpeId npeId,
+				    IxNpeMhMessage msg);
+
+PRIVATE void
+ixEthAccMulticastAddressSet(IxEthAccPortId portId);
+
+PRIVATE BOOL
+ixEthAccMacEqual(IxEthAccMacAddr *macAddr1,
+		 IxEthAccMacAddr *macAddr2);
+
+PRIVATE void
+ixEthAccMacPrint(IxEthAccMacAddr *m);
+
+PRIVATE void
+ixEthAccMacStateUpdate(IxEthAccPortId portId);
+
+IxEthAccStatus
+ixEthAccMacMemInit(void)
+{
+    ixEthAccMacBase[IX_ETH_PORT_1] =
+	(UINT32) IX_OSAL_MEM_MAP(IX_ETH_ACC_MAC_0_BASE,
+				 IX_OSAL_IXP400_ETHA_MAP_SIZE);
+    ixEthAccMacBase[IX_ETH_PORT_2] =
+	(UINT32) IX_OSAL_MEM_MAP(IX_ETH_ACC_MAC_1_BASE,
+				 IX_OSAL_IXP400_ETHB_MAP_SIZE);
+#ifdef __ixp46X
+    ixEthAccMacBase[IX_ETH_PORT_3] =
+	(UINT32) IX_OSAL_MEM_MAP(IX_ETH_ACC_MAC_2_BASE,
+				 IX_OSAL_IXP400_ETH_NPEA_MAP_SIZE);
+    if (ixEthAccMacBase[IX_ETH_PORT_3] == 0)
+    {
+	ixOsalLog(IX_OSAL_LOG_LVL_FATAL,
+		  IX_OSAL_LOG_DEV_STDOUT,
+		  "EthAcc: Could not map MAC I/O memory\n",
+		  0, 0, 0, 0, 0 ,0);
+
+	return IX_ETH_ACC_FAIL;
+    }
+#endif
+
+    if (ixEthAccMacBase[IX_ETH_PORT_1] == 0
+	|| ixEthAccMacBase[IX_ETH_PORT_2] == 0)
+    {
+	ixOsalLog(IX_OSAL_LOG_LVL_FATAL,
+		  IX_OSAL_LOG_DEV_STDOUT,
+		  "EthAcc: Could not map MAC I/O memory\n",
+		  0, 0, 0, 0, 0 ,0);
+
+	return IX_ETH_ACC_FAIL;
+    }
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+void
+ixEthAccMacUnload(void)
+{
+    IX_OSAL_MEM_UNMAP(ixEthAccMacBase[IX_ETH_PORT_1]);
+    IX_OSAL_MEM_UNMAP(ixEthAccMacBase[IX_ETH_PORT_2]);
+#ifdef __ixp46X
+    IX_OSAL_MEM_UNMAP(ixEthAccMacBase[IX_ETH_PORT_3]);
+    ixEthAccMacBase[IX_ETH_PORT_3] = 0;
+#endif
+    ixEthAccMacBase[IX_ETH_PORT_2] = 0;
+    ixEthAccMacBase[IX_ETH_PORT_1] = 0;
+}
+
+IxEthAccStatus
+ixEthAccPortEnablePriv(IxEthAccPortId portId)
+{
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot enable port.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+        printf("EthAcc: (Mac) cannot enable port %d, port not initialized\n", portId);
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    if (ixEthAccPortData[portId].ixEthAccTxData.txBufferDoneCallbackFn == NULL)
+    {
+        /* TxDone callback not registered */
+        printf("EthAcc: (Mac) cannot enable port %d, TxDone callback not registered\n", portId);
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    if ((ixEthAccPortData[portId].ixEthAccRxData.rxCallbackFn == NULL)
+	&& (ixEthAccPortData[portId].ixEthAccRxData.rxMultiBufferCallbackFn == NULL))
+    {
+        /* Receive callback not registered */
+        printf("EthAcc: (Mac) cannot enable port %d, Rx callback not registered\n", portId);
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    if(!ixEthAccMacState[portId].initDone)
+    {
+        printf("EthAcc: (Mac) cannot enable port %d, MAC address not set\n", portId);
+	return (IX_ETH_ACC_MAC_UNINITIALIZED);
+    }
+
+    /* if the state is being set to what it is already at, do nothing*/
+    if (ixEthAccMacState[portId].enabled)
+    {
+        return IX_ETH_ACC_SUCCESS;
+    }
+
+#ifdef CONFIG_IXP425_COMPONENT_ETHDB
+    /* enable ethernet database for this port */
+    if (ixEthDBPortEnable(portId) != IX_ETH_DB_SUCCESS)
+    {
+        printf("EthAcc: (Mac) cannot enable port %d, EthDB failure\n", portId);
+        return IX_ETH_ACC_FAIL;
+    }
+#endif
+
+    /* set the MAC core registers */
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_TX_CNTRL2,
+	      IX_ETH_ACC_TX_CNTRL2_RETRIES_MASK);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_RANDOM_SEED,
+	      IX_ETH_ACC_RANDOM_SEED_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_THRESH_P_EMPTY,
+	      IX_ETH_ACC_MAC_THRESH_P_EMPTY_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_THRESH_P_FULL,
+	      IX_ETH_ACC_MAC_THRESH_P_FULL_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_TX_DEFER,
+	      IX_ETH_ACC_MAC_TX_DEFER_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_TX_TWO_DEFER_1,
+	      IX_ETH_ACC_MAC_TX_TWO_DEFER_1_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_TX_TWO_DEFER_2,
+	      IX_ETH_ACC_MAC_TX_TWO_DEFER_2_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_SLOT_TIME,
+	      IX_ETH_ACC_MAC_SLOT_TIME_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_INT_CLK_THRESH,
+	      IX_ETH_ACC_MAC_INT_CLK_THRESH_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_BUF_SIZE_TX,
+	      IX_ETH_ACC_MAC_BUF_SIZE_TX_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_TX_CNTRL1,
+	      IX_ETH_ACC_TX_CNTRL1_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_RX_CNTRL1,
+	      IX_ETH_ACC_RX_CNTRL1_DEFAULT);
+
+    /* set the global state */
+    ixEthAccMacState[portId].portDisableState = ACTIVE;
+    ixEthAccMacState[portId].enabled = TRUE;
+
+    /* rewrite the setup (including mac filtering) depending
+     * on current options
+     */
+    ixEthAccMacStateUpdate(portId);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+/*
+ * PortDisable local variables. They contain the intermediate steps
+ * while the port is being disabled and the buffers being drained out
+ * of the NPE.
+ */
+typedef void (*IxEthAccPortDisableRx)(IxEthAccPortId portId,
+				      IX_OSAL_MBUF * mBufPtr,
+				      BOOL useMultiBufferCallback);
+static IxEthAccPortRxCallback
+ixEthAccPortDisableFn[IX_ETH_ACC_NUMBER_OF_PORTS];
+static IxEthAccPortMultiBufferRxCallback
+ixEthAccPortDisableMultiBufferFn[IX_ETH_ACC_NUMBER_OF_PORTS];
+static IxEthAccPortDisableRx
+ixEthAccPortDisableRxTable[IX_ETH_ACC_NUMBER_OF_PORTS];
+static UINT32
+ixEthAccPortDisableCbTag[IX_ETH_ACC_NUMBER_OF_PORTS];
+static UINT32
+ixEthAccPortDisableMultiBufferCbTag[IX_ETH_ACC_NUMBER_OF_PORTS];
+
+static IxEthAccPortTxDoneCallback
+ixEthAccPortDisableTxDoneFn[IX_ETH_ACC_NUMBER_OF_PORTS];
+static UINT32
+ixEthAccPortDisableTxDoneCbTag[IX_ETH_ACC_NUMBER_OF_PORTS];
+
+static UINT32
+ixEthAccPortDisableUserBufferCount[IX_ETH_ACC_NUMBER_OF_PORTS];
+
+/*
+ * PortDisable private callbacks functions. They handle the user
+ * traffic, and the special buffers (one for tx, one for rx) used
+ * in portDisable.
+ */
+PRIVATE void
+ixEthAccPortDisableTxDone(UINT32 cbTag,
+			  IX_OSAL_MBUF *mbuf)
+{
+    IxEthAccPortId portId = (IxEthAccPortId)cbTag;
+    volatile IxEthAccPortDisableState *txState = &ixEthAccMacState[portId].txState;
+
+    /* check for the special mbuf used in portDisable */
+    if (mbuf == ixEthAccMacState[portId].portDisableTxMbufPtr)
+    {
+        *txState = TRANSMIT_DONE;
+    }
+    else
+    {
+	/* increment the count of user traffic during portDisable */
+	ixEthAccPortDisableUserBufferCount[portId]++;
+
+       /* call client TxDone function */
+        ixEthAccPortDisableTxDoneFn[portId](ixEthAccPortDisableTxDoneCbTag[portId], mbuf);
+    }
+}
+
+PRIVATE IxEthAccStatus
+ixEthAccPortDisableTryTransmit(UINT32 portId)
+{
+    int key;
+    IxEthAccStatus status = IX_ETH_ACC_SUCCESS;
+    volatile IxEthAccPortDisableState *txState = &ixEthAccMacState[portId].txState;
+    /* transmit the special buffer again if it is transmitted
+     * and update the txState
+     * This section is protected because the portDisable context
+     * run an identical code, so the system keeps transmitting at the
+     * maximum rate.
+     */
+    key = ixOsalIrqLock();
+    if (*txState == TRANSMIT_DONE)
+    {
+	IX_OSAL_MBUF *mbufTxPtr = ixEthAccMacState[portId].portDisableTxMbufPtr;
+	*txState = TRANSMIT;
+	status = ixEthAccPortTxFrameSubmit(portId,
+					   mbufTxPtr,
+					   IX_ETH_ACC_TX_DEFAULT_PRIORITY);
+    }
+    ixOsalIrqUnlock(key);
+
+    return status;
+}
+
+PRIVATE void
+ixEthAccPortDisableTxDoneAndSubmit(UINT32 cbTag,
+				   IX_OSAL_MBUF *mbuf)
+{
+    IxEthAccPortId portId = (IxEthAccPortId)cbTag;
+
+    /* call the callback which forwards the traffic to the client */
+    ixEthAccPortDisableTxDone(cbTag, mbuf);
+
+    /* try to transmit the buffer used in portDisable
+     * if seen in TxDone
+     */
+    ixEthAccPortDisableTryTransmit(portId);
+}
+
+PRIVATE void
+ixEthAccPortDisableRx (IxEthAccPortId portId,
+		       IX_OSAL_MBUF * mBufPtr,
+		       BOOL useMultiBufferCallback)
+{
+    volatile IxEthAccPortDisableState *rxState = &ixEthAccMacState[portId].rxState;
+    IX_OSAL_MBUF *mNextPtr;
+
+    while (mBufPtr)
+    {
+	mNextPtr = IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(mBufPtr);
+	IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(mBufPtr) = NULL;
+
+	/* check for the special mbuf used in portDisable */
+	if (mBufPtr == ixEthAccMacState[portId].portDisableRxMbufPtr)
+	{
+            *rxState = RECEIVE;
+	}
+	else
+	{
+	    /* increment the count of user traffic during portDisable */
+	    ixEthAccPortDisableUserBufferCount[portId]++;
+
+	    /* reset the received payload length during portDisable */
+	    IX_OSAL_MBUF_MLEN(mBufPtr)    = 0;
+	    IX_OSAL_MBUF_PKT_LEN(mBufPtr) = 0;
+
+	    if (useMultiBufferCallback)
+	    {
+		/* call the user callback with one unchained
+		 * buffer, without payload. A small array is built
+		 * to be used as a parameter (the user callback expects
+		 * to receive an array ended by a NULL pointer.
+		 */
+		IX_OSAL_MBUF *mBufPtrArray[2];
+
+		mBufPtrArray[0] = mBufPtr;
+		mBufPtrArray[1] = NULL;
+		ixEthAccPortDisableMultiBufferFn[portId](
+			 ixEthAccPortDisableMultiBufferCbTag[portId],
+			 mBufPtrArray);
+	    }
+	    else
+	    {
+		/* call the user callback with a unchained
+		 * buffer, without payload and the destination port is
+		 * unknown.
+		 */
+		ixEthAccPortDisableFn[portId](
+		      ixEthAccPortDisableCbTag[portId],
+		      mBufPtr,
+		      IX_ETH_DB_UNKNOWN_PORT /* port not found */);
+	    }
+        }
+
+        mBufPtr = mNextPtr;
+    }
+}
+
+PRIVATE IxEthAccStatus
+ixEthAccPortDisableTryReplenish(UINT32 portId)
+{
+    int key;
+    IxEthAccStatus status = IX_ETH_ACC_SUCCESS;
+    volatile IxEthAccPortDisableState *rxState = &ixEthAccMacState[portId].rxState;
+    /* replenish with the special buffer again if it is received
+     * and update the rxState
+     * This section is protected because the portDisable context
+     * run an identical code, so the system keeps replenishing at the
+     * maximum rate.
+     */
+    key = ixOsalIrqLock();
+    if (*rxState == RECEIVE)
+    {
+	IX_OSAL_MBUF *mbufRxPtr = ixEthAccMacState[portId].portDisableRxMbufPtr;
+	*rxState = REPLENISH;
+	IX_OSAL_MBUF_MLEN(mbufRxPtr) = IX_ETHACC_RX_MBUF_MIN_SIZE;
+	status = ixEthAccPortRxFreeReplenish(portId, mbufRxPtr);
+    }
+    ixOsalIrqUnlock(key);
+
+    return status;
+}
+
+PRIVATE void
+ixEthAccPortDisableRxAndReplenish (IxEthAccPortId portId,
+				   IX_OSAL_MBUF * mBufPtr,
+				   BOOL useMultiBufferCallback)
+{
+    /* call the callback which forwards the traffic to the client */
+    ixEthAccPortDisableRx(portId, mBufPtr, useMultiBufferCallback);
+
+    /* try to replenish with the buffer used in portDisable
+     * if seen in Rx
+     */
+    ixEthAccPortDisableTryReplenish(portId);
+}
+
+PRIVATE void
+ixEthAccPortDisableRxCallback (UINT32 cbTag,
+			       IX_OSAL_MBUF * mBufPtr,
+			       UINT32 learnedPortId)
+{
+    IxEthAccPortId portId = (IxEthAccPortId)cbTag;
+
+    /* call the portDisable receive callback */
+   (ixEthAccPortDisableRxTable[portId])(portId, mBufPtr, FALSE);
+}
+
+PRIVATE void
+ixEthAccPortDisableMultiBufferRxCallback (UINT32 cbTag,
+					  IX_OSAL_MBUF **mBufPtr)
+{
+    IxEthAccPortId portId = (IxEthAccPortId)cbTag;
+
+    while (*mBufPtr)
+    {
+	/* call the portDisable receive callback with one buffer at a time */
+	(ixEthAccPortDisableRxTable[portId])(portId, *mBufPtr++, TRUE);
+    }
+}
+
+IxEthAccStatus
+ixEthAccPortDisablePriv(IxEthAccPortId portId)
+{
+    IxEthAccStatus status = IX_ETH_ACC_SUCCESS;
+    int key;
+    int retry, retryTimeout;
+    volatile IxEthAccPortDisableState *state = &ixEthAccMacState[portId].portDisableState;
+    volatile IxEthAccPortDisableState *rxState = &ixEthAccMacState[portId].rxState;
+    volatile IxEthAccPortDisableState *txState = &ixEthAccMacState[portId].txState;
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot disable port.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /* if the state is being set to what it is already at, do nothing */
+    if (!ixEthAccMacState[portId].enabled)
+    {
+        return IX_ETH_ACC_SUCCESS;
+    }
+
+    *state = DISABLED;
+
+    /* disable MAC receive first */
+    ixEthAccPortRxDisablePriv(portId);
+
+#ifdef CONFIG_IXP425_COMPONENT_ETHDB
+    /* disable ethernet database for this port - It is done now to avoid
+     * issuing ELT maintenance after requesting 'port disable' in an NPE
+     */
+    if (ixEthDBPortDisable(portId) != IX_ETH_DB_SUCCESS)
+    {
+	status = IX_ETH_ACC_FAIL;
+        IX_ETH_ACC_FATAL_LOG("ixEthAccPortDisable: failed to disable EthDB for this port\n", 0, 0, 0, 0, 0, 0);
+    }
+#endif
+
+    /* enter the critical section */
+    key = ixOsalIrqLock();
+
+    /* swap the Rx and TxDone callbacks */
+    ixEthAccPortDisableFn[portId]            = ixEthAccPortData[portId].ixEthAccRxData.rxCallbackFn;
+    ixEthAccPortDisableMultiBufferFn[portId] = ixEthAccPortData[portId].ixEthAccRxData.rxMultiBufferCallbackFn;
+    ixEthAccPortDisableCbTag[portId]         = ixEthAccPortData[portId].ixEthAccRxData.rxCallbackTag;
+    ixEthAccPortDisableMultiBufferCbTag[portId] = ixEthAccPortData[portId].ixEthAccRxData.rxMultiBufferCallbackTag;
+    ixEthAccPortDisableTxDoneFn[portId]      = ixEthAccPortData[portId].ixEthAccTxData.txBufferDoneCallbackFn;
+    ixEthAccPortDisableTxDoneCbTag[portId]   = ixEthAccPortData[portId].ixEthAccTxData.txCallbackTag;
+    ixEthAccPortDisableRxTable[portId]       =  ixEthAccPortDisableRx;
+
+    /* register temporary callbacks */
+    ixEthAccPortData[portId].ixEthAccRxData.rxCallbackFn            = ixEthAccPortDisableRxCallback;
+    ixEthAccPortData[portId].ixEthAccRxData.rxCallbackTag           = portId;
+
+    ixEthAccPortData[portId].ixEthAccRxData.rxMultiBufferCallbackFn = ixEthAccPortDisableMultiBufferRxCallback;
+    ixEthAccPortData[portId].ixEthAccRxData.rxMultiBufferCallbackTag = portId;
+
+    ixEthAccPortData[portId].ixEthAccTxData.txBufferDoneCallbackFn  = ixEthAccPortDisableTxDone;
+    ixEthAccPortData[portId].ixEthAccTxData.txCallbackTag           = portId;
+
+    /* initialise the Rx state and Tx states */
+    *txState = TRANSMIT_DONE;
+    *rxState = RECEIVE;
+
+    /* exit the critical section */
+    ixOsalIrqUnlock(key);
+
+    /* enable a NPE loopback */
+    if (ixEthAccNpeLoopbackEnablePriv(portId) != IX_ETH_ACC_SUCCESS)
+    {
+	status = IX_ETH_ACC_FAIL;
+    }
+
+    if (status == IX_ETH_ACC_SUCCESS)
+    {
+	retry = 0;
+
+	/* Step 1 : Drain Tx traffic and TxDone queues :
+	 *
+	 * Transmit and replenish at least once with the
+	 * special buffers until both of them are seen
+	 * in the callback hook
+	 *
+	 * (the receive callback keeps replenishing, so once we see
+	 * the special Tx buffer, we can be sure that Tx drain is complete)
+	 */
+	ixEthAccPortDisableRxTable[portId]
+	    =  ixEthAccPortDisableRxAndReplenish;
+	ixEthAccPortData[portId].ixEthAccTxData.txBufferDoneCallbackFn
+	    = ixEthAccPortDisableTxDone;
+
+	do
+	{
+	    /* keep replenishing */
+	    status = ixEthAccPortDisableTryReplenish(portId);
+	    if (status == IX_ETH_ACC_SUCCESS)
+	    {
+		/* keep transmitting */
+		status = ixEthAccPortDisableTryTransmit(portId);
+	    }
+	    if (status == IX_ETH_ACC_SUCCESS)
+	    {
+		/* wait for some traffic being processed */
+		ixOsalSleep(IX_ETH_ACC_PORT_DISABLE_DELAY_MSECS);
+	    }
+	}
+	while ((status == IX_ETH_ACC_SUCCESS)
+	       && (retry++ < IX_ETH_ACC_MAX_RETRY)
+	       && (*txState == TRANSMIT));
+
+	/* Step 2 : Drain Rx traffic, RxFree and Rx queues :
+	 *
+	 * Transmit and replenish at least once with the
+	 * special buffers until both of them are seen
+	 * in the callback hook
+	 * (the transmit callback keeps transmitting, and when we see
+	 * the special Rx buffer, we can be sure that rxFree drain
+	 * is complete)
+	 *
+	 * The nested loop helps to retry if the user was keeping
+	 * replenishing or transmitting during portDisable.
+	 *
+	 * The 2 nested loops ensure more retries if user traffic is
+	 * seen during portDisable : the user should not replenish
+	 * or transmit while portDisable is running. However, because of
+	 * the queueing possibilities in ethAcc dataplane, it is possible
+	 * that a lot of traffic is left in the queues (e.g. when
+	 * transmitting over a low speed link) and therefore, more
+	 * retries are allowed to help flushing the buffers out.
+	 */
+	ixEthAccPortDisableRxTable[portId]
+	    =  ixEthAccPortDisableRx;
+	ixEthAccPortData[portId].ixEthAccTxData.txBufferDoneCallbackFn
+	    = ixEthAccPortDisableTxDoneAndSubmit;
+
+	do
+	{
+	    do
+	    {
+		ixEthAccPortDisableUserBufferCount[portId] = 0;
+
+		/* keep replenishing */
+		status = ixEthAccPortDisableTryReplenish(portId);
+		if (status == IX_ETH_ACC_SUCCESS)
+		{
+		    /* keep transmitting */
+		    status = ixEthAccPortDisableTryTransmit(portId);
+		}
+		if (status == IX_ETH_ACC_SUCCESS)
+		{
+		    /* wait for some traffic being processed */
+		    ixOsalSleep(IX_ETH_ACC_PORT_DISABLE_DELAY_MSECS);
+		}
+	    }
+	    while ((status == IX_ETH_ACC_SUCCESS)
+		   && (retry++ < IX_ETH_ACC_MAX_RETRY)
+		   && ((ixEthAccPortDisableUserBufferCount[portId] != 0)
+		       || (*rxState == REPLENISH)));
+
+	    /* After the first iteration, change the receive callbacks,
+	     * to process only 1 buffer at a time
+	     */
+	    ixEthAccPortDisableRxTable[portId]
+		= ixEthAccPortDisableRx;
+	    ixEthAccPortData[portId].ixEthAccTxData.txBufferDoneCallbackFn
+		= ixEthAccPortDisableTxDone;
+
+	    /* repeat the whole process while user traffic is seen in TxDone
+	     *
+	     * The conditions to stop the loop are
+	     * - Xscale has both Rx and Tx special buffers
+	     *   (txState = transmit, rxState = receive)
+	     * - any error in txSubmit or rxReplenish
+	     * - no user traffic seen
+	     * - an excessive amount of retries
+	     */
+	}
+	while ((status == IX_ETH_ACC_SUCCESS)
+	       && (retry < IX_ETH_ACC_MAX_RETRY)
+	       && (*txState == TRANSMIT));
+
+	/* check the loop exit conditions. The NPE should not hold
+	 * the special buffers.
+	 */
+	if ((*rxState == REPLENISH) || (*txState == TRANSMIT))
+	{
+	    status = IX_ETH_ACC_FAIL;
+	}
+
+	if (status == IX_ETH_ACC_SUCCESS)
+	{
+	    /* Step 3 : Replenish without transmitting until a timeout
+	     * occurs, in order to drain the internal NPE fifos
+	     *
+	     * we can expect a few frames srill held
+	     * in the NPE.
+	     *
+	     * The 2 nested loops take care about the NPE dropping traffic
+	     * (including loopback traffic) when the Rx queue is full.
+	     *
+	     * The timeout value is very conservative
+	     * since the loopback used keeps replenishhing.
+	     *
+	     */
+	    do
+	    {
+		ixEthAccPortDisableRxTable[portId] = ixEthAccPortDisableRxAndReplenish;
+		ixEthAccPortDisableUserBufferCount[portId] = 0;
+		retryTimeout = 0;
+		do
+		{
+		    /* keep replenishing */
+		    status = ixEthAccPortDisableTryReplenish(portId);
+		    if (status == IX_ETH_ACC_SUCCESS)
+		    {
+			/* wait for some traffic being processed */
+			ixOsalSleep(IX_ETH_ACC_PORT_DISABLE_DELAY_MSECS);
+		    }
+		}
+		while ((status == IX_ETH_ACC_SUCCESS)
+		       && (retryTimeout++ < IX_ETH_ACC_MAX_RETRY_TIMEOUT));
+
+		/* Step 4 : Transmit once. Stop replenish
+		 *
+		 * After the Rx timeout, we are sure that the NPE does not
+		 * hold any frame in its internal NPE fifos.
+		 *
+		 * At this point, the NPE still holds the last rxFree buffer.
+		 * By transmitting a single frame, this should unblock the
+		 * last rxFree buffer. This code just transmit once and
+		 * wait for both frames seen in TxDone and in rxFree.
+		 *
+		 */
+		ixEthAccPortDisableRxTable[portId] =  ixEthAccPortDisableRx;
+		status = ixEthAccPortDisableTryTransmit(portId);
+
+		/* the NPE should immediatelyt release
+		 * the last Rx buffer and the last transmitted buffer
+		 * unless the last Tx frame was dropped (rx queue full)
+		 */
+		if (status == IX_ETH_ACC_SUCCESS)
+		{
+		    retryTimeout = 0;
+		    do
+		    {
+			ixOsalSleep(IX_ETH_ACC_PORT_DISABLE_DELAY_MSECS);
+		    }
+		    while ((*rxState == REPLENISH)
+			   && (retryTimeout++ < IX_ETH_ACC_MAX_RETRY_TIMEOUT));
+		}
+
+		/* the NPE may have dropped the traffic because of Rx
+		 * queue being full. This code ensures that the last
+		 * Tx and Rx frames are both received.
+		 */
+	    }
+	    while ((status == IX_ETH_ACC_SUCCESS)
+		   && (retry++ < IX_ETH_ACC_MAX_RETRY)
+		   && ((*txState == TRANSMIT)
+		       || (*rxState == REPLENISH)
+		       || (ixEthAccPortDisableUserBufferCount[portId] != 0)));
+
+	    /* Step 5 : check the final states : the NPE has
+	     * no buffer left, nor in Tx , nor in Rx directions.
+	     */
+	    if ((*rxState == REPLENISH) || (*txState == TRANSMIT))
+	    {
+		status = IX_ETH_ACC_FAIL;
+	    }
+	}
+
+        /* now all the buffers are drained, disable NPE loopback
+	 * This is done regardless of the logic to drain the queues and
+	 * the internal buffers held by the NPE.
+	 */
+	if (ixEthAccNpeLoopbackDisablePriv(portId) != IX_ETH_ACC_SUCCESS)
+	{
+	    status = IX_ETH_ACC_FAIL;
+	}
+    }
+
+    /* disable MAC Tx and Rx services */
+    ixEthAccMacState[portId].enabled = FALSE;
+    ixEthAccMacStateUpdate(portId);
+
+    /* restore the Rx and TxDone callbacks (within a critical section) */
+    key = ixOsalIrqLock();
+
+    ixEthAccPortData[portId].ixEthAccRxData.rxCallbackFn            = ixEthAccPortDisableFn[portId];
+    ixEthAccPortData[portId].ixEthAccRxData.rxCallbackTag           = ixEthAccPortDisableCbTag[portId];
+    ixEthAccPortData[portId].ixEthAccRxData.rxMultiBufferCallbackFn = ixEthAccPortDisableMultiBufferFn[portId];
+    ixEthAccPortData[portId].ixEthAccRxData.rxMultiBufferCallbackTag = ixEthAccPortDisableMultiBufferCbTag[portId];
+    ixEthAccPortData[portId].ixEthAccTxData.txBufferDoneCallbackFn  = ixEthAccPortDisableTxDoneFn[portId];
+    ixEthAccPortData[portId].ixEthAccTxData.txCallbackTag           = ixEthAccPortDisableTxDoneCbTag[portId];
+
+    ixOsalIrqUnlock(key);
+
+    /* the MAC core rx/tx disable may left the MAC hardware in an
+     * unpredictable state. A hw reset is executed before resetting
+     * all the MAC parameters to a known value.
+     */
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_CORE_CNTRL,
+	      IX_ETH_ACC_CORE_RESET);
+
+    ixOsalSleep(IX_ETH_ACC_MAC_RESET_DELAY);
+
+    /* rewrite all parameters to their current value */
+    ixEthAccMacStateUpdate(portId);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_INT_CLK_THRESH,
+	      IX_ETH_ACC_MAC_INT_CLK_THRESH_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_CORE_CNTRL,
+	      IX_ETH_ACC_CORE_MDC_EN);
+
+    return status;
+}
+
+IxEthAccStatus
+ixEthAccPortEnabledQueryPriv(IxEthAccPortId portId, BOOL *enabled)
+{
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot enable port.\n",(INT32)portId,0,0,0,0,0);
+
+        /* Since Eth NPE is not available, port must be disabled */
+        *enabled = FALSE ;
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+        /* Since Eth NPE is not available, port must be disabled */
+        *enabled = FALSE ;
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    *enabled = ixEthAccMacState[portId].enabled;
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortMacResetPriv(IxEthAccPortId portId)
+{
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Eth %d: Cannot reset Ethernet coprocessor.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_CORE_CNTRL,
+	      IX_ETH_ACC_CORE_RESET);
+
+    ixOsalSleep(IX_ETH_ACC_MAC_RESET_DELAY);
+
+    /* rewrite all parameters to their current value */
+    ixEthAccMacStateUpdate(portId);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_INT_CLK_THRESH,
+	      IX_ETH_ACC_MAC_INT_CLK_THRESH_DEFAULT);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_CORE_CNTRL,
+	      IX_ETH_ACC_CORE_MDC_EN);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortLoopbackEnable(IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    /* Turn off promiscuous mode */
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Eth %d: Cannot enable loopback.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /* read register */
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_RX_CNTRL1,
+	     regval);
+
+    /* update register */
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_RX_CNTRL1,
+	      regval | IX_ETH_ACC_RX_CNTRL1_LOOP_EN);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+PRIVATE void
+ixEthAccNpeLoopbackMessageCallback (IxNpeMhNpeId npeId,
+				    IxNpeMhMessage msg)
+{
+    IxEthAccPortId portId = IX_ETH_ACC_NPE_TO_PORT_ID(npeId);
+
+#ifndef NDEBUG
+    /* Prudent to at least check the port is within range */
+    if (portId >= IX_ETH_ACC_NUMBER_OF_PORTS)
+    {
+	IX_ETH_ACC_FATAL_LOG("IXETHACC:ixEthAccPortDisableMessageCallback: Illegal port: %u\n",
+            (UINT32) portId, 0, 0, 0, 0, 0);
+
+	return;
+    }
+#endif
+
+    /* unlock message reception mutex */
+    ixOsalMutexUnlock(&ixEthAccMacState[portId].npeLoopbackMessageLock);
+}
+
+IxEthAccStatus
+ixEthAccNpeLoopbackEnablePriv(IxEthAccPortId portId)
+{
+    IX_STATUS npeMhStatus;
+    IxNpeMhMessage message;
+    IxEthAccStatus status = IX_ETH_ACC_SUCCESS;
+
+    /* Turn off promiscuous mode */
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Eth %d: Cannot enable NPE loopback.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /* enable NPE loopback (lsb of the message contains the value 1) */
+    message.data[0] = (IX_ETHNPE_SETLOOPBACK_MODE << IX_ETH_ACC_MAC_MSGID_SHL)
+	| 0x01;
+    message.data[1] = 0;
+
+    npeMhStatus = ixNpeMhMessageWithResponseSend(IX_ETH_ACC_PORT_TO_NPE_ID(portId),
+		message,
+		IX_ETHNPE_SETLOOPBACK_MODE_ACK,
+		ixEthAccNpeLoopbackMessageCallback,
+		IX_NPEMH_SEND_RETRIES_DEFAULT);
+
+    if (npeMhStatus != IX_SUCCESS)
+    {
+        status = IX_ETH_ACC_FAIL;
+    }
+    else
+    {
+	/* wait for NPE loopbackEnable response */
+        if (ixOsalMutexLock(&ixEthAccMacState[portId]. npeLoopbackMessageLock,
+			    IX_ETH_ACC_PORT_DISABLE_DELAY_MSECS)
+	    != IX_SUCCESS)
+        {
+            status = IX_ETH_ACC_FAIL;
+        }
+    }
+
+    return status;
+}
+
+IxEthAccStatus
+ixEthAccPortTxEnablePriv(IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    /* Turn off promiscuous mode */
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Eth %d: Cannot enable TX.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /* read register */
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_TX_CNTRL1,
+	     regval);
+
+    /* update register */
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_TX_CNTRL1,
+	      regval | IX_ETH_ACC_TX_CNTRL1_TX_EN);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortRxEnablePriv(IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    /* Turn off promiscuous mode */
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Eth %d: Cannot enable RX.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /* read register */
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_RX_CNTRL1,
+	     regval);
+
+    /* update register */
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_RX_CNTRL1,
+	      regval | IX_ETH_ACC_RX_CNTRL1_RX_EN);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortLoopbackDisable(IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Eth %d: Cannot disable loopback.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /*disable MAC loopabck */
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_RX_CNTRL1,
+	     regval);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_RX_CNTRL1,
+	      (regval & ~IX_ETH_ACC_RX_CNTRL1_LOOP_EN));
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccNpeLoopbackDisablePriv(IxEthAccPortId portId)
+{
+    IX_STATUS npeMhStatus;
+    IxNpeMhMessage message;
+    IxEthAccStatus status = IX_ETH_ACC_SUCCESS;
+
+    /* Turn off promiscuous mode */
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Eth %d: Cannot enable NPE loopback.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /* disable NPE loopback (lsb of the message contains the value 0) */
+    message.data[0] = (IX_ETHNPE_SETLOOPBACK_MODE << IX_ETH_ACC_MAC_MSGID_SHL);
+    message.data[1] = 0;
+
+    npeMhStatus = ixNpeMhMessageWithResponseSend(IX_ETH_ACC_PORT_TO_NPE_ID(portId),
+		message,
+		IX_ETHNPE_SETLOOPBACK_MODE_ACK,
+		ixEthAccNpeLoopbackMessageCallback,
+		IX_NPEMH_SEND_RETRIES_DEFAULT);
+
+    if (npeMhStatus != IX_SUCCESS)
+    {
+        status = IX_ETH_ACC_FAIL;
+    }
+    else
+    {
+	/* wait for NPE loopbackEnable response */
+        if (ixOsalMutexLock(&ixEthAccMacState[portId].npeLoopbackMessageLock,
+			    IX_ETH_ACC_PORT_DISABLE_DELAY_MSECS)
+	    != IX_SUCCESS)
+        {
+            status = IX_ETH_ACC_FAIL;
+        }
+    }
+
+    return status;
+}
+
+IxEthAccStatus
+ixEthAccPortTxDisablePriv(IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    /* Turn off promiscuous mode */
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Eth %d: Cannot disable TX.\n", (INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /* read register */
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_TX_CNTRL1,
+	     regval);
+
+    /* update register */
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_TX_CNTRL1,
+	      (regval & ~IX_ETH_ACC_TX_CNTRL1_TX_EN));
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortRxDisablePriv(IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    /* Turn off promiscuous mode */
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Eth %d: Cannot disable RX.\n", (INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /* read register */
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_RX_CNTRL1,
+	     regval);
+
+    /* update register */
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_RX_CNTRL1,
+	      (regval & ~IX_ETH_ACC_RX_CNTRL1_RX_EN));
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortPromiscuousModeClearPriv(IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    /* Turn off promiscuous mode */
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot clear promiscuous mode.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /*set bit 5 of Rx control 1 - enable address filtering*/
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_RX_CNTRL1,
+	     regval);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_RX_CNTRL1,
+	      regval | IX_ETH_ACC_RX_CNTRL1_ADDR_FLTR_EN);
+
+    ixEthAccMacState[portId].promiscuous = FALSE;
+
+    ixEthAccMulticastAddressSet(portId);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortPromiscuousModeSetPriv(IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot set promiscuous mode.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /*
+     * Set bit 5 of Rx control 1 - We enable address filtering even in
+     * promiscuous mode because we want the MAC to set the appropriate
+     * bits in m_flags which doesn't happen if we turn off filtering.
+     */
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_RX_CNTRL1,
+	     regval);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_RX_CNTRL1,
+	      regval | IX_ETH_ACC_RX_CNTRL1_ADDR_FLTR_EN);
+
+    ixEthAccMacState[portId].promiscuous = TRUE;
+
+    ixEthAccMulticastAddressSet(portId);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortUnicastMacAddressSetPriv (IxEthAccPortId portId,
+				  IxEthAccMacAddr *macAddr)
+{
+    UINT32 i;
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot set Unicast Mac Address.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+
+    if (macAddr == NULL)
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+    if ( macAddr->macAddress[0] & IX_ETH_ACC_ETH_MAC_BCAST_MCAST_BIT )
+    {
+	/* This is a multicast/broadcast address cant set it ! */
+	return IX_ETH_ACC_FAIL;
+    }
+
+    if ( macAddr->macAddress[0] == 0 &&
+	 macAddr->macAddress[1] == 0 &&
+	 macAddr->macAddress[2] == 0 &&
+	 macAddr->macAddress[3] == 0 &&
+	 macAddr->macAddress[4] == 0 &&
+	 macAddr->macAddress[5] == 0  )
+    {
+	/* This is an invalid mac address cant set it ! */
+	return IX_ETH_ACC_FAIL;
+    }
+
+#ifdef CONFIG_IXP425_COMPONENT_ETHDB
+    /* update the MAC address in the ethernet database */
+    if (ixEthDBPortAddressSet(portId, (IxEthDBMacAddr *) macAddr) != IX_ETH_DB_SUCCESS)
+    {
+        return IX_ETH_ACC_FAIL;
+    }
+#endif
+
+    /*Set the Unicast MAC to the specified value*/
+    for(i=0;i<IX_IEEE803_MAC_ADDRESS_SIZE;i++)
+    {
+	REG_WRITE(ixEthAccMacBase[portId],
+		  IX_ETH_ACC_MAC_UNI_ADDR_1 + i*sizeof(UINT32),
+		  macAddr->macAddress[i]);
+    }
+    ixEthAccMacState[portId].initDone = TRUE;
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortUnicastMacAddressGetPriv (IxEthAccPortId portId,
+				  IxEthAccMacAddr *macAddr)
+{
+    /*Return the current value of the Unicast MAC from h/w
+      for the specified port*/
+    UINT32 i;
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot get Unicast Mac Address.\n",(INT32)portId,0,0,0,0,0);
+        /* Since Eth Npe is unavailable, return invalid MAC Address = 00:00:00:00:00:00 */
+        for(i=0;i<IX_IEEE803_MAC_ADDRESS_SIZE;i++)
+        {
+	    macAddr->macAddress[i] = 0;
+        }
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if(!ixEthAccMacState[portId].initDone)
+    {
+	return (IX_ETH_ACC_MAC_UNINITIALIZED);
+    }
+
+    if (macAddr == NULL)
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+
+    for(i=0;i<IX_IEEE803_MAC_ADDRESS_SIZE;i++)
+    {
+	REG_READ(ixEthAccMacBase[portId],
+		 IX_ETH_ACC_MAC_UNI_ADDR_1 + i*sizeof(UINT32),
+		 macAddr->macAddress[i]);
+    }
+    return IX_ETH_ACC_SUCCESS;
+}
+
+PRIVATE IxEthAccStatus
+ixEthAccPortMulticastMacAddressGet (IxEthAccPortId portId,
+				    IxEthAccMacAddr *macAddr)
+{
+    /*Return the current value of the Multicast MAC from h/w
+      for the specified port*/
+    UINT32 i;
+
+    for(i=0;i<IX_IEEE803_MAC_ADDRESS_SIZE;i++)
+    {
+
+	REG_READ(ixEthAccMacBase[portId],
+		 IX_ETH_ACC_MAC_ADDR_1 + i*sizeof(UINT32),
+		 macAddr->macAddress[i]);
+    }
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+PRIVATE IxEthAccStatus
+ixEthAccPortMulticastMacFilterGet (IxEthAccPortId portId,
+				   IxEthAccMacAddr *macAddr)
+{
+    /*Return the current value of the Multicast MAC from h/w
+      for the specified port*/
+    UINT32 i;
+
+    for(i=0;i<IX_IEEE803_MAC_ADDRESS_SIZE;i++)
+    {
+
+	REG_READ(ixEthAccMacBase[portId],
+		 IX_ETH_ACC_MAC_ADDR_MASK_1 + i*sizeof(UINT32),
+		 macAddr->macAddress[i]);
+    }
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortMulticastAddressJoinPriv (IxEthAccPortId portId,
+				  IxEthAccMacAddr *macAddr)
+{
+    UINT32 i;
+    IxEthAccMacAddr broadcastAddr = {{0xff,0xff,0xff,0xff,0xff,0xff}};
+
+    /*Check that the port parameter is valid*/
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot join Multicast Mac Address.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /*Check that the mac address is valid*/
+    if(macAddr == NULL)
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+    /* Check that this is a multicast address */
+    if (!(macAddr->macAddress[0] & IX_ETH_ACC_ETH_MAC_BCAST_MCAST_BIT))
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+    /* We don't add the Broadcast address */
+    if(ixEthAccMacEqual(&broadcastAddr, macAddr))
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+    for (i = 0;
+	 i<ixEthAccMacState[portId].mcastAddrIndex;
+	 i++)
+    {
+	/*Check if the current entry already match an existing matches*/
+	if(ixEthAccMacEqual(&ixEthAccMacState[portId].mcastAddrsTable[i], macAddr))
+	{
+	    /* Address found in the list and already configured,
+	     * return a success status
+	     */
+	    return IX_ETH_ACC_SUCCESS;
+	}
+    }
+
+    /* check for availability at the end of the current table */
+    if(ixEthAccMacState[portId].mcastAddrIndex >= IX_ETH_ACC_MAX_MULTICAST_ADDRESSES)
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+    /*First add the address to the multicast table for the
+      specified port*/
+    i=ixEthAccMacState[portId].mcastAddrIndex;
+
+    memcpy(&ixEthAccMacState[portId].mcastAddrsTable[i],
+	   &macAddr->macAddress,
+	   IX_IEEE803_MAC_ADDRESS_SIZE);
+
+    /*Increment the index into the table, this must be done here
+     as MulticastAddressSet below needs to know about the latest
+     entry.
+    */
+    ixEthAccMacState[portId].mcastAddrIndex++;
+
+    /*Then calculate the new value to be written to the address and
+      address mask registers*/
+    ixEthAccMulticastAddressSet(portId);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+
+IxEthAccStatus
+ixEthAccPortMulticastAddressJoinAllPriv (IxEthAccPortId portId)
+{
+    IxEthAccMacAddr mcastMacAddr = {{0x1,0x0,0x0,0x0,0x0,0x0}};
+
+    /*Check that the port parameter is valid*/
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot join all Multicast Address.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /* remove all entries from the database and
+    *  insert a multicast entry
+    */
+    memcpy(&ixEthAccMacState[portId].mcastAddrsTable[0],
+	   &mcastMacAddr.macAddress,
+	   IX_IEEE803_MAC_ADDRESS_SIZE);
+
+    ixEthAccMacState[portId].mcastAddrIndex = 1;
+    ixEthAccMacState[portId].joinAll = TRUE;
+
+    ixEthAccMulticastAddressSet(portId);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortMulticastAddressLeavePriv (IxEthAccPortId portId,
+				   IxEthAccMacAddr *macAddr)
+{
+    UINT32 i;
+    IxEthAccMacAddr mcastMacAddr = {{0x1,0x0,0x0,0x0,0x0,0x0}};
+
+    /*Check that the port parameter is valid*/
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot leave Multicast Address.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /*Check that the mac address is valid*/
+    if(macAddr == NULL)
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+    /* Remove this mac address from the mask for the specified port
+     * we copy down all entries above the blanked entry, and
+     * decrement the index
+     */
+    i=0;
+
+    while(i<ixEthAccMacState[portId].mcastAddrIndex)
+    {
+	/*Check if the current entry matches*/
+	if(ixEthAccMacEqual(&ixEthAccMacState[portId].mcastAddrsTable[i],
+			    macAddr))
+	{
+	    if(ixEthAccMacEqual(macAddr, &mcastMacAddr))
+	    {
+		ixEthAccMacState[portId].joinAll = FALSE;
+	    }
+	    /*Decrement the index into the multicast address table
+	      for the current port*/
+	    ixEthAccMacState[portId].mcastAddrIndex--;
+
+	    /*Copy down all entries above the current entry*/
+	    while(i<ixEthAccMacState[portId].mcastAddrIndex)
+	    {
+		memcpy(&ixEthAccMacState[portId].mcastAddrsTable[i],
+		       &ixEthAccMacState[portId].mcastAddrsTable[i+1],
+		       IX_IEEE803_MAC_ADDRESS_SIZE);
+                i++;
+	    }
+	    /*recalculate the mask and write it to the MAC*/
+	    ixEthAccMulticastAddressSet(portId);
+
+	    return IX_ETH_ACC_SUCCESS;
+	}
+	/* search the next entry */
+	i++;
+    }
+    /* no matching entry found */
+    return IX_ETH_ACC_NO_SUCH_ADDR;
+}
+
+IxEthAccStatus
+ixEthAccPortMulticastAddressLeaveAllPriv (IxEthAccPortId portId)
+{
+    /*Check that the port parameter is valid*/
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot leave all Multicast Address.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    ixEthAccMacState[portId].mcastAddrIndex = 0;
+    ixEthAccMacState[portId].joinAll = FALSE;
+
+    ixEthAccMulticastAddressSet(portId);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+
+IxEthAccStatus
+ixEthAccPortUnicastAddressShowPriv (IxEthAccPortId portId)
+{
+    IxEthAccMacAddr macAddr;
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot show Unicast Address.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    /*Get the MAC (UINICAST) address from hardware*/
+    if(ixEthAccPortUnicastMacAddressGetPriv(portId, &macAddr) != IX_ETH_ACC_SUCCESS)
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: MAC address uninitialised port %u\n",
+			       (INT32)portId,0,0,0,0,0);
+	return IX_ETH_ACC_MAC_UNINITIALIZED;
+    }
+
+    /*print it out*/
+    ixEthAccMacPrint(&macAddr);
+    printf("\n");
+    return IX_ETH_ACC_SUCCESS;
+}
+
+
+
+void
+ixEthAccPortMulticastAddressShowPriv(IxEthAccPortId portId)
+{
+    IxEthAccMacAddr macAddr;
+    UINT32 i;
+
+    if(!IX_ETH_ACC_IS_PORT_VALID(portId))
+    {
+        return;
+    }
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot show Multicast Address.\n",(INT32)portId,0,0,0,0,0);
+        return ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return;
+    }
+
+    printf("Multicast MAC: ");
+    /*Get the MAC (MULTICAST) address from hardware*/
+    ixEthAccPortMulticastMacAddressGet(portId, &macAddr);
+    /*print it out*/
+    ixEthAccMacPrint(&macAddr);
+    /*Get the MAC (MULTICAST) filter from hardware*/
+    ixEthAccPortMulticastMacFilterGet(portId, &macAddr);
+    /*print it out*/
+    printf(" ( ");
+    ixEthAccMacPrint(&macAddr);
+    printf(" )\n");
+    printf("Constituent Addresses:\n");
+    for(i=0;i<ixEthAccMacState[portId].mcastAddrIndex;i++)
+    {
+	ixEthAccMacPrint(&ixEthAccMacState[portId].mcastAddrsTable[i]);
+	printf("\n");
+    }
+    return;
+}
+
+/*Set the duplex mode*/
+IxEthAccStatus
+ixEthAccPortDuplexModeSetPriv (IxEthAccPortId portId,
+			   IxEthAccDuplexMode mode)
+{
+    UINT32 txregval;
+    UINT32 rxregval;
+
+    /*This is bit 1 of the transmit control reg, set to 1 for half
+      duplex, 0 for full duplex*/
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot set Duplex Mode.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_TX_CNTRL1,
+	     txregval);
+
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_RX_CNTRL1,
+	     rxregval);
+
+    if (mode ==  IX_ETH_ACC_FULL_DUPLEX)
+    {
+	/*Clear half duplex bit in TX*/
+	REG_WRITE(ixEthAccMacBase[portId],
+		  IX_ETH_ACC_MAC_TX_CNTRL1,
+		  txregval & ~IX_ETH_ACC_TX_CNTRL1_DUPLEX);
+
+	/*We must set the pause enable in the receive logic when in
+	  full duplex mode*/
+	REG_WRITE(ixEthAccMacBase[portId],
+		  IX_ETH_ACC_MAC_RX_CNTRL1,
+		  rxregval | IX_ETH_ACC_RX_CNTRL1_PAUSE_EN);
+	ixEthAccMacState[portId].fullDuplex = TRUE;
+
+    }
+    else if (mode ==  IX_ETH_ACC_HALF_DUPLEX)
+    {
+	/*Set half duplex bit in TX*/
+	REG_WRITE(ixEthAccMacBase[portId],
+		  IX_ETH_ACC_MAC_TX_CNTRL1,
+		  txregval | IX_ETH_ACC_TX_CNTRL1_DUPLEX);
+
+	/*We must clear pause enable in the receive logic when in
+	  half duplex mode*/
+	REG_WRITE(ixEthAccMacBase[portId],
+		  IX_ETH_ACC_MAC_RX_CNTRL1,
+		  rxregval & ~IX_ETH_ACC_RX_CNTRL1_PAUSE_EN);
+
+	ixEthAccMacState[portId].fullDuplex = FALSE;
+    }
+    else
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+
+    return IX_ETH_ACC_SUCCESS;
+
+}
+
+
+
+IxEthAccStatus
+ixEthAccPortDuplexModeGetPriv (IxEthAccPortId portId,
+			   IxEthAccDuplexMode *mode)
+{
+    /*Return the duplex mode for the specified port*/
+    UINT32 regval;
+
+    /*This is bit 1 of the transmit control reg, set to 1 for half
+      duplex, 0 for full duplex*/
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot get Duplex Mode.\n",(INT32)portId,0,0,0,0,0);
+        /* return hald duplex */
+        *mode = IX_ETH_ACC_HALF_DUPLEX ;
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    if (mode == NULL)
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_TX_CNTRL1,
+	     regval);
+
+    if( regval & IX_ETH_ACC_TX_CNTRL1_DUPLEX)
+    {
+	*mode = IX_ETH_ACC_HALF_DUPLEX;
+    }
+    else
+    {
+	*mode = IX_ETH_ACC_FULL_DUPLEX;
+    }
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+
+
+IxEthAccStatus
+ixEthAccPortTxFrameAppendPaddingEnablePriv (IxEthAccPortId portId)
+{
+    UINT32 regval;
+    /*Enable FCS computation by the MAC and appending to the
+      frame*/
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot enable Tx Frame Append Padding.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_TX_CNTRL1,
+	     regval);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_TX_CNTRL1,
+	      regval |
+	      IX_ETH_ACC_TX_CNTRL1_PAD_EN);
+
+    ixEthAccMacState[portId].txPADAppend = TRUE;
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortTxFrameAppendPaddingDisablePriv (IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    /*disable FCS computation and appending*/
+    /*Set bit 4 of Tx control register one to zero*/
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot disble Tx Frame Append Padding.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_TX_CNTRL1,
+	     regval);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_TX_CNTRL1,
+	      regval & ~IX_ETH_ACC_TX_CNTRL1_PAD_EN);
+
+    ixEthAccMacState[portId].txPADAppend = FALSE;
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortTxFrameAppendFCSEnablePriv (IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    /*Enable FCS computation by the MAC and appending to the
+      frame*/
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot enable Tx Frame Append FCS.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_TX_CNTRL1,
+	     regval);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_TX_CNTRL1,
+	      regval | IX_ETH_ACC_TX_CNTRL1_FCS_EN);
+
+    ixEthAccMacState[portId].txFCSAppend = TRUE;
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortTxFrameAppendFCSDisablePriv (IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    /*disable FCS computation and appending*/
+    /*Set bit 4 of Tx control register one to zero*/
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot disable Tx Frame Append FCS.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_TX_CNTRL1,
+	     regval);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_TX_CNTRL1,
+	      regval & ~IX_ETH_ACC_TX_CNTRL1_FCS_EN);
+
+    ixEthAccMacState[portId].txFCSAppend = FALSE;
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortRxFrameAppendFCSEnablePriv (IxEthAccPortId portId)
+{
+    /*Set bit 2 of Rx control 1*/
+    UINT32 regval;
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot enable Rx Frame Append FCS.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_RX_CNTRL1,
+	     regval);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_RX_CNTRL1,
+	      regval | IX_ETH_ACC_RX_CNTRL1_CRC_EN);
+
+    ixEthAccMacState[portId].rxFCSAppend = TRUE;
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccPortRxFrameAppendFCSDisablePriv (IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    /*Clear bit 2 of Rx control 1*/
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot disable Rx Frame Append FCS.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    REG_READ(ixEthAccMacBase[portId],
+	     IX_ETH_ACC_MAC_RX_CNTRL1,
+	     regval);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_RX_CNTRL1,
+	      regval & ~IX_ETH_ACC_RX_CNTRL1_CRC_EN);
+
+    ixEthAccMacState[portId].rxFCSAppend = FALSE;
+    return IX_ETH_ACC_SUCCESS;
+}
+
+
+
+PRIVATE void
+ixEthAccMacNpeStatsMessageCallback (IxNpeMhNpeId npeId,
+				    IxNpeMhMessage msg)
+{
+    IxEthAccPortId portId = IX_ETH_ACC_NPE_TO_PORT_ID(npeId);
+
+#ifndef NDEBUG
+    /* Prudent to at least check the port is within range */
+    if (portId >= IX_ETH_ACC_NUMBER_OF_PORTS)
+    {
+	IX_ETH_ACC_FATAL_LOG(
+     "IXETHACC:ixEthAccMacNpeStatsMessageCallback: Illegal port: %u\n",
+     (UINT32)portId, 0, 0, 0, 0, 0);
+	return;
+    }
+#endif
+
+    /*Unblock Stats Get call*/
+    ixOsalMutexUnlock(&ixEthAccMacState[portId].ackMIBStatsLock);
+
+}
+
+PRIVATE void
+ixEthAccMibIIStatsEndianConvert (IxEthEthObjStats *retStats)
+{
+    /* endianness conversion */
+
+    /* Rx stats */
+    retStats->dot3StatsAlignmentErrors =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->dot3StatsAlignmentErrors);
+    retStats->dot3StatsFCSErrors =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->dot3StatsFCSErrors);
+    retStats->dot3StatsInternalMacReceiveErrors =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->dot3StatsInternalMacReceiveErrors);
+    retStats->RxOverrunDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->RxOverrunDiscards);
+    retStats->RxLearnedEntryDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->RxLearnedEntryDiscards);
+    retStats->RxLargeFramesDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->RxLargeFramesDiscards);
+    retStats->RxSTPBlockedDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->RxSTPBlockedDiscards);
+    retStats->RxVLANTypeFilterDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->RxVLANTypeFilterDiscards);
+    retStats->RxVLANIdFilterDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->RxVLANIdFilterDiscards);
+    retStats->RxInvalidSourceDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->RxInvalidSourceDiscards);
+    retStats->RxBlackListDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->RxBlackListDiscards);
+    retStats->RxWhiteListDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->RxWhiteListDiscards);
+    retStats->RxUnderflowEntryDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->RxUnderflowEntryDiscards);
+
+    /* Tx stats */
+    retStats->dot3StatsSingleCollisionFrames =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->dot3StatsSingleCollisionFrames);
+    retStats->dot3StatsMultipleCollisionFrames =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->dot3StatsMultipleCollisionFrames);
+    retStats->dot3StatsDeferredTransmissions =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->dot3StatsDeferredTransmissions);
+    retStats->dot3StatsLateCollisions =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->dot3StatsLateCollisions);
+    retStats->dot3StatsExcessiveCollsions =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->dot3StatsExcessiveCollsions);
+    retStats->dot3StatsInternalMacTransmitErrors =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->dot3StatsInternalMacTransmitErrors);
+    retStats->dot3StatsCarrierSenseErrors =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->dot3StatsCarrierSenseErrors);
+    retStats->TxLargeFrameDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->TxLargeFrameDiscards);
+    retStats->TxVLANIdFilterDiscards =
+	IX_OSAL_SWAP_BE_SHARED_LONG(retStats->TxVLANIdFilterDiscards);
+}
+
+IxEthAccStatus
+ixEthAccMibIIStatsGet (IxEthAccPortId portId,
+		       IxEthEthObjStats *retStats )
+{
+    IxNpeMhMessage message;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+        printf("EthAcc: ixEthAccMibIIStatsGet (Mac) EthAcc service is not initialized\n");
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (retStats == NULL)
+    {
+        printf("EthAcc: ixEthAccMibIIStatsGet (Mac) NULL argument\n");
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        printf("EthAcc: ixEthAccMibIIStatsGet (Mac) NPE for port %d is not available\n", portId);
+
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot get MIB II Stats.\n",(INT32)portId,0,0,0,0,0);
+
+        /* Return all zero stats */
+        IX_ETH_ACC_MEMSET(retStats, 0, sizeof(IxEthEthObjStats));
+
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+        printf("EthAcc: ixEthAccMibIIStatsGet (Mac) port %d is not initialized\n", portId);
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    IX_OSAL_CACHE_INVALIDATE(retStats, sizeof(IxEthEthObjStats));
+
+    message.data[0] = IX_ETHNPE_GETSTATS << IX_ETH_ACC_MAC_MSGID_SHL;
+    message.data[1] = (UINT32) IX_OSAL_MMU_VIRT_TO_PHYS(retStats);
+
+    /* Permit only one task to request MIB statistics Get operation
+       at a time */
+    ixOsalMutexLock(&ixEthAccMacState[portId].MIBStatsGetAccessLock, IX_OSAL_WAIT_FOREVER);
+
+    if(ixNpeMhMessageWithResponseSend(IX_ETH_ACC_PORT_TO_NPE_ID(portId),
+				      message,
+				      IX_ETHNPE_GETSTATS,
+				      ixEthAccMacNpeStatsMessageCallback,
+				      IX_NPEMH_SEND_RETRIES_DEFAULT)
+       != IX_SUCCESS)
+    {
+	ixOsalMutexUnlock(&ixEthAccMacState[portId].MIBStatsGetAccessLock);
+
+        printf("EthAcc: (Mac) StatsGet failed to send NPE message\n");
+
+	return IX_ETH_ACC_FAIL;
+    }
+
+    /* Wait for callback invocation indicating response to
+       this request - we need this mutex in order to ensure
+       that the return from this function is synchronous */
+    ixOsalMutexLock(&ixEthAccMacState[portId].ackMIBStatsLock, IX_ETH_ACC_MIB_STATS_DELAY_MSECS);
+
+    /* Permit other tasks to perform MIB statistics Get operation */
+    ixOsalMutexUnlock(&ixEthAccMacState[portId].MIBStatsGetAccessLock);
+
+    ixEthAccMibIIStatsEndianConvert (retStats);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+
+PRIVATE void
+ixEthAccMacNpeStatsResetMessageCallback (IxNpeMhNpeId npeId,
+					 IxNpeMhMessage msg)
+{
+    IxEthAccPortId portId = IX_ETH_ACC_NPE_TO_PORT_ID(npeId);
+
+#ifndef NDEBUG
+    /* Prudent to at least check the port is within range */
+    if (portId >= IX_ETH_ACC_NUMBER_OF_PORTS)
+    {
+	IX_ETH_ACC_FATAL_LOG(
+     "IXETHACC:ixEthAccMacNpeStatsResetMessageCallback: Illegal port: %u\n",
+     (UINT32)portId, 0, 0, 0, 0, 0);
+	return;
+    }
+#endif
+
+    /*Unblock Stats Get & reset call*/
+    ixOsalMutexUnlock(&ixEthAccMacState[portId].ackMIBStatsResetLock);
+
+}
+
+
+
+IxEthAccStatus
+ixEthAccMibIIStatsGetClear (IxEthAccPortId portId,
+			    IxEthEthObjStats *retStats)
+{
+    IxNpeMhMessage message;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+        printf("EthAcc: ixEthAccMibIIStatsGetClear (Mac) EthAcc service is not initialized\n");
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (retStats == NULL)
+    {
+        printf("EthAcc: ixEthAccMibIIStatsGetClear (Mac) NULL argument\n");
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        printf("EthAcc: ixEthAccMibIIStatsGetClear (Mac) NPE for port %d is not available\n", portId);
+
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot get and clear MIB II Stats.\n", (INT32)portId, 0, 0, 0, 0, 0);
+
+        /* Return all zero stats */
+        IX_ETH_ACC_MEMSET(retStats, 0, sizeof(IxEthEthObjStats));
+
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if (!IX_ETH_IS_PORT_INITIALIZED(portId))
+    {
+        printf("EthAcc: ixEthAccMibIIStatsGetClear (Mac) port %d is not initialized\n", portId);
+	return (IX_ETH_ACC_PORT_UNINITIALIZED);
+    }
+
+    IX_OSAL_CACHE_INVALIDATE(retStats, sizeof(IxEthEthObjStats));
+
+    message.data[0] = IX_ETHNPE_RESETSTATS << IX_ETH_ACC_MAC_MSGID_SHL;
+    message.data[1] = (UINT32) IX_OSAL_MMU_VIRT_TO_PHYS(retStats);
+
+    /* Permit only one task to request MIB statistics Get-Reset operation at a time */
+    ixOsalMutexLock(&ixEthAccMacState[portId].MIBStatsGetResetAccessLock, IX_OSAL_WAIT_FOREVER);
+
+    if(ixNpeMhMessageWithResponseSend(IX_ETH_ACC_PORT_TO_NPE_ID(portId),
+				      message,
+				      IX_ETHNPE_RESETSTATS,
+				      ixEthAccMacNpeStatsResetMessageCallback,
+				      IX_NPEMH_SEND_RETRIES_DEFAULT)
+       != IX_SUCCESS)
+    {
+	ixOsalMutexUnlock(&ixEthAccMacState[portId].MIBStatsGetResetAccessLock);
+
+        printf("EthAcc: (Mac) ixEthAccMibIIStatsGetClear failed to send NPE message\n");
+
+	return IX_ETH_ACC_FAIL;
+    }
+
+    /* Wait for callback invocation indicating response to this request */
+    ixOsalMutexLock(&ixEthAccMacState[portId].ackMIBStatsResetLock, IX_ETH_ACC_MIB_STATS_DELAY_MSECS);
+
+    /* permit other tasks to get and reset MIB stats*/
+    ixOsalMutexUnlock(&ixEthAccMacState[portId].MIBStatsGetResetAccessLock);
+
+    ixEthAccMibIIStatsEndianConvert(retStats);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+IxEthAccStatus
+ixEthAccMibIIStatsClear (IxEthAccPortId portId)
+{
+    static IxEthEthObjStats retStats;
+    IxEthAccStatus status;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot clear MIB II Stats.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    /* there is no reset operation without a corresponding Get */
+    status = ixEthAccMibIIStatsGetClear(portId, &retStats);
+
+    return status;
+}
+
+/* Initialize the ethernet MAC settings */
+IxEthAccStatus
+ixEthAccMacInit(IxEthAccPortId portId)
+{
+    IX_OSAL_MBUF_POOL* portDisablePool;
+    UINT8 *data;
+
+    IX_ETH_ACC_VALIDATE_PORT_ID(portId);
+
+    if (IX_ETH_ACC_SUCCESS != ixEthAccSingleEthNpeCheck(portId))
+    {
+        IX_ETH_ACC_WARNING_LOG("EthAcc: Unavailable Eth %d: Cannot initialize Mac.\n",(INT32)portId,0,0,0,0,0);
+        return IX_ETH_ACC_SUCCESS ;
+    }
+
+    if(ixEthAccMacState[portId].macInitialised == FALSE)
+    {
+	ixEthAccMacState[portId].fullDuplex  = TRUE;
+	ixEthAccMacState[portId].rxFCSAppend = TRUE;
+	ixEthAccMacState[portId].txFCSAppend = TRUE;
+	ixEthAccMacState[portId].txPADAppend = TRUE;
+	ixEthAccMacState[portId].enabled     = FALSE;
+	ixEthAccMacState[portId].promiscuous = TRUE;
+	ixEthAccMacState[portId].joinAll     = FALSE;
+	ixEthAccMacState[portId].initDone    = FALSE;
+	ixEthAccMacState[portId].macInitialised = TRUE;
+
+        /* initialize MIB stats mutexes */
+        ixOsalMutexInit(&ixEthAccMacState[portId].ackMIBStatsLock);
+        ixOsalMutexLock(&ixEthAccMacState[portId].ackMIBStatsLock, IX_OSAL_WAIT_FOREVER);
+
+        ixOsalMutexInit(&ixEthAccMacState[portId].ackMIBStatsResetLock);
+        ixOsalMutexLock(&ixEthAccMacState[portId].ackMIBStatsResetLock, IX_OSAL_WAIT_FOREVER);
+
+        ixOsalMutexInit(&ixEthAccMacState[portId].MIBStatsGetAccessLock);
+
+        ixOsalMutexInit(&ixEthAccMacState[portId].MIBStatsGetResetAccessLock);
+
+        ixOsalMutexInit(&ixEthAccMacState[portId].npeLoopbackMessageLock);
+
+	ixEthAccMacState[portId].portDisableRxMbufPtr = NULL;
+        ixEthAccMacState[portId].portDisableTxMbufPtr = NULL;
+
+	portDisablePool = IX_OSAL_MBUF_POOL_INIT(2,
+			  IX_ETHACC_RX_MBUF_MIN_SIZE,
+			  "portDisable Pool");
+
+        IX_OSAL_ENSURE(portDisablePool != NULL, "Failed to initialize PortDisable pool");
+
+	ixEthAccMacState[portId].portDisableRxMbufPtr = IX_OSAL_MBUF_POOL_GET(portDisablePool);
+        ixEthAccMacState[portId].portDisableTxMbufPtr = IX_OSAL_MBUF_POOL_GET(portDisablePool);
+
+	IX_OSAL_ENSURE(ixEthAccMacState[portId].portDisableRxMbufPtr != NULL,
+		  "Pool allocation failed");
+        IX_OSAL_ENSURE(ixEthAccMacState[portId].portDisableTxMbufPtr != NULL,
+		  "Pool allocation failed");
+	/* fill the payload of the Rx mbuf used in portDisable */
+        IX_OSAL_MBUF_MLEN(ixEthAccMacState[portId].portDisableRxMbufPtr) = IX_ETHACC_RX_MBUF_MIN_SIZE;
+
+        memset(IX_OSAL_MBUF_MDATA(ixEthAccMacState[portId].portDisableRxMbufPtr),
+	       0xAA,
+	       IX_ETHACC_RX_MBUF_MIN_SIZE);
+
+	/* fill the payload of the Tx mbuf used in portDisable (64 bytes) */
+        IX_OSAL_MBUF_MLEN(ixEthAccMacState[portId].portDisableTxMbufPtr) = 64;
+        IX_OSAL_MBUF_PKT_LEN(ixEthAccMacState[portId].portDisableTxMbufPtr) = 64;
+
+        data = (UINT8 *) IX_OSAL_MBUF_MDATA(ixEthAccMacState[portId].portDisableTxMbufPtr);
+        memset(data, 0xBB, 64);
+        data[0] = 0x00; /* unicast destination MAC address */
+        data[6] = 0x00; /* unicast source MAC address */
+        data[12] = 0x08; /* typelength : IP frame */
+        data[13] = 0x00; /* typelength : IP frame */
+
+	IX_OSAL_CACHE_FLUSH(data, 64);
+    }
+
+    IX_OSAL_ASSERT (ixEthAccMacBase[portId] != 0);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_CORE_CNTRL,
+	      IX_ETH_ACC_CORE_RESET);
+
+    ixOsalSleep(IX_ETH_ACC_MAC_RESET_DELAY);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_CORE_CNTRL,
+	      IX_ETH_ACC_CORE_MDC_EN);
+
+    REG_WRITE(ixEthAccMacBase[portId],
+	      IX_ETH_ACC_MAC_INT_CLK_THRESH,
+	      IX_ETH_ACC_MAC_INT_CLK_THRESH_DEFAULT);
+
+    ixEthAccMacStateUpdate(portId);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+/* PRIVATE Functions*/
+
+PRIVATE void
+ixEthAccMacStateUpdate(IxEthAccPortId portId)
+{
+    UINT32 regval;
+
+    if ( ixEthAccMacState[portId].enabled == FALSE )
+    {
+	/*  Just disable both the transmitter and reciver in the MAC.  */
+        REG_READ(ixEthAccMacBase[portId],
+		 IX_ETH_ACC_MAC_RX_CNTRL1,
+		 regval);
+ 	REG_WRITE(ixEthAccMacBase[portId],
+		  IX_ETH_ACC_MAC_RX_CNTRL1,
+		  regval & ~IX_ETH_ACC_RX_CNTRL1_RX_EN);
+
+        REG_READ(ixEthAccMacBase[portId],
+		 IX_ETH_ACC_MAC_TX_CNTRL1,
+		 regval);
+ 	REG_WRITE(ixEthAccMacBase[portId],
+		  IX_ETH_ACC_MAC_TX_CNTRL1,
+		  regval & ~IX_ETH_ACC_TX_CNTRL1_TX_EN);
+    }
+
+    if(ixEthAccMacState[portId].fullDuplex)
+    {
+	ixEthAccPortDuplexModeSetPriv (portId, IX_ETH_ACC_FULL_DUPLEX);
+    }
+    else
+    {
+	ixEthAccPortDuplexModeSetPriv (portId, IX_ETH_ACC_HALF_DUPLEX);
+    }
+
+    if(ixEthAccMacState[portId].rxFCSAppend)
+    {
+	ixEthAccPortRxFrameAppendFCSEnablePriv (portId);
+    }
+    else
+    {
+	ixEthAccPortRxFrameAppendFCSDisablePriv (portId);
+    }
+
+    if(ixEthAccMacState[portId].txFCSAppend)
+    {
+	ixEthAccPortTxFrameAppendFCSEnablePriv (portId);
+    }
+    else
+    {
+	ixEthAccPortTxFrameAppendFCSDisablePriv (portId);
+    }
+
+    if(ixEthAccMacState[portId].txPADAppend)
+    {
+	ixEthAccPortTxFrameAppendPaddingEnablePriv (portId);
+    }
+    else
+    {
+	ixEthAccPortTxFrameAppendPaddingDisablePriv (portId);
+    }
+
+    if(ixEthAccMacState[portId].promiscuous)
+    {
+	ixEthAccPortPromiscuousModeSetPriv(portId);
+    }
+    else
+    {
+	ixEthAccPortPromiscuousModeClearPriv(portId);
+    }
+
+    if ( ixEthAccMacState[portId].enabled == TRUE )
+    {
+        /*   Enable both the transmitter and reciver in the MAC.  */
+        REG_READ(ixEthAccMacBase[portId],
+		 IX_ETH_ACC_MAC_RX_CNTRL1,
+		 regval);
+        REG_WRITE(ixEthAccMacBase[portId],
+		  IX_ETH_ACC_MAC_RX_CNTRL1,
+		  regval | IX_ETH_ACC_RX_CNTRL1_RX_EN);
+
+        REG_READ(ixEthAccMacBase[portId],
+		 IX_ETH_ACC_MAC_TX_CNTRL1,
+		 regval);
+ 	REG_WRITE(ixEthAccMacBase[portId],
+		  IX_ETH_ACC_MAC_TX_CNTRL1,
+		  regval | IX_ETH_ACC_TX_CNTRL1_TX_EN);
+    }
+}
+
+
+PRIVATE BOOL
+ixEthAccMacEqual(IxEthAccMacAddr *macAddr1,
+		 IxEthAccMacAddr *macAddr2)
+{
+    UINT32 i;
+    for(i=0;i<IX_IEEE803_MAC_ADDRESS_SIZE; i++)
+    {
+	if(macAddr1->macAddress[i] != macAddr2->macAddress[i])
+	{
+	    return FALSE;
+	}
+    }
+    return TRUE;
+}
+
+PRIVATE void
+ixEthAccMacPrint(IxEthAccMacAddr *m)
+{
+    printf("%2.2x:%2.2x:%2.2x:%2.2x:%2.2x:%2.2x",
+	   m->macAddress[0], m->macAddress[1],
+	   m->macAddress[2], m->macAddress[3],
+	   m->macAddress[4], m->macAddress[5]);
+}
+
+/* Set the multicast address and address mask registers
+ *
+ * A bit in the address mask register must be set if
+ * all multicast addresses always have that bit set, or if
+ * all multicast addresses always have that bit cleared.
+ *
+ * A bit in the address register must be set if all multicast
+ * addresses have that bit set, otherwise, it should be cleared
+ */
+
+PRIVATE void
+ixEthAccMulticastAddressSet(IxEthAccPortId portId)
+{
+    UINT32 i;
+    UINT32 j;
+    IxEthAccMacAddr addressMask;
+    IxEthAccMacAddr address;
+    IxEthAccMacAddr alwaysClearBits;
+    IxEthAccMacAddr alwaysSetBits;
+
+    /* calculate alwaysClearBits and alwaysSetBits:
+     * alwaysClearBits is calculated by ORing all
+     * multicast addresses, those bits that are always
+     * clear are clear in the result
+     *
+     * alwaysSetBits is calculated by ANDing all
+     * multicast addresses, those bits that are always set
+     * are set in the result
+     */
+
+    if (ixEthAccMacState[portId].promiscuous == TRUE)
+    {
+	/* Promiscuous Mode is set, and filtering
+	 * allow all packets, and enable the mcast and
+	 * bcast detection.
+	 */
+	memset(&addressMask.macAddress,
+	       0,
+	       IX_IEEE803_MAC_ADDRESS_SIZE);
+	memset(&address.macAddress,
+	       0,
+	       IX_IEEE803_MAC_ADDRESS_SIZE);
+    }
+    else
+    {
+	if(ixEthAccMacState[portId].joinAll == TRUE)
+	{
+	    /* Join all is set. The mask and address are
+	     * the multicast settings.
+	     */
+	    IxEthAccMacAddr macAddr = {{0x1,0x0,0x0,0x0,0x0,0x0}};
+
+	    memcpy(addressMask.macAddress,
+		   macAddr.macAddress,
+		   IX_IEEE803_MAC_ADDRESS_SIZE);
+	    memcpy(address.macAddress,
+		   macAddr.macAddress,
+		   IX_IEEE803_MAC_ADDRESS_SIZE);
+	}
+	else if(ixEthAccMacState[portId].mcastAddrIndex == 0)
+	{
+	    /* No entry in the filtering database,
+	     * Promiscuous Mode is cleared, Broadcast filtering
+	     * is configured.
+	     */
+	    memset(addressMask.macAddress,
+		   IX_ETH_ACC_MAC_ALL_BITS_SET,
+		   IX_IEEE803_MAC_ADDRESS_SIZE);
+	    memset(address.macAddress,
+		   IX_ETH_ACC_MAC_ALL_BITS_SET,
+		   IX_IEEE803_MAC_ADDRESS_SIZE);
+	}
+	else
+	{
+	    /* build a mask and an address which mix all entreis
+	     * from the list of multicast addresses
+	     */
+	    memset(alwaysClearBits.macAddress,
+		   0,
+		   IX_IEEE803_MAC_ADDRESS_SIZE);
+	    memset(alwaysSetBits.macAddress,
+		   IX_ETH_ACC_MAC_ALL_BITS_SET,
+		   IX_IEEE803_MAC_ADDRESS_SIZE);
+
+	    for(i=0;i<ixEthAccMacState[portId].mcastAddrIndex;i++)
+	    {
+		for(j=0;j<IX_IEEE803_MAC_ADDRESS_SIZE;j++)
+		{
+		    alwaysClearBits.macAddress[j] |=
+			ixEthAccMacState[portId].mcastAddrsTable[i].macAddress[j];
+		    alwaysSetBits.macAddress[j] &=
+			ixEthAccMacState[portId].mcastAddrsTable[i].macAddress[j];
+		}
+	    }
+
+	    for(i=0;i<IX_IEEE803_MAC_ADDRESS_SIZE;i++)
+	    {
+		addressMask.macAddress[i] = alwaysSetBits.macAddress[i]
+		    | ~alwaysClearBits.macAddress[i];
+		address.macAddress[i] = alwaysSetBits.macAddress[i];
+	    }
+	}
+    }
+
+    /*write the new addr filtering to h/w*/
+    for(i=0;i<IX_IEEE803_MAC_ADDRESS_SIZE;i++)
+    {
+	REG_WRITE(ixEthAccMacBase[portId],
+		  IX_ETH_ACC_MAC_ADDR_MASK_1+i*sizeof(UINT32),
+		  addressMask.macAddress[i]);
+	REG_WRITE(ixEthAccMacBase[portId],
+		  IX_ETH_ACC_MAC_ADDR_1+i*sizeof(UINT32),
+		  address.macAddress[i]);
+    }
+}

cpu/ixp/npe/IxEthAccMii.c

@@ -0,0 +1,410 @@
+/**
+ * @file IxEthAccMii.c
+ *
+ * @author Intel Corporation
+ * @date 
+ *
+ * @brief  MII control functions
+ *
+ * Design Notes:
+ *
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxOsal.h"
+#include "IxEthAcc.h"
+#include "IxEthAcc_p.h"
+#include "IxEthAccMac_p.h"
+#include "IxEthAccMii_p.h"
+
+
+PRIVATE UINT32 miiBaseAddressVirt;
+PRIVATE IxOsalMutex miiAccessLock;
+
+PUBLIC UINT32 ixEthAccMiiRetryCount    = IX_ETH_ACC_MII_TIMEOUT_10TH_SECS;
+PUBLIC UINT32 ixEthAccMiiAccessTimeout = IX_ETH_ACC_MII_10TH_SEC_IN_MILLIS;
+
+/* -----------------------------------
+ * private function prototypes
+ */
+PRIVATE void
+ixEthAccMdioCmdWrite(UINT32 mdioCommand);
+
+PRIVATE void
+ixEthAccMdioCmdRead(UINT32 *data);
+
+PRIVATE void
+ixEthAccMdioStatusRead(UINT32 *data);
+
+
+PRIVATE void
+ixEthAccMdioCmdWrite(UINT32 mdioCommand)
+{
+    REG_WRITE(miiBaseAddressVirt,
+	      IX_ETH_ACC_MAC_MDIO_CMD_1,
+	      mdioCommand & 0xff);
+
+    REG_WRITE(miiBaseAddressVirt,
+	      IX_ETH_ACC_MAC_MDIO_CMD_2,
+	      (mdioCommand >> 8) & 0xff);
+
+    REG_WRITE(miiBaseAddressVirt,
+	      IX_ETH_ACC_MAC_MDIO_CMD_3,
+	      (mdioCommand >> 16) & 0xff);
+
+    REG_WRITE(miiBaseAddressVirt,
+	      IX_ETH_ACC_MAC_MDIO_CMD_4,
+	      (mdioCommand >> 24) & 0xff);
+}
+
+PRIVATE void
+ixEthAccMdioCmdRead(UINT32 *data)
+{
+    UINT32 regval;
+
+    REG_READ(miiBaseAddressVirt,
+	     IX_ETH_ACC_MAC_MDIO_CMD_1,
+	     regval);
+
+    *data = regval & 0xff;
+
+    REG_READ(miiBaseAddressVirt,
+	     IX_ETH_ACC_MAC_MDIO_CMD_2,
+	     regval);
+
+    *data |= (regval & 0xff) << 8;
+
+    REG_READ(miiBaseAddressVirt,
+	     IX_ETH_ACC_MAC_MDIO_CMD_3,
+	     regval);
+
+    *data |= (regval & 0xff) << 16;
+
+    REG_READ(miiBaseAddressVirt,
+	     IX_ETH_ACC_MAC_MDIO_CMD_4,
+	     regval);
+
+    *data |= (regval & 0xff) << 24;
+
+}
+
+PRIVATE void
+ixEthAccMdioStatusRead(UINT32 *data)
+{
+    UINT32 regval;
+
+    REG_READ(miiBaseAddressVirt,
+	     IX_ETH_ACC_MAC_MDIO_STS_1,
+	     regval);
+
+    *data = regval & 0xff;
+
+    REG_READ(miiBaseAddressVirt,
+	     IX_ETH_ACC_MAC_MDIO_STS_2,
+	     regval);
+
+    *data |= (regval & 0xff) << 8;
+
+    REG_READ(miiBaseAddressVirt,
+	     IX_ETH_ACC_MAC_MDIO_STS_3,
+	     regval);
+
+    *data |= (regval & 0xff) << 16;
+
+    REG_READ(miiBaseAddressVirt,
+	     IX_ETH_ACC_MAC_MDIO_STS_4,
+	     regval);
+    
+    *data |= (regval & 0xff) << 24;
+    
+}
+
+
+/********************************************************************
+ * ixEthAccMiiInit
+ */
+IxEthAccStatus
+ixEthAccMiiInit()
+{
+    if(ixOsalMutexInit(&miiAccessLock)!= IX_SUCCESS)
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+
+    miiBaseAddressVirt = (UINT32) IX_OSAL_MEM_MAP(IX_ETH_ACC_MAC_0_BASE, IX_OSAL_IXP400_ETHA_MAP_SIZE);
+    
+    if (miiBaseAddressVirt == 0)
+    {
+      ixOsalLog(IX_OSAL_LOG_LVL_FATAL, 
+		IX_OSAL_LOG_DEV_STDOUT, 
+		"EthAcc: Could not map MII I/O mapped memory\n", 
+		0, 0, 0, 0, 0, 0);
+      
+      return IX_ETH_ACC_FAIL;
+    }
+    
+    return IX_ETH_ACC_SUCCESS;
+}
+
+void
+ixEthAccMiiUnload(void)
+{
+    IX_OSAL_MEM_UNMAP(miiBaseAddressVirt);
+  
+    miiBaseAddressVirt = 0;
+}
+
+PUBLIC IxEthAccStatus
+ixEthAccMiiAccessTimeoutSet(UINT32 timeout, UINT32 retryCount)
+{
+    if (retryCount < 1) return IX_ETH_ACC_FAIL;
+
+    ixEthAccMiiRetryCount    = retryCount;
+    ixEthAccMiiAccessTimeout = timeout;
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+/*********************************************************************
+ * ixEthAccMiiReadRtn - read a 16 bit value from a PHY
+ */
+IxEthAccStatus      
+ixEthAccMiiReadRtn (UINT8 phyAddr, 
+		    UINT8 phyReg, 
+		    UINT16 *value)
+{
+    UINT32 mdioCommand;
+    UINT32 regval;
+    UINT32 miiTimeout;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    if ((phyAddr >= IXP425_ETH_ACC_MII_MAX_ADDR) 
+	|| (phyReg >= IXP425_ETH_ACC_MII_MAX_REG))
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    if (value == NULL)
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&miiAccessLock, IX_OSAL_WAIT_FOREVER);
+    mdioCommand = phyReg <<  IX_ETH_ACC_MII_REG_SHL 
+	| phyAddr << IX_ETH_ACC_MII_ADDR_SHL;
+    mdioCommand |= IX_ETH_ACC_MII_GO;
+
+    ixEthAccMdioCmdWrite(mdioCommand);
+    
+    miiTimeout = ixEthAccMiiRetryCount;
+
+    while(miiTimeout)
+    {
+	
+	ixEthAccMdioCmdRead(&regval);
+     
+	if((regval & IX_ETH_ACC_MII_GO) == 0x0)
+	{	    
+	    break;
+	}
+	/* Sleep for a while */
+	ixOsalSleep(ixEthAccMiiAccessTimeout);
+	miiTimeout--;
+    }
+    
+
+    
+    if(miiTimeout == 0)
+    {	
+	ixOsalMutexUnlock(&miiAccessLock);
+	*value = 0xffff;
+	return IX_ETH_ACC_FAIL;
+    }
+    
+    
+    ixEthAccMdioStatusRead(&regval);
+    if(regval & IX_ETH_ACC_MII_READ_FAIL)
+    {
+	ixOsalMutexUnlock(&miiAccessLock);
+	*value = 0xffff;
+	return IX_ETH_ACC_FAIL;
+    }
+
+    *value = regval & 0xffff;
+    ixOsalMutexUnlock(&miiAccessLock);
+    return IX_ETH_ACC_SUCCESS;
+    
+}
+
+
+/*********************************************************************
+ * ixEthAccMiiWriteRtn - write a 16 bit value to a PHY
+ */
+IxEthAccStatus
+ixEthAccMiiWriteRtn (UINT8 phyAddr, 
+		     UINT8 phyReg, 
+		     UINT16 value)
+{
+    UINT32 mdioCommand;
+    UINT32 regval;
+    UINT16 readVal;
+    UINT32 miiTimeout;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    if ((phyAddr >= IXP425_ETH_ACC_MII_MAX_ADDR) 
+	|| (phyReg >= IXP425_ETH_ACC_MII_MAX_REG))
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+   
+    /* ensure that a PHY is present at this address */
+    if(ixEthAccMiiReadRtn(phyAddr,
+                          IX_ETH_ACC_MII_CTRL_REG,
+                          &readVal) != IX_ETH_ACC_SUCCESS)
+    {
+        return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&miiAccessLock, IX_OSAL_WAIT_FOREVER);
+    mdioCommand = phyReg << IX_ETH_ACC_MII_REG_SHL
+	| phyAddr << IX_ETH_ACC_MII_ADDR_SHL ;
+    mdioCommand |= IX_ETH_ACC_MII_GO | IX_ETH_ACC_MII_WRITE | value;
+
+    ixEthAccMdioCmdWrite(mdioCommand);
+    
+    miiTimeout = ixEthAccMiiRetryCount;
+
+    while(miiTimeout)
+    {
+	
+	ixEthAccMdioCmdRead(&regval);
+
+	/*The "GO" bit is reset to 0 when the write completes*/
+	if((regval & IX_ETH_ACC_MII_GO) == 0x0)
+	{	    	    
+	    break;
+	}
+	/* Sleep for a while */
+	ixOsalSleep(ixEthAccMiiAccessTimeout);
+        miiTimeout--;
+    }
+    
+    ixOsalMutexUnlock(&miiAccessLock);
+    if(miiTimeout == 0)
+    {
+	return IX_ETH_ACC_FAIL;
+    }
+    return IX_ETH_ACC_SUCCESS;
+}
+
+
+/*****************************************************************
+ *
+ *  Phy query functions
+ *
+ */
+IxEthAccStatus
+ixEthAccMiiStatsShow (UINT32 phyAddr)
+{
+    UINT16 regval;
+    printf("Regisers on PHY at address 0x%x\n", phyAddr);
+    ixEthAccMiiReadRtn(phyAddr, IX_ETH_ACC_MII_CTRL_REG, &regval);
+    printf("    Control Register                  :      0x%4.4x\n", regval);
+    ixEthAccMiiReadRtn(phyAddr,  IX_ETH_ACC_MII_STAT_REG, &regval);
+    printf("    Status Register                   :      0x%4.4x\n", regval);
+    ixEthAccMiiReadRtn(phyAddr,  IX_ETH_ACC_MII_PHY_ID1_REG, &regval);
+    printf("    PHY ID1 Register                  :      0x%4.4x\n", regval);
+    ixEthAccMiiReadRtn(phyAddr,  IX_ETH_ACC_MII_PHY_ID2_REG, &regval);
+    printf("    PHY ID2 Register                  :      0x%4.4x\n", regval);
+    ixEthAccMiiReadRtn(phyAddr,  IX_ETH_ACC_MII_AN_ADS_REG, &regval);
+    printf("    Auto Neg ADS Register             :      0x%4.4x\n", regval);
+    ixEthAccMiiReadRtn(phyAddr,  IX_ETH_ACC_MII_AN_PRTN_REG, &regval);
+    printf("    Auto Neg Partner Ability Register :      0x%4.4x\n", regval);
+    ixEthAccMiiReadRtn(phyAddr,  IX_ETH_ACC_MII_AN_EXP_REG, &regval);
+    printf("    Auto Neg Expansion Register       :      0x%4.4x\n", regval);
+    ixEthAccMiiReadRtn(phyAddr,  IX_ETH_ACC_MII_AN_NEXT_REG, &regval);
+    printf("    Auto Neg Next Register            :      0x%4.4x\n", regval);
+
+    return IX_ETH_ACC_SUCCESS;
+}
+
+
+/*****************************************************************
+ *
+ *  Interface query functions
+ *
+ */
+IxEthAccStatus
+ixEthAccMdioShow (void)
+{
+    UINT32 regval;
+
+    if (!IX_ETH_ACC_IS_SERVICE_INITIALIZED())
+    {
+	return (IX_ETH_ACC_FAIL);
+    }
+
+    ixOsalMutexLock(&miiAccessLock, IX_OSAL_WAIT_FOREVER);
+    ixEthAccMdioCmdRead(&regval);
+    ixOsalMutexUnlock(&miiAccessLock);
+    
+    printf("MDIO command register\n");
+    printf("    Go bit      : 0x%x\n", (regval & BIT(31)) >> 31);
+    printf("    MDIO Write  : 0x%x\n", (regval & BIT(26)) >> 26);
+    printf("    PHY address : 0x%x\n", (regval >> 21) & 0x1f);
+    printf("    Reg address : 0x%x\n", (regval >> 16) & 0x1f);
+	
+    ixOsalMutexLock(&miiAccessLock, IX_OSAL_WAIT_FOREVER);
+    ixEthAccMdioStatusRead(&regval);
+    ixOsalMutexUnlock(&miiAccessLock);
+
+    printf("MDIO status register\n");
+    printf("    Read OK     : 0x%x\n", (regval & BIT(31)) >> 31);
+    printf("    Read Data   : 0x%x\n", (regval >> 16) & 0xff);
+
+    return IX_ETH_ACC_SUCCESS;   
+}
+

cpu/ixp/npe/IxEthDBAPI.c

@@ -0,0 +1,448 @@
+/**
+ * @file IxEthDBAPI.c
+ *
+ * @brief Implementation of the public API
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxEthDB_p.h"
+#include "IxFeatureCtrl.h"
+
+extern HashTable dbHashtable;
+extern IxEthDBPortMap overflowUpdatePortList;
+extern BOOL ixEthDBPortUpdateRequired[IX_ETH_DB_MAX_RECORD_TYPE_INDEX + 1];
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFilteringStaticEntryProvision(IxEthDBPortId portID, IxEthDBMacAddr *macAddr)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_LEARNING);
+
+    return ixEthDBTriggerAddPortUpdate(macAddr, portID, TRUE);
+}
+    
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFilteringDynamicEntryProvision(IxEthDBPortId portID, IxEthDBMacAddr *macAddr)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+        
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_LEARNING);
+
+    return ixEthDBTriggerAddPortUpdate(macAddr, portID, FALSE);
+}
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFilteringEntryDelete(IxEthDBMacAddr *macAddr)
+{
+    HashNode *searchResult;
+
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+
+    searchResult = ixEthDBSearch(macAddr, IX_ETH_DB_ALL_FILTERING_RECORDS);
+
+    if (searchResult == NULL)
+    {
+        return IX_ETH_DB_NO_SUCH_ADDR; /* not found */
+    }
+    
+    ixEthDBReleaseHashNode(searchResult);
+    
+    /* build a remove event and place it on the event queue */
+    return ixEthDBTriggerRemovePortUpdate(macAddr, ((MacDescriptor *) searchResult->data)->portID);
+}
+       
+IX_ETH_DB_PUBLIC
+void ixEthDBDatabaseMaintenance()
+{
+    HashIterator iterator;
+    UINT32 portIndex;
+    BOOL agingRequired = FALSE;
+
+    /* ports who will have deleted records and therefore will need updating */
+    IxEthDBPortMap triggerPorts;
+
+    if (IX_FEATURE_CTRL_SWCONFIG_ENABLED !=
+        ixFeatureCtrlSwConfigurationCheck (IX_FEATURECTRL_ETH_LEARNING))
+    {
+        return;
+    }
+
+    SET_EMPTY_DEPENDENCY_MAP(triggerPorts);
+
+    /* check if there's at least a port that needs aging */
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        if (ixEthDBPortInfo[portIndex].agingEnabled && ixEthDBPortInfo[portIndex].enabled)
+        {
+            agingRequired = TRUE;
+        }
+    }
+
+    if (agingRequired)
+    {
+        /* ask each NPE port to write back the database for aging inspection */
+        for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+        {
+            if (ixEthDBPortDefinitions[portIndex].type == IX_ETH_NPE
+                && ixEthDBPortInfo[portIndex].agingEnabled
+                && ixEthDBPortInfo[portIndex].enabled)
+            {
+                IxNpeMhMessage message;
+                IX_STATUS result;
+                
+                /* send EDB_GetMACAddressDatabase message */
+                FILL_GETMACADDRESSDATABASE(message, 
+                    0 /* unused */, 
+                    IX_OSAL_MMU_VIRT_TO_PHYS(ixEthDBPortInfo[portIndex].updateMethod.npeUpdateZone));
+
+                IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portIndex), message, result);
+
+                if (result == IX_SUCCESS)
+                {
+                    /* analyze NPE copy */
+                    ixEthDBNPESyncScan(portIndex, ixEthDBPortInfo[portIndex].updateMethod.npeUpdateZone, FULL_ELT_BYTE_SIZE);
+
+                    IX_ETH_DB_SUPPORT_TRACE("DB: (API) Finished scanning NPE tree on port %d\n", portIndex);
+                }
+                else
+                {
+                    ixEthDBPortInfo[portIndex].agingEnabled                = FALSE;
+                    ixEthDBPortInfo[portIndex].updateMethod.updateEnabled  = FALSE;
+                    ixEthDBPortInfo[portIndex].updateMethod.userControlled = TRUE;
+
+                    ixOsalLog(IX_OSAL_LOG_LVL_FATAL,
+                        IX_OSAL_LOG_DEV_STDOUT, 
+                        "EthDB: (Maintenance) warning, disabling aging and updates for port %d (assumed dead)\n",
+                        portIndex, 0, 0, 0, 0, 0);
+
+                    ixEthDBDatabaseClear(portIndex, IX_ETH_DB_ALL_RECORD_TYPES);
+                }
+            }
+        }
+
+        /* browse database and age entries */
+        BUSY_RETRY(ixEthDBInitHashIterator(&dbHashtable, &iterator));
+
+        while (IS_ITERATOR_VALID(&iterator))
+        {
+            MacDescriptor *descriptor = (MacDescriptor *) iterator.node->data;
+            UINT32 *age               = NULL;
+            BOOL staticEntry          = TRUE;
+
+            if (descriptor->type == IX_ETH_DB_FILTERING_RECORD)
+            {
+                age         = &descriptor->recordData.filteringData.age;
+                staticEntry = descriptor->recordData.filteringData.staticEntry;
+            }
+            else if (descriptor->type == IX_ETH_DB_FILTERING_VLAN_RECORD)
+            {
+                age         = &descriptor->recordData.filteringVlanData.age;
+                staticEntry = descriptor->recordData.filteringVlanData.staticEntry;
+            }
+            else
+            {
+                staticEntry = TRUE;
+            }
+
+            if (ixEthDBPortInfo[descriptor->portID].agingEnabled && (staticEntry == FALSE))
+            {
+                /* manually increment the age if the port has no such capability */
+                if ((ixEthDBPortDefinitions[descriptor->portID].capabilities & IX_ETH_ENTRY_AGING) == 0)
+                {
+                    *age += (IX_ETH_DB_MAINTENANCE_TIME / 60);
+                }
+
+                /* age entry if it exceeded the maximum time to live */
+                if (*age >= (IX_ETH_DB_LEARNING_ENTRY_AGE_TIME / 60))
+                {
+                    /* add port to the set of update trigger ports */
+                    JOIN_PORT_TO_MAP(triggerPorts, descriptor->portID);
+
+                    /* delete entry */
+                    BUSY_RETRY(ixEthDBRemoveEntryAtHashIterator(&dbHashtable, &iterator));
+                }
+                else
+                {
+                    /* move to the next record */
+                    BUSY_RETRY(ixEthDBIncrementHashIterator(&dbHashtable, &iterator));
+                }
+            }
+            else
+            {
+                /* move to the next record */
+                BUSY_RETRY(ixEthDBIncrementHashIterator(&dbHashtable, &iterator));
+            }
+        }
+
+        /* update ports which lost records */
+        ixEthDBUpdatePortLearningTrees(triggerPorts);
+    }
+}
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBDatabaseClear(IxEthDBPortId portID, IxEthDBRecordType recordType)
+{
+    IxEthDBPortMap triggerPorts;
+    HashIterator iterator;
+
+    if (portID >= IX_ETH_DB_NUMBER_OF_PORTS && portID != IX_ETH_DB_ALL_PORTS)
+    {
+        return IX_ETH_DB_INVALID_PORT;
+    }
+
+    /* check if the user passes some extra bits */
+    if ((recordType | IX_ETH_DB_ALL_RECORD_TYPES) != IX_ETH_DB_ALL_RECORD_TYPES)
+    {
+        return IX_ETH_DB_INVALID_ARG;
+    }
+
+    SET_EMPTY_DEPENDENCY_MAP(triggerPorts);
+    
+    /* browse database and age entries */
+    BUSY_RETRY(ixEthDBInitHashIterator(&dbHashtable, &iterator));
+
+    while (IS_ITERATOR_VALID(&iterator))
+    {
+        MacDescriptor *descriptor = (MacDescriptor *) iterator.node->data;
+
+        if (((descriptor->portID == portID) || (portID == IX_ETH_DB_ALL_PORTS))
+            && ((descriptor->type & recordType) != 0))
+        {
+            /* add to trigger if automatic updates are required */
+            if (ixEthDBPortUpdateRequired[descriptor->type])
+            {
+                /* add port to the set of update trigger ports */
+                JOIN_PORT_TO_MAP(triggerPorts, descriptor->portID);
+            }
+
+            /* delete entry */
+            BUSY_RETRY(ixEthDBRemoveEntryAtHashIterator(&dbHashtable, &iterator));
+        }
+        else
+        {
+            /* move to the next record */
+            BUSY_RETRY(ixEthDBIncrementHashIterator(&dbHashtable, &iterator));
+        }
+    }
+
+    /* update ports which lost records */
+    ixEthDBUpdatePortLearningTrees(triggerPorts);
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFilteringPortSearch(IxEthDBPortId portID, IxEthDBMacAddr *macAddr)
+{
+    HashNode *searchResult;
+    IxEthDBStatus result = IX_ETH_DB_NO_SUCH_ADDR;
+
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_LEARNING);
+
+    searchResult = ixEthDBSearch(macAddr, IX_ETH_DB_ALL_FILTERING_RECORDS);
+
+    if (searchResult == NULL)
+    {
+        return IX_ETH_DB_NO_SUCH_ADDR; /* not found */
+    }
+
+    if (((MacDescriptor *) (searchResult->data))->portID == portID)
+    {
+        result = IX_ETH_DB_SUCCESS; /* address and port match */
+    }
+
+    ixEthDBReleaseHashNode(searchResult);
+
+    return result;
+}
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFilteringDatabaseSearch(IxEthDBPortId *portID, IxEthDBMacAddr *macAddr)
+{
+    HashNode *searchResult;
+
+    IX_ETH_DB_CHECK_REFERENCE(portID);
+    
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+
+    searchResult = ixEthDBSearch(macAddr, IX_ETH_DB_ALL_FILTERING_RECORDS);
+
+    if (searchResult == NULL)
+    {
+        return IX_ETH_DB_NO_SUCH_ADDR; /* not found */
+    }
+
+    /* return the port ID */
+    *portID = ((MacDescriptor *) searchResult->data)->portID;
+
+    ixEthDBReleaseHashNode(searchResult);
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBPortAgingDisable(IxEthDBPortId portID)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_LEARNING);
+
+    ixEthDBPortInfo[portID].agingEnabled = FALSE;
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBPortAgingEnable(IxEthDBPortId portID)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_LEARNING);
+
+    ixEthDBPortInfo[portID].agingEnabled = TRUE;
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFilteringPortUpdatingSearch(IxEthDBPortId *portID, IxEthDBMacAddr *macAddr)
+{
+    HashNode *searchResult;
+    MacDescriptor *descriptor;
+
+    IX_ETH_DB_CHECK_REFERENCE(portID);
+    
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+
+    searchResult = ixEthDBSearch(macAddr, IX_ETH_DB_ALL_FILTERING_RECORDS);
+
+    if (searchResult == NULL)
+    {
+        return IX_ETH_DB_NO_SUCH_ADDR; /* not found */
+    }
+    
+    descriptor = (MacDescriptor *) searchResult->data;
+
+    /* return the port ID */
+    *portID = descriptor->portID;
+
+    /* reset entry age */
+    if (descriptor->type == IX_ETH_DB_FILTERING_RECORD)
+    {
+        descriptor->recordData.filteringData.age = 0;
+    }
+    else
+    {
+        descriptor->recordData.filteringVlanData.age = 0;
+    }
+
+    ixEthDBReleaseHashNode(searchResult);
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBPortDependencyMapSet(IxEthDBPortId portID, IxEthDBPortMap dependencyPortMap)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_REFERENCE(dependencyPortMap);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_FILTERING);
+
+    /* force bit at offset 255 to 0 (reserved) */
+    dependencyPortMap[31] &= 0xFE;
+
+    COPY_DEPENDENCY_MAP(ixEthDBPortInfo[portID].dependencyPortMap, dependencyPortMap);
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBPortDependencyMapGet(IxEthDBPortId portID, IxEthDBPortMap dependencyPortMap)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_REFERENCE(dependencyPortMap);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_FILTERING);
+
+    COPY_DEPENDENCY_MAP(dependencyPortMap, ixEthDBPortInfo[portID].dependencyPortMap);
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBPortUpdateEnableSet(IxEthDBPortId portID, BOOL enableUpdate)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);    
+
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_FILTERING);
+
+    ixEthDBPortInfo[portID].updateMethod.updateEnabled  = enableUpdate;
+    ixEthDBPortInfo[portID].updateMethod.userControlled = TRUE;
+
+    return IX_ETH_DB_SUCCESS;
+}

cpu/ixp/npe/IxEthDBAPISupport.c

@@ -0,0 +1,678 @@
+/**
+ * @file IxEthDBAPISupport.c
+ *
+ * @brief Public API support functions
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include <IxEthDB.h>
+#include <IxNpeMh.h>
+#include <IxFeatureCtrl.h>
+
+#include "IxEthDB_p.h"
+#include "IxEthDBMessages_p.h"
+#include "IxEthDB_p.h"
+#include "IxEthDBLog_p.h"
+
+#ifdef IX_UNIT_TEST
+
+int dbAccessCounter = 0;
+int overflowEvent   = 0;
+
+#endif
+
+/*
+ * External declaration
+ */
+extern HashTable dbHashtable;
+
+/*
+ * Internal declaration
+ */
+IX_ETH_DB_PUBLIC
+PortInfo ixEthDBPortInfo[IX_ETH_DB_NUMBER_OF_PORTS];
+
+IX_ETH_DB_PRIVATE
+struct
+{
+    BOOL saved;
+    IxEthDBPriorityTable priorityTable;
+    IxEthDBVlanSet vlanMembership;
+    IxEthDBVlanSet transmitTaggingInfo;
+    IxEthDBFrameFilter frameFilter;
+    IxEthDBTaggingAction taggingAction;
+    IxEthDBFirewallMode firewallMode;
+    BOOL stpBlocked;
+    BOOL srcAddressFilterEnabled;
+    UINT32 maxRxFrameSize;
+    UINT32 maxTxFrameSize;
+} ixEthDBPortState[IX_ETH_DB_NUMBER_OF_PORTS];
+
+#define IX_ETH_DB_DEFAULT_FRAME_SIZE (1518)
+
+/**
+ * @brief initializes a port
+ *
+ * @param portID ID of the port to be initialized
+ *
+ * Note that redundant initializations are silently
+ * dealt with and do not constitute an error
+ *
+ * This function is fully documented in the main
+ * header file, IxEthDB.h
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBPortInit(IxEthDBPortId portID)
+{
+    PortInfo *portInfo;
+
+    if (portID >= IX_ETH_DB_NUMBER_OF_PORTS)
+    {
+        return;
+    }
+
+    portInfo = &ixEthDBPortInfo[portID];
+
+    if (ixEthDBSingleEthNpeCheck(portID) != IX_ETH_DB_SUCCESS)
+    {
+        WARNING_LOG("EthDB: Unavailable Eth %d: Cannot initialize EthDB Port.\n", (UINT32) portID);
+
+        return;
+    }
+
+    if (portInfo->initialized)
+    {
+        /* redundant */
+        return;
+    }
+
+    /* initialize core fields */
+    portInfo->portID = portID;
+    SET_DEPENDENCY_MAP(portInfo->dependencyPortMap, portID);
+
+    /* default values */
+    portInfo->agingEnabled       = FALSE;
+    portInfo->enabled            = FALSE;
+    portInfo->macAddressUploaded = FALSE;
+    portInfo->maxRxFrameSize     = IX_ETHDB_DEFAULT_FRAME_SIZE;
+    portInfo->maxTxFrameSize     = IX_ETHDB_DEFAULT_FRAME_SIZE;
+
+    /* default update control values */
+    portInfo->updateMethod.searchTree              = NULL;
+    portInfo->updateMethod.searchTreePendingWrite  = FALSE;
+    portInfo->updateMethod.treeInitialized         = FALSE;
+    portInfo->updateMethod.updateEnabled           = FALSE;
+    portInfo->updateMethod.userControlled          = FALSE;
+
+    /* default WiFi parameters */
+    memset(portInfo->bbsid, 0, sizeof (portInfo->bbsid));
+    portInfo->frameControlDurationID = 0;
+
+    /* Ethernet NPE-specific initializations */
+    if (ixEthDBPortDefinitions[portID].type == IX_ETH_NPE)
+    {
+        /* update handler */
+        portInfo->updateMethod.updateHandler = ixEthDBNPEUpdateHandler;
+    }
+
+    /* initialize state save */
+    ixEthDBPortState[portID].saved = FALSE;
+
+    portInfo->initialized = TRUE;
+}
+
+/**
+ * @brief enables a port
+ *
+ * @param portID ID of the port to enable
+ *
+ * This function is fully documented in the main
+ * header file, IxEthDB.h
+ *
+ * @return IX_ETH_DB_SUCCESS if enabling was
+ * accomplished, or a meaningful error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBPortEnable(IxEthDBPortId portID)
+{
+    IxEthDBPortMap triggerPorts;
+    PortInfo *portInfo;
+
+    IX_ETH_DB_CHECK_PORT_EXISTS(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    portInfo = &ixEthDBPortInfo[portID];
+
+    if (portInfo->enabled)
+    {
+        /* redundant */
+        return IX_ETH_DB_SUCCESS;
+    }
+
+    SET_DEPENDENCY_MAP(triggerPorts, portID);
+
+    /* mark as enabled */
+    portInfo->enabled = TRUE;
+
+    /* Operation stops here when Ethernet Learning is not enabled */
+    if(IX_FEATURE_CTRL_SWCONFIG_DISABLED ==
+       ixFeatureCtrlSwConfigurationCheck(IX_FEATURECTRL_ETH_LEARNING))
+    {
+        return IX_ETH_DB_SUCCESS;
+    }
+
+    if (ixEthDBPortDefinitions[portID].type == IX_ETH_NPE && !portInfo->macAddressUploaded)
+    {
+        IX_ETH_DB_SUPPORT_TRACE("DB: (Support) MAC address not set on port %d, enable failed\n", portID);
+
+        /* must use UnicastAddressSet() before enabling an NPE port */
+        return IX_ETH_DB_MAC_UNINITIALIZED;
+    }
+
+    if (ixEthDBPortDefinitions[portID].type == IX_ETH_NPE)
+    {
+        IX_ETH_DB_SUPPORT_TRACE("DB: (Support) Attempting to enable the NPE callback for port %d...\n", portID);
+
+        if (!portInfo->updateMethod.userControlled
+                && ((portInfo->featureCapability & IX_ETH_DB_FILTERING) != 0))
+        {
+            portInfo->updateMethod.updateEnabled = TRUE;
+        }
+
+        /* if this is first time initialization then we already have
+           write access to the tree and can AccessRelease directly */
+        if (!portInfo->updateMethod.treeInitialized)
+        {
+            IX_ETH_DB_SUPPORT_TRACE("DB: (Support) Initializing tree for port %d\n", portID);
+
+            /* create an initial tree and release access into it */
+            ixEthDBUpdatePortLearningTrees(triggerPorts);
+
+            /* mark tree as being initialized */
+            portInfo->updateMethod.treeInitialized = TRUE;
+        }
+    }
+
+    if (ixEthDBPortState[portID].saved)
+    {
+        /* previous configuration data stored, restore state */
+        if ((portInfo->featureCapability & IX_ETH_DB_FIREWALL) != 0)
+        {
+            ixEthDBFirewallModeSet(portID, ixEthDBPortState[portID].firewallMode);
+            ixEthDBFirewallInvalidAddressFilterEnable(portID, ixEthDBPortState[portID].srcAddressFilterEnabled);
+        }
+
+#if 0 /* test-only */
+        if ((portInfo->featureCapability & IX_ETH_DB_VLAN_QOS) != 0)
+        {
+            ixEthDBAcceptableFrameTypeSet(portID, ixEthDBPortState[portID].frameFilter);
+            ixEthDBIngressVlanTaggingEnabledSet(portID, ixEthDBPortState[portID].taggingAction);
+
+            ixEthDBEgressVlanTaggingEnabledSet(portID, ixEthDBPortState[portID].transmitTaggingInfo);
+            ixEthDBPortVlanMembershipSet(portID, ixEthDBPortState[portID].vlanMembership);
+
+            ixEthDBPriorityMappingTableSet(portID, ixEthDBPortState[portID].priorityTable);
+        }
+#endif
+
+        if ((portInfo->featureCapability & IX_ETH_DB_SPANNING_TREE_PROTOCOL) != 0)
+        {
+            ixEthDBSpanningTreeBlockingStateSet(portID, ixEthDBPortState[portID].stpBlocked);
+        }
+
+        ixEthDBFilteringPortMaximumRxFrameSizeSet(portID, ixEthDBPortState[portID].maxRxFrameSize);
+        ixEthDBFilteringPortMaximumTxFrameSizeSet(portID, ixEthDBPortState[portID].maxTxFrameSize);
+
+        /* discard previous save */
+        ixEthDBPortState[portID].saved = FALSE;
+    }
+
+    IX_ETH_DB_SUPPORT_TRACE("DB: (Support) Enabling succeeded for port %d\n", portID);
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief disables a port
+ *
+ * @param portID ID of the port to disable
+ *
+ * This function is fully documented in the
+ * main header file, IxEthDB.h
+ *
+ * @return IX_ETH_DB_SUCCESS if disabling was
+ * successful or an appropriate error message
+ * otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBPortDisable(IxEthDBPortId portID)
+{
+    HashIterator iterator;
+    IxEthDBPortMap triggerPorts; /* ports who will have deleted records and therefore will need updating */
+    BOOL result;
+    PortInfo *portInfo;
+    IxEthDBFeature learningEnabled;
+#if 0 /* test-only */
+    IxEthDBPriorityTable classZeroTable;
+#endif
+
+    IX_ETH_DB_CHECK_PORT_EXISTS(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    portInfo = &ixEthDBPortInfo[portID];
+
+    if (!portInfo->enabled)
+    {
+        /* redundant */
+        return IX_ETH_DB_SUCCESS;
+    }
+
+    if (ixEthDBPortDefinitions[portID].type == IX_ETH_NPE)
+    {
+        /* save filtering state */
+        ixEthDBPortState[portID].firewallMode            = portInfo->firewallMode;
+        ixEthDBPortState[portID].frameFilter             = portInfo->frameFilter;
+        ixEthDBPortState[portID].taggingAction           = portInfo->taggingAction;
+        ixEthDBPortState[portID].stpBlocked              = portInfo->stpBlocked;
+        ixEthDBPortState[portID].srcAddressFilterEnabled = portInfo->srcAddressFilterEnabled;
+        ixEthDBPortState[portID].maxRxFrameSize          = portInfo->maxRxFrameSize;
+        ixEthDBPortState[portID].maxTxFrameSize          = portInfo->maxTxFrameSize;
+
+        memcpy(ixEthDBPortState[portID].vlanMembership, portInfo->vlanMembership, sizeof (IxEthDBVlanSet));
+        memcpy(ixEthDBPortState[portID].transmitTaggingInfo, portInfo->transmitTaggingInfo, sizeof (IxEthDBVlanSet));
+        memcpy(ixEthDBPortState[portID].priorityTable, portInfo->priorityTable, sizeof (IxEthDBPriorityTable));
+
+        ixEthDBPortState[portID].saved = TRUE;
+
+        /* now turn off all EthDB filtering features on the port */
+
+#if 0 /* test-only */
+        /* VLAN & QoS */
+        if ((portInfo->featureCapability & IX_ETH_DB_VLAN_QOS) != 0)
+        {
+            ixEthDBPortVlanMembershipRangeAdd((IxEthDBPortId) portID, 0, IX_ETH_DB_802_1Q_MAX_VLAN_ID);
+            ixEthDBEgressVlanRangeTaggingEnabledSet((IxEthDBPortId) portID, 0, IX_ETH_DB_802_1Q_MAX_VLAN_ID, FALSE);
+            ixEthDBAcceptableFrameTypeSet((IxEthDBPortId) portID, IX_ETH_DB_ACCEPT_ALL_FRAMES);
+            ixEthDBIngressVlanTaggingEnabledSet((IxEthDBPortId) portID, IX_ETH_DB_PASS_THROUGH);
+
+            memset(classZeroTable, 0, sizeof (classZeroTable));
+            ixEthDBPriorityMappingTableSet((IxEthDBPortId) portID, classZeroTable);
+        }
+#endif
+
+        /* STP */
+        if ((portInfo->featureCapability & IX_ETH_DB_SPANNING_TREE_PROTOCOL) != 0)
+        {
+            ixEthDBSpanningTreeBlockingStateSet((IxEthDBPortId) portID, FALSE);
+        }
+
+        /* Firewall */
+        if ((portInfo->featureCapability & IX_ETH_DB_FIREWALL) != 0)
+        {
+            ixEthDBFirewallModeSet((IxEthDBPortId) portID, IX_ETH_DB_FIREWALL_BLACK_LIST);
+            ixEthDBFirewallTableDownload((IxEthDBPortId) portID);
+            ixEthDBFirewallInvalidAddressFilterEnable((IxEthDBPortId) portID, FALSE);
+        }
+
+        /* Frame size filter */
+        ixEthDBFilteringPortMaximumFrameSizeSet((IxEthDBPortId) portID, IX_ETH_DB_DEFAULT_FRAME_SIZE);
+
+        /* WiFi */
+        if ((portInfo->featureCapability & IX_ETH_DB_WIFI_HEADER_CONVERSION) != 0)
+        {
+            ixEthDBWiFiConversionTableDownload((IxEthDBPortId) portID);
+        }
+
+        /* save and disable the learning feature bit */
+        learningEnabled          = portInfo->featureStatus & IX_ETH_DB_LEARNING;
+        portInfo->featureStatus &= ~IX_ETH_DB_LEARNING;
+    }
+    else
+    {
+        /* save the learning feature bit */
+        learningEnabled          = portInfo->featureStatus & IX_ETH_DB_LEARNING;
+    }
+
+    SET_EMPTY_DEPENDENCY_MAP(triggerPorts);
+
+    ixEthDBUpdateLock();
+
+    /* wipe out current entries for this port */
+    BUSY_RETRY(ixEthDBInitHashIterator(&dbHashtable, &iterator));
+
+    while (IS_ITERATOR_VALID(&iterator))
+    {
+        MacDescriptor *descriptor =  (MacDescriptor *) iterator.node->data;
+
+        /* check if the port match. If so, remove the entry  */
+        if (descriptor->portID == portID
+                && (descriptor->type == IX_ETH_DB_FILTERING_RECORD || descriptor->type == IX_ETH_DB_FILTERING_VLAN_RECORD)
+                && !descriptor->recordData.filteringData.staticEntry)
+        {
+            /* delete entry */
+            BUSY_RETRY(ixEthDBRemoveEntryAtHashIterator(&dbHashtable, &iterator));
+
+            /* add port to the set of update trigger ports */
+            JOIN_PORT_TO_MAP(triggerPorts, portID);
+        }
+        else
+        {
+            /* move to the next record */
+            BUSY_RETRY(ixEthDBIncrementHashIterator(&dbHashtable, &iterator));
+        }
+    }
+
+    if (ixEthDBPortDefinitions[portID].type == IX_ETH_NPE)
+    {
+        if (portInfo->updateMethod.searchTree != NULL)
+        {
+            ixEthDBFreeMacTreeNode(portInfo->updateMethod.searchTree);
+            portInfo->updateMethod.searchTree = NULL;
+        }
+
+        ixEthDBNPEUpdateHandler(portID, IX_ETH_DB_FILTERING_RECORD);
+    }
+
+    /* mark as disabled */
+    portInfo->enabled = FALSE;
+
+    /* disable updates unless the user has specifically altered the default behavior */
+    if (ixEthDBPortDefinitions[portID].type == IX_ETH_NPE)
+    {
+        if (!portInfo->updateMethod.userControlled)
+        {
+            portInfo->updateMethod.updateEnabled = FALSE;
+        }
+
+        /* make sure we re-initialize the NPE learning tree when the port is re-enabled */
+        portInfo->updateMethod.treeInitialized = FALSE;
+    }
+
+    ixEthDBUpdateUnlock();
+
+    /* restore learning feature bit */
+    portInfo->featureStatus |= learningEnabled;
+
+    /* if we've removed any records or lost any events make sure to force an update */
+    IS_EMPTY_DEPENDENCY_MAP(result, triggerPorts);
+
+    if (!result)
+    {
+        ixEthDBUpdatePortLearningTrees(triggerPorts);
+    }
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief sends the updated maximum Tx/Rx frame lengths to the NPE
+ *
+ * @param portID ID of the port to update
+ *
+ * @return IX_ETH_DB_SUCCESS if the update completed
+ * successfully or an appropriate error message otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+IxEthDBStatus ixEthDBPortFrameLengthsUpdate(IxEthDBPortId portID)
+{
+    IxNpeMhMessage message;
+    PortInfo *portInfo = &ixEthDBPortInfo[portID];
+    IX_STATUS result;
+
+    FILL_SETMAXFRAMELENGTHS_MSG(message, portID, portInfo->maxRxFrameSize, portInfo->maxTxFrameSize);
+
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+
+    return result;
+}
+
+/**
+ * @brief sets the port maximum Rx frame size
+ *
+ * @param portID ID of the port to set the frame size on
+ * @param maximumRxFrameSize maximum Rx frame size
+ *
+ * This function updates the internal data structures and
+ * calls ixEthDBPortFrameLengthsUpdate() for NPE update.
+ *
+ * This function is fully documented in the main header
+ * file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation was
+ * successfull or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFilteringPortMaximumRxFrameSizeSet(IxEthDBPortId portID, UINT32 maximumRxFrameSize)
+{
+    IX_ETH_DB_CHECK_PORT_EXISTS(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+	if (!ixEthDBPortInfo[portID].initialized)
+	{
+		return IX_ETH_DB_PORT_UNINITIALIZED;
+	}
+
+    if (ixEthDBPortDefinitions[portID].type == IX_ETH_NPE)
+    {
+	if ((maximumRxFrameSize < IX_ETHDB_MIN_NPE_FRAME_SIZE) ||
+	    (maximumRxFrameSize > IX_ETHDB_MAX_NPE_FRAME_SIZE))
+	{
+	    return IX_ETH_DB_INVALID_ARG;
+	}
+    }
+    else
+    {
+        return IX_ETH_DB_NO_PERMISSION;
+    }
+
+    /* update internal structure */
+    ixEthDBPortInfo[portID].maxRxFrameSize = maximumRxFrameSize;
+
+    /* update the maximum frame size in the NPE */
+    return ixEthDBPortFrameLengthsUpdate(portID);
+}
+
+/**
+ * @brief sets the port maximum Tx frame size
+ *
+ * @param portID ID of the port to set the frame size on
+ * @param maximumTxFrameSize maximum Tx frame size
+ *
+ * This function updates the internal data structures and
+ * calls ixEthDBPortFrameLengthsUpdate() for NPE update.
+ *
+ * This function is fully documented in the main header
+ * file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation was
+ * successfull or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFilteringPortMaximumTxFrameSizeSet(IxEthDBPortId portID, UINT32 maximumTxFrameSize)
+{
+    IX_ETH_DB_CHECK_PORT_EXISTS(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+	if (!ixEthDBPortInfo[portID].initialized)
+	{
+		return IX_ETH_DB_PORT_UNINITIALIZED;
+	}
+
+    if (ixEthDBPortDefinitions[portID].type == IX_ETH_NPE)
+    {
+	if ((maximumTxFrameSize < IX_ETHDB_MIN_NPE_FRAME_SIZE) ||
+	    (maximumTxFrameSize > IX_ETHDB_MAX_NPE_FRAME_SIZE))
+	{
+	    return IX_ETH_DB_INVALID_ARG;
+	}
+    }
+    else
+    {
+        return IX_ETH_DB_NO_PERMISSION;
+    }
+
+    /* update internal structure */
+    ixEthDBPortInfo[portID].maxTxFrameSize = maximumTxFrameSize;
+
+    /* update the maximum frame size in the NPE */
+    return ixEthDBPortFrameLengthsUpdate(portID);
+}
+
+/**
+ * @brief sets the port maximum Tx and Rx frame sizes
+ *
+ * @param portID ID of the port to set the frame size on
+ * @param maximumFrameSize maximum Tx and Rx frame sizes
+ *
+ * This function updates the internal data structures and
+ * calls ixEthDBPortFrameLengthsUpdate() for NPE update.
+ *
+ * Note that both the maximum Tx and Rx frame size are set
+ * to the same value. This function is kept for compatibility
+ * reasons.
+ *
+ * This function is fully documented in the main header
+ * file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation was
+ * successfull or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFilteringPortMaximumFrameSizeSet(IxEthDBPortId portID, UINT32 maximumFrameSize)
+{
+    IX_ETH_DB_CHECK_PORT_EXISTS(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    if (!ixEthDBPortInfo[portID].initialized)
+    {
+        return IX_ETH_DB_PORT_UNINITIALIZED;
+    }
+
+    if (ixEthDBPortDefinitions[portID].type == IX_ETH_NPE)
+    {
+	if ((maximumFrameSize < IX_ETHDB_MIN_NPE_FRAME_SIZE) ||
+	    (maximumFrameSize > IX_ETHDB_MAX_NPE_FRAME_SIZE))
+	{
+	    return IX_ETH_DB_INVALID_ARG;
+	}
+    }
+    else
+    {
+        return IX_ETH_DB_NO_PERMISSION;
+    }
+
+    /* update internal structure */
+    ixEthDBPortInfo[portID].maxRxFrameSize = maximumFrameSize;
+    ixEthDBPortInfo[portID].maxTxFrameSize = maximumFrameSize;
+
+    /* update the maximum frame size in the NPE */
+    return ixEthDBPortFrameLengthsUpdate(portID);
+}
+
+/**
+ * @brief sets the MAC address of an NPE port
+ *
+ * @param portID port ID to set the MAC address on
+ * @param macAddr pointer to the 6-byte MAC address
+ *
+ * This function is called by the EthAcc
+ * ixEthAccUnicastMacAddressSet() and should not be
+ * manually invoked unless required by special circumstances.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation succeeded
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBPortAddressSet(IxEthDBPortId portID, IxEthDBMacAddr *macAddr)
+{
+    IxNpeMhMessage message;
+    IxOsalMutex *ackPortAddressLock;
+    IX_STATUS result;
+
+    /* use this macro instead CHECK_PORT
+       as the port doesn't need to be enabled */
+    IX_ETH_DB_CHECK_PORT_EXISTS(portID);
+
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+
+    if (!ixEthDBPortInfo[portID].initialized)
+    {
+        return IX_ETH_DB_PORT_UNINITIALIZED;
+    }
+
+    ackPortAddressLock = &ixEthDBPortInfo[portID].npeAckLock;
+
+    /* Operation stops here when Ethernet Learning is not enabled */
+    if(IX_FEATURE_CTRL_SWCONFIG_DISABLED ==
+       ixFeatureCtrlSwConfigurationCheck(IX_FEATURECTRL_ETH_LEARNING))
+    {
+        return IX_ETH_DB_SUCCESS;
+    }
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    /* exit if the port is not an Ethernet NPE */
+    if (ixEthDBPortDefinitions[portID].type != IX_ETH_NPE)
+    {
+        return IX_ETH_DB_INVALID_PORT;
+    }
+
+    /* populate message */
+    FILL_SETPORTADDRESS_MSG(message, portID, macAddr->macAddress);
+
+    IX_ETH_DB_SUPPORT_TRACE("DB: (Support) Sending SetPortAddress on port %d...\n", portID);
+
+    /* send a SetPortAddress message */
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+
+    if (result == IX_SUCCESS)
+    {
+        ixEthDBPortInfo[portID].macAddressUploaded = TRUE;
+    }
+
+    return result;
+}

cpu/ixp/npe/IxEthDBCore.c

@@ -0,0 +1,463 @@
+/**
+ * @file IxEthDBDBCore.c
+ *
+ * @brief Database support functions
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxEthDB_p.h"
+
+/* list of database hashtables */
+IX_ETH_DB_PUBLIC HashTable dbHashtable;
+IX_ETH_DB_PUBLIC MatchFunction matchFunctions[IX_ETH_DB_MAX_KEY_INDEX + 1];
+IX_ETH_DB_PUBLIC BOOL ixEthDBPortUpdateRequired[IX_ETH_DB_MAX_RECORD_TYPE_INDEX + 1];
+IX_ETH_DB_PUBLIC UINT32 ixEthDBKeyType[IX_ETH_DB_MAX_RECORD_TYPE_INDEX + 1];
+
+/* private initialization flag */
+IX_ETH_DB_PRIVATE BOOL ethDBInitializationComplete = FALSE;
+
+/**
+ * @brief initializes EthDB
+ *
+ * This function must be called to initialize the component.
+ *
+ * It does the following things:
+ * - checks the port definition structure
+ * - scans the capabilities of the NPE images and sets the
+ *   capabilities of the ports accordingly
+ * - initializes the memory pools internally used in EthDB
+ *   for storing database records and handling data
+ * - registers automatic update handlers for add and remove
+ *   operations
+ * - registers hashing match functions, depending on key sets
+ * - initializes the main database hashtable
+ * - allocates contiguous memory zones to be used for NPE
+ *   updates
+ * - registers the serialize methods used to convert data
+ *   into NPE-readable format
+ * - starts the event processor
+ *
+ * Note that this function is documented in the public
+ * component header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS or an appropriate error if the
+ * component failed to initialize correctly
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBInit(void)
+{
+    IxEthDBStatus result;
+
+    if (ethDBInitializationComplete)
+    {
+        /* redundant */
+        return IX_ETH_DB_SUCCESS;
+    }
+
+    /* trap an invalid port definition structure */
+    IX_ETH_DB_PORTS_ASSERTION;
+
+    /* memory management */
+    ixEthDBInitMemoryPools();
+
+    /* register hashing search methods */
+    ixEthDBMatchMethodsRegister(matchFunctions);
+
+    /* register type-based automatic port updates */
+    ixEthDBUpdateTypeRegister(ixEthDBPortUpdateRequired);
+
+    /* register record to key type mappings */
+    ixEthDBKeyTypeRegister(ixEthDBKeyType);
+
+    /* hash table */
+    ixEthDBInitHash(&dbHashtable, NUM_BUCKETS, ixEthDBEntryXORHash, matchFunctions, (FreeFunction) ixEthDBFreeMacDescriptor);
+
+    /* NPE update zones */
+    ixEthDBNPEUpdateAreasInit();
+
+    /* register record serialization methods */
+    ixEthDBRecordSerializeMethodsRegister();
+
+    /* start the event processor */
+    result = ixEthDBEventProcessorInit();
+
+    /* scan NPE features */
+    if (result == IX_ETH_DB_SUCCESS)
+    {
+        ixEthDBFeatureCapabilityScan();
+    }
+
+    ethDBInitializationComplete = TRUE;
+
+    return result;
+}
+
+/**
+ * @brief prepares EthDB for unloading
+ *
+ * This function must be called before removing the
+ * EthDB component from memory (e.g. doing rmmod in
+ * Linux) if the component is to be re-initialized again
+ * without rebooting the platform.
+ *
+ * All the EthDB ports must be disabled before this
+ * function is to be called. Failure to disable all
+ * the ports will return the IX_ETH_DB_BUSY error.
+ *
+ * This function will destroy mutexes, deallocate
+ * memory and stop the event processor.
+ *
+ * Note that this function is fully documented in the
+ * main component header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if de-initialization
+ * completed successfully or an appropriate error
+ * message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBUnload(void)
+{
+    IxEthDBPortId portIndex;
+
+    if (!ethDBInitializationComplete)
+    {
+        /* redundant */
+        return IX_ETH_DB_SUCCESS;
+    }
+
+    /* check if any ports are enabled */
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        if (ixEthDBPortInfo[portIndex].enabled)
+        {
+            return IX_ETH_DB_BUSY;
+        }
+    }
+
+    /* free port resources */
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        if (ixEthDBPortDefinitions[portIndex].type == IX_ETH_NPE)
+        {
+            ixOsalMutexDestroy(&ixEthDBPortInfo[portIndex].npeAckLock);
+        }
+
+        ixEthDBPortInfo[portIndex].initialized = FALSE;
+    }
+
+    /* shutdown event processor */
+    ixEthDBStopLearningFunction();
+
+    /* deallocate NPE update zones */
+    ixEthDBNPEUpdateAreasUnload();
+
+    ethDBInitializationComplete = FALSE;
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief adds a new entry to the Ethernet database
+ *
+ * @param newRecordTemplate address of the record template to use
+ * @param updateTrigger port map containing the update triggers
+ * resulting from this update operation
+ *
+ * Creates a new database entry, populates it with the data
+ * copied from the given template and adds the record to the
+ * database hash table.
+ * It also checks whether the new record type is registered to trigger
+ * automatic updates; if it is, the update trigger will contain the
+ * port on which the record insertion was performed, as well as the
+ * old port in case the addition was a record migration (from one port
+ * to the other). The caller can use the updateTrigger to trigger
+ * automatic updates on the ports changed as a result of this addition.
+ *
+ * @retval IX_ETH_DB_SUCCESS addition successful
+ * @retval IX_ETH_DB_NOMEM insertion failed, no memory left in the mac descriptor memory pool
+ * @retval IX_ETH_DB_BUSY database busy, cannot insert due to locking
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBAdd(MacDescriptor *newRecordTemplate, IxEthDBPortMap updateTrigger)
+{
+    IxEthDBStatus result;
+    MacDescriptor *newDescriptor;
+    IxEthDBPortId originalPortID;
+    HashNode *node = NULL;
+
+    BUSY_RETRY(ixEthDBSearchHashEntry(&dbHashtable, ixEthDBKeyType[newRecordTemplate->type], newRecordTemplate, &node));
+
+    TEST_FIXTURE_INCREMENT_DB_CORE_ACCESS_COUNTER;
+
+    if (node == NULL)
+    {
+        /* not found, create a new one */
+        newDescriptor = ixEthDBAllocMacDescriptor();
+
+        if (newDescriptor == NULL)
+        {
+            return IX_ETH_DB_NOMEM; /* no memory */
+        }
+
+        /* old port does not exist, avoid unnecessary updates */
+        originalPortID = newRecordTemplate->portID;
+    }
+    else
+    {
+        /* a node with the same key exists, will update node */
+        newDescriptor = (MacDescriptor *) node->data;
+
+        /* save original port id */
+        originalPortID = newDescriptor->portID;
+    }
+
+    /* copy/update fields into new record */
+    memcpy(newDescriptor->macAddress, newRecordTemplate->macAddress, sizeof (IxEthDBMacAddr));
+    memcpy(&newDescriptor->recordData, &newRecordTemplate->recordData, sizeof (IxEthDBRecordData));
+
+    newDescriptor->type   = newRecordTemplate->type;
+    newDescriptor->portID = newRecordTemplate->portID;
+    newDescriptor->user   = newRecordTemplate->user;
+
+    if (node == NULL)
+    {
+        /* new record, insert into hashtable */
+        BUSY_RETRY_WITH_RESULT(ixEthDBAddHashEntry(&dbHashtable, newDescriptor), result);
+
+        if (result != IX_ETH_DB_SUCCESS)
+        {
+            ixEthDBFreeMacDescriptor(newDescriptor);
+
+            return result; /* insertion failed */
+        }
+    }
+
+    if (node != NULL)
+    {
+        /* release access */
+        ixEthDBReleaseHashNode(node);
+    }
+
+    /* trigger add/remove update if required by type */
+    if (updateTrigger != NULL &&
+        ixEthDBPortUpdateRequired[newRecordTemplate->type])
+    {
+        /* add new port to update list */
+        JOIN_PORT_TO_MAP(updateTrigger, newRecordTemplate->portID);
+
+        /* check if record has moved, we'll need to update the old port as well */
+        if (originalPortID != newDescriptor->portID)
+        {
+            JOIN_PORT_TO_MAP(updateTrigger, originalPortID);
+        }
+    }
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief remove a record from the Ethernet database
+ *
+ * @param templateRecord template record used to determine
+ * what record is to be removed
+ * @param updateTrigger port map containing the update triggers
+ * resulting from this update operation
+ *
+ * This function will examine the template record it receives
+ * and attempts to delete a record of the same type and containing
+ * the same keys as the template record. If deletion is successful
+ * and the record type is registered for automatic port updates the
+ * port will also be set in the updateTrigger port map, so that the
+ * client can perform an update of the port.
+ *
+ * @retval IX_ETH_DB_SUCCESS removal was successful
+ * @retval IX_ETH_DB_NO_SUCH_ADDR the record with the given MAC address was not found
+ * @retval IX_ETH_DB_BUSY database busy, cannot remove due to locking
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBRemove(MacDescriptor *templateRecord, IxEthDBPortMap updateTrigger)
+{
+    IxEthDBStatus result;
+
+    TEST_FIXTURE_INCREMENT_DB_CORE_ACCESS_COUNTER;
+
+    BUSY_RETRY_WITH_RESULT(ixEthDBRemoveHashEntry(&dbHashtable, ixEthDBKeyType[templateRecord->type], templateRecord), result);
+
+    if (result != IX_ETH_DB_SUCCESS)
+    {
+        return IX_ETH_DB_NO_SUCH_ADDR; /* not found */
+    }
+
+    /* trigger add/remove update if required by type */
+    if (updateTrigger != NULL
+        &&ixEthDBPortUpdateRequired[templateRecord->type])
+    {
+        /* add new port to update list */
+        JOIN_PORT_TO_MAP(updateTrigger, templateRecord->portID);
+    }
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief register record key types
+ *
+ * This function registers the appropriate key types,
+ * depending on record types.
+ *
+ * All filtering records use the MAC address as the key.
+ * WiFi and Firewall records use a compound key consisting
+ * in both the MAC address and the port ID.
+ *
+ * @return the number of registered record types
+ */
+IX_ETH_DB_PUBLIC
+UINT32 ixEthDBKeyTypeRegister(UINT32 *keyType)
+{
+    /* safety */
+    memset(keyType, 0, sizeof (keyType));
+
+    /* register all known record types */
+    keyType[IX_ETH_DB_FILTERING_RECORD]      = IX_ETH_DB_MAC_KEY;
+    keyType[IX_ETH_DB_FILTERING_VLAN_RECORD] = IX_ETH_DB_MAC_KEY;
+    keyType[IX_ETH_DB_ALL_FILTERING_RECORDS] = IX_ETH_DB_MAC_KEY;
+    keyType[IX_ETH_DB_WIFI_RECORD]           = IX_ETH_DB_MAC_PORT_KEY;
+    keyType[IX_ETH_DB_FIREWALL_RECORD]       = IX_ETH_DB_MAC_PORT_KEY;
+
+    return 5;
+}
+
+/**
+ * @brief Sets a user-defined field into a database record
+ *
+ * Note that this function is fully documented in the main component
+ * header file.
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBUserFieldSet(IxEthDBRecordType recordType, IxEthDBMacAddr *macAddr, IxEthDBPortId portID, IxEthDBVlanId vlanID, void *field)
+{
+    HashNode *result = NULL;
+
+    if (macAddr == NULL)
+    {
+        return IX_ETH_DB_INVALID_ARG;
+    }
+
+    if (recordType == IX_ETH_DB_FILTERING_RECORD)
+    {
+        result = ixEthDBSearch(macAddr, recordType);
+    }
+    else if (recordType == IX_ETH_DB_FILTERING_VLAN_RECORD)
+    {
+        result = ixEthDBVlanSearch(macAddr, vlanID, recordType);
+    }
+    else if (recordType == IX_ETH_DB_WIFI_RECORD || recordType == IX_ETH_DB_FIREWALL_RECORD)
+    {
+        IX_ETH_DB_CHECK_PORT_EXISTS(portID);
+
+        result = ixEthDBPortSearch(macAddr, portID, recordType);
+    }
+    else
+    {
+        return IX_ETH_DB_INVALID_ARG;
+    }
+
+    if (result == NULL)
+    {
+        return IX_ETH_DB_NO_SUCH_ADDR;
+    }
+
+    ((MacDescriptor *) result->data)->user = field;
+
+    ixEthDBReleaseHashNode(result);
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief Retrieves a user-defined field from a database record
+ *
+ * Note that this function is fully documented in the main component
+ * header file.
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBUserFieldGet(IxEthDBRecordType recordType, IxEthDBMacAddr *macAddr, IxEthDBPortId portID, IxEthDBVlanId vlanID, void **field)
+{
+    HashNode *result = NULL;
+
+    if (macAddr == NULL || field == NULL)
+    {
+        return IX_ETH_DB_INVALID_ARG;
+    }
+
+    if (recordType == IX_ETH_DB_FILTERING_RECORD)
+    {
+        result = ixEthDBSearch(macAddr, recordType);
+    }
+    else if (recordType == IX_ETH_DB_FILTERING_VLAN_RECORD)
+    {
+        result = ixEthDBVlanSearch(macAddr, vlanID, recordType);
+    }
+    else if (recordType == IX_ETH_DB_WIFI_RECORD || recordType == IX_ETH_DB_FIREWALL_RECORD)
+    {
+        IX_ETH_DB_CHECK_PORT_EXISTS(portID);
+
+        result = ixEthDBPortSearch(macAddr, portID, recordType);
+    }
+    else
+    {
+        return IX_ETH_DB_INVALID_ARG;
+    }
+
+    if (result == NULL)
+    {
+        return IX_ETH_DB_NO_SUCH_ADDR;
+    }
+
+    *field = ((MacDescriptor *) result->data)->user;
+
+    ixEthDBReleaseHashNode(result);
+
+    return IX_ETH_DB_SUCCESS;
+}

cpu/ixp/npe/IxEthDBEvents.c

@@ -0,0 +1,520 @@
+/**
+ * @file IxEthDBEvents.c
+ *
+ * @brief Implementation of the event processor component
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include <IxNpeMh.h>
+#include <IxFeatureCtrl.h>
+
+#include "IxEthDB_p.h"
+
+/* forward prototype declarations */
+IX_ETH_DB_PUBLIC void ixEthDBEventProcessorLoop(void *); 
+IX_ETH_DB_PUBLIC void ixEthDBNPEEventCallback(IxNpeMhNpeId npeID, IxNpeMhMessage msg);
+IX_ETH_DB_PRIVATE void ixEthDBProcessEvent(PortEvent *local_event, IxEthDBPortMap triggerPorts);
+IX_ETH_DB_PRIVATE IxEthDBStatus ixEthDBTriggerPortUpdate(UINT32 eventType, IxEthDBMacAddr *macAddr, IxEthDBPortId portID, BOOL staticEntry);
+IX_ETH_DB_PUBLIC IxEthDBStatus ixEthDBStartLearningFunction(void);
+IX_ETH_DB_PUBLIC IxEthDBStatus ixEthDBStopLearningFunction(void);
+
+/* data */
+IX_ETH_DB_PRIVATE IxOsalSemaphore eventQueueSemaphore;
+IX_ETH_DB_PRIVATE PortEventQueue eventQueue;
+IX_ETH_DB_PRIVATE IxOsalMutex eventQueueLock;
+IX_ETH_DB_PRIVATE IxOsalMutex portUpdateLock;
+
+IX_ETH_DB_PRIVATE BOOL ixEthDBLearningShutdown      = FALSE;
+IX_ETH_DB_PRIVATE BOOL ixEthDBEventProcessorRunning = FALSE;
+
+/* imported data */
+extern HashTable dbHashtable;
+
+/**
+ * @brief initializes the event processor
+ *
+ * Initializes the event processor queue and processing thread.
+ * Called from ixEthDBInit() DB-subcomponent master init function.
+ *
+ * @warning do not call directly
+ *
+ * @retval IX_ETH_DB_SUCCESS initialization was successful
+ * @retval IX_ETH_DB_FAIL initialization failed (OSAL or mutex init failure)
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBEventProcessorInit(void)
+{
+    if (ixOsalMutexInit(&portUpdateLock) != IX_SUCCESS)
+    {
+        return IX_ETH_DB_FAIL;
+    }
+
+    if (ixOsalMutexInit(&eventQueueLock) != IX_SUCCESS)
+    {
+        return IX_ETH_DB_FAIL;
+    }
+
+    if (IX_FEATURE_CTRL_SWCONFIG_ENABLED ==
+        ixFeatureCtrlSwConfigurationCheck (IX_FEATURECTRL_ETH_LEARNING))
+    {
+
+        /* start processor loop thread */
+        if (ixEthDBStartLearningFunction() != IX_ETH_DB_SUCCESS)
+        {
+            return IX_ETH_DB_FAIL;
+        }
+    }
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief initializes the event queue and the event processor
+ *
+ * This function is called by the component initialization
+ * function, ixEthDBInit().
+ *
+ * @warning do not call directly
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or IX_ETH_DB_FAIL otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBStartLearningFunction(void)
+{
+    IxOsalThread eventProcessorThread;
+    IxOsalThreadAttr threadAttr;
+
+    threadAttr.name      = "EthDB event thread";
+    threadAttr.stackSize = 32 * 1024; /* 32kbytes */
+    threadAttr.priority  = 128;
+
+    /* reset event queue */
+    ixOsalMutexLock(&eventQueueLock, IX_OSAL_WAIT_FOREVER);
+
+    RESET_QUEUE(&eventQueue);
+
+    ixOsalMutexUnlock(&eventQueueLock);
+
+    /* init event queue semaphore */
+    if (ixOsalSemaphoreInit(&eventQueueSemaphore, 0) != IX_SUCCESS)
+    {
+        return IX_ETH_DB_FAIL;
+    }
+
+    ixEthDBLearningShutdown = FALSE;
+
+    /* create processor loop thread */
+    if (ixOsalThreadCreate(&eventProcessorThread, &threadAttr, ixEthDBEventProcessorLoop, NULL) != IX_SUCCESS)
+    {
+        return IX_ETH_DB_FAIL;
+    }
+
+    /* start event processor */
+    ixOsalThreadStart(&eventProcessorThread);
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief stops the event processor
+ *
+ * Stops the event processor and frees the event queue semaphore
+ * Called by the component de-initialization function, ixEthDBUnload()
+ *
+ * @warning do not call directly
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed 
+ * successfully or IX_ETH_DB_FAIL otherwise;
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBStopLearningFunction(void)
+{
+    ixEthDBLearningShutdown = TRUE;
+
+    /* wake up event processing loop to actually process the shutdown event */
+    ixOsalSemaphorePost(&eventQueueSemaphore);
+
+    if (ixOsalSemaphoreDestroy(&eventQueueSemaphore) != IX_SUCCESS)
+    {
+        return IX_ETH_DB_FAIL;
+    }
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief default NPE event processing callback
+ *
+ * @param npeID ID of the NPE that generated the event
+ * @param msg NPE message (encapsulated event)
+ *
+ * Creates an event object on the Ethernet event processor queue
+ * and signals the new event by incrementing the event queue semaphore.
+ * Events are processed by @ref ixEthDBEventProcessorLoop() which runs
+ * at user level.
+ *
+ * @see ixEthDBEventProcessorLoop()
+ *
+ * @warning do not call directly
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBNPEEventCallback(IxNpeMhNpeId npeID, IxNpeMhMessage msg)
+{
+    PortEvent *local_event;
+
+    IX_ETH_DB_IRQ_EVENTS_TRACE("DB: (Events) new event received by processor callback from port %d, id 0x%X\n", IX_ETH_DB_NPE_TO_PORT_ID(npeID), NPE_MSG_ID(msg), 0, 0, 0, 0);
+
+    if (CAN_ENQUEUE(&eventQueue))
+    {
+        TEST_FIXTURE_LOCK_EVENT_QUEUE;
+
+        local_event = QUEUE_HEAD(&eventQueue);
+
+        /* create event structure on queue */
+        local_event->eventType = NPE_MSG_ID(msg);
+        local_event->portID    = IX_ETH_DB_NPE_TO_PORT_ID(npeID);
+        
+        /* update queue */
+        PUSH_UPDATE_QUEUE(&eventQueue);
+
+        TEST_FIXTURE_UNLOCK_EVENT_QUEUE;
+
+        IX_ETH_DB_IRQ_EVENTS_TRACE("DB: (Events) Waking up main processor loop...\n", 0, 0, 0, 0, 0, 0);
+
+        /* increment event queue semaphore */
+        ixOsalSemaphorePost(&eventQueueSemaphore);
+    }
+    else
+    {
+        IX_ETH_DB_IRQ_EVENTS_TRACE("DB: (Events) Warning: could not enqueue event (overflow)\n", 0, 0, 0, 0, 0, 0);
+    }
+}
+
+/**
+ * @brief Ethernet event processor loop
+ *
+ * Extracts at most EVENT_PROCESSING_LIMIT batches of events and
+ * sends them for processing to @ref ixEthDBProcessEvent().
+ * Triggers port updates which normally follow learning events.
+ *
+ * @warning do not call directly, executes in separate thread
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBEventProcessorLoop(void *unused1)
+{
+    IxEthDBPortMap triggerPorts;
+    IxEthDBPortId portIndex;
+
+    ixEthDBEventProcessorRunning = TRUE;
+
+    IX_ETH_DB_EVENTS_TRACE("DB: (Events) Event processor loop was started\n");
+
+    while (!ixEthDBLearningShutdown)
+    {
+        BOOL keepProcessing    = TRUE;
+        UINT32 processedEvents = 0;
+
+        IX_ETH_DB_EVENTS_VERBOSE_TRACE("DB: (Events) Waiting for new learning event...\n");
+
+        ixOsalSemaphoreWait(&eventQueueSemaphore, IX_OSAL_WAIT_FOREVER);
+
+        IX_ETH_DB_EVENTS_VERBOSE_TRACE("DB: (Events) Received new event\n");
+
+        if (!ixEthDBLearningShutdown)
+        {
+            /* port update handling */
+            SET_EMPTY_DEPENDENCY_MAP(triggerPorts);
+
+            while (keepProcessing)
+            {
+                PortEvent local_event;
+                UINT32 intLockKey;
+
+                /* lock queue */
+                ixOsalMutexLock(&eventQueueLock, IX_OSAL_WAIT_FOREVER);
+
+                /* lock NPE interrupts */
+                intLockKey = ixOsalIrqLock();
+
+                /* extract event */
+                local_event = *(QUEUE_TAIL(&eventQueue));
+
+                SHIFT_UPDATE_QUEUE(&eventQueue);
+
+                ixOsalIrqUnlock(intLockKey);
+
+                ixOsalMutexUnlock(&eventQueueLock);
+
+                IX_ETH_DB_EVENTS_TRACE("DB: (Events) Processing event with ID 0x%X\n", local_event.eventType);
+
+                ixEthDBProcessEvent(&local_event, triggerPorts);
+
+                processedEvents++;
+
+                if (processedEvents > EVENT_PROCESSING_LIMIT /* maximum burst reached? */
+                    || ixOsalSemaphoreTryWait(&eventQueueSemaphore) != IX_SUCCESS) /* or empty queue? */
+                {
+                    keepProcessing = FALSE;
+                }
+            }
+
+            ixEthDBUpdatePortLearningTrees(triggerPorts);
+        }
+    }
+
+    /* turn off automatic updates */
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        ixEthDBPortInfo[portIndex].updateMethod.updateEnabled = FALSE;
+    }
+
+    ixEthDBEventProcessorRunning = FALSE;
+}
+
+/**
+ * @brief event processor routine
+ *
+ * @param event event to be processed
+ * @param triggerPorts port map accumulating ports to be updated
+ *
+ * Processes learning events by synchronizing the database with
+ * newly learnt data. Called only by @ref ixEthDBEventProcessorLoop().
+ *
+ * @warning do not call directly
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBProcessEvent(PortEvent *local_event, IxEthDBPortMap triggerPorts)
+{
+    MacDescriptor recordTemplate;
+
+    switch (local_event->eventType)
+    {
+        case IX_ETH_DB_ADD_FILTERING_RECORD:
+            /* add record */
+            memset(&recordTemplate, 0, sizeof (recordTemplate));
+            memcpy(recordTemplate.macAddress, local_event->macAddr.macAddress, sizeof (IxEthDBMacAddr));
+            
+            recordTemplate.type   = IX_ETH_DB_FILTERING_RECORD;
+            recordTemplate.portID = local_event->portID;
+            recordTemplate.recordData.filteringData.staticEntry = local_event->staticEntry;
+            
+            ixEthDBAdd(&recordTemplate, triggerPorts);
+
+            IX_ETH_DB_EVENTS_TRACE("DB: (Events) Added record on port %d\n", local_event->portID);
+
+            break;
+
+        case IX_ETH_DB_REMOVE_FILTERING_RECORD:
+            /* remove record */
+            memset(&recordTemplate, 0, sizeof (recordTemplate));
+            memcpy(recordTemplate.macAddress, local_event->macAddr.macAddress, sizeof (IxEthDBMacAddr));
+            
+            recordTemplate.type = IX_ETH_DB_FILTERING_RECORD | IX_ETH_DB_FILTERING_VLAN_RECORD;
+            
+            ixEthDBRemove(&recordTemplate, triggerPorts);
+            
+            IX_ETH_DB_EVENTS_TRACE("DB: (Events) Removed record on port %d\n", local_event->portID);
+
+            break;
+
+        default:
+            /* can't handle/not interested in this event type */
+            ERROR_LOG("DB: (Events) Event processor received an unknown event type (0x%X)\n", local_event->eventType);
+
+            return;
+    }
+}
+
+/**
+ * @brief asynchronously adds a filtering record
+ * by posting an ADD_FILTERING_RECORD event to the event queue
+ *
+ * @param macAddr MAC address of the new record
+ * @param portID port ID of the new record
+ * @param staticEntry TRUE if record is static, FALSE if dynamic
+ *
+ * @return IX_ETH_DB_SUCCESS if the event creation was
+ * successfull or IX_ETH_DB_BUSY if the event queue is full
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBTriggerAddPortUpdate(IxEthDBMacAddr *macAddr, IxEthDBPortId portID, BOOL staticEntry)
+{
+    MacDescriptor reference;
+    
+    TEST_FIXTURE_INCREMENT_DB_CORE_ACCESS_COUNTER;
+
+    /* fill search fields */
+    memcpy(reference.macAddress, macAddr, sizeof (IxEthDBMacAddr));
+    reference.portID = portID;
+    
+    /* set acceptable record types */
+    reference.type = IX_ETH_DB_ALL_FILTERING_RECORDS;
+
+    if (ixEthDBPeekHashEntry(&dbHashtable, IX_ETH_DB_MAC_PORT_KEY, &reference) == IX_ETH_DB_SUCCESS)
+    {
+        /* already have an identical record */
+        return IX_ETH_DB_SUCCESS;
+    }
+    else
+    {
+        return ixEthDBTriggerPortUpdate(IX_ETH_DB_ADD_FILTERING_RECORD, macAddr, portID, staticEntry);
+    }
+}
+
+/**
+ * @brief asynchronously removes a filtering record
+ * by posting a REMOVE_FILTERING_RECORD event to the event queue
+ *
+ * @param macAddr MAC address of the record to remove
+ * @param portID port ID of the record to remove
+ *
+ * @return IX_ETH_DB_SUCCESS if the event creation was
+ * successfull or IX_ETH_DB_BUSY if the event queue is full
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBTriggerRemovePortUpdate(IxEthDBMacAddr *macAddr, IxEthDBPortId portID)
+{
+    if (ixEthDBPeek(macAddr, IX_ETH_DB_ALL_FILTERING_RECORDS) != IX_ETH_DB_NO_SUCH_ADDR)
+    {
+        return ixEthDBTriggerPortUpdate(IX_ETH_DB_REMOVE_FILTERING_RECORD, macAddr, portID, FALSE);
+    }
+    else
+    {
+        return IX_ETH_DB_NO_SUCH_ADDR;
+    }
+}
+
+/**
+ * @brief adds an ADD or REMOVE event to the main event queue
+ *
+ * @param eventType event type - IX_ETH_DB_ADD_FILTERING_RECORD 
+ * to add and IX_ETH_DB_REMOVE_FILTERING_RECORD to remove a
+ * record.
+ *
+ * @return IX_ETH_DB_SUCCESS if the event was successfully
+ * sent or IX_ETH_DB_BUSY if the event queue is full
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+IxEthDBStatus ixEthDBTriggerPortUpdate(UINT32 eventType, IxEthDBMacAddr *macAddr, IxEthDBPortId portID, BOOL staticEntry)
+{
+    UINT32 intLockKey;
+
+    /* lock interrupts to protect queue */
+    intLockKey = ixOsalIrqLock();
+
+    if (CAN_ENQUEUE(&eventQueue))
+    {
+        PortEvent *queueEvent = QUEUE_HEAD(&eventQueue);
+
+        /* update fields on the queue */
+        memcpy(queueEvent->macAddr.macAddress, macAddr->macAddress, sizeof (IxEthDBMacAddr));
+        
+        queueEvent->eventType     = eventType;
+        queueEvent->portID        = portID;
+        queueEvent->staticEntry   = staticEntry;
+
+        PUSH_UPDATE_QUEUE(&eventQueue);
+
+        /* imcrement event queue semaphore */
+        ixOsalSemaphorePost(&eventQueueSemaphore);
+        
+        /* unlock interrupts */
+        ixOsalIrqUnlock(intLockKey);
+
+        return IX_ETH_DB_SUCCESS;
+    }
+    else /* event queue full */
+    {
+        /* unlock interrupts */
+        ixOsalIrqUnlock(intLockKey);
+
+        return IX_ETH_DB_BUSY;
+    }
+}
+
+/**
+ * @brief Locks learning tree updates and port disable
+ *
+ *
+ * This function locks portUpdateLock single mutex. It is primarily used
+ * to avoid executing 'port disable' during ELT maintenance.
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBUpdateLock(void)
+{
+    ixOsalMutexLock(&portUpdateLock, IX_OSAL_WAIT_FOREVER);
+}
+
+/**
+ * @brief Unlocks learning tree updates and port disable
+ *
+ *
+ * This function unlocks a portUpdateLock mutex. It is primarily used
+ * to avoid executing 'port disable' during ELT maintenance.
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBUpdateUnlock(void)
+{
+    ixOsalMutexUnlock(&portUpdateLock);
+}
+

cpu/ixp/npe/IxEthDBFeatures.c

@@ -0,0 +1,662 @@
+/**
+ * @file IxEthDBFeatures.c
+ *
+ * @brief Implementation of the EthDB feature control API
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxNpeDl.h"
+#include "IxEthDBQoS.h"
+#include "IxEthDB_p.h"
+
+/**
+ * @brief scans the capabilities of the loaded NPE images
+ *
+ * This function MUST be called by the ixEthDBInit() function.
+ * No EthDB features (including learning and filtering) are enabled
+ * before this function is called.
+ *
+ * @return none
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBFeatureCapabilityScan(void)
+{
+    IxNpeDlImageId imageId, npeAImageId;
+    IxEthDBPortId portIndex;
+    PortInfo *portInfo;
+    IxEthDBPriorityTable defaultPriorityTable;
+    IX_STATUS result;
+    UINT32 queueIndex;
+    UINT32 queueStructureIndex;
+    UINT32 trafficClassDefinitionIndex;
+
+    /* read version of NPE A - required to set the AQM queues for B and C */
+    npeAImageId.functionalityId = 0;
+    ixNpeDlLoadedImageGet(IX_NPEDL_NPEID_NPEA, &npeAImageId);
+
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        IxNpeMhMessage msg;
+
+        portInfo = &ixEthDBPortInfo[portIndex];
+
+        /* check and bypass if NPE B or C is fused out */
+        if (ixEthDBSingleEthNpeCheck(portIndex) != IX_ETH_DB_SUCCESS) continue;
+
+        /* all ports are capable of LEARNING by default */
+        portInfo->featureCapability |= IX_ETH_DB_LEARNING;
+        portInfo->featureStatus     |= IX_ETH_DB_LEARNING;
+
+        if (ixEthDBPortDefinitions[portIndex].type == IX_ETH_NPE)
+        {
+
+            if (ixNpeDlLoadedImageGet(IX_ETH_DB_PORT_ID_TO_NPE(portIndex), &imageId) != IX_SUCCESS)
+            {
+                WARNING_LOG("DB: (FeatureScan) NpeDl did not provide the image ID for NPE port %d\n", portIndex);
+            }
+            else
+            {
+                /* initialize and empty NPE response mutex */
+                ixOsalMutexInit(&portInfo->npeAckLock);
+                ixOsalMutexLock(&portInfo->npeAckLock, IX_OSAL_WAIT_FOREVER);
+
+                /* check NPE response to GetStatus */
+                msg.data[0] = IX_ETHNPE_NPE_GETSTATUS << 24;
+                msg.data[1] = 0;
+                IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portIndex), msg, result);
+                if (result != IX_SUCCESS)
+                {
+                    WARNING_LOG("DB: (FeatureScan) warning, could not send message to the NPE\n");
+                    continue;
+                }
+
+
+                if (imageId.functionalityId == 0x00
+                    || imageId.functionalityId == 0x03
+                    || imageId.functionalityId == 0x04
+                    || imageId.functionalityId == 0x80)
+                {
+                    portInfo->featureCapability |= IX_ETH_DB_FILTERING;
+                    portInfo->featureCapability |= IX_ETH_DB_FIREWALL;
+                    portInfo->featureCapability |= IX_ETH_DB_SPANNING_TREE_PROTOCOL;
+                }
+                else if (imageId.functionalityId == 0x01
+                         || imageId.functionalityId == 0x81)
+                {
+                    portInfo->featureCapability |= IX_ETH_DB_FILTERING;
+                    portInfo->featureCapability |= IX_ETH_DB_FIREWALL;
+                    portInfo->featureCapability |= IX_ETH_DB_SPANNING_TREE_PROTOCOL;
+                    portInfo->featureCapability |= IX_ETH_DB_VLAN_QOS;
+                }
+                else if (imageId.functionalityId == 0x02
+                         || imageId.functionalityId == 0x82)
+                {
+                    portInfo->featureCapability |= IX_ETH_DB_WIFI_HEADER_CONVERSION;
+                    portInfo->featureCapability |= IX_ETH_DB_FIREWALL;
+                    portInfo->featureCapability |= IX_ETH_DB_SPANNING_TREE_PROTOCOL;
+                    portInfo->featureCapability |= IX_ETH_DB_VLAN_QOS;
+                }
+
+                /* reset AQM queues */
+                memset(portInfo->ixEthDBTrafficClassAQMAssignments, 0, sizeof (portInfo->ixEthDBTrafficClassAQMAssignments));
+
+                /* ensure there's at least one traffic class record in the definition table, otherwise we have no default case, hence no queues */
+                IX_ENSURE(sizeof (ixEthDBTrafficClassDefinitions) != 0, "DB: no traffic class definitions found, check IxEthDBQoS.h");
+
+                /* find the traffic class definition index compatible with the current NPE A functionality ID */
+                for (trafficClassDefinitionIndex = 0 ;
+                    trafficClassDefinitionIndex < sizeof (ixEthDBTrafficClassDefinitions) / sizeof (ixEthDBTrafficClassDefinitions[0]);
+                    trafficClassDefinitionIndex++)
+                {
+                    if (ixEthDBTrafficClassDefinitions[trafficClassDefinitionIndex][IX_ETH_DB_NPE_A_FUNCTIONALITY_ID_INDEX] == npeAImageId.functionalityId)
+                    {
+                        /* found it */
+                        break;
+                    }
+                }
+
+                /* select the default case if we went over the array boundary */
+                if (trafficClassDefinitionIndex == sizeof (ixEthDBTrafficClassDefinitions) / sizeof (ixEthDBTrafficClassDefinitions[0]))
+                {
+                    trafficClassDefinitionIndex = 0; /* the first record is the default case */
+                }
+
+                /* select queue assignment structure based on the traffic class configuration index */
+                queueStructureIndex = ixEthDBTrafficClassDefinitions[trafficClassDefinitionIndex][IX_ETH_DB_QUEUE_ASSIGNMENT_INDEX];
+
+                /* only traffic class 0 is active at initialization time */
+                portInfo->ixEthDBTrafficClassCount = 1;
+
+                /* enable port, VLAN and Firewall feature bits to initialize QoS/VLAN/Firewall configuration */
+                portInfo->featureStatus |= IX_ETH_DB_VLAN_QOS;
+                portInfo->featureStatus |= IX_ETH_DB_FIREWALL;
+                portInfo->enabled        = TRUE;
+
+#define CONFIG_WITH_VLAN  /* test-only: VLAN support not included to save space!!! */
+#ifdef CONFIG_WITH_VLAN /* test-only: VLAN support not included to save space!!! */
+                /* set VLAN initial configuration (permissive) */
+                if ((portInfo->featureCapability & IX_ETH_DB_VLAN_QOS) != 0) /* QoS-enabled image */
+                {
+                    /* QoS capable */
+                    portInfo->ixEthDBTrafficClassAvailable = ixEthDBTrafficClassDefinitions[trafficClassDefinitionIndex][IX_ETH_DB_TRAFFIC_CLASS_COUNT_INDEX];
+
+                    /* set AQM queues */
+                    for (queueIndex = 0 ; queueIndex < IX_IEEE802_1Q_QOS_PRIORITY_COUNT ; queueIndex++)
+                    {
+                        portInfo->ixEthDBTrafficClassAQMAssignments[queueIndex] = ixEthDBQueueAssignments[queueStructureIndex][queueIndex];
+                    }
+
+                    /* set default PVID (0) and default traffic class 0 */
+                    ixEthDBPortVlanTagSet(portIndex, 0);
+
+                    /* enable reception of all frames */
+                    ixEthDBAcceptableFrameTypeSet(portIndex, IX_ETH_DB_ACCEPT_ALL_FRAMES);
+
+                    /* clear full VLAN membership */
+                    ixEthDBPortVlanMembershipRangeRemove(portIndex, 0, IX_ETH_DB_802_1Q_MAX_VLAN_ID);
+
+                    /* clear TTI table - no VLAN tagged frames will be transmitted */
+                    ixEthDBEgressVlanRangeTaggingEnabledSet(portIndex, 0, 4094, FALSE);
+
+                    /* set membership on 0, otherwise no Tx or Rx is working */
+                    ixEthDBPortVlanMembershipAdd(portIndex, 0);
+                }
+                else /* QoS not available in this image */
+#endif /* test-only */
+                {
+                    /* initialize traffic class availability (only class 0 is available) */
+                    portInfo->ixEthDBTrafficClassAvailable = 1;
+
+                    /* point all AQM queues to traffic class 0 */
+                    for (queueIndex = 0 ; queueIndex < IX_IEEE802_1Q_QOS_PRIORITY_COUNT ; queueIndex++)
+                    {
+                        portInfo->ixEthDBTrafficClassAQMAssignments[queueIndex] =
+                            ixEthDBQueueAssignments[queueStructureIndex][0];
+                    }
+                }
+
+#ifdef CONFIG_WITH_VLAN /* test-only: VLAN support not included to save space!!! */
+                /* download priority mapping table and Rx queue configuration */
+                memset (defaultPriorityTable, 0, sizeof (defaultPriorityTable));
+                ixEthDBPriorityMappingTableSet(portIndex, defaultPriorityTable);
+#endif
+
+                /* by default we turn off invalid source MAC address filtering */
+                ixEthDBFirewallInvalidAddressFilterEnable(portIndex, FALSE);
+
+                /* disable port, VLAN, Firewall feature bits */
+                portInfo->featureStatus &= ~IX_ETH_DB_VLAN_QOS;
+                portInfo->featureStatus &= ~IX_ETH_DB_FIREWALL;
+                portInfo->enabled        = FALSE;
+
+                /* enable filtering by default if present */
+                if ((portInfo->featureCapability & IX_ETH_DB_FILTERING) != 0)
+                {
+                    portInfo->featureStatus |= IX_ETH_DB_FILTERING;
+                }
+            }
+        }
+    }
+}
+
+/**
+ * @brief returns the capability of a port
+ *
+ * @param portID ID of the port
+ * @param featureSet location to store the port capability in
+ *
+ * This function will save the capability set of the given port
+ * into the given location. Capabilities are bit-ORed, each representing
+ * a bit of the feature set.
+ *
+ * Note that this function is documented in the main component
+ * public header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or IX_ETH_DB_INVALID_PORT if the given port is invalid
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFeatureCapabilityGet(IxEthDBPortId portID, IxEthDBFeature *featureSet)
+{
+    IX_ETH_DB_CHECK_PORT_INITIALIZED(portID);
+
+    IX_ETH_DB_CHECK_REFERENCE(featureSet);
+
+    *featureSet = ixEthDBPortInfo[portID].featureCapability;
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief enables or disables a port capability
+ *
+ * @param portID ID of the port
+ * @param feature feature to enable or disable
+ * @param enabled TRUE to enable the selected feature or FALSE to disable it
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFeatureEnable(IxEthDBPortId portID, IxEthDBFeature feature, BOOL enable)
+{
+    PortInfo *portInfo;
+    IxEthDBPriorityTable defaultPriorityTable;
+    IxEthDBVlanSet vlanSet;
+    IxEthDBStatus status = IX_ETH_DB_SUCCESS;
+    BOOL portEnabled;
+
+    IX_ETH_DB_CHECK_PORT_INITIALIZED(portID);
+
+    portInfo    = &ixEthDBPortInfo[portID];
+    portEnabled = portInfo->enabled;
+
+    /* check that only one feature is selected */
+    if (!ixEthDBCheckSingleBitValue(feature))
+    {
+        return IX_ETH_DB_FEATURE_UNAVAILABLE;
+    }
+
+    /* port capable of this feature? */
+    if ((portInfo->featureCapability & feature) == 0)
+    {
+        return IX_ETH_DB_FEATURE_UNAVAILABLE;
+    }
+
+    /* mutual exclusion between learning and WiFi header conversion */
+    if (enable && ((feature | portInfo->featureStatus) & (IX_ETH_DB_FILTERING | IX_ETH_DB_WIFI_HEADER_CONVERSION))
+            == (IX_ETH_DB_FILTERING | IX_ETH_DB_WIFI_HEADER_CONVERSION))
+    {
+        return IX_ETH_DB_NO_PERMISSION;
+    }
+
+    /* learning must be enabled before filtering */
+    if (enable && (feature == IX_ETH_DB_FILTERING) && ((portInfo->featureStatus & IX_ETH_DB_LEARNING) == 0))
+    {
+        return IX_ETH_DB_NO_PERMISSION;
+    }
+
+    /* filtering must be disabled before learning */
+    if (!enable && (feature == IX_ETH_DB_LEARNING) && ((portInfo->featureStatus & IX_ETH_DB_FILTERING) != 0))
+    {
+        return IX_ETH_DB_NO_PERMISSION;
+    }
+
+    /* redundant enabling or disabling */
+    if ((!enable && ((portInfo->featureStatus & feature) == 0))
+        || (enable && ((portInfo->featureStatus & feature) != 0)))
+    {
+        /* do nothing */
+        return IX_ETH_DB_SUCCESS;
+    }
+
+    /* force port enabled */
+    portInfo->enabled = TRUE;
+
+    if (enable)
+    {
+        /* turn on enable bit */
+        portInfo->featureStatus |= feature;
+
+#ifdef CONFIG_WITH_VLAN /* test-only: VLAN support not included to save space!!! */
+        /* if this is VLAN/QoS set the default priority table */
+        if (feature == IX_ETH_DB_VLAN_QOS)
+        {
+            /* turn on VLAN/QoS (most permissive mode):
+                - set default 802.1Q priority mapping table, in accordance to the
+                  availability of traffic classes
+                - set the acceptable frame filter to accept all
+                - set the Ingress tagging mode to pass-through
+                - set full VLAN membership list
+                - set full TTI table
+                - set the default 802.1Q tag to 0 (VLAN ID 0, Pri 0, CFI 0)
+                - enable TPID port extraction
+            */
+
+            portInfo->ixEthDBTrafficClassCount = portInfo->ixEthDBTrafficClassAvailable;
+
+            /* set default 802.1Q priority mapping table - note that C indexing starts from 0, so we substract 1 here */
+            memcpy (defaultPriorityTable,
+                (const void *) ixEthIEEE802_1QUserPriorityToTrafficClassMapping[portInfo->ixEthDBTrafficClassCount - 1],
+                sizeof (defaultPriorityTable));
+
+            /* update priority mapping and AQM queue assignments */
+            status = ixEthDBPriorityMappingTableSet(portID, defaultPriorityTable);
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBAcceptableFrameTypeSet(portID, IX_ETH_DB_ACCEPT_ALL_FRAMES);
+            }
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBIngressVlanTaggingEnabledSet(portID, IX_ETH_DB_PASS_THROUGH);
+            }
+
+            /* set membership and TTI tables */
+            memset (vlanSet, 0xFF, sizeof (vlanSet));
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                /* use the internal function to bypass PVID check */
+                status = ixEthDBPortVlanTableSet(portID, portInfo->vlanMembership, vlanSet);
+            }
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                /* use the internal function to bypass PVID check */
+                status = ixEthDBPortVlanTableSet(portID, portInfo->transmitTaggingInfo, vlanSet);
+            }
+
+            /* reset the PVID */
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBPortVlanTagSet(portID, 0);
+            }
+
+            /* enable TPID port extraction */
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBVlanPortExtractionEnable(portID, TRUE);
+            }
+        }
+        else if (feature == IX_ETH_DB_FIREWALL)
+#endif
+        {
+            /* firewall starts in black-list mode unless otherwise configured before *
+             * note that invalid source MAC address filtering is disabled by default */
+            if (portInfo->firewallMode != IX_ETH_DB_FIREWALL_BLACK_LIST
+                && portInfo->firewallMode != IX_ETH_DB_FIREWALL_WHITE_LIST)
+            {
+                status = ixEthDBFirewallModeSet(portID, IX_ETH_DB_FIREWALL_BLACK_LIST);
+
+                if (status == IX_ETH_DB_SUCCESS)
+                {
+                    status = ixEthDBFirewallInvalidAddressFilterEnable(portID, FALSE);
+                }
+            }
+        }
+
+        if (status != IX_ETH_DB_SUCCESS)
+        {
+            /* checks failed, disable */
+            portInfo->featureStatus &= ~feature;
+        }
+    }
+    else
+    {
+        /* turn off features */
+        if (feature == IX_ETH_DB_FIREWALL)
+        {
+            /* turning off the firewall is equivalent to:
+                - set to black-list mode
+                - clear all the entries and download the new table
+                - turn off the invalid source address checking
+            */
+
+            status = ixEthDBDatabaseClear(portID, IX_ETH_DB_FIREWALL_RECORD);
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBFirewallModeSet(portID, IX_ETH_DB_FIREWALL_BLACK_LIST);
+            }
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBFirewallInvalidAddressFilterEnable(portID, FALSE);
+            }
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBFirewallTableDownload(portID);
+            }
+        }
+        else if (feature == IX_ETH_DB_WIFI_HEADER_CONVERSION)
+        {
+            /* turn off header conversion */
+            status = ixEthDBDatabaseClear(portID, IX_ETH_DB_WIFI_RECORD);
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBWiFiConversionTableDownload(portID);
+            }
+        }
+#ifdef CONFIG_WITH_VLAN /* test-only: VLAN support not included to save space!!! */
+        else if (feature == IX_ETH_DB_VLAN_QOS)
+        {
+            /* turn off VLAN/QoS:
+                - set a priority mapping table with one traffic class
+                - set the acceptable frame filter to accept all
+                - set the Ingress tagging mode to pass-through
+                - clear the VLAN membership list
+                - clear the TTI table
+                - set the default 802.1Q tag to 0 (VLAN ID 0, Pri 0, CFI 0)
+                - disable TPID port extraction
+            */
+
+            /* initialize all => traffic class 0 priority mapping table */
+            memset (defaultPriorityTable, 0, sizeof (defaultPriorityTable));
+            portInfo->ixEthDBTrafficClassCount = 1;
+            status = ixEthDBPriorityMappingTableSet(portID, defaultPriorityTable);
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBAcceptableFrameTypeSet(portID, IX_ETH_DB_ACCEPT_ALL_FRAMES);
+            }
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBIngressVlanTaggingEnabledSet(portID, IX_ETH_DB_PASS_THROUGH);
+            }
+
+            /* clear membership and TTI tables */
+            memset (vlanSet, 0, sizeof (vlanSet));
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                /* use the internal function to bypass PVID check */
+                status = ixEthDBPortVlanTableSet(portID, portInfo->vlanMembership, vlanSet);
+            }
+
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                /* use the internal function to bypass PVID check */
+                status = ixEthDBPortVlanTableSet(portID, portInfo->transmitTaggingInfo, vlanSet);
+            }
+
+            /* reset the PVID */
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBPortVlanTagSet(portID, 0);
+            }
+
+            /* disable TPID port extraction */
+            if (status == IX_ETH_DB_SUCCESS)
+            {
+                status = ixEthDBVlanPortExtractionEnable(portID, FALSE);
+            }
+        }
+#endif
+
+        if (status == IX_ETH_DB_SUCCESS)
+        {
+            /* checks passed, disable */
+            portInfo->featureStatus &= ~feature;
+        }
+    }
+
+    /* restore port enabled state */
+    portInfo->enabled = portEnabled;
+
+    return status;
+}
+
+/**
+ * @brief returns the status of a feature
+ *
+ * @param portID port ID
+ * @param present location to store a boolean value indicating
+ * if the feature is present (TRUE) or not (FALSE)
+ * @param enabled location to store a booleam value indicating
+ * if the feature is present (TRUE) or not (FALSE)
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFeatureStatusGet(IxEthDBPortId portID, IxEthDBFeature feature, BOOL *present, BOOL *enabled)
+{
+    PortInfo *portInfo;
+
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_REFERENCE(present);
+
+    IX_ETH_DB_CHECK_REFERENCE(enabled);
+
+    portInfo = &ixEthDBPortInfo[portID];
+
+    *present = (portInfo->featureCapability & feature) != 0;
+    *enabled = (portInfo->featureStatus & feature) != 0;
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief returns the value of an EthDB property
+ *
+ * @param portID ID of the port
+ * @param feature feature owning the property
+ * @param property ID of the property
+ * @param type location to store the property type into
+ * @param value location to store the property value into
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFeaturePropertyGet(IxEthDBPortId portID, IxEthDBFeature feature, IxEthDBProperty property, IxEthDBPropertyType *type, void *value)
+{
+    IX_ETH_DB_CHECK_PORT_EXISTS(portID);
+
+    IX_ETH_DB_CHECK_REFERENCE(type);
+
+    IX_ETH_DB_CHECK_REFERENCE(value);
+
+    if (feature == IX_ETH_DB_VLAN_QOS)
+    {
+        if (property == IX_ETH_DB_QOS_TRAFFIC_CLASS_COUNT_PROPERTY)
+        {
+            * (UINT32 *) value = ixEthDBPortInfo[portID].ixEthDBTrafficClassCount;
+            *type              = IX_ETH_DB_INTEGER_PROPERTY;
+
+            return IX_ETH_DB_SUCCESS;
+        }
+        else if (property >= IX_ETH_DB_QOS_TRAFFIC_CLASS_0_RX_QUEUE_PROPERTY
+            && property <= IX_ETH_DB_QOS_TRAFFIC_CLASS_7_RX_QUEUE_PROPERTY)
+        {
+            UINT32 classDelta = property - IX_ETH_DB_QOS_TRAFFIC_CLASS_0_RX_QUEUE_PROPERTY;
+
+            if (classDelta >= ixEthDBPortInfo[portID].ixEthDBTrafficClassCount)
+            {
+                return IX_ETH_DB_FAIL;
+            }
+
+            * (UINT32 *) value = ixEthDBPortInfo[portID].ixEthDBTrafficClassAQMAssignments[classDelta];
+            *type              = IX_ETH_DB_INTEGER_PROPERTY;
+
+            return IX_ETH_DB_SUCCESS;
+        }
+    }
+
+    return IX_ETH_DB_INVALID_ARG;
+}
+
+/**
+ * @brief sets the value of an EthDB property
+ *
+ * @param portID ID of the port
+ * @param feature feature owning the property
+ * @param property ID of the property
+ * @param value location containing the property value
+ *
+ * This function implements a private property intended
+ * only for EthAcc usage. Upon setting the IX_ETH_DB_QOS_QUEUE_CONFIGURATION_COMPLETE
+ * property (the value is ignored), the availability of traffic classes is
+ * frozen to whatever traffic class structure is currently in use.
+ * This means that if VLAN_QOS has been enabled before EthAcc
+ * initialization then all the defined traffic classes will be available;
+ * otherwise only one traffic class (0) will be available.
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h as not accepting any parameters. The
+ * current implementation is only intended for the private use of EthAcc.
+ *
+ * Also note that once this function is called the effect is irreversible,
+ * unless EthDB is complete unloaded and re-initialized.
+ *
+ * @return IX_ETH_DB_INVALID_ARG (no read-write properties are
+ * supported in this release)
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFeaturePropertySet(IxEthDBPortId portID, IxEthDBFeature feature, IxEthDBProperty property, void *value)
+{
+    IX_ETH_DB_CHECK_PORT_EXISTS(portID);
+
+    if ((feature == IX_ETH_DB_VLAN_QOS) && (property == IX_ETH_DB_QOS_QUEUE_CONFIGURATION_COMPLETE))
+    {
+        ixEthDBPortInfo[portID].ixEthDBTrafficClassAvailable = ixEthDBPortInfo[portID].ixEthDBTrafficClassCount;
+
+        return IX_ETH_DB_SUCCESS;
+    }
+
+    return IX_ETH_DB_INVALID_ARG;
+}

cpu/ixp/npe/IxEthDBFirewall.c

@@ -0,0 +1,266 @@
+/**
+ * @file IxEthDBFirewall.c
+ *
+ * @brief Implementation of the firewall API
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+
+#include "IxEthDB_p.h"
+
+/**
+ * @brief updates the NPE firewall operating mode and 
+ * firewall address table
+ *
+ * @param portID ID of the port
+ * @param epDelta initial entry point for binary searches (NPE optimization)
+ * @param address address of the firewall MAC address table
+ *
+ * This function will send a message to the NPE configuring the
+ * firewall mode (white list or black list), invalid source 
+ * address filtering and downloading a new MAC address database 
+ * to be used for firewall matching.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed 
+ * successfully or IX_ETH_DB_FAIL otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFirewallUpdate(IxEthDBPortId portID, void *address, UINT32 epDelta)
+{
+    IxNpeMhMessage message;
+    IX_STATUS result;
+    
+    UINT32 mode        = 0;    
+    PortInfo *portInfo = &ixEthDBPortInfo[portID];
+
+    mode = (portInfo->srcAddressFilterEnabled != FALSE) << 1 | (portInfo->firewallMode == IX_ETH_DB_FIREWALL_WHITE_LIST);
+
+    FILL_SETFIREWALLMODE_MSG(message, 
+        IX_ETH_DB_PORT_ID_TO_NPE_LOGICAL_ID(portID), 
+        epDelta, 
+        mode, 
+        IX_OSAL_MMU_VIRT_TO_PHYS(address));
+
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+    
+    return result;
+}
+
+/**
+ * @brief configures the firewall white list/black list
+ * access mode
+ *
+ * @param portID ID of the port
+ * @param mode firewall filtering mode (IX_ETH_DB_FIREWALL_WHITE_LIST
+ * or IX_ETH_DB_FIREWALL_BLACK_LIST)
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBFirewallModeSet(IxEthDBPortId portID, IxEthDBFirewallMode mode)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+     
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_FIREWALL);
+    
+    if (mode != IX_ETH_DB_FIREWALL_WHITE_LIST
+        && mode != IX_ETH_DB_FIREWALL_BLACK_LIST)
+    {
+        return IX_ETH_DB_INVALID_ARG;
+    }    
+    
+    ixEthDBPortInfo[portID].firewallMode = mode;
+    
+    return ixEthDBFirewallTableDownload(portID);
+}
+
+/**
+ * @brief enables or disables the invalid source MAC address filter
+ *
+ * @param portID ID of the port
+ * @param enable TRUE to enable invalid source MAC address filtering
+ * or FALSE to disable it
+ *
+ * The invalid source MAC address filter will discard, when enabled,
+ * frames whose source MAC address is a multicast or the broadcast MAC
+ * address.
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed 
+ * successfully or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBFirewallInvalidAddressFilterEnable(IxEthDBPortId portID, BOOL enable)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_FIREWALL);
+
+    ixEthDBPortInfo[portID].srcAddressFilterEnabled = enable;
+    
+    return ixEthDBFirewallTableDownload(portID);
+}
+
+/**
+ * @brief adds a firewall record
+ *
+ * @param portID ID of the port
+ * @param macAddr MAC address of the new record
+ *
+ * This function will add a new firewall record
+ * on the specified port, using the specified 
+ * MAC address. If the record already exists this
+ * function will silently return IX_ETH_DB_SUCCESS,
+ * although no duplicate records are added.
+ *
+ * Note that this function is documented in the main
+ * component header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBFirewallEntryAdd(IxEthDBPortId portID, IxEthDBMacAddr *macAddr)
+{
+    MacDescriptor recordTemplate;
+
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_FIREWALL);
+    
+    memcpy(recordTemplate.macAddress, macAddr, sizeof (IxEthDBMacAddr));
+    
+    recordTemplate.type   = IX_ETH_DB_FIREWALL_RECORD;
+    recordTemplate.portID = portID;
+    
+    return ixEthDBAdd(&recordTemplate, NULL);
+}
+
+/**
+ * @brief removes a firewall record
+ *
+ * @param portID ID of the port
+ * @param macAddr MAC address of the record to remove
+ *
+ * This function will attempt to remove a firewall
+ * record from the given port, using the specified
+ * MAC address.
+ *
+ * Note that this function is documented in the main
+ * component header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully of an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBFirewallEntryRemove(IxEthDBPortId portID, IxEthDBMacAddr *macAddr)
+{
+    MacDescriptor recordTemplate;
+    
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_FIREWALL);
+    
+    memcpy(recordTemplate.macAddress, macAddr, sizeof (IxEthDBMacAddr));
+    
+    recordTemplate.type   = IX_ETH_DB_FIREWALL_RECORD;
+    recordTemplate.portID = portID;
+    
+    return ixEthDBRemove(&recordTemplate, NULL);
+}
+
+/**
+ * @brief downloads the firewall address table to an NPE
+ *
+ * @param portID ID of the port
+ *
+ * This function will download the firewall address table to
+ * an NPE port.
+ *
+ * Note that this function is documented in the main 
+ * component header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or IX_ETH_DB_FAIL otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBFirewallTableDownload(IxEthDBPortId portID)
+{
+    IxEthDBPortMap query;
+    IxEthDBStatus result;
+    
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_FIREWALL);
+    
+    SET_DEPENDENCY_MAP(query, portID);
+
+    ixEthDBUpdateLock();
+    
+    ixEthDBPortInfo[portID].updateMethod.searchTree = ixEthDBQuery(NULL, query, IX_ETH_DB_FIREWALL_RECORD, MAX_FW_SIZE);
+    
+    result = ixEthDBNPEUpdateHandler(portID, IX_ETH_DB_FIREWALL_RECORD);
+
+    ixEthDBUpdateUnlock();
+
+    return result;
+}

cpu/ixp/npe/IxEthDBHashtable.c

@@ -0,0 +1,642 @@
+/**
+ * @file ethHash.c
+ *
+ * @brief Hashtable implementation
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+
+#include "IxEthDB_p.h"
+#include "IxEthDBLocks_p.h"
+
+/**
+ * @addtogroup EthDB
+ *
+ * @{
+ */
+
+/**
+ * @brief initializes a hash table object
+ *
+ * @param hashTable uninitialized hash table structure
+ * @param numBuckets number of buckets to use
+ * @param entryHashFunction hash function used 
+ * to hash entire hash node data block (for adding)
+ * @param matchFunctions array of match functions, indexed on type,
+ * used to differentiate records with the same hash value
+ * @param freeFunction function used to free node data blocks
+ *
+ * Initializes the given hash table object.
+ *
+ * @internal
+ */
+void ixEthDBInitHash(HashTable *hashTable, 
+                     UINT32 numBuckets, 
+                     HashFunction entryHashFunction, 
+                     MatchFunction *matchFunctions, 
+                     FreeFunction freeFunction)
+{
+    UINT32 bucketIndex;
+    UINT32 hashSize = numBuckets * sizeof(HashNode *);
+
+    /* entry hashing, matching and freeing methods */
+    hashTable->entryHashFunction  = entryHashFunction;
+    hashTable->matchFunctions     = matchFunctions;
+    hashTable->freeFunction       = freeFunction;
+
+    /* buckets */
+    hashTable->numBuckets = numBuckets;
+
+    /* set to 0 all buckets */
+    memset(hashTable->hashBuckets, 0, hashSize);
+
+    /* init bucket locks - note that initially all mutexes are unlocked after MutexInit()*/
+    for (bucketIndex = 0 ; bucketIndex < numBuckets ; bucketIndex++)
+    {
+        ixOsalFastMutexInit(&hashTable->bucketLocks[bucketIndex]);
+    }
+}
+
+/**
+ * @brief adds an entry to the hash table
+ *
+ * @param hashTable hash table to add the entry to
+ * @param entry entry to add
+ *
+ * The entry will be hashed using the entry hashing function and added to the
+ * hash table, unless a locking blockage occurs, in which case the caller
+ * should retry.
+ *
+ * @retval IX_ETH_DB_SUCCESS if adding <i>entry</i> has succeeded
+ * @retval IX_ETH_DB_NOMEM if there's no memory left in the hash node pool
+ * @retval IX_ETH_DB_BUSY if there's a locking failure on the insertion path
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC IxEthDBStatus ixEthDBAddHashEntry(HashTable *hashTable, void *entry)
+{
+    UINT32 hashValue   = hashTable->entryHashFunction(entry);
+    UINT32 bucketIndex = hashValue % hashTable->numBuckets;
+    HashNode *bucket   = hashTable->hashBuckets[bucketIndex];
+    HashNode *newNode;
+
+    LockStack locks;
+
+    INIT_STACK(&locks);
+
+    /* lock bucket */
+    PUSH_LOCK(&locks, &hashTable->bucketLocks[bucketIndex]);
+
+    /* lock insertion element (first in chain), if any */
+    if (bucket != NULL)
+    {
+        PUSH_LOCK(&locks, &bucket->lock);
+    }
+
+    /* get new node */
+    newNode = ixEthDBAllocHashNode();
+
+    if (newNode == NULL)
+    {
+        /* unlock everything */
+        UNROLL_STACK(&locks);
+
+        return IX_ETH_DB_NOMEM;
+    }
+
+    /* init lock - note that mutexes are unlocked after MutexInit */
+    ixOsalFastMutexInit(&newNode->lock);
+
+    /* populate new link */
+    newNode->data = entry;
+
+    /* add to bucket */
+    newNode->next                       = bucket;
+    hashTable->hashBuckets[bucketIndex] = newNode;
+
+    /* unlock bucket and insertion point */
+    UNROLL_STACK(&locks);
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief removes an entry from the hashtable
+ *
+ * @param hashTable hash table to remove entry from
+ * @param keyType type of record key used for matching
+ * @param reference reference key used to identify the entry
+ *
+ * The reference key will be hashed using the key hashing function,
+ * the entry is searched using the hashed value and then examined
+ * against the reference entry using the match function. A positive
+ * match will trigger the deletion of the entry.
+ * Locking failures are reported and the caller should retry.
+ *
+ * @retval IX_ETH_DB_SUCCESS if the removal was successful
+ * @retval IX_ETH_DB_NO_SUCH_ADDR if the entry was not found
+ * @retval IX_ETH_DB_BUSY if a locking failure occured during the process
+ *
+ * @internal
+ */
+IxEthDBStatus ixEthDBRemoveHashEntry(HashTable *hashTable, int keyType, void *reference)
+{
+    UINT32 hashValue       = hashTable->entryHashFunction(reference);
+    UINT32 bucketIndex     = hashValue % hashTable->numBuckets;
+    HashNode *node         = hashTable->hashBuckets[bucketIndex];
+    HashNode *previousNode = NULL;
+    
+    LockStack locks;
+
+    INIT_STACK(&locks);
+
+    while (node != NULL)
+    {
+        /* try to lock node */
+        PUSH_LOCK(&locks, &node->lock);
+
+        if (hashTable->matchFunctions[keyType](reference, node->data))
+        {
+            /* found entry */
+            if (node->next != NULL)
+            {
+                PUSH_LOCK(&locks, &node->next->lock);
+            }
+
+            if (previousNode == NULL)
+            {
+                /* node is head of chain */
+                PUSH_LOCK(&locks, &hashTable->bucketLocks[bucketIndex]);
+
+                hashTable->hashBuckets[bucketIndex] = node->next;
+
+                POP_LOCK(&locks);
+            }
+            else
+            {
+                /* relink */
+                previousNode->next = node->next;
+            }
+
+            UNROLL_STACK(&locks);
+
+            /* free node */
+            hashTable->freeFunction(node->data);
+            ixEthDBFreeHashNode(node);
+
+            return IX_ETH_DB_SUCCESS;
+        }
+        else
+        {
+            if (previousNode != NULL)
+            {
+                /* unlock previous node */
+                SHIFT_STACK(&locks);
+            }
+
+            /* advance to next element in chain */
+            previousNode = node;
+            node         = node->next;
+        }
+    }
+
+    UNROLL_STACK(&locks);
+
+    /* not found */
+    return IX_ETH_DB_NO_SUCH_ADDR;
+}
+
+/**
+ * @brief retrieves an entry from the hash table
+ *
+ * @param hashTable hash table to perform the search into
+ * @param reference search key (a MAC address)
+ * @param keyType type of record key used for matching
+ * @param searchResult pointer where a reference to the located hash node 
+ * is placed
+ *
+ * Searches the entry with the same key as <i>reference</i> and places the
+ * pointer to the resulting node in <i>searchResult</i>.
+ * An implicit write access lock is granted after a search, which gives the 
+ * caller the opportunity to modify the entry.
+ * Access should be released as soon as possible using @ref ixEthDBReleaseHashNode().
+ *
+ * @see ixEthDBReleaseHashNode()
+ *
+ * @retval IX_ETH_DB_SUCCESS if the search was completed successfully
+ * @retval IX_ETH_DB_NO_SUCH_ADDRESS if no entry with the given key was found
+ * @retval IX_ETH_DB_BUSY if a locking failure has occured, in which case
+ * the caller should retry
+ *
+ * @warning unless the return value is <b>IX_ETH_DB_SUCCESS</b> the searchResult
+ * location is NOT modified and therefore using a NULL comparison test when the
+ * value was not properly initialized would be an error
+ *
+ * @internal
+ */
+IxEthDBStatus ixEthDBSearchHashEntry(HashTable *hashTable, int keyType, void *reference, HashNode **searchResult)
+{
+    UINT32 hashValue;
+    HashNode *node;
+
+    hashValue = hashTable->entryHashFunction(reference);
+    node      = hashTable->hashBuckets[hashValue % hashTable->numBuckets];
+
+    while (node != NULL)
+    {
+        TRY_LOCK(&node->lock);
+
+        if (hashTable->matchFunctions[keyType](reference, node->data))
+        {
+            *searchResult = node;
+
+            return IX_ETH_DB_SUCCESS;
+        }
+        else
+        {
+            UNLOCK(&node->lock);
+
+            node = node->next;
+        }
+    }
+
+    /* not found */
+    return IX_ETH_DB_NO_SUCH_ADDR;
+}
+
+/**
+ * @brief reports the existence of an entry in the hash table
+ *
+ * @param hashTable hash table to perform the search into
+ * @param reference search key (a MAC address)
+ * @param keyType type of record key used for matching
+ *
+ * Searches the entry with the same key as <i>reference</i>.
+ * No implicit write access lock is granted after a search, hence the 
+ * caller cannot access or modify the entry. The result is only temporary.
+ *
+ * @see ixEthDBReleaseHashNode()
+ *
+ * @retval IX_ETH_DB_SUCCESS if the search was completed successfully
+ * @retval IX_ETH_DB_NO_SUCH_ADDRESS if no entry with the given key was found
+ * @retval IX_ETH_DB_BUSY if a locking failure has occured, in which case
+ * the caller should retry
+ *
+ * @internal
+ */
+IxEthDBStatus ixEthDBPeekHashEntry(HashTable *hashTable, int keyType, void *reference)
+{
+    UINT32 hashValue;
+    HashNode *node;
+
+    hashValue = hashTable->entryHashFunction(reference);
+    node      = hashTable->hashBuckets[hashValue % hashTable->numBuckets];
+
+    while (node != NULL)
+    {
+        TRY_LOCK(&node->lock);
+
+        if (hashTable->matchFunctions[keyType](reference, node->data))
+        {
+            UNLOCK(&node->lock);
+
+            return IX_ETH_DB_SUCCESS;
+        }
+        else
+        {
+            UNLOCK(&node->lock);
+
+            node = node->next;
+        }
+    }
+
+    /* not found */
+    return IX_ETH_DB_NO_SUCH_ADDR;
+}
+
+/**
+ * @brief releases the write access lock
+ *
+ * @pre the node should have been obtained via @ref ixEthDBSearchHashEntry()
+ *
+ * @see ixEthDBSearchHashEntry()
+ *
+ * @internal
+ */
+void ixEthDBReleaseHashNode(HashNode *node)
+{
+    UNLOCK(&node->lock);
+}
+
+/**
+ * @brief initializes a hash iterator
+ *
+ * @param hashTable hash table to be iterated
+ * @param iterator iterator object
+ *
+ * If the initialization is successful the iterator will point to the
+ * first hash table record (if any).
+ * Testing if the iterator has not passed the end of the table should be
+ * done using the IS_ITERATOR_VALID(iteratorPtr) macro.
+ * An implicit write access lock is granted on the entry pointed by the iterator.
+ * The access is automatically revoked when the iterator is incremented.
+ * If the caller decides to terminate the iteration before the end of the table is
+ * passed then the manual access release method, @ref ixEthDBReleaseHashIterator,
+ * must be called.
+ *
+ * @see ixEthDBReleaseHashIterator()
+ *
+ * @retval IX_ETH_DB_SUCCESS if initialization was successful and the iterator points
+ * to the first valid table node
+ * @retval IX_ETH_DB_FAIL if the table is empty
+ * @retval IX_ETH_DB_BUSY if a locking failure has occured, in which case the caller
+ * should retry
+ *
+ * @warning do not use ixEthDBReleaseHashNode() on entries pointed by the iterator, as this
+ * might place the database in a permanent invalid lock state
+ *
+ * @internal
+ */
+IxEthDBStatus ixEthDBInitHashIterator(HashTable *hashTable, HashIterator *iterator)
+{
+    iterator->bucketIndex  = 0;
+    iterator->node         = NULL;
+    iterator->previousNode = NULL;
+
+    return ixEthDBIncrementHashIterator(hashTable, iterator);
+}
+
+/**
+ * @brief releases the write access locks of the iterator nodes
+ *
+ * @warning use of this function is required only when the caller terminates an iteration
+ * before reaching the end of the table
+ *
+ * @see ixEthDBInitHashIterator()
+ * @see ixEthDBIncrementHashIterator()
+ *
+ * @param iterator iterator whose node(s) should be unlocked
+ *
+ * @internal
+ */
+void ixEthDBReleaseHashIterator(HashIterator *iterator)
+{
+    if (iterator->previousNode != NULL)
+    {
+        UNLOCK(&iterator->previousNode->lock);
+    }
+
+    if (iterator->node != NULL)
+    {
+        UNLOCK(&iterator->node->lock);
+    }
+}
+
+/**
+ * @brief incremenents an iterator so that it points to the next valid entry of the table
+ * (if any)
+ *
+ * @param hashTable hash table to iterate
+ * @param iterator iterator object
+ *
+ * @pre the iterator object must be initialized using @ref ixEthDBInitHashIterator()
+ *
+ * If the increment operation is successful the iterator will point to the
+ * next hash table record (if any).
+ * Testing if the iterator has not passed the end of the table should be
+ * done using the IS_ITERATOR_VALID(iteratorPtr) macro.
+ * An implicit write access lock is granted on the entry pointed by the iterator.
+ * The access is automatically revoked when the iterator is re-incremented.
+ * If the caller decides to terminate the iteration before the end of the table is
+ * passed then the manual access release method, @ref ixEthDBReleaseHashIterator,
+ * must be called.
+ * Is is guaranteed that no other thread can remove or change the iterated entry until
+ * the iterator is incremented successfully.
+ *
+ * @see ixEthDBReleaseHashIterator()
+ *
+ * @retval IX_ETH_DB_SUCCESS if the operation was successful and the iterator points
+ * to the next valid table node
+ * @retval IX_ETH_DB_FAIL if the iterator has passed the end of the table
+ * @retval IX_ETH_DB_BUSY if a locking failure has occured, in which case the caller
+ * should retry
+ *
+ * @warning do not use ixEthDBReleaseHashNode() on entries pointed by the iterator, as this
+ * might place the database in a permanent invalid lock state
+ *
+ * @internal
+ */
+IxEthDBStatus ixEthDBIncrementHashIterator(HashTable *hashTable, HashIterator *iterator)
+{
+    /* unless iterator is just initialized... */
+    if (iterator->node != NULL)
+    {
+        /* try next in chain */
+        if (iterator->node->next != NULL)
+        {
+            TRY_LOCK(&iterator->node->next->lock);
+
+            if (iterator->previousNode != NULL)
+            {
+                UNLOCK(&iterator->previousNode->lock);
+            }
+
+            iterator->previousNode = iterator->node;
+            iterator->node         = iterator->node->next;
+
+            return IX_ETH_DB_SUCCESS;
+        }
+        else
+        {
+            /* last in chain, prepare for next bucket */
+            iterator->bucketIndex++;
+        }
+    }
+
+   /* try next used bucket */
+    for (; iterator->bucketIndex < hashTable->numBuckets ; iterator->bucketIndex++)
+    {
+        HashNode **nodePtr = &(hashTable->hashBuckets[iterator->bucketIndex]);
+        HashNode *node = *nodePtr;
+#if (CPU!=SIMSPARCSOLARIS) && !defined (__wince)
+        if (((iterator->bucketIndex & IX_ETHDB_BUCKET_INDEX_MASK) == 0) &&
+            (iterator->bucketIndex < (hashTable->numBuckets - IX_ETHDB_BUCKETPTR_AHEAD)))
+        {
+            /* preload next cache line (2 cache line ahead) */
+            nodePtr += IX_ETHDB_BUCKETPTR_AHEAD;
+            __asm__ ("pld [%0];\n": : "r" (nodePtr));
+        }
+#endif
+        if (node != NULL)
+        {
+            TRY_LOCK(&node->lock);
+
+            /* unlock last one or two nodes in the previous chain */
+            if (iterator->node != NULL)
+            {
+                UNLOCK(&iterator->node->lock);
+
+                if (iterator->previousNode != NULL)
+                {
+                    UNLOCK(&iterator->previousNode->lock);
+                }
+            }
+            
+            /* redirect iterator */
+            iterator->previousNode = NULL;
+            iterator->node         = node;
+
+            return IX_ETH_DB_SUCCESS;
+        }
+    }
+
+    /* could not advance iterator */
+    if (iterator->node != NULL)
+    {
+        UNLOCK(&iterator->node->lock);
+
+        if (iterator->previousNode != NULL)
+        {
+            UNLOCK(&iterator->previousNode->lock);
+        }
+
+        iterator->node = NULL;
+    }
+
+    return IX_ETH_DB_END;
+}
+
+/**
+ * @brief removes an entry pointed by an iterator
+ *
+ * @param hashTable iterated hash table
+ * @param iterator iterator object
+ *
+ * Removes the entry currently pointed by the iterator and repositions the iterator
+ * on the next valid entry (if any). Handles locking issues automatically and
+ * implicitely grants write access lock to the new pointed entry.
+ * Failures due to concurrent threads having write access locks in the same region
+ * preserve the state of the database and the iterator object, leaving the caller
+ * free to retry without loss of access. It is guaranteed that only the thread owning
+ * the iterator can remove the object pointed by the iterator.
+ *
+ * @retval IX_ETH_DB_SUCCESS if removal has succeeded
+ * @retval IX_ETH_DB_BUSY if a locking failure has occured, in which case the caller
+ * should retry
+ *
+ * @internal
+ */
+IxEthDBStatus ixEthDBRemoveEntryAtHashIterator(HashTable *hashTable, HashIterator *iterator)
+{
+    HashIterator nextIteratorPos;
+    LockStack locks;
+
+    INIT_STACK(&locks);
+
+    /* set initial bucket index for next position */
+    nextIteratorPos.bucketIndex = iterator->bucketIndex;
+
+    /* compute iterator position before removing anything and lock ahead */
+    if (iterator->node->next != NULL)
+    {
+        PUSH_LOCK(&locks, &iterator->node->next->lock);
+
+        /* reposition on the next node in the chain */
+        nextIteratorPos.node         = iterator->node->next;
+        nextIteratorPos.previousNode = iterator->previousNode;
+    }
+    else
+    {
+        /* try next chain - don't know yet if we'll find anything */
+        nextIteratorPos.node = NULL;
+
+        /* if we find something it's a chain head */
+        nextIteratorPos.previousNode = NULL;
+
+        /* browse up in the buckets to find a non-null chain */
+        while (++nextIteratorPos.bucketIndex < hashTable->numBuckets)
+        {
+            nextIteratorPos.node = hashTable->hashBuckets[nextIteratorPos.bucketIndex];
+
+            if (nextIteratorPos.node != NULL)
+            {
+                /* found a non-empty chain, try to lock head */
+                PUSH_LOCK(&locks, &nextIteratorPos.node->lock);
+
+                break;
+            }
+        }
+    }
+
+    /* restore links over the to-be-deleted item */
+    if (iterator->previousNode == NULL)
+    {
+        /* first in chain, lock bucket */
+        PUSH_LOCK(&locks, &hashTable->bucketLocks[iterator->bucketIndex]);
+
+        hashTable->hashBuckets[iterator->bucketIndex] = iterator->node->next;
+
+        POP_LOCK(&locks);
+    }
+    else
+    {
+        /* relink */
+        iterator->previousNode->next = iterator->node->next;
+
+        /* unlock last remaining node in current chain when moving between chains */
+        if (iterator->node->next == NULL)
+        {
+            UNLOCK(&iterator->previousNode->lock);
+        }
+    }
+
+    /* delete entry */
+    hashTable->freeFunction(iterator->node->data);
+    ixEthDBFreeHashNode(iterator->node);
+
+    /* reposition iterator */
+    *iterator = nextIteratorPos;
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @}
+ */

cpu/ixp/npe/IxEthDBLearning.c

@@ -0,0 +1,149 @@
+/**
+ * @file IxEthDBLearning.c
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxEthDB_p.h"
+
+/**
+ * @brief hashes the mac address in a mac descriptor with a XOR function
+ *
+ * @param entry pointer to a mac descriptor to be hashed
+ * 
+ * This function only extracts the mac address and employs ixEthDBKeyXORHash()
+ * to do the actual hashing.
+ * Used only to add a whole entry to a hash table, as opposed to searching which
+ * takes only a key and uses the key hashing directly.
+ *
+ * @see ixEthDBKeyXORHash()
+ *
+ * @return the hash value
+ *
+ * @internal
+ */
+UINT32 ixEthDBEntryXORHash(void *entry)
+{
+    MacDescriptor *descriptor = (MacDescriptor *) entry;
+
+    return ixEthDBKeyXORHash(descriptor->macAddress);
+}
+
+/**
+ * @brief hashes a mac address
+ *
+ * @param key pointer to a 6 byte structure (typically an IxEthDBMacAddr pointer)
+ * to be hashed
+ *
+ * Given a 6 bytes MAC address, the hash used is:
+ *
+ * hash(MAC[0:5]) = MAC[0:1] ^ MAC[2:3] ^ MAC[4:5]
+ *
+ * Used by the hash table to search and remove entries based
+ * solely on their keys (mac addresses).
+ *
+ * @return the hash value
+ *
+ * @internal
+ */
+UINT32 ixEthDBKeyXORHash(void *key)
+{
+    UINT32 hashValue;
+    UINT8 *value = (UINT8 *) key;
+    
+    hashValue  = (value[5] << 8) | value[4];
+    hashValue ^= (value[3] << 8) | value[2];
+    hashValue ^= (value[1] << 8) | value[0];
+
+    return hashValue;
+}
+
+/**
+ * @brief mac descriptor match function
+ *
+ * @param reference mac address (typically an IxEthDBMacAddr pointer) structure
+ * @param entry pointer to a mac descriptor whose key (mac address) is to be 
+ * matched against the reference key
+ *
+ * Used by the hash table to retrieve entries. Hashing entries can produce
+ * collisions, i.e. descriptors with different mac addresses and the same
+ * hash value, where this function is used to differentiate entries.
+ *
+ * @retval TRUE if the entry matches the reference key (equal addresses)
+ * @retval FALSE if the entry does not match the reference key
+ *
+ * @internal
+ */
+BOOL ixEthDBAddressMatch(void *reference, void *entry)
+{
+    return (ixEthDBAddressCompare(reference, ((MacDescriptor *) entry)->macAddress) == 0);
+}
+
+/**
+ * @brief compares two mac addresses
+ *
+ * @param mac1 first mac address to compare
+ * @param mac2 second mac address to compare
+ * 
+ * This comparison works in a similar way to strcmp, producing similar results.
+ * Used to insert values keyed on mac addresses into binary search trees.
+ *
+ * @retval -1 if mac1 < mac2
+ * @retval 0 if ma1 == mac2
+ * @retval 1 if mac1 > mac2
+ */
+UINT32 ixEthDBAddressCompare(UINT8 *mac1, UINT8 *mac2)
+{
+    UINT32 local_index;
+
+    for (local_index = 0 ; local_index < IX_IEEE803_MAC_ADDRESS_SIZE ; local_index++)
+    {
+        if (mac1[local_index] > mac2[local_index])
+        {
+            return 1;
+        }
+        else if (mac1[local_index] < mac2[local_index])
+        {
+            return -1;
+        }
+    }
+
+    return 0;
+}
+

cpu/ixp/npe/IxEthDBMem.c

@@ -0,0 +1,649 @@
+/**
+ * @file IxEthDBDBMem.c
+ *
+ * @brief Memory handling routines for the MAC address database
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+
+#include "IxEthDB_p.h"
+
+IX_ETH_DB_PRIVATE HashNode *nodePool     = NULL;
+IX_ETH_DB_PRIVATE MacDescriptor *macPool = NULL;
+IX_ETH_DB_PRIVATE MacTreeNode *treePool  = NULL;
+
+IX_ETH_DB_PRIVATE HashNode nodePoolArea[NODE_POOL_SIZE];
+IX_ETH_DB_PRIVATE MacDescriptor macPoolArea[MAC_POOL_SIZE];
+IX_ETH_DB_PRIVATE MacTreeNode treePoolArea[TREE_POOL_SIZE];
+
+IX_ETH_DB_PRIVATE IxOsalMutex nodePoolLock;
+IX_ETH_DB_PRIVATE IxOsalMutex macPoolLock;
+IX_ETH_DB_PRIVATE IxOsalMutex treePoolLock;
+
+#define LOCK_NODE_POOL   { ixOsalMutexLock(&nodePoolLock, IX_OSAL_WAIT_FOREVER); }
+#define UNLOCK_NODE_POOL { ixOsalMutexUnlock(&nodePoolLock); }
+
+#define LOCK_MAC_POOL    { ixOsalMutexLock(&macPoolLock, IX_OSAL_WAIT_FOREVER); }
+#define UNLOCK_MAC_POOL  { ixOsalMutexUnlock(&macPoolLock); }
+
+#define LOCK_TREE_POOL   { ixOsalMutexLock(&treePoolLock, IX_OSAL_WAIT_FOREVER); }
+#define UNLOCK_TREE_POOL { ixOsalMutexUnlock(&treePoolLock); }
+
+/* private function prototypes */
+IX_ETH_DB_PRIVATE MacDescriptor* ixEthDBPoolAllocMacDescriptor(void);
+IX_ETH_DB_PRIVATE void ixEthDBPoolFreeMacDescriptor(MacDescriptor *macDescriptor);
+
+/**
+ * @addtogroup EthMemoryManagement
+ *
+ * @{
+ */
+
+/**
+ * @brief initializes the memory pools used by the ethernet database component
+ *
+ * Initializes the hash table node, mac descriptor and mac tree node pools.
+ * Called at initialization time by @ref ixEthDBInit().
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBInitMemoryPools(void)
+{
+    int local_index;
+
+    /* HashNode pool */
+    ixOsalMutexInit(&nodePoolLock);
+
+    for (local_index = 0 ; local_index < NODE_POOL_SIZE ; local_index++)
+    {
+        HashNode *freeNode = &nodePoolArea[local_index];
+
+        freeNode->nextFree = nodePool;
+        nodePool           = freeNode;
+    }
+
+    /* MacDescriptor pool */
+    ixOsalMutexInit(&macPoolLock);
+
+    for (local_index = 0 ; local_index < MAC_POOL_SIZE ; local_index++)
+    {
+        MacDescriptor *freeDescriptor = &macPoolArea[local_index];
+
+        freeDescriptor->nextFree = macPool;
+        macPool                  = freeDescriptor;
+    }
+
+    /* MacTreeNode pool */
+    ixOsalMutexInit(&treePoolLock);
+
+    for (local_index = 0 ; local_index < TREE_POOL_SIZE ; local_index++)
+    {
+        MacTreeNode *freeNode = &treePoolArea[local_index];
+
+        freeNode->nextFree = treePool;
+        treePool           = freeNode;
+    }
+}
+
+/**
+ * @brief allocates a hash node from the pool
+ *
+ * Allocates a hash node and resets its value.
+ *
+ * @return the allocated hash node or NULL if the pool is empty
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+HashNode* ixEthDBAllocHashNode(void)
+{
+    HashNode *allocatedNode = NULL;
+
+    if (nodePool != NULL)
+    {
+        LOCK_NODE_POOL;
+
+        allocatedNode = nodePool;
+        nodePool      = nodePool->nextFree;
+
+        UNLOCK_NODE_POOL;
+
+        memset(allocatedNode, 0, sizeof(HashNode));
+    }
+
+    return allocatedNode;
+}
+
+/**
+ * @brief frees a hash node into the pool
+ *
+ * @param hashNode node to be freed
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBFreeHashNode(HashNode *hashNode)
+{
+    if (hashNode != NULL)
+    {
+        LOCK_NODE_POOL;
+
+        hashNode->nextFree = nodePool;
+        nodePool           = hashNode;
+
+        UNLOCK_NODE_POOL;
+    }
+}
+
+/**
+ * @brief allocates a mac descriptor from the pool
+ *
+ * Allocates a mac descriptor and resets its value.
+ * This function is not used directly, instead @ref ixEthDBAllocMacDescriptor()
+ * is used, which keeps track of the pointer reference count.
+ *
+ * @see ixEthDBAllocMacDescriptor()
+ * 
+ * @warning this function is not used directly by any other function
+ * apart from ixEthDBAllocMacDescriptor()
+ *
+ * @return the allocated mac descriptor or NULL if the pool is empty
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+MacDescriptor* ixEthDBPoolAllocMacDescriptor(void)
+{
+    MacDescriptor *allocatedDescriptor = NULL;
+
+    if (macPool != NULL)
+    {
+        LOCK_MAC_POOL;
+
+        allocatedDescriptor = macPool;
+        macPool             = macPool->nextFree;
+
+        UNLOCK_MAC_POOL;
+
+        memset(allocatedDescriptor, 0, sizeof(MacDescriptor));
+    }
+
+    return allocatedDescriptor;
+}
+
+/**
+ * @brief allocates and initializes a mac descriptor smart pointer
+ *
+ * Uses @ref ixEthDBPoolAllocMacDescriptor() to allocate a mac descriptor
+ * from the pool and initializes its reference count.
+ *
+ * @see ixEthDBPoolAllocMacDescriptor()
+ *
+ * @return the allocated mac descriptor or NULL if the pool is empty
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+MacDescriptor* ixEthDBAllocMacDescriptor(void)
+{
+    MacDescriptor *allocatedDescriptor = ixEthDBPoolAllocMacDescriptor();
+
+    if (allocatedDescriptor != NULL)
+    {
+        LOCK_MAC_POOL;
+
+        allocatedDescriptor->refCount++;
+
+        UNLOCK_MAC_POOL;
+    }
+
+    return allocatedDescriptor;
+}
+
+/**
+ * @brief frees a mac descriptor back into the pool
+ *
+ * @param macDescriptor mac descriptor to be freed
+ *
+ * @warning this function is not to be called by anyone but
+ * ixEthDBFreeMacDescriptor()
+ *
+ * @see ixEthDBFreeMacDescriptor()
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBPoolFreeMacDescriptor(MacDescriptor *macDescriptor)
+{
+    LOCK_MAC_POOL;
+
+    macDescriptor->nextFree = macPool;
+    macPool                 = macDescriptor;
+
+    UNLOCK_MAC_POOL;
+}
+
+/**
+ * @brief frees or reduces the usage count of a mac descriptor smart pointer
+ *
+ * If the reference count reaches 0 (structure is no longer used anywhere)
+ * then the descriptor is freed back into the pool using ixEthDBPoolFreeMacDescriptor().
+ *
+ * @see ixEthDBPoolFreeMacDescriptor()
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBFreeMacDescriptor(MacDescriptor *macDescriptor)
+{
+    if (macDescriptor != NULL)
+    {
+        LOCK_MAC_POOL;
+
+        if (macDescriptor->refCount > 0)
+        {
+            macDescriptor->refCount--;
+
+            if (macDescriptor->refCount == 0)
+            {
+                UNLOCK_MAC_POOL;
+
+                ixEthDBPoolFreeMacDescriptor(macDescriptor);
+            }
+            else
+            {
+                UNLOCK_MAC_POOL;
+            }
+        }
+        else
+        {
+            UNLOCK_MAC_POOL;
+        }
+    }
+}
+
+/**
+ * @brief clones a mac descriptor smart pointer
+ *
+ * @param macDescriptor mac descriptor to clone
+ *
+ * Increments the usage count of the smart pointer
+ *
+ * @returns the cloned smart pointer
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+MacDescriptor* ixEthDBCloneMacDescriptor(MacDescriptor *macDescriptor)
+{
+    LOCK_MAC_POOL;
+
+    if (macDescriptor->refCount == 0)
+    {
+        UNLOCK_MAC_POOL;
+
+        return NULL;
+    }
+
+    macDescriptor->refCount++;
+
+    UNLOCK_MAC_POOL;
+
+    return macDescriptor;
+}
+
+/**
+ * @brief allocates a mac tree node from the pool
+ *
+ * Allocates and initializes a mac tree node from the pool.
+ *
+ * @return the allocated mac tree node or NULL if the pool is empty
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+MacTreeNode* ixEthDBAllocMacTreeNode(void)
+{
+    MacTreeNode *allocatedNode = NULL;
+
+    if (treePool != NULL)
+    {
+        LOCK_TREE_POOL;
+
+        allocatedNode = treePool;
+        treePool      = treePool->nextFree;
+
+        UNLOCK_TREE_POOL;
+
+        memset(allocatedNode, 0, sizeof(MacTreeNode));
+    }
+
+    return allocatedNode;
+}
+
+/**
+ * @brief frees a mac tree node back into the pool
+ *
+ * @param macNode mac tree node to be freed
+ *
+ * @warning not to be used except from ixEthDBFreeMacTreeNode().
+ *
+ * @see ixEthDBFreeMacTreeNode()
+ *
+ * @internal
+ */
+void ixEthDBPoolFreeMacTreeNode(MacTreeNode *macNode)
+{
+    if (macNode != NULL)
+    {
+        LOCK_TREE_POOL;
+
+        macNode->nextFree = treePool;
+        treePool          = macNode;
+
+        UNLOCK_TREE_POOL;
+    }
+}
+
+/**
+ * @brief frees or reduces the usage count of a mac tree node smart pointer
+ *
+ * @param macNode mac tree node to free
+ *
+ * Reduces the usage count of the given mac node. If the usage count
+ * reaches 0 the node is freed back into the pool using ixEthDBPoolFreeMacTreeNode()
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBFreeMacTreeNode(MacTreeNode *macNode)
+{
+    if (macNode->descriptor != NULL)
+    {
+        ixEthDBFreeMacDescriptor(macNode->descriptor);
+    }
+
+    if (macNode->left != NULL)
+    {
+        ixEthDBFreeMacTreeNode(macNode->left);
+    }
+
+    if (macNode->right != NULL)
+    {
+        ixEthDBFreeMacTreeNode(macNode->right);
+    }
+
+    ixEthDBPoolFreeMacTreeNode(macNode);
+}
+
+/**
+ * @brief clones a mac tree node
+ *
+ * @param macNode mac tree node to be cloned
+ *
+ * Increments the usage count of the node, <i>its associated descriptor 
+ * and <b>recursively</b> of all its child nodes</i>.
+ *
+ * @warning this function is recursive and clones whole trees/subtrees, use only for
+ * root nodes
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+MacTreeNode* ixEthDBCloneMacTreeNode(MacTreeNode *macNode)
+{
+    if (macNode != NULL)
+    {
+        MacTreeNode *clonedMacNode = ixEthDBAllocMacTreeNode();
+
+        if (clonedMacNode != NULL)
+        {
+            if (macNode->right != NULL)
+            {
+                clonedMacNode->right = ixEthDBCloneMacTreeNode(macNode->right);
+            }
+
+            if (macNode->left != NULL)
+            {
+                clonedMacNode->left = ixEthDBCloneMacTreeNode(macNode->left);
+            }
+
+            if (macNode->descriptor != NULL)
+            {
+                clonedMacNode->descriptor = ixEthDBCloneMacDescriptor(macNode->descriptor);
+            }
+        }
+
+        return clonedMacNode;
+    }
+    else
+    {
+        return NULL;
+    }
+}
+
+#ifndef NDEBUG
+/* Debug statistical functions for memory usage */
+
+extern HashTable dbHashtable;
+int ixEthDBNumHashElements(void);
+
+int ixEthDBNumHashElements(void)
+{   
+    UINT32 bucketIndex;
+    int numElements = 0;
+    HashTable *hashTable = &dbHashtable;
+
+    for (bucketIndex = 0 ; bucketIndex < hashTable->numBuckets ; bucketIndex++)
+    {
+        if (hashTable->hashBuckets[bucketIndex] != NULL)
+        {
+            HashNode *node = hashTable->hashBuckets[bucketIndex];
+
+            while (node != NULL)
+            {
+                numElements++;
+
+                node = node->next;
+            }
+        }
+    }
+
+    return numElements;
+}
+
+UINT32 ixEthDBSearchTreeUsageGet(MacTreeNode *tree)
+{
+    if (tree == NULL)
+    {
+        return 0;
+    }
+    else
+    {
+        return 1 /* this node */ + ixEthDBSearchTreeUsageGet(tree->left) + ixEthDBSearchTreeUsageGet(tree->right);
+    }
+}
+
+int ixEthDBShowMemoryStatus(void)
+{
+    MacDescriptor *mac;
+    MacTreeNode *tree;
+    HashNode *node;
+
+    int macCounter  = 0;
+    int treeCounter = 0;
+    int nodeCounter = 0;
+
+    int totalTreeUsage            = 0;
+    int totalDescriptorUsage      = 0;
+    int totalCloneDescriptorUsage = 0;
+    int totalNodeUsage            = 0;
+
+    UINT32 portIndex;
+
+    LOCK_NODE_POOL;
+    LOCK_MAC_POOL;
+    LOCK_TREE_POOL;
+
+    mac  = macPool;
+    tree = treePool;
+    node = nodePool;
+
+    while (mac != NULL)
+    {
+        macCounter++;
+
+        mac = mac->nextFree;
+
+        if (macCounter > MAC_POOL_SIZE)
+        {
+            break;
+        }
+    }
+
+    while (tree != NULL)
+    {
+        treeCounter++;
+
+        tree = tree->nextFree;
+
+        if (treeCounter > TREE_POOL_SIZE)
+        {
+            break;
+        }
+    }
+
+    while (node != NULL)
+    {
+        nodeCounter++;
+
+        node = node->nextFree;
+
+        if (nodeCounter > NODE_POOL_SIZE)
+        {
+            break;
+        }
+    }
+
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        int treeUsage = ixEthDBSearchTreeUsageGet(ixEthDBPortInfo[portIndex].updateMethod.searchTree);
+
+        totalTreeUsage            += treeUsage;
+        totalCloneDescriptorUsage += treeUsage; /* each tree node contains a descriptor */
+    }
+
+    totalNodeUsage        = ixEthDBNumHashElements();
+    totalDescriptorUsage += totalNodeUsage; /* each hash table entry contains a descriptor */
+
+    UNLOCK_NODE_POOL;
+    UNLOCK_MAC_POOL;
+    UNLOCK_TREE_POOL;
+
+    printf("Ethernet database memory usage stats:\n\n");
+
+    if (macCounter <= MAC_POOL_SIZE)
+    {
+        printf("\tMAC descriptor pool  : %d free out of %d entries (%d%%)\n", macCounter, MAC_POOL_SIZE, macCounter * 100 / MAC_POOL_SIZE);
+    }
+    else
+    {
+        printf("\tMAC descriptor pool  : invalid state (ring within the pool), normally %d entries\n", MAC_POOL_SIZE);
+    }
+
+    if (treeCounter <= TREE_POOL_SIZE)
+    {
+	printf("\tTree node pool       : %d free out of %d entries (%d%%)\n", treeCounter, TREE_POOL_SIZE, treeCounter * 100 / TREE_POOL_SIZE);
+    }
+    else
+    {
+        printf("\tTREE descriptor pool  : invalid state (ring within the pool), normally %d entries\n", TREE_POOL_SIZE);
+    }
+
+    if (nodeCounter <= NODE_POOL_SIZE)
+    {
+	printf("\tHash node pool       : %d free out of %d entries (%d%%)\n", nodeCounter, NODE_POOL_SIZE, nodeCounter * 100 / NODE_POOL_SIZE);
+    }
+    else
+    {
+        printf("\tNODE descriptor pool  : invalid state (ring within the pool), normally %d entries\n", NODE_POOL_SIZE);
+    }
+
+    printf("\n");
+    printf("\tMAC descriptor usage : %d entries, %d cloned\n", totalDescriptorUsage, totalCloneDescriptorUsage);
+    printf("\tTree node usage      : %d entries\n", totalTreeUsage);
+    printf("\tHash node usage      : %d entries\n", totalNodeUsage);
+    printf("\n");
+
+    /* search for duplicate nodes in the mac pool */
+    {
+        MacDescriptor *reference = macPool;
+
+        while (reference != NULL)
+        {
+            MacDescriptor *comparison = reference->nextFree;
+
+            while (comparison != NULL)
+            {
+                if (reference == comparison)
+                {
+                    printf("Warning: reached a duplicate (%p), invalid MAC pool state\n", reference);
+
+                    return 1;
+                }
+
+                comparison = comparison->nextFree;
+            }
+
+            reference = reference->nextFree;
+        }
+    }
+
+    printf("No duplicates found in the MAC pool (sanity check ok)\n");
+
+    return 0;
+}
+
+#endif /* NDEBUG */
+
+/**
+ * @} EthMemoryManagement
+ */

cpu/ixp/npe/IxEthDBNPEAdaptor.c

@@ -0,0 +1,719 @@
+/**
+ * @file IxEthDBDBNPEAdaptor.c
+ *
+ * @brief Routines that read and write learning/search trees in NPE-specific format
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxEthDB_p.h"
+#include "IxEthDBLog_p.h"
+
+/* forward prototype declarations */
+IX_ETH_DB_PUBLIC void ixEthDBELTShow(IxEthDBPortId portID);
+IX_ETH_DB_PUBLIC void ixEthDBShowNpeMsgHistory(void);
+
+/* data */
+UINT8* ixEthDBNPEUpdateArea[IX_ETH_DB_NUMBER_OF_PORTS];
+UINT32 dumpEltSize;
+
+/* private data */
+IX_ETH_DB_PRIVATE IxEthDBNoteWriteFn ixEthDBNPENodeWrite[IX_ETH_DB_MAX_RECORD_TYPE_INDEX + 1];
+
+#define IX_ETH_DB_MAX_DELTA_ZONES (6) /* at most 6 EP Delta zones, according to NPE FS */
+IX_ETH_DB_PRIVATE UINT32 ixEthDBEPDeltaOffset[IX_ETH_DB_MAX_RECORD_TYPE_INDEX + 1][IX_ETH_DB_MAX_DELTA_ZONES]; 
+IX_ETH_DB_PRIVATE UINT32 ixEthDBEPDelta[IX_ETH_DB_MAX_RECORD_TYPE_INDEX + 1][IX_ETH_DB_MAX_DELTA_ZONES];
+
+/**
+ * @brief allocates non-cached or contiguous NPE tree update areas for all the ports
+ *
+ * This function is called only once at initialization time from
+ * @ref ixEthDBInit().
+ *
+ * @warning do not call manually
+ *
+ * @see ixEthDBInit()
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBNPEUpdateAreasInit(void)
+{
+    UINT32 portIndex;
+    PortUpdateMethod *update;
+
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        update = &ixEthDBPortInfo[portIndex].updateMethod;
+
+        if (ixEthDBPortDefinitions[portIndex].type == IX_ETH_NPE)
+        {
+            update->npeUpdateZone   = IX_OSAL_CACHE_DMA_MALLOC(FULL_ELT_BYTE_SIZE);
+            update->npeGwUpdateZone = IX_OSAL_CACHE_DMA_MALLOC(FULL_GW_BYTE_SIZE);
+            update->vlanUpdateZone  = IX_OSAL_CACHE_DMA_MALLOC(FULL_VLAN_BYTE_SIZE);
+
+            if (update->npeUpdateZone == NULL
+                || update->npeGwUpdateZone == NULL
+                || update->vlanUpdateZone == NULL)
+            {
+                ERROR_LOG("Fatal error: IX_ACC_DRV_DMA_MALLOC() returned NULL, no NPE update zones available\n");
+            }
+            else
+            {
+                memset(update->npeUpdateZone, 0, FULL_ELT_BYTE_SIZE);
+                memset(update->npeGwUpdateZone, 0, FULL_GW_BYTE_SIZE);
+                memset(update->vlanUpdateZone, 0, FULL_VLAN_BYTE_SIZE);
+            }
+        }
+        else
+        {
+            /* unused */
+            update->npeUpdateZone   = NULL;
+            update->npeGwUpdateZone = NULL;
+            update->vlanUpdateZone  = NULL;
+        }
+    }
+}
+
+/**
+ * @brief deallocates the NPE update areas for all the ports
+ *
+ * This function is called at component de-initialization time
+ * by @ref ixEthDBUnload().
+ *
+ * @warning do not call manually
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBNPEUpdateAreasUnload(void)
+{
+    UINT32 portIndex;
+
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        if (ixEthDBPortDefinitions[portIndex].type == IX_ETH_NPE)
+        {
+            IX_OSAL_CACHE_DMA_FREE(ixEthDBPortInfo[portIndex].updateMethod.npeUpdateZone);
+            IX_OSAL_CACHE_DMA_FREE(ixEthDBPortInfo[portIndex].updateMethod.npeGwUpdateZone);
+            IX_OSAL_CACHE_DMA_FREE(ixEthDBPortInfo[portIndex].updateMethod.vlanUpdateZone);
+        }
+    }
+}
+
+/**
+ * @brief general-purpose NPE callback function
+ *
+ * @param npeID NPE ID
+ * @param msg NPE message
+ *
+ * This function will unblock the caller by unlocking
+ * the npeAckLock mutex defined for each NPE port
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBNpeMsgAck(IxNpeMhNpeId npeID, IxNpeMhMessage msg)
+{
+    IxEthDBPortId portID = IX_ETH_DB_NPE_TO_PORT_ID(npeID);
+    PortInfo *portInfo;
+
+    if (portID >= IX_ETH_DB_NUMBER_OF_PORTS)
+    {
+        /* invalid port */
+        return;
+    }
+
+    if (ixEthDBPortDefinitions[portID].type != IX_ETH_NPE)
+    {
+        /* not an NPE */
+        return;
+    }
+
+    portInfo = &ixEthDBPortInfo[portID];
+    
+    ixOsalMutexUnlock(&portInfo->npeAckLock);
+}
+
+/**
+ * @brief synchronizes the database with tree
+ *
+ * @param portID port ID of the NPE whose tree is to be scanned
+ * @param eltBaseAddress memory base address of the NPE serialized tree
+ * @param eltSize size in bytes of the NPE serialized tree
+ *
+ * Scans the NPE learning tree and resets the age of active database records.
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBNPESyncScan(IxEthDBPortId portID, void *eltBaseAddress, UINT32 eltSize)
+{
+    UINT32 eltEntryOffset;
+    UINT32 entryPortID;
+
+    /* invalidate cache */
+    IX_OSAL_CACHE_INVALIDATE(eltBaseAddress, eltSize);
+
+    for (eltEntryOffset = ELT_ROOT_OFFSET ; eltEntryOffset < eltSize ; eltEntryOffset += ELT_ENTRY_SIZE)
+    {
+        /* (eltBaseAddress + eltEntryOffset) points to a valid NPE tree node
+         *
+         * the format of the node is MAC[6 bytes]:PortID[1 byte]:Reserved[6 bits]:Active[1 bit]:Valid[1 bit]
+         * therefore we can just use the pointer for database searches as only the first 6 bytes are checked
+         */
+        void *eltNodeAddress       = (void *) ((UINT32) eltBaseAddress + eltEntryOffset);
+
+        /* debug */
+        IX_ETH_DB_NPE_VERBOSE_TRACE("DB: (NPEAdaptor) checking node at offset %d...\n", eltEntryOffset / ELT_ENTRY_SIZE);
+
+        if (IX_EDB_NPE_NODE_VALID(eltNodeAddress) != TRUE)
+        {
+            IX_ETH_DB_NPE_VERBOSE_TRACE("\t... node is empty\n");
+        }
+        else if (eltEntryOffset == ELT_ROOT_OFFSET)
+        {
+            IX_ETH_DB_NPE_VERBOSE_TRACE("\t... node is root\n");
+        }
+
+        if (IX_EDB_NPE_NODE_VALID(eltNodeAddress))
+        {
+            entryPortID = IX_ETH_DB_NPE_LOGICAL_ID_TO_PORT_ID(IX_EDB_NPE_NODE_PORT_ID(eltNodeAddress));
+
+            /* check only active entries belonging to this port */
+            if (ixEthDBPortInfo[portID].agingEnabled && IX_EDB_NPE_NODE_ACTIVE(eltNodeAddress) && (portID == entryPortID)
+                && ((ixEthDBPortDefinitions[portID].capabilities & IX_ETH_ENTRY_AGING) == 0))
+            {
+                /* search record */
+                HashNode *node = ixEthDBSearch((IxEthDBMacAddr *) eltNodeAddress, IX_ETH_DB_ALL_FILTERING_RECORDS);
+
+                /* safety check, maybe user deleted record right before sync? */
+                if (node != NULL)
+                {
+                    /* found record */
+                    MacDescriptor *descriptor = (MacDescriptor *) node->data;
+
+                    IX_ETH_DB_NPE_VERBOSE_TRACE("DB: (NPEAdaptor) synced entry [%s] already in the database, updating fields\n", mac2string(eltNodeAddress));
+
+                    /* reset age - set to -1 so that maintenance will restore it to 0 (or more) when incrementing */
+                    if (!descriptor->recordData.filteringData.staticEntry)
+                    {
+                        if (descriptor->type == IX_ETH_DB_FILTERING_RECORD)
+                        {
+                            descriptor->recordData.filteringData.age = AGE_RESET;
+                        }
+                        else if (descriptor->type == IX_ETH_DB_FILTERING_VLAN_RECORD)
+                        {
+                            descriptor->recordData.filteringVlanData.age = AGE_RESET;
+                        }
+                    }
+
+                    /* end transaction */
+                    ixEthDBReleaseHashNode(node);
+                }
+            }
+            else
+            {
+                IX_ETH_DB_NPE_VERBOSE_TRACE("\t... found portID %d, we check only port %d\n", entryPortID, portID);
+            }
+        }
+    }
+}
+
+/**
+ * @brief writes a search tree in NPE format
+ *
+ * @param type type of records to be written into the NPE update zone
+ * @param totalSize maximum size of the linearized tree
+ * @param baseAddress memory base address where to write the NPE tree into
+ * @param tree search tree to write in NPE format
+ * @param blocks number of written 64-byte blocks
+ * @param startIndex optimal binary search start index
+ *
+ * Serializes the given tree in NPE linear format
+ *
+ * @return none
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBNPETreeWrite(IxEthDBRecordType type, UINT32 totalSize, void *baseAddress, MacTreeNode *tree, UINT32 *epDelta, UINT32 *blocks)
+{
+    MacTreeNodeStack *stack;
+    UINT32 maxOffset = 0;
+    UINT32 emptyOffset;
+
+    stack = ixOsalCacheDmaMalloc(sizeof (MacTreeNodeStack));
+    
+    if (stack == NULL)
+    {
+        ERROR_LOG("DB: (NPEAdaptor) failed to allocate the node stack for learning tree linearization, out of memory?\n");
+        return;
+    }
+
+    /* zero out empty root */
+    memset(baseAddress, 0, ELT_ENTRY_SIZE);
+
+    NODE_STACK_INIT(stack);
+
+    if (tree != NULL)
+    {
+        /* push tree root at offset 1 */
+        NODE_STACK_PUSH(stack, tree, 1);
+
+        maxOffset = 1;
+    }
+
+    while (NODE_STACK_NONEMPTY(stack))
+    {
+        MacTreeNode *node;
+        UINT32 offset;
+
+        NODE_STACK_POP(stack, node, offset);
+
+        /* update maximum offset */
+        if (offset > maxOffset)
+        {
+            maxOffset = offset;
+        }
+
+        IX_ETH_DB_NPE_VERBOSE_TRACE("DB: (NPEAdaptor) writing MAC [%s] at offset %d\n", mac2string(node->descriptor->macAddress), offset);
+
+        /* add node to NPE ELT at position indicated by offset */
+        if (offset < MAX_ELT_SIZE)
+        {
+            ixEthDBNPENodeWrite[type]((void *) (((UINT32) baseAddress) + offset * ELT_ENTRY_SIZE), node);
+        }
+
+        if (node->left != NULL)
+        {
+            NODE_STACK_PUSH(stack, node->left, LEFT_CHILD_OFFSET(offset));
+        }
+        else
+        {
+            /* ensure this entry is zeroed */
+            memset((void *) ((UINT32) baseAddress + LEFT_CHILD_OFFSET(offset) * ELT_ENTRY_SIZE), 0, ELT_ENTRY_SIZE);
+        }
+
+        if (node->right != NULL)
+        {
+            NODE_STACK_PUSH(stack, node->right, RIGHT_CHILD_OFFSET(offset));
+        }
+        else
+        {
+            /* ensure this entry is zeroed */
+            memset((void *) ((UINT32) baseAddress + RIGHT_CHILD_OFFSET(offset) * ELT_ENTRY_SIZE), 0, ELT_ENTRY_SIZE);
+        }
+    }
+    
+    emptyOffset = maxOffset + 1;
+
+    /* zero out rest of the tree */
+    IX_ETH_DB_NPE_TRACE("DB: (NPEAdaptor) Emptying tree from offset %d, address 0x%08X, %d bytes\n", 
+        emptyOffset, ((UINT32) baseAddress) + emptyOffset * ELT_ENTRY_SIZE, totalSize - (emptyOffset * ELT_ENTRY_SIZE));
+
+    if (emptyOffset < MAX_ELT_SIZE - 1)
+    {
+        memset((void *) (((UINT32) baseAddress) + (emptyOffset * ELT_ENTRY_SIZE)), 0, totalSize - (emptyOffset * ELT_ENTRY_SIZE));
+    }
+
+    /* flush cache */
+    IX_OSAL_CACHE_FLUSH(baseAddress, totalSize);
+
+    /* debug */
+    IX_ETH_DB_NPE_TRACE("DB: (NPEAdaptor) Ethernet learning/filtering tree XScale wrote at address 0x%08X (max %d bytes):\n\n",
+        (UINT32) baseAddress, FULL_ELT_BYTE_SIZE);
+
+    IX_ETH_DB_NPE_DUMP_ELT(baseAddress, FULL_ELT_BYTE_SIZE);
+    
+    /* compute number of 64-byte blocks */
+    if (blocks != NULL)
+    {
+        *blocks = maxOffset != 0 ? 1 + maxOffset / 8 : 0;
+
+        IX_ETH_DB_NPE_TRACE("DB: (NPEAdaptor) Wrote %d 64-byte blocks\n", *blocks);
+    }
+    
+    /* compute epDelta - start index for binary search */
+    if (epDelta != NULL)
+    {
+        UINT32 deltaIndex = 0;
+
+        *epDelta = 0;
+        
+        for (; deltaIndex < IX_ETH_DB_MAX_DELTA_ZONES ; deltaIndex ++)
+        {
+            if (ixEthDBEPDeltaOffset[type][deltaIndex] >= maxOffset)
+            {
+                *epDelta = ixEthDBEPDelta[type][deltaIndex];
+                break;
+            }
+        }
+
+        IX_ETH_DB_NPE_TRACE("DB: (NPEAdaptor) Computed epDelta %d (based on maxOffset %d)\n", *epDelta, maxOffset);
+    }
+
+    ixOsalCacheDmaFree(stack);
+}
+
+/**
+ * @brief implements a dummy node serialization function
+ *
+ * @param address address of where the node is to be serialized (unused)
+ * @param node tree node to be serialized (unused)
+ *
+ * This function is registered for safety reasons and should
+ * never be called. It will display an error message if this
+ * function is called.
+ *
+ * @return none
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBNullSerialize(void *address, MacTreeNode *node)
+{
+    IX_ETH_DB_NPE_TRACE("DB: (NPEAdaptor) Warning, the NullSerialize function was called, wrong record type?\n");
+}
+
+/**
+ * @brief writes a filtering entry in NPE linear format
+ *
+ * @param address memory address to write node to
+ * @param node node to be written
+ *
+ * Used by @ref ixEthDBNPETreeWrite to liniarize a search tree
+ * in NPE-readable format.
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBNPELearningNodeWrite(void *address, MacTreeNode *node)
+{
+    /* copy mac address */
+    memcpy(address, node->descriptor->macAddress, IX_IEEE803_MAC_ADDRESS_SIZE);
+
+    /* copy port ID */
+    NPE_NODE_BYTE(address, IX_EDB_NPE_NODE_ELT_PORT_ID_OFFSET) = IX_ETH_DB_PORT_ID_TO_NPE_LOGICAL_ID(node->descriptor->portID);
+
+    /* copy flags (valid and not active, as the NPE sets it to active) and clear reserved section (bits 2-7) */
+    NPE_NODE_BYTE(address, IX_EDB_NPE_NODE_ELT_FLAGS_OFFSET) = (UINT8) IX_EDB_FLAGS_INACTIVE_VALID;
+
+    IX_ETH_DB_NPE_VERBOSE_TRACE("DB: (NPEAdaptor) writing ELT node 0x%08x:0x%08x\n", * (UINT32 *) address, * (((UINT32 *) (address)) + 1));
+}
+
+/**
+ * @brief writes a WiFi header conversion record in
+ * NPE linear format
+ *
+ * @param address memory address to write node to
+ * @param node node to be written
+ *
+ * Used by @ref ixEthDBNPETreeWrite to liniarize a search tree
+ * in NPE-readable format.
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBNPEWiFiNodeWrite(void *address, MacTreeNode *node)
+{
+    /* copy mac address */
+    memcpy(address, node->descriptor->macAddress, IX_IEEE803_MAC_ADDRESS_SIZE);
+
+    /* copy index */
+    NPE_NODE_BYTE(address, IX_EDB_NPE_NODE_WIFI_INDEX_OFFSET) = node->descriptor->recordData.wifiData.gwAddressIndex;
+
+    /* copy flags (type and valid) */
+    NPE_NODE_BYTE(address, IX_EDB_NPE_NODE_WIFI_FLAGS_OFFSET) = node->descriptor->recordData.wifiData.type << 1 | IX_EDB_FLAGS_VALID;
+}
+
+/**
+ * @brief writes a WiFi gateway header conversion record in
+ * NPE linear format
+ *
+ * @param address memory address to write node to
+ * @param node node to be written
+ *
+ * Used by @ref ixEthDBNPETreeWrite to liniarize a search tree
+ * in NPE-readable format.
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBNPEGatewayNodeWrite(void *address, MacTreeNode *node)
+{
+    /* copy mac address */
+    memcpy(address, node->descriptor->recordData.wifiData.gwMacAddress, IX_IEEE803_MAC_ADDRESS_SIZE);
+
+    /* set reserved field, two bytes */
+    NPE_NODE_BYTE(address, IX_EDB_NPE_NODE_FW_RESERVED_OFFSET)     = 0;
+    NPE_NODE_BYTE(address, IX_EDB_NPE_NODE_FW_RESERVED_OFFSET + 1) = 0;
+}
+
+/**
+ * @brief writes a firewall record in
+ * NPE linear format
+ *
+ * @param address memory address to write node to
+ * @param node node to be written
+ *
+ * Used by @ref ixEthDBNPETreeWrite to liniarize a search tree
+ * in NPE-readable format.
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBNPEFirewallNodeWrite(void *address, MacTreeNode *node)
+{
+    /* set reserved field */
+    NPE_NODE_BYTE(address, IX_EDB_NPE_NODE_FW_RESERVED_OFFSET) = 0;
+
+    /* set flags */
+    NPE_NODE_BYTE(address, IX_EDB_NPE_NODE_FW_FLAGS_OFFSET) = IX_EDB_FLAGS_VALID;
+
+    /* copy mac address */
+    memcpy((void *) ((UINT32) address + IX_EDB_NPE_NODE_FW_ADDR_OFFSET), node->descriptor->macAddress, IX_IEEE803_MAC_ADDRESS_SIZE);
+}
+
+/**
+ * @brief registers the NPE serialization methods
+ *
+ * This functions registers NPE serialization methods
+ * for writing the following types of records in NPE
+ * readable linear format:
+ * - filtering records
+ * - WiFi header conversion records
+ * - WiFi gateway header conversion records
+ * - firewall records
+ *
+ * Note that this function should be called by the
+ * component initialization function.
+ *
+ * @return number of registered record types
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+UINT32 ixEthDBRecordSerializeMethodsRegister()
+{
+    int i;
+    
+    /* safety - register a blank method for everybody first */
+    for ( i = 0 ; i < IX_ETH_DB_MAX_RECORD_TYPE_INDEX + 1 ; i++)
+    {
+        ixEthDBNPENodeWrite[i] = ixEthDBNullSerialize;
+    }
+    
+    /* register real methods */
+    ixEthDBNPENodeWrite[IX_ETH_DB_FILTERING_RECORD]      = ixEthDBNPELearningNodeWrite;
+    ixEthDBNPENodeWrite[IX_ETH_DB_FILTERING_VLAN_RECORD] = ixEthDBNPELearningNodeWrite;
+    ixEthDBNPENodeWrite[IX_ETH_DB_WIFI_RECORD]           = ixEthDBNPEWiFiNodeWrite;
+    ixEthDBNPENodeWrite[IX_ETH_DB_FIREWALL_RECORD]       = ixEthDBNPEFirewallNodeWrite;
+    ixEthDBNPENodeWrite[IX_ETH_DB_GATEWAY_RECORD]        = ixEthDBNPEGatewayNodeWrite;
+    
+    /* EP Delta arrays */
+    memset(ixEthDBEPDeltaOffset, 0, sizeof (ixEthDBEPDeltaOffset));
+    memset(ixEthDBEPDelta, 0, sizeof (ixEthDBEPDelta));
+    
+    /* filtering records */
+    ixEthDBEPDeltaOffset[IX_ETH_DB_FILTERING_RECORD][0] = 1;
+    ixEthDBEPDelta[IX_ETH_DB_FILTERING_RECORD][0]       = 0;
+    
+    ixEthDBEPDeltaOffset[IX_ETH_DB_FILTERING_RECORD][1] = 3;
+    ixEthDBEPDelta[IX_ETH_DB_FILTERING_RECORD][1]       = 7;
+    
+    ixEthDBEPDeltaOffset[IX_ETH_DB_FILTERING_RECORD][2] = 511;
+    ixEthDBEPDelta[IX_ETH_DB_FILTERING_RECORD][2]       = 14;
+    
+    /* wifi records */
+    ixEthDBEPDeltaOffset[IX_ETH_DB_WIFI_RECORD][0] = 1;
+    ixEthDBEPDelta[IX_ETH_DB_WIFI_RECORD][0]       = 0;
+    
+    ixEthDBEPDeltaOffset[IX_ETH_DB_WIFI_RECORD][1] = 3;
+    ixEthDBEPDelta[IX_ETH_DB_WIFI_RECORD][1]       = 7;
+    
+    ixEthDBEPDeltaOffset[IX_ETH_DB_WIFI_RECORD][2] = 511;
+    ixEthDBEPDelta[IX_ETH_DB_WIFI_RECORD][2]       = 14;
+
+    /* firewall records */
+    ixEthDBEPDeltaOffset[IX_ETH_DB_FIREWALL_RECORD][0] = 0;
+    ixEthDBEPDelta[IX_ETH_DB_FIREWALL_RECORD][0]       = 0;
+    
+    ixEthDBEPDeltaOffset[IX_ETH_DB_FIREWALL_RECORD][1] = 1;
+    ixEthDBEPDelta[IX_ETH_DB_FIREWALL_RECORD][1]       = 5;
+    
+    ixEthDBEPDeltaOffset[IX_ETH_DB_FIREWALL_RECORD][2] = 3;
+    ixEthDBEPDelta[IX_ETH_DB_FIREWALL_RECORD][2]       = 13;
+    
+    ixEthDBEPDeltaOffset[IX_ETH_DB_FIREWALL_RECORD][3] = 7;
+    ixEthDBEPDelta[IX_ETH_DB_FIREWALL_RECORD][3]       = 21;
+    
+    ixEthDBEPDeltaOffset[IX_ETH_DB_FIREWALL_RECORD][4] = 15;
+    ixEthDBEPDelta[IX_ETH_DB_FIREWALL_RECORD][4]       = 29;
+    
+    ixEthDBEPDeltaOffset[IX_ETH_DB_FIREWALL_RECORD][5] = 31;
+    ixEthDBEPDelta[IX_ETH_DB_FIREWALL_RECORD][5]       = 37;
+    
+    return 5; /* 5 methods registered */
+}
+
+#ifndef IX_NDEBUG
+
+IX_ETH_DB_PUBLIC UINT32 npeMsgHistory[IX_ETH_DB_NPE_MSG_HISTORY_DEPTH][2];
+IX_ETH_DB_PUBLIC UINT32 npeMsgHistoryLen = 0;
+
+/**
+ * When compiled in DEBUG mode, this function can be used to display
+ * the history of messages sent to the NPEs (up to 100).
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBShowNpeMsgHistory()
+{
+    UINT32 i = 0;
+    UINT32 base, len;
+
+    if (npeMsgHistoryLen <= IX_ETH_DB_NPE_MSG_HISTORY_DEPTH)
+    {
+        base = 0;
+        len  = npeMsgHistoryLen;
+    }
+    else
+    {
+        base = npeMsgHistoryLen % IX_ETH_DB_NPE_MSG_HISTORY_DEPTH;
+        len  = IX_ETH_DB_NPE_MSG_HISTORY_DEPTH;
+    }
+
+    printf("NPE message history [last %d messages, from least to most recent]:\n", len);
+
+    for (; i < len ; i++)
+    {
+        UINT32 pos = (base + i) % IX_ETH_DB_NPE_MSG_HISTORY_DEPTH;
+        printf("msg[%d]: 0x%08x:0x%08x\n", i, npeMsgHistory[pos][0], npeMsgHistory[pos][1]);
+    }
+}
+
+IX_ETH_DB_PUBLIC
+void ixEthDBELTShow(IxEthDBPortId portID)
+{
+    IxNpeMhMessage message;
+    IX_STATUS result;
+    
+    /* send EDB_GetMACAddressDatabase message */
+    FILL_GETMACADDRESSDATABASE(message, 
+        0 /* reserved */, 
+        IX_OSAL_MMU_VIRT_TO_PHYS(ixEthDBPortInfo[portID].updateMethod.npeUpdateZone));
+
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+
+    if (result == IX_SUCCESS)
+    {
+        /* analyze NPE copy */
+        UINT32 eltEntryOffset;
+        UINT32 entryPortID;
+
+        UINT32 eltBaseAddress = (UINT32) ixEthDBPortInfo[portID].updateMethod.npeUpdateZone;
+        UINT32 eltSize        = FULL_ELT_BYTE_SIZE;
+
+        /* invalidate cache */
+        IX_OSAL_CACHE_INVALIDATE((void *) eltBaseAddress, eltSize);
+
+        printf("Listing records in main learning tree for port %d\n", portID);
+
+        for (eltEntryOffset = ELT_ROOT_OFFSET ; eltEntryOffset < eltSize ; eltEntryOffset += ELT_ENTRY_SIZE)
+        {
+            /* (eltBaseAddress + eltEntryOffset) points to a valid NPE tree node
+            *
+            * the format of the node is MAC[6 bytes]:PortID[1 byte]:Reserved[6 bits]:Active[1 bit]:Valid[1 bit]
+            * therefore we can just use the pointer for database searches as only the first 6 bytes are checked
+            */
+            void *eltNodeAddress = (void *) ((UINT32) eltBaseAddress + eltEntryOffset);
+
+            if (IX_EDB_NPE_NODE_VALID(eltNodeAddress))
+            {
+                HashNode *node;
+
+                entryPortID = IX_ETH_DB_NPE_LOGICAL_ID_TO_PORT_ID(IX_EDB_NPE_NODE_PORT_ID(eltNodeAddress));
+
+                /* search record */
+                node = ixEthDBSearch((IxEthDBMacAddr *) eltNodeAddress, IX_ETH_DB_ALL_RECORD_TYPES);
+
+                printf("%s - port %d - %s ", mac2string((unsigned char *) eltNodeAddress), entryPortID, 
+                    IX_EDB_NPE_NODE_ACTIVE(eltNodeAddress) ? "active" : "inactive");
+
+                /* safety check, maybe user deleted record right before sync? */
+                if (node != NULL)
+                {
+                    /* found record */
+                    MacDescriptor *descriptor = (MacDescriptor *) node->data;
+
+                    printf("- %s ",
+                        descriptor->type == IX_ETH_DB_FILTERING_RECORD ? "filtering" :
+                        descriptor->type == IX_ETH_DB_FILTERING_VLAN_RECORD ? "vlan" :
+                        descriptor->type == IX_ETH_DB_WIFI_RECORD ? "wifi" : "other (check main DB)");
+
+                    if (descriptor->type == IX_ETH_DB_FILTERING_RECORD) printf("- age %d - %s ", 
+                        descriptor->recordData.filteringData.age,
+                        descriptor->recordData.filteringData.staticEntry ? "static" : "dynamic");
+
+                    if (descriptor->type == IX_ETH_DB_FILTERING_VLAN_RECORD) printf("- age %d - %s - tci %d ",
+                        descriptor->recordData.filteringVlanData.age,
+                        descriptor->recordData.filteringVlanData.staticEntry ? "static" : "dynamic",
+                        descriptor->recordData.filteringVlanData.ieee802_1qTag);
+
+                    /* end transaction */
+                    ixEthDBReleaseHashNode(node);
+                }
+                else
+                {
+                    printf("- not synced");
+                }
+
+                printf("\n");
+            }
+        }
+    }
+    else
+    {
+        ixOsalLog(IX_OSAL_LOG_LVL_FATAL, IX_OSAL_LOG_DEV_STDOUT, 
+            "EthDB: (ShowELT) Could not complete action (communication failure)\n",
+            portID, 0, 0, 0, 0, 0);
+    }
+}
+
+#endif

cpu/ixp/npe/IxEthDBPortUpdate.c

@@ -0,0 +1,740 @@
+/**
+ * @file IxEthDBDBPortUpdate.c
+ *
+ * @brief Implementation of dependency and port update handling
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxEthDB_p.h"
+
+/* forward prototype declarations */
+IX_ETH_DB_PRIVATE MacTreeNode* ixEthDBTreeInsert(MacTreeNode *searchTree, MacDescriptor *descriptor);
+IX_ETH_DB_PRIVATE void ixEthDBCreateTrees(IxEthDBPortMap updatePorts);
+IX_ETH_DB_PRIVATE MacTreeNode* ixEthDBTreeRebalance(MacTreeNode *searchTree);
+IX_ETH_DB_PRIVATE void ixEthDBRebalanceTreeToVine(MacTreeNode *root, UINT32 *size);
+IX_ETH_DB_PRIVATE void ixEthDBRebalanceVineToTree(MacTreeNode *root, UINT32 size);
+IX_ETH_DB_PRIVATE void ixEthDBRebalanceCompression(MacTreeNode *root, UINT32 count);
+IX_ETH_DB_PRIVATE UINT32 ixEthDBRebalanceLog2Floor(UINT32 x);
+
+extern HashTable dbHashtable;
+
+/**
+ * @brief register types requiring automatic updates
+ *
+ * @param typeArray array indexed on record types, each
+ * element indicating whether the record type requires an
+ * automatic update (TRUE) or not (FALSE)
+ * 
+ * Automatic updates are done for registered record types
+ * upon adding, updating (that is, updating the record portID) 
+ * and removing records. Whenever an automatic update is triggered
+ * the appropriate ports will be provided with new database
+ * information.
+ *
+ * It is assumed that the typeArray parameter is allocated large
+ * enough to hold all the user defined types. Also, the type
+ * array should be initialized to FALSE as this function only
+ * caters for types which do require automatic updates.
+ *
+ * Note that this function should be called by the component
+ * initialization function.
+ *
+ * @return number of record types registered for automatic
+ * updates
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+UINT32 ixEthDBUpdateTypeRegister(BOOL *typeArray)
+{
+    typeArray[IX_ETH_DB_FILTERING_RECORD]      = TRUE;
+    typeArray[IX_ETH_DB_FILTERING_VLAN_RECORD] = TRUE;
+
+    return 2;
+}
+
+/**
+ * @brief computes dependencies and triggers port learning tree updates
+ *
+ * @param triggerPorts port map consisting in the ports which triggered the update
+ *
+ * This function browses through all the ports and determines how to waterfall the update
+ * event from the trigger ports to all other ports depending on them.
+ *
+ * Once the list of ports to be updated is determined this function 
+ * calls @ref ixEthDBCreateTrees.
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBUpdatePortLearningTrees(IxEthDBPortMap triggerPorts)
+{
+    IxEthDBPortMap updatePorts;
+    UINT32 portIndex;
+    
+    ixEthDBUpdateLock();
+    
+    SET_EMPTY_DEPENDENCY_MAP(updatePorts);
+    
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        PortInfo *port   = &ixEthDBPortInfo[portIndex];
+        BOOL mapsCollide;
+        
+        MAPS_COLLIDE(mapsCollide, triggerPorts, port->dependencyPortMap);
+
+        if (mapsCollide                                   /* do triggers influence this port? */
+            && !IS_PORT_INCLUDED(portIndex, updatePorts)  /* and it's not already in the update list */
+            && port->updateMethod.updateEnabled)          /* and we're allowed to update it */
+        {
+            IX_ETH_DB_UPDATE_TRACE("DB: (Update) Adding port %d to update set\n", portIndex);
+
+            JOIN_PORT_TO_MAP(updatePorts, portIndex);
+        }
+        else
+        {
+            IX_ETH_DB_UPDATE_TRACE("DB: (Update) Didn't add port %d to update set, reasons follow:\n", portIndex);
+
+            if (!mapsCollide)
+            {
+                IX_ETH_DB_UPDATE_TRACE("\tMaps don't collide on port %d\n", portIndex);
+            }
+
+            if (IS_PORT_INCLUDED(portIndex, updatePorts))
+            {
+                IX_ETH_DB_UPDATE_TRACE("\tPort %d is already in the update set\n", portIndex);
+            }
+
+            if (!port->updateMethod.updateEnabled)
+            {
+                IX_ETH_DB_UPDATE_TRACE("\tPort %d doesn't have updateEnabled set\n", portIndex);
+            }
+        }
+    }
+    
+    IX_ETH_DB_UPDATE_TRACE("DB: (Update) Updating port set\n");
+
+    ixEthDBCreateTrees(updatePorts);
+        
+    ixEthDBUpdateUnlock();
+}
+
+/**
+ * @brief creates learning trees and calls the port update handlers
+ *
+ * @param updatePorts set of ports in need of learning trees
+ *
+ * This function determines the optimal method of creating learning
+ * trees using a minimal number of database queries, keeping in mind
+ * that different ports can either use the same learning trees or they
+ * can partially share them. The actual tree building routine is
+ * @ref ixEthDBQuery.
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBCreateTrees(IxEthDBPortMap updatePorts)
+{
+    UINT32 portIndex;
+    BOOL result;
+    BOOL portsLeft = TRUE;
+
+    while (portsLeft)
+    {
+        /* get port with minimal dependency map and NULL search tree */
+        UINT32 minPortIndex = MAX_PORT_SIZE;
+        UINT32 minimalSize  = MAX_PORT_SIZE;
+
+        for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+        {
+            UINT32 size;
+            PortInfo *port = &ixEthDBPortInfo[portIndex];
+
+            /* generate trees only for ports that need them */
+            if (!port->updateMethod.searchTreePendingWrite && IS_PORT_INCLUDED(portIndex, updatePorts))
+            {
+                GET_MAP_SIZE(port->dependencyPortMap, size);
+                
+                IX_ETH_DB_UPDATE_TRACE("DB: (Update) Dependency map for port %d: size %d\n",
+                    portIndex, size);
+
+                if (size < minimalSize)
+                {
+                    minPortIndex = portIndex;
+                    minimalSize  = size;
+                }
+            }
+            else
+            {
+                IX_ETH_DB_UPDATE_TRACE("DB: (Update) Skipped port %d from tree diff (%s)\n", portIndex,
+                    port->updateMethod.searchTreePendingWrite ? "pending write access" : "ignored by query");
+            }            
+        }
+
+        /* if a port was found than minimalSize is not MAX_PORT_SIZE */
+        if (minimalSize != MAX_PORT_SIZE)
+        {
+            /* minPortIndex is the port we seek */
+            PortInfo *port = &ixEthDBPortInfo[minPortIndex];
+
+            IxEthDBPortMap query;
+            MacTreeNode *baseTree;
+
+            /* now try to find a port with minimal map difference */
+            PortInfo *minimalDiffPort = NULL;
+            UINT32 minimalDiff        = MAX_PORT_SIZE;
+            
+            IX_ETH_DB_UPDATE_TRACE("DB: (Update) Minimal size port is %d\n", minPortIndex);
+
+            for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+            {   
+                PortInfo *diffPort = &ixEthDBPortInfo[portIndex];
+                BOOL mapIsSubset;
+                
+                IS_MAP_SUBSET(mapIsSubset, diffPort->dependencyPortMap, port->dependencyPortMap);
+                
+
+                if (portIndex != minPortIndex
+                    && diffPort->updateMethod.searchTree != NULL
+                    && mapIsSubset)
+                {
+                    /* compute size and pick only minimal size difference */
+                    UINT32 diffPortSize;
+                    UINT32 sizeDifference;
+
+                    GET_MAP_SIZE(diffPort->dependencyPortMap, diffPortSize);
+                     
+                    IX_ETH_DB_UPDATE_TRACE("DB: (Update) Checking port %d for differences...\n", portIndex);
+
+                    sizeDifference = minimalSize - diffPortSize;
+
+                    if (sizeDifference < minimalDiff)
+                    {
+                        minimalDiffPort = diffPort;
+                        minimalDiff     = sizeDifference;
+                        
+                        IX_ETH_DB_UPDATE_TRACE("DB: (Update) Minimal difference 0x%x was found on port %d\n",
+                            minimalDiff, portIndex);
+                    }
+                }
+            }
+
+            /* check if filtering is enabled on this port */
+            if ((port->featureStatus & IX_ETH_DB_FILTERING) != 0)
+            {
+                /* if minimalDiff is not MAX_PORT_SIZE minimalDiffPort points to the most similar port */
+                if (minimalDiff != MAX_PORT_SIZE)
+                {
+                    baseTree = ixEthDBCloneMacTreeNode(minimalDiffPort->updateMethod.searchTree);
+                    DIFF_MAPS(query, port->dependencyPortMap , minimalDiffPort->dependencyPortMap);
+                    
+                    IX_ETH_DB_UPDATE_TRACE("DB: (Update) Found minimal diff, extending tree %d on query\n",
+                        minimalDiffPort->portID);
+                }
+                else /* .. otherwise no similar port was found, build tree from scratch */
+                {
+                    baseTree = NULL;
+                    
+                    COPY_DEPENDENCY_MAP(query, port->dependencyPortMap);
+                    
+                    IX_ETH_DB_UPDATE_TRACE("DB: (Update) No similar diff, creating tree from query\n");
+                }
+
+                IS_EMPTY_DEPENDENCY_MAP(result, query);
+                
+                if (!result) /* otherwise we don't need anything more on top of the cloned tree */
+                {
+                    IX_ETH_DB_UPDATE_TRACE("DB: (Update) Adding query tree to port %d\n", minPortIndex);
+                        
+                    /* build learning tree */
+                    port->updateMethod.searchTree = ixEthDBQuery(baseTree, query, IX_ETH_DB_ALL_FILTERING_RECORDS, MAX_ELT_SIZE);
+                }
+                else
+                {
+                    IX_ETH_DB_UPDATE_TRACE("DB: (Update) Query is empty, assuming identical nearest tree\n");
+                      
+                    port->updateMethod.searchTree = baseTree;
+                }
+            }
+            else
+            {
+                /* filtering is not enabled, will download an empty tree */
+                if (port->updateMethod.searchTree != NULL)
+                {
+                    ixEthDBFreeMacTreeNode(port->updateMethod.searchTree);
+                }
+
+                port->updateMethod.searchTree = NULL;
+            }
+
+            /* mark tree as valid */
+            port->updateMethod.searchTreePendingWrite = TRUE;
+        }
+        else
+        {
+            portsLeft = FALSE;
+
+            IX_ETH_DB_UPDATE_TRACE("DB: (Update) No trees to create this round\n");            
+        }
+    }
+    
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        PortInfo *updatePort = &ixEthDBPortInfo[portIndex];
+
+        if (updatePort->updateMethod.searchTreePendingWrite)
+        {
+            IX_ETH_DB_UPDATE_TRACE("DB: (PortUpdate) Starting procedure to upload new search tree (%snull) into NPE %d\n", 
+                updatePort->updateMethod.searchTree != NULL ? "not " : "",
+                portIndex);
+
+            updatePort->updateMethod.updateHandler(portIndex, IX_ETH_DB_FILTERING_RECORD);
+        }
+    }
+}
+
+/**
+ * @brief standard NPE update handler
+ *
+ * @param portID id of the port to be updated
+ * @param type record type to be pushed during this update
+ *
+ * The NPE update handler manages updating the NPE databases
+ * given a certain record type.
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBNPEUpdateHandler(IxEthDBPortId portID, IxEthDBRecordType type)
+{
+    UINT32 epDelta, blockCount;
+    IxNpeMhMessage message;
+    UINT32 treeSize = 0;
+    PortInfo *port = &ixEthDBPortInfo[portID];
+
+    /* size selection and type check */
+    if (type == IX_ETH_DB_FILTERING_RECORD || type == IX_ETH_DB_WIFI_RECORD)
+    {
+        treeSize = FULL_ELT_BYTE_SIZE;
+    }
+    else if (type == IX_ETH_DB_FIREWALL_RECORD)
+    {
+        treeSize = FULL_FW_BYTE_SIZE;
+    }
+    else
+    {
+        return IX_ETH_DB_INVALID_ARG;
+    }
+    
+    /* serialize tree into memory */
+    ixEthDBNPETreeWrite(type, treeSize, port->updateMethod.npeUpdateZone, port->updateMethod.searchTree, &epDelta, &blockCount);
+
+    /* free internal copy */
+    if (port->updateMethod.searchTree != NULL)
+    {
+        ixEthDBFreeMacTreeNode(port->updateMethod.searchTree);
+    }
+
+    /* forget last used search tree */
+    port->updateMethod.searchTree             = NULL;
+    port->updateMethod.searchTreePendingWrite = FALSE;
+
+    /* dependending on the update type we do different things */
+    if (type == IX_ETH_DB_FILTERING_RECORD || type == IX_ETH_DB_WIFI_RECORD)
+    {
+        IX_STATUS result;
+
+        FILL_SETMACADDRESSDATABASE_MSG(message, IX_ETH_DB_PORT_ID_TO_NPE_LOGICAL_ID(portID), 
+            epDelta, blockCount, 
+            IX_OSAL_MMU_VIRT_TO_PHYS(port->updateMethod.npeUpdateZone));
+
+        IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+
+        if (result == IX_SUCCESS)
+	{
+            IX_ETH_DB_UPDATE_TRACE("DB: (PortUpdate) Finished downloading NPE tree on port %d\n", portID);
+        }
+        else
+        {
+            ixEthDBPortInfo[portID].agingEnabled                = FALSE;
+            ixEthDBPortInfo[portID].updateMethod.updateEnabled  = FALSE;
+            ixEthDBPortInfo[portID].updateMethod.userControlled = TRUE;
+
+            ERROR_LOG("EthDB: (PortUpdate) disabling aging and updates on port %d (assumed dead)\n", portID);
+
+            ixEthDBDatabaseClear(portID, IX_ETH_DB_ALL_RECORD_TYPES);
+
+            return IX_ETH_DB_FAIL;
+        }
+
+	return IX_ETH_DB_SUCCESS;
+    }
+    else if (type == IX_ETH_DB_FIREWALL_RECORD)
+    {
+        return ixEthDBFirewallUpdate(portID, port->updateMethod.npeUpdateZone, epDelta);
+    }
+    
+    return IX_ETH_DB_INVALID_ARG;
+}
+
+/**
+ * @brief queries the database for a set of records to be inserted into a given tree
+ *
+ * @param searchTree pointer to a tree where insertions will be performed; can be NULL
+ * @param query set of ports that a database record must match to be inserted into the tree
+ *
+ * The query method browses through the database, extracts all the descriptors matching
+ * the given query parameter and inserts them into the given learning tree.
+ * Note that this is an append procedure, the given tree needs not to be empty.
+ * A "descriptor matching the query" is a descriptor whose port id is in the query map.
+ * If the given tree is empty (NULL) a new tree is created and returned.
+ * 
+ * @return the tree root
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+MacTreeNode* ixEthDBQuery(MacTreeNode *searchTree, IxEthDBPortMap query, IxEthDBRecordType recordFilter, UINT32 maxEntries)
+{
+    HashIterator iterator;
+    UINT32 entryCount = 0;
+
+    /* browse database */
+    BUSY_RETRY(ixEthDBInitHashIterator(&dbHashtable, &iterator));
+
+    while (IS_ITERATOR_VALID(&iterator))
+    {
+        MacDescriptor *descriptor = (MacDescriptor *) iterator.node->data;
+
+        IX_ETH_DB_UPDATE_TRACE("DB: (PortUpdate) querying [%s]:%d on port map ... ", 
+            mac2string(descriptor->macAddress), 
+            descriptor->portID);
+
+	if ((descriptor->type & recordFilter) != 0 
+            && IS_PORT_INCLUDED(descriptor->portID, query))
+	{
+            MacDescriptor *descriptorClone = ixEthDBCloneMacDescriptor(descriptor);
+
+            IX_ETH_DB_UPDATE_TRACE("match\n");
+
+            if (descriptorClone != NULL)
+            {
+                /* add descriptor to tree */
+                searchTree = ixEthDBTreeInsert(searchTree, descriptorClone);
+
+                entryCount++;
+            }
+        }
+        else
+        {
+            IX_ETH_DB_UPDATE_TRACE("no match\n");
+        }
+
+        if (entryCount < maxEntries)
+        {
+            /* advance to the next record */
+	        BUSY_RETRY(ixEthDBIncrementHashIterator(&dbHashtable, &iterator));
+        }
+        else
+        {
+            /* the NPE won't accept more entries so we can stop now */
+            ixEthDBReleaseHashIterator(&iterator);
+
+            IX_ETH_DB_UPDATE_TRACE("DB: (PortUpdate) number of elements reached maximum supported by port\n");
+
+            break;
+        }
+    }
+
+    IX_ETH_DB_UPDATE_TRACE("DB: (PortUpdate) query inserted %d records in the search tree\n", entryCount);
+
+    return ixEthDBTreeRebalance(searchTree);
+}
+
+/**
+ * @brief inserts a mac descriptor into an tree
+ *
+ * @param searchTree tree where the insertion is to be performed (may be NULL)
+ * @param descriptor descriptor to insert into tree
+ *
+ * @return the tree root
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+MacTreeNode* ixEthDBTreeInsert(MacTreeNode *searchTree, MacDescriptor *descriptor)
+{
+    MacTreeNode *currentNode    = searchTree;
+    MacTreeNode *insertLocation = NULL;
+    MacTreeNode *newNode;
+    INT32 insertPosition = RIGHT;
+
+    if (descriptor == NULL)
+    {
+        return searchTree;
+    }
+
+    /* create a new node */
+    newNode = ixEthDBAllocMacTreeNode();
+
+    if (newNode == NULL)
+    {
+        /* out of memory */
+        ERROR_LOG("Warning: ixEthDBAllocMacTreeNode returned NULL in file %s:%d (out of memory?)\n", __FILE__, __LINE__);
+
+        ixEthDBFreeMacDescriptor(descriptor);
+
+        return NULL;
+    }
+
+    /* populate node */
+    newNode->descriptor = descriptor;
+
+    /* an empty initial tree is a special case */
+    if (searchTree == NULL)
+    {
+        return newNode;
+    }
+
+    /* get insertion location */
+    while (insertLocation == NULL)
+    {
+        MacTreeNode *nextNode;
+
+        /* compare given key with current node key */
+        insertPosition = ixEthDBAddressCompare(descriptor->macAddress, currentNode->descriptor->macAddress);
+
+        /* navigate down */
+        if (insertPosition == RIGHT)
+        {
+            nextNode = currentNode->right;
+        }
+        else if (insertPosition == LEFT)
+        {
+            nextNode = currentNode->left;
+        }
+        else
+        {
+            /* error, duplicate key */
+            ERROR_LOG("Warning: trapped insertion of a duplicate MAC address in an NPE search tree\n");
+
+            /* this will free the MAC descriptor as well */
+            ixEthDBFreeMacTreeNode(newNode);
+
+            return searchTree;
+        }
+
+        /* when we can no longer dive through the tree we found the insertion place */
+        if (nextNode != NULL)
+        {
+            currentNode = nextNode;
+        }
+        else
+        {
+            insertLocation = currentNode;
+        }
+    }
+
+    /* insert node */
+    if (insertPosition == RIGHT)
+    {
+        insertLocation->right = newNode;
+    }
+    else
+    {
+        insertLocation->left = newNode;
+    }
+
+    return searchTree;
+}
+
+/**
+ * @brief balance a tree
+ *
+ * @param searchTree tree to balance
+ *
+ * Converts a tree into a balanced tree and returns the root of
+ * the balanced tree. The resulting tree is <i>route balanced</i>
+ * not <i>perfectly balanced</i>. This makes no difference to the
+ * average tree search time which is the same in both cases, O(log2(n)).
+ *
+ * @return root of the balanced tree or NULL if there's no memory left
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+MacTreeNode* ixEthDBTreeRebalance(MacTreeNode *searchTree)
+{
+    MacTreeNode *pseudoRoot = ixEthDBAllocMacTreeNode();
+    UINT32 size;
+
+    if (pseudoRoot == NULL)
+    {
+        /* out of memory */
+        return NULL;
+    }
+
+    pseudoRoot->right = searchTree;
+
+    ixEthDBRebalanceTreeToVine(pseudoRoot, &size);
+    ixEthDBRebalanceVineToTree(pseudoRoot, size);
+
+    searchTree = pseudoRoot->right;
+
+    /* remove pseudoRoot right branch, otherwise it will free the entire tree */
+    pseudoRoot->right = NULL;
+
+    ixEthDBFreeMacTreeNode(pseudoRoot);
+
+    return searchTree;
+}
+
+/**
+ * @brief converts a tree into a vine
+ *
+ * @param root root of tree to convert
+ * @param size depth of vine (equal to the number of nodes in the tree)
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBRebalanceTreeToVine(MacTreeNode *root, UINT32 *size)
+{
+    MacTreeNode *vineTail  = root;
+    MacTreeNode *remainder = vineTail->right;
+    MacTreeNode *tempPtr;
+
+    *size = 0;
+
+    while (remainder != NULL)
+    {
+        if (remainder->left == NULL)
+        {
+            /* move tail down one */
+            vineTail  = remainder;
+            remainder = remainder->right;
+            (*size)++;
+        }
+        else
+        {
+            /* rotate around remainder */
+            tempPtr         = remainder->left;
+            remainder->left = tempPtr->right;
+            tempPtr->right  = remainder;
+            remainder       = tempPtr;
+            vineTail->right = tempPtr;
+        }
+    }
+}
+
+/**
+ * @brief converts a vine into a balanced tree
+ *
+ * @param root vine to convert
+ * @param size depth of vine
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBRebalanceVineToTree(MacTreeNode *root, UINT32 size)
+{
+    UINT32 leafCount = size + 1 - (1 << ixEthDBRebalanceLog2Floor(size + 1));
+
+    ixEthDBRebalanceCompression(root, leafCount);
+
+    size = size - leafCount;
+
+    while (size > 1)
+    {
+        ixEthDBRebalanceCompression(root, size / 2);
+
+        size /= 2;
+    }
+}
+
+/**
+ * @brief compresses a vine/tree stage into a more balanced vine/tree
+ *
+ * @param root root of the tree to compress
+ * @param count number of "spine" nodes
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBRebalanceCompression(MacTreeNode *root, UINT32 count)
+{
+    MacTreeNode *scanner = root;
+    MacTreeNode *child;
+    UINT32 local_index;
+
+    for (local_index = 0 ; local_index < count ; local_index++)
+    {
+        child          = scanner->right;
+        scanner->right = child->right;
+        scanner        = scanner->right;
+        child->right   = scanner->left;
+        scanner->left  = child;
+    }
+}
+
+/**
+ * @brief computes |_log2(x)_| (a.k.a. floor(log2(x)))
+ *
+ * @param x number to compute |_log2(x)_| for
+ *
+ * @return |_log2(x)_|
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+UINT32 ixEthDBRebalanceLog2Floor(UINT32 x)
+{
+    UINT32 log = 0;
+    UINT32 val = 1;
+
+    while (val < x)
+    {
+        log++;
+        val <<= 1;
+    }
+
+    return val == x ? log : log - 1;
+}
+

cpu/ixp/npe/IxEthDBReports.c

@@ -0,0 +1,652 @@
+/**
+ * @file IxEthDBAPI.c
+ *
+ * @brief Implementation of the public API
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxEthDB_p.h"
+
+extern HashTable dbHashtable;
+IX_ETH_DB_PRIVATE void ixEthDBPortInfoShow(IxEthDBPortId portID, IxEthDBRecordType recordFilter);
+IX_ETH_DB_PRIVATE IxEthDBStatus ixEthDBHeaderShow(IxEthDBRecordType recordFilter);
+IX_ETH_DB_PUBLIC IxEthDBStatus ixEthDBDependencyPortMapShow(IxEthDBPortId portID, IxEthDBPortMap map);
+
+/**
+ * @brief displays a port dependency map
+ *
+ * @param portID ID of the port
+ * @param map port map to display
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully
+ */ 
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBDependencyPortMapShow(IxEthDBPortId portID, IxEthDBPortMap map)
+{
+    UINT32 portIndex;
+    BOOL mapSelf = TRUE, mapNone = TRUE, firstPort = TRUE;
+    
+    /* dependency port maps */
+    printf("Dependency port map: ");
+    
+    /* browse the port map */
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        if (IS_PORT_INCLUDED(portIndex, map))
+        {
+            mapNone   = FALSE;
+            
+            if (portIndex != portID)
+            {
+                mapSelf = FALSE;
+            }
+            
+            printf("%s%d", firstPort ? "{" : ", ", portIndex);
+            
+            firstPort = FALSE;
+        }
+    }
+    
+    if (mapNone)
+    {
+        mapSelf = FALSE;
+    }
+    
+    printf("%s (%s)\n", firstPort ? "" : "}", mapSelf ? "self" : mapNone ? "none" : "group");
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief displays all the filtering records belonging to a port
+ *
+ * @param portID ID of the port to display
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @warning deprecated, use @ref ixEthDBFilteringDatabaseShowRecords() 
+ * instead. Calling this function is equivalent to calling
+ * ixEthDBFilteringDatabaseShowRecords(portID, IX_ETH_DB_FILTERING_RECORD)
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFilteringDatabaseShow(IxEthDBPortId portID)
+{
+    IxEthDBStatus local_result;
+    HashIterator iterator;
+    PortInfo *portInfo;
+    UINT32 recordCount = 0;
+
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    portInfo = &ixEthDBPortInfo[portID];
+
+    /* display table header */
+    printf("Ethernet database records for port ID [%d]\n", portID);
+    
+    ixEthDBDependencyPortMapShow(portID, portInfo->dependencyPortMap);
+    
+    if (ixEthDBPortDefinitions[portID].type == IX_ETH_NPE)
+    {
+        printf("NPE updates are %s\n\n", portInfo->updateMethod.updateEnabled ? "enabled" : "disabled");
+    }
+    else
+    {
+        printf("updates disabled (not an NPE)\n\n");
+    }
+
+    printf("    MAC address    |   Age  | Type \n");
+    printf("___________________________________\n");
+
+    /* browse database */
+    BUSY_RETRY(ixEthDBInitHashIterator(&dbHashtable, &iterator));
+
+    while (IS_ITERATOR_VALID(&iterator))
+    {
+      MacDescriptor *descriptor = (MacDescriptor *) iterator.node->data;
+
+      if (descriptor->portID == portID && descriptor->type == IX_ETH_DB_FILTERING_RECORD)
+      {
+          recordCount++;
+
+          /* display entry */
+          printf(" %02X:%02X:%02X:%02X:%02X:%02X | %5d  | %s\n",
+              descriptor->macAddress[0],
+              descriptor->macAddress[1],
+              descriptor->macAddress[2],
+              descriptor->macAddress[3],
+              descriptor->macAddress[4],
+              descriptor->macAddress[5],
+              descriptor->recordData.filteringData.age,
+              descriptor->recordData.filteringData.staticEntry ? "static" : "dynamic");
+      }
+
+      /* move to the next record */
+      BUSY_RETRY_WITH_RESULT(ixEthDBIncrementHashIterator(&dbHashtable, &iterator), local_result);
+
+      /* debug */
+      if (local_result == IX_ETH_DB_BUSY)
+      {
+          return IX_ETH_DB_FAIL;
+      }
+    }
+
+    /* display number of records */
+    printf("\nFound %d records\n", recordCount);
+
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief displays all the filtering records belonging to all the ports
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @warning deprecated, use @ref ixEthDBFilteringDatabaseShowRecords() 
+ * instead. Calling this function is equivalent to calling
+ * ixEthDBFilteringDatabaseShowRecords(IX_ETH_DB_ALL_PORTS, IX_ETH_DB_FILTERING_RECORD)
+ */
+IX_ETH_DB_PUBLIC
+void ixEthDBFilteringDatabaseShowAll()
+{
+    IxEthDBPortId portIndex;
+
+    printf("\nEthernet learning/filtering database: listing %d ports\n\n", (UINT32) IX_ETH_DB_NUMBER_OF_PORTS);
+
+    for (portIndex = 0 ; portIndex < IX_ETH_DB_NUMBER_OF_PORTS ; portIndex++)
+    {
+        ixEthDBFilteringDatabaseShow(portIndex);
+
+        if (portIndex < IX_ETH_DB_NUMBER_OF_PORTS - 1)
+        {
+            printf("\n");
+        }
+    }
+}
+
+/**
+ * @brief displays one record in a format depending on the record filter
+ *
+ * @param descriptor pointer to the record
+ * @param recordFilter format filter
+ *
+ * This function will display the fields in a record depending on the
+ * selected record filter.
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBRecordShow(MacDescriptor *descriptor, IxEthDBRecordType recordFilter)
+{
+    if (recordFilter == IX_ETH_DB_FILTERING_VLAN_RECORD
+        || recordFilter == (IX_ETH_DB_FILTERING_RECORD | IX_ETH_DB_FILTERING_VLAN_RECORD))
+    {
+        /* display VLAN record header - leave this commented code in place, its purpose is to align the print format with the header
+        printf("    MAC address    |   Age  |   Type   | VLAN ID | CFI | QoS class \n");
+        printf("___________________________________________________________________\n"); */
+
+        if (descriptor->type == IX_ETH_DB_FILTERING_VLAN_RECORD)
+        {
+            printf("%02X:%02X:%02X:%02X:%02X:%02X | %3d | %s | %d | %d | %d\n",
+                descriptor->macAddress[0],
+                descriptor->macAddress[1],
+                descriptor->macAddress[2],
+                descriptor->macAddress[3],
+                descriptor->macAddress[4],
+                descriptor->macAddress[5],
+                descriptor->recordData.filteringVlanData.age,
+                descriptor->recordData.filteringVlanData.staticEntry ? "static" : "dynamic",
+                IX_ETH_DB_GET_VLAN_ID(descriptor->recordData.filteringVlanData.ieee802_1qTag),
+                (descriptor->recordData.filteringVlanData.ieee802_1qTag & 0x1000) >> 12,
+                IX_ETH_DB_GET_QOS_PRIORITY(descriptor->recordData.filteringVlanData.ieee802_1qTag));
+         }
+         else if (descriptor->type == IX_ETH_DB_FILTERING_RECORD)
+         {
+            printf("%02X:%02X:%02X:%02X:%02X:%02X | %3d | %s | - | - | -\n",
+                descriptor->macAddress[0],
+                descriptor->macAddress[1],
+                descriptor->macAddress[2],
+                descriptor->macAddress[3],
+                descriptor->macAddress[4],
+                descriptor->macAddress[5],
+                descriptor->recordData.filteringData.age,
+                descriptor->recordData.filteringData.staticEntry ? "static" : "dynamic");
+         }
+    }
+    else if (recordFilter == IX_ETH_DB_FILTERING_RECORD)
+    {
+        /* display filtering record header - leave this commented code in place, its purpose is to align the print format with the header
+        printf("    MAC address    |   Age  |   Type   \n");
+        printf("_______________________________________\n");  */
+
+        if (descriptor->type == IX_ETH_DB_FILTERING_RECORD)
+        {
+         printf("%02X:%02X:%02X:%02X:%02X:%02X | %3d | %s \n",
+             descriptor->macAddress[0],
+             descriptor->macAddress[1],
+             descriptor->macAddress[2],
+             descriptor->macAddress[3],
+             descriptor->macAddress[4],
+             descriptor->macAddress[5],
+             descriptor->recordData.filteringData.age,
+             descriptor->recordData.filteringData.staticEntry ? "static" : "dynamic");
+        }
+    }
+    else if (recordFilter == IX_ETH_DB_WIFI_RECORD)
+    {
+        /* display WiFi record header - leave this commented code in place, its purpose is to align the print format with the header 
+        printf("    MAC address    |   GW MAC address  \n");
+        printf("_______________________________________\n"); */
+
+        if (descriptor->type == IX_ETH_DB_WIFI_RECORD)
+        {
+            if (descriptor->recordData.wifiData.type == IX_ETH_DB_WIFI_AP_TO_AP)
+            {
+                /* gateway address present */
+                printf("%02X:%02X:%02X:%02X:%02X:%02X | %02X:%02X:%02X:%02X:%02X:%02X \n",
+                    descriptor->macAddress[0],
+                    descriptor->macAddress[1],
+                    descriptor->macAddress[2],
+                    descriptor->macAddress[3],
+                    descriptor->macAddress[4],
+                    descriptor->macAddress[5],
+                    descriptor->recordData.wifiData.gwMacAddress[0],
+                    descriptor->recordData.wifiData.gwMacAddress[1],
+                    descriptor->recordData.wifiData.gwMacAddress[2],
+                    descriptor->recordData.wifiData.gwMacAddress[3],
+                    descriptor->recordData.wifiData.gwMacAddress[4],
+                    descriptor->recordData.wifiData.gwMacAddress[5]);
+            }
+            else
+            {
+                /* no gateway */
+                printf("%02X:%02X:%02X:%02X:%02X:%02X | ----no gateway----- \n",
+                    descriptor->macAddress[0],
+                    descriptor->macAddress[1],
+                    descriptor->macAddress[2],
+                    descriptor->macAddress[3],
+                    descriptor->macAddress[4],
+                    descriptor->macAddress[5]);
+            }
+        }
+    }
+    else if (recordFilter == IX_ETH_DB_FIREWALL_RECORD)
+    {
+        /* display Firewall record header - leave this commented code in place, its purpose is to align the print format with the header 
+        printf("    MAC address   \n");
+        printf("__________________\n"); */
+
+        if (descriptor->type == IX_ETH_DB_FIREWALL_RECORD)
+        {
+            printf("%02X:%02X:%02X:%02X:%02X:%02X \n",
+                descriptor->macAddress[0],
+                descriptor->macAddress[1],
+                descriptor->macAddress[2],
+                descriptor->macAddress[3],
+                descriptor->macAddress[4],
+                descriptor->macAddress[5]);
+        }
+    }
+    else if (recordFilter == IX_ETH_DB_ALL_RECORD_TYPES)
+    {
+        /* display composite record header - leave this commented code in place, its purpose is to align the print format with the header 
+        printf("    MAC address   | Record | Age|  Type   | VLAN |CFI| QoS |  GW MAC address   \n");
+        printf("_______________________________________________________________________________\n"); */
+
+        if (descriptor->type == IX_ETH_DB_FILTERING_VLAN_RECORD)
+        {
+            printf("%02X:%02X:%02X:%02X:%02X:%02X |  VLAN  | %2d | %s | %4d | %1d |  %1d  | -----------------\n",
+                descriptor->macAddress[0],
+                descriptor->macAddress[1],
+                descriptor->macAddress[2],
+                descriptor->macAddress[3],
+                descriptor->macAddress[4],
+                descriptor->macAddress[5],
+                descriptor->recordData.filteringVlanData.age,
+                descriptor->recordData.filteringVlanData.staticEntry ? "static " : "dynamic",
+                IX_ETH_DB_GET_VLAN_ID(descriptor->recordData.filteringVlanData.ieee802_1qTag),
+                (descriptor->recordData.filteringVlanData.ieee802_1qTag & 0x1000) >> 12,
+                IX_ETH_DB_GET_QOS_PRIORITY(descriptor->recordData.filteringVlanData.ieee802_1qTag));
+         }
+         else if (descriptor->type == IX_ETH_DB_FILTERING_RECORD)
+         {
+            printf("%02X:%02X:%02X:%02X:%02X:%02X | Filter | %2d | %s | ---- | - | --- | -----------------\n",
+                descriptor->macAddress[0],
+                descriptor->macAddress[1],
+                descriptor->macAddress[2],
+                descriptor->macAddress[3],
+                descriptor->macAddress[4],
+                descriptor->macAddress[5],
+                descriptor->recordData.filteringData.age,
+                descriptor->recordData.filteringData.staticEntry ? "static " : "dynamic");
+         }
+        else if (descriptor->type == IX_ETH_DB_WIFI_RECORD)
+        {
+            if (descriptor->recordData.wifiData.type == IX_ETH_DB_WIFI_AP_TO_AP)
+            {
+                /* gateway address present */
+                printf("%02X:%02X:%02X:%02X:%02X:%02X |  WiFi  | -- | AP=>AP  | ---- | - | --- | %02X:%02X:%02X:%02X:%02X:%02X\n",
+                    descriptor->macAddress[0],
+                    descriptor->macAddress[1],
+                    descriptor->macAddress[2],
+                    descriptor->macAddress[3],
+                    descriptor->macAddress[4],
+                    descriptor->macAddress[5],
+                    descriptor->recordData.wifiData.gwMacAddress[0],
+                    descriptor->recordData.wifiData.gwMacAddress[1],
+                    descriptor->recordData.wifiData.gwMacAddress[2],
+                    descriptor->recordData.wifiData.gwMacAddress[3],
+                    descriptor->recordData.wifiData.gwMacAddress[4],
+                    descriptor->recordData.wifiData.gwMacAddress[5]);
+            }
+            else
+            {
+                /* no gateway */
+                printf("%02X:%02X:%02X:%02X:%02X:%02X |  WiFi  | -- | AP=>ST  | ---- | - | --- | -- no gateway -- \n",
+                    descriptor->macAddress[0],
+                    descriptor->macAddress[1],
+                    descriptor->macAddress[2],
+                    descriptor->macAddress[3],
+                    descriptor->macAddress[4],
+                    descriptor->macAddress[5]);
+            }
+        }
+        else if (descriptor->type == IX_ETH_DB_FIREWALL_RECORD)
+        {
+            printf("%02X:%02X:%02X:%02X:%02X:%02X |   FW   | -- | ------- | ---- | - | --- | -----------------\n",
+                descriptor->macAddress[0],
+                descriptor->macAddress[1],
+                descriptor->macAddress[2],
+                descriptor->macAddress[3],
+                descriptor->macAddress[4],
+                descriptor->macAddress[5]);
+        }
+    }
+    else
+    {
+        printf("invalid record filter\n");
+    }
+}
+
+/**
+ * @brief displays the status, records and configuration information of a port
+ *
+ * @param portID ID of the port
+ * @param recordFilter record filter to display
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBPortInfoShow(IxEthDBPortId portID, IxEthDBRecordType recordFilter)
+{
+    PortInfo *portInfo = &ixEthDBPortInfo[portID];
+    UINT32 recordCount = 0;
+    HashIterator iterator;
+    IxEthDBStatus local_result;
+
+    /* display port status */
+    printf("== Port ID %d ==\n", portID);
+
+    /* display capabilities */
+    printf("- Capabilities: ");
+
+    if ((portInfo->featureCapability & IX_ETH_DB_LEARNING) != 0)
+    {
+        printf("Learning (%s) ", ((portInfo->featureStatus & IX_ETH_DB_LEARNING) != 0) ? "on" : "off");
+    }
+
+    if ((portInfo->featureCapability & IX_ETH_DB_VLAN_QOS) != 0)
+    {
+        printf("VLAN/QoS (%s) ", ((portInfo->featureStatus & IX_ETH_DB_VLAN_QOS) != 0) ? "on" : "off");
+    }
+
+    if ((portInfo->featureCapability & IX_ETH_DB_FIREWALL) != 0)
+    {
+        printf("Firewall (%s) ", ((portInfo->featureStatus & IX_ETH_DB_FIREWALL) != 0) ? "on" : "off");
+    }
+
+    if ((portInfo->featureCapability & IX_ETH_DB_WIFI_HEADER_CONVERSION) != 0)
+    {
+        printf("WiFi (%s) ", ((portInfo->featureStatus & IX_ETH_DB_WIFI_HEADER_CONVERSION) != 0) ? "on" : "off");
+    }
+
+    if ((portInfo->featureCapability & IX_ETH_DB_SPANNING_TREE_PROTOCOL) != 0)
+    {
+        printf("STP (%s) ", ((portInfo->featureStatus & IX_ETH_DB_SPANNING_TREE_PROTOCOL) != 0) ? "on" : "off");
+    }
+
+    printf("\n");
+
+    /* dependency map */
+    ixEthDBDependencyPortMapShow(portID, portInfo->dependencyPortMap);
+
+    /* NPE dynamic updates */
+    if (ixEthDBPortDefinitions[portID].type == IX_ETH_NPE) 
+    {
+        printf(" - NPE dynamic update is %s\n", portInfo->updateMethod.updateEnabled ? "enabled" : "disabled");
+    }
+    else
+    {
+        printf(" - dynamic update disabled (not an NPE)\n");
+    }
+
+    if ((portInfo->featureCapability & IX_ETH_DB_WIFI_HEADER_CONVERSION) != 0)
+    {
+        if ((portInfo->featureStatus & IX_ETH_DB_WIFI_HEADER_CONVERSION) != 0)
+        {
+            /* WiFi header conversion */
+            if ((portInfo->frameControlDurationID 
+                + portInfo->bbsid[0] 
+                + portInfo->bbsid[1] 
+                + portInfo->bbsid[2] 
+                + portInfo->bbsid[3] 
+                + portInfo->bbsid[4] 
+                + portInfo->bbsid[5]) == 0)
+            {
+                printf(" - WiFi header conversion not configured\n");
+            }
+            else
+            {  
+                printf(" - WiFi header conversion: BBSID [%02X:%02X:%02X:%02X:%02X:%02X], Frame Control 0x%X, Duration/ID 0x%X\n", 
+                    portInfo->bbsid[0],
+                    portInfo->bbsid[1],
+                    portInfo->bbsid[2],
+                    portInfo->bbsid[3],
+                    portInfo->bbsid[4],
+                    portInfo->bbsid[5],
+                    portInfo->frameControlDurationID >> 16,
+                    portInfo->frameControlDurationID & 0xFFFF);
+            }
+        }
+        else
+        {
+            printf(" - WiFi header conversion not enabled\n");
+        }
+    }
+
+    /* Firewall */
+    if ((portInfo->featureCapability & IX_ETH_DB_FIREWALL) != 0)
+    {
+        if ((portInfo->featureStatus & IX_ETH_DB_FIREWALL) != 0)
+        {
+            printf(" - Firewall is in %s-list mode\n", portInfo->firewallMode == IX_ETH_DB_FIREWALL_BLACK_LIST ? "black" : "white");
+            printf(" - Invalid source MAC address filtering is %s\n", portInfo->srcAddressFilterEnabled ? "enabled" : "disabled");
+        }
+        else
+        {
+            printf(" - Firewall not enabled\n");
+        }
+    }
+  
+    /* browse database if asked to display records */
+    if (recordFilter != IX_ETH_DB_NO_RECORD_TYPE)
+    {
+        printf("\n");
+        ixEthDBHeaderShow(recordFilter);
+
+        BUSY_RETRY(ixEthDBInitHashIterator(&dbHashtable, &iterator));
+
+        while (IS_ITERATOR_VALID(&iterator))
+        {
+            MacDescriptor *descriptor = (MacDescriptor *) iterator.node->data;
+
+            if (descriptor->portID == portID && (descriptor->type & recordFilter) != 0)
+            {
+                recordCount++;
+
+                /* display entry */
+                ixEthDBRecordShow(descriptor, recordFilter);
+            }
+
+            /* move to the next record */
+            BUSY_RETRY_WITH_RESULT(ixEthDBIncrementHashIterator(&dbHashtable, &iterator), local_result);
+
+            /* debug */
+            if (local_result == IX_ETH_DB_BUSY)
+            {
+                printf("EthDB (API): Error, database browser failed (no access), giving up\n");
+            }
+        }
+        
+        printf("\nFound %d records\n\n", recordCount);
+    }
+}
+
+/**
+ * @brief displays a record header
+ *
+ * @param recordFilter record type filter
+ *
+ * This function displays a record header, depending on
+ * the given record type filter. It is useful when used
+ * in conjunction with ixEthDBRecordShow which will display
+ * record fields formatted for the header, provided the same
+ * record filter is used.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or IX_ETH_DB_INVALID_ARG if the recordFilter
+ * parameter is invalid or not supported
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+IxEthDBStatus ixEthDBHeaderShow(IxEthDBRecordType recordFilter)
+{
+  if (recordFilter == IX_ETH_DB_FILTERING_VLAN_RECORD
+      || recordFilter == (IX_ETH_DB_FILTERING_RECORD | IX_ETH_DB_FILTERING_VLAN_RECORD))
+  {
+    /* display VLAN record header */
+    printf("    MAC address    |   Age  |   Type   | VLAN ID | CFI | QoS class \n");
+    printf("___________________________________________________________________\n");
+  }
+  else if (recordFilter == IX_ETH_DB_FILTERING_RECORD)
+  {
+    /* display filtering record header */
+    printf("    MAC address    |   Age  |   Type   \n");
+    printf("_______________________________________\n");
+  }
+  else if (recordFilter == IX_ETH_DB_WIFI_RECORD)
+  {
+    /* display WiFi record header */
+    printf("    MAC address    |   GW MAC address  \n");
+    printf("_______________________________________\n");
+  }
+  else if (recordFilter == IX_ETH_DB_FIREWALL_RECORD)
+  {
+    /* display Firewall record header */
+    printf("    MAC address   \n");
+    printf("__________________\n");
+  }
+  else if (recordFilter == IX_ETH_DB_ALL_RECORD_TYPES)
+  {
+    /* display composite record header */
+    printf("    MAC address   | Record | Age|  Type   | VLAN |CFI| QoS |  GW MAC address   \n");
+    printf("_______________________________________________________________________________\n");
+  }
+  else
+  {
+    return IX_ETH_DB_INVALID_ARG;
+  }
+  
+  return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief displays database information (records and port information)
+ *
+ * @param portID ID of the port to display (or IX_ETH_DB_ALL_PORTS for all the ports)
+ * @param recordFilter record filter (use IX_ETH_DB_NO_RECORD_TYPE to display only
+ * port information)
+ * 
+ * Note that this function is documented in the main component header
+ * file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully or
+ * an appropriate error code otherwise
+ * 
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBFilteringDatabaseShowRecords(IxEthDBPortId portID, IxEthDBRecordType recordFilter)
+{
+  IxEthDBPortId currentPort;
+  BOOL showAllPorts = (portID == IX_ETH_DB_ALL_PORTS);
+  
+  IX_ETH_DB_CHECK_PORT_ALL(portID);
+
+  printf("\nEthernet learning/filtering database: listing %d port(s)\n\n", showAllPorts ? (UINT32) IX_ETH_DB_NUMBER_OF_PORTS : 1);
+  
+  currentPort = showAllPorts ? 0 : portID;
+  
+  while (currentPort != IX_ETH_DB_NUMBER_OF_PORTS)
+  {
+    /* display port info */
+    ixEthDBPortInfoShow(currentPort, recordFilter);
+    
+    /* next port */
+    currentPort = showAllPorts ? currentPort + 1 : IX_ETH_DB_NUMBER_OF_PORTS;
+  }
+  
+  return IX_ETH_DB_SUCCESS;
+}
+

cpu/ixp/npe/IxEthDBSearch.c

@@ -0,0 +1,327 @@
+/**
+ * @file IxEthDBSearch.c
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxEthDB_p.h"
+
+extern HashTable dbHashtable;
+
+/**
+ * @brief matches two database records based on their MAC addresses
+ *
+ * @param untypedReference record to match against
+ * @param untypedEntry record to match
+ *
+ * @return TRUE if the match is successful or FALSE otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+BOOL ixEthDBAddressRecordMatch(void *untypedReference, void *untypedEntry)
+{
+    MacDescriptor *entry     = (MacDescriptor *) untypedEntry;
+    MacDescriptor *reference = (MacDescriptor *) untypedReference;
+    
+    /* check accepted record types */
+    if ((entry->type & reference->type) == 0) return FALSE;
+       
+    return (ixEthDBAddressCompare((UINT8 *) entry->macAddress, (UINT8 *) reference->macAddress) == 0);
+}
+
+/**
+ * @brief matches two database records based on their MAC addresses
+ * and VLAN IDs
+ *
+ * @param untypedReference record to match against
+ * @param untypedEntry record to match
+ *
+ * @return TRUE if the match is successful or FALSE otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+BOOL ixEthDBVlanRecordMatch(void *untypedReference, void *untypedEntry)
+{
+    MacDescriptor *entry     = (MacDescriptor *) untypedEntry;
+    MacDescriptor *reference = (MacDescriptor *) untypedReference;
+    
+    /* check accepted record types */
+    if ((entry->type & reference->type) == 0) return FALSE;
+    
+    return (IX_ETH_DB_GET_VLAN_ID(entry->recordData.filteringVlanData.ieee802_1qTag) ==
+        IX_ETH_DB_GET_VLAN_ID(reference->recordData.filteringVlanData.ieee802_1qTag)) &&
+        (ixEthDBAddressCompare(entry->macAddress, reference->macAddress) == 0);
+}
+
+/**
+ * @brief matches two database records based on their MAC addresses
+ * and port IDs
+ *
+ * @param untypedReference record to match against
+ * @param untypedEntry record to match
+ *
+ * @return TRUE if the match is successful or FALSE otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+BOOL ixEthDBPortRecordMatch(void *untypedReference, void *untypedEntry)
+{
+    MacDescriptor *entry     = (MacDescriptor *) untypedEntry;
+    MacDescriptor *reference = (MacDescriptor *) untypedReference;
+    
+    /* check accepted record types */
+    if ((entry->type & reference->type) == 0) return FALSE;
+    
+    return (entry->portID == reference->portID) &&
+        (ixEthDBAddressCompare(entry->macAddress, reference->macAddress) == 0);
+}
+
+/**
+ * @brief dummy matching function, registered for safety
+ *
+ * @param reference record to match against (unused)
+ * @param entry record to match (unused)
+ *
+ * This function is registered in the matching functions
+ * array on invalid types. Calling it will display an 
+ * error message, indicating an error in the component logic.
+ *
+ * @return FALSE
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+BOOL ixEthDBNullMatch(void *reference, void *entry)
+{
+    /* display an error message */
+
+    ixOsalLog(IX_OSAL_LOG_LVL_WARNING, IX_OSAL_LOG_DEV_STDOUT, "DB: (Search) The NullMatch function was called, wrong key type?\n", 0, 0, 0, 0, 0, 0);
+
+
+    return FALSE;
+}
+
+/**
+ * @brief registers hash matching methods
+ *
+ * @param matchFunctions table of match functions to be populated
+ *
+ * This function registers the available record matching functions
+ * by indexing them on record types into the given function array.
+ * 
+ * Note that it is compulsory to call this in ixEthDBInit(), 
+ * otherwise hashtable searching and removal will not work
+ *
+ * @return number of registered functions
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+UINT32 ixEthDBMatchMethodsRegister(MatchFunction *matchFunctions)
+{
+    UINT32 i;
+    
+    /* safety first */
+    for ( i = 0 ; i < IX_ETH_DB_MAX_KEY_INDEX + 1 ; i++)
+    {
+        matchFunctions[i] = ixEthDBNullMatch;
+    }
+    
+    /* register MAC search method */
+    matchFunctions[IX_ETH_DB_MAC_KEY] = ixEthDBAddressRecordMatch;
+    
+    /* register MAC/PortID search method */
+    matchFunctions[IX_ETH_DB_MAC_PORT_KEY] = ixEthDBPortRecordMatch;
+    
+    /* register MAC/VLAN ID search method */
+    matchFunctions[IX_ETH_DB_MAC_VLAN_KEY] = ixEthDBVlanRecordMatch;
+    
+    return 3; /* three methods */
+}
+
+/**
+ * @brief search a record in the Ethernet datbase
+ *
+ * @param macAddress MAC address to perform the search on
+ * @param typeFilter type of records to consider for matching
+ *
+ * @warning if searching is successful an implicit write lock
+ * to the search result is granted, therefore unlock the 
+ * entry using @ref ixEthDBReleaseHashNode() as soon as possible.
+ *
+ * @see ixEthDBReleaseHashNode()
+ *
+ * @return the search result, or NULL if a record with the given
+ * MAC address was not found
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+HashNode* ixEthDBSearch(IxEthDBMacAddr *macAddress, IxEthDBRecordType typeFilter)
+{
+    HashNode *searchResult = NULL;
+    MacDescriptor reference;
+    
+    TEST_FIXTURE_INCREMENT_DB_CORE_ACCESS_COUNTER;
+
+    if (macAddress == NULL)
+    {
+        return NULL;
+    }
+
+    /* fill search fields */
+    memcpy(reference.macAddress, macAddress, sizeof (IxEthDBMacAddr));
+    
+    /* set acceptable record types */
+    reference.type = typeFilter;
+    
+    BUSY_RETRY(ixEthDBSearchHashEntry(&dbHashtable, IX_ETH_DB_MAC_KEY, &reference, &searchResult));
+
+    return searchResult;
+}
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBPeek(IxEthDBMacAddr *macAddress, IxEthDBRecordType typeFilter)
+{
+    MacDescriptor reference;
+    IxEthDBStatus result;
+    
+    TEST_FIXTURE_INCREMENT_DB_CORE_ACCESS_COUNTER;
+
+    if (macAddress == NULL)
+    {
+        return IX_ETH_DB_INVALID_ARG;
+    }
+
+    /* fill search fields */
+    memcpy(reference.macAddress, macAddress, sizeof (IxEthDBMacAddr));
+    
+    /* set acceptable record types */
+    reference.type = typeFilter;
+    
+    result = ixEthDBPeekHashEntry(&dbHashtable, IX_ETH_DB_MAC_KEY, &reference);
+
+    return result;
+}
+
+/**
+ * @brief search a record in the Ethernet datbase
+ *
+ * @param macAddress MAC address to perform the search on
+ * @param portID port ID to perform the search on
+ * @param typeFilter type of records to consider for matching
+ *
+ * @warning if searching is successful an implicit write lock
+ * to the search result is granted, therefore unlock the 
+ * entry using @ref ixEthDBReleaseHashNode() as soon as possible.
+ *
+ * @see ixEthDBReleaseHashNode()
+ *
+ * @return the search result, or NULL if a record with the given
+ * MAC address/port ID combination was not found
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+HashNode* ixEthDBPortSearch(IxEthDBMacAddr *macAddress, IxEthDBPortId portID, IxEthDBRecordType typeFilter)
+{
+    HashNode *searchResult = NULL;
+    MacDescriptor reference;
+    
+    if (macAddress == NULL)
+    {
+        return NULL;
+    }
+    
+    /* fill search fields */
+    memcpy(reference.macAddress, macAddress, sizeof (IxEthDBMacAddr));
+    reference.portID = portID;
+    
+    /* set acceptable record types */
+    reference.type = typeFilter;
+
+    BUSY_RETRY(ixEthDBSearchHashEntry(&dbHashtable, IX_ETH_DB_MAC_PORT_KEY, &reference, &searchResult));
+
+    return searchResult;
+}
+
+/**
+ * @brief search a record in the Ethernet datbase
+ *
+ * @param macAddress MAC address to perform the search on
+ * @param vlanID VLAN ID to perform the search on
+ * @param typeFilter type of records to consider for matching
+ *
+ * @warning if searching is successful an implicit write lock
+ * to the search result is granted, therefore unlock the 
+ * entry using @ref ixEthDBReleaseHashNode() as soon as possible.
+ *
+ * @see ixEthDBReleaseHashNode()
+ *
+ * @return the search result, or NULL if a record with the given
+ * MAC address/VLAN ID combination was not found
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+HashNode* ixEthDBVlanSearch(IxEthDBMacAddr *macAddress, IxEthDBVlanId vlanID, IxEthDBRecordType typeFilter)
+{
+    HashNode *searchResult = NULL;
+    MacDescriptor reference;
+    
+    if (macAddress == NULL)
+    {
+        return NULL;
+    }
+    
+    /* fill search fields */
+    memcpy(reference.macAddress, macAddress, sizeof (IxEthDBMacAddr));
+    reference.recordData.filteringVlanData.ieee802_1qTag = 
+            IX_ETH_DB_SET_VLAN_ID(reference.recordData.filteringVlanData.ieee802_1qTag, vlanID);
+    
+    /* set acceptable record types */
+    reference.type = typeFilter;
+
+    BUSY_RETRY(ixEthDBSearchHashEntry(&dbHashtable, IX_ETH_DB_MAC_VLAN_KEY, &reference, &searchResult));
+
+    return searchResult;
+}

cpu/ixp/npe/IxEthDBSpanningTree.c

@@ -0,0 +1,107 @@
+/**
+ * @file IxEthDBSpanningTree.c
+ *
+ * @brief Implementation of the STP API
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+
+#include "IxEthDB_p.h"
+
+/**
+ * @brief sets the STP blocking state of a port
+ *
+ * @param portID ID of the port
+ * @param blocked TRUE to block the port or FALSE to unblock it
+ * 
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBSpanningTreeBlockingStateSet(IxEthDBPortId portID, BOOL blocked)
+{
+    IxNpeMhMessage message;
+    IX_STATUS result;
+    
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+ 
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_SPANNING_TREE_PROTOCOL);
+    
+    ixEthDBPortInfo[portID].stpBlocked = blocked;
+
+    FILL_SETBLOCKINGSTATE_MSG(message, portID, blocked);
+    
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+    
+    return result;
+}
+
+/**
+ * @brief retrieves the STP blocking state of a port
+ *
+ * @param portID ID of the port
+ * @param blocked address to write the blocked status into
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBSpanningTreeBlockingStateGet(IxEthDBPortId portID, BOOL *blocked)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+ 
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_SPANNING_TREE_PROTOCOL);
+    
+    IX_ETH_DB_CHECK_REFERENCE(blocked);
+    
+    *blocked = ixEthDBPortInfo[portID].stpBlocked;
+    
+    return IX_ETH_DB_SUCCESS;
+}

cpu/ixp/npe/IxEthDBUtil.c

@@ -0,0 +1,120 @@
+/**
+ * @file ethUtil.c
+ *
+ * @brief Utility functions
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+
+#include "IxFeatureCtrl.h"
+#include "IxEthDB_p.h"
+
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBSingleEthNpeCheck(IxEthDBPortId portID)
+{
+    /* If not IXP42X A0 stepping, proceed to check for existence of coprocessors */ 
+    if ((IX_FEATURE_CTRL_SILICON_TYPE_A0 != 
+        (ixFeatureCtrlProductIdRead() & IX_FEATURE_CTRL_SILICON_STEPPING_MASK))
+        || (IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X != ixFeatureCtrlDeviceRead ()))
+    {  
+        if ((portID == 0) && 
+            (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_ETH0) ==
+             IX_FEATURE_CTRL_COMPONENT_DISABLED))
+        {
+	    return IX_ETH_DB_FAIL;
+        }
+
+        if ((portID == 1) &&
+            (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_ETH1) ==
+             IX_FEATURE_CTRL_COMPONENT_DISABLED))
+        {
+	    return IX_ETH_DB_FAIL;	    
+        }
+
+        if ((portID == 2) &&
+            (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEA_ETH) ==
+             IX_FEATURE_CTRL_COMPONENT_DISABLED))
+        {
+	    return IX_ETH_DB_FAIL;	    
+        }
+    }
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+IX_ETH_DB_PUBLIC
+BOOL ixEthDBCheckSingleBitValue(UINT32 value)
+{
+#if (CPU != SIMSPARCSOLARIS) && !defined (__wince)
+    UINT32 shift;
+    
+    /* use the count-leading-zeros XScale instruction */
+    __asm__ ("clz %0, %1\n" : "=r" (shift) : "r" (value));
+    
+    return ((value << shift) == 0x80000000UL);
+    
+#else
+	
+    while (value != 0)
+    {
+        if (value == 1) return TRUE;
+        else if ((value & 1) == 1) return FALSE;
+
+        value >>= 1;
+    }
+    
+    return FALSE;
+
+#endif
+}
+
+const char *mac2string(const unsigned char *mac)
+{
+  static char str[19];
+  
+  if (mac == NULL)
+  {
+    return NULL;
+  }  
+
+  sprintf(str, "%02X:%02X:%02X:%02X:%02X:%02X", mac[0], mac[1], mac[2], mac[3], mac[4], mac[5]);
+
+  return str;
+}

cpu/ixp/npe/IxEthDBVlan.c

@@ -0,0 +1,1179 @@
+/**
+ * @file IxEthDBVlan.c
+ *
+ * @brief Implementation of the VLAN API
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxEthDB.h"
+#include "IxEthDB_p.h"
+
+/* forward prototypes */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBUpdateTrafficClass(IxEthDBPortId portID, UINT32 classIndex);
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBVlanTableGet(IxEthDBPortId portID, IxEthDBVlanSet portVlanTable, IxEthDBVlanSet vlanSet);
+
+/* contants used by various functions as "action" parameter */
+#define ADD_VLAN    (0x1)
+#define REMOVE_VLAN (0x2)
+
+/**
+ * @brief adds or removes a VLAN from a VLAN set
+ *
+ * @param vlanID VLAN ID to add or remove
+ * @param table VLAN set to add into or remove from
+ * @param action ADD_VLAN or REMOVE_VLAN
+ * 
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+void ixEthDBLocalVlanMembershipChange(UINT32 vlanID, IxEthDBVlanSet table, UINT32 action)
+{
+    UINT32 setOffset;
+    
+    /* add/remove VID to membership table */
+    setOffset = VLAN_SET_OFFSET(vlanID); /* we need 9 bits to index the 512 byte membership array */
+
+    if (action == ADD_VLAN)
+    {
+        table[setOffset] |= 1 << VLAN_SET_MASK(vlanID);
+    }
+    else if (action == REMOVE_VLAN)
+    {
+        table[setOffset] &= ~(1 << VLAN_SET_MASK(vlanID));
+    }
+}
+
+/**
+ * @brief updates a set of 8 VLANs in an NPE
+ *
+ * @param portID ID of the port
+ * @param setOffset offset of the 8 VLANs
+ *
+ * This function updates the VLAN membership table
+ * and Transmit Tagging Info table for 8 consecutive
+ * VLAN IDs indexed by setOffset.
+ *
+ * For example, a setOffset of 0 indexes VLAN IDs 0
+ * through 7, 1 indexes VLAN IDs 8 through 9 etc.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+IxEthDBStatus ixEthDBVlanTableEntryUpdate(IxEthDBPortId portID, UINT32 setOffset)
+{
+    PortInfo *portInfo = &ixEthDBPortInfo[portID];
+    IxNpeMhMessage message;
+    IX_STATUS result;
+        
+    FILL_SETPORTVLANTABLEENTRY_MSG(message, IX_ETH_DB_PORT_ID_TO_NPE_LOGICAL_ID(portID), 
+        2 * setOffset, 
+        portInfo->vlanMembership[setOffset], 
+        portInfo->transmitTaggingInfo[setOffset]);
+    
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+    
+    return result;
+}
+
+/**
+ * @brief updates a VLAN range in an NPE
+ *
+ * @param portID ID of the port
+ *
+ * This function is similar to @ref ixEthDBVlanTableEntryUpdate
+ * except that it can update more than one VLAN set (up to
+ * the entire VLAN membership and TTI tables if the offset is 0
+ * and length is sizeof (IxEthDBVlanSet) (512 bytes).
+ *
+ * Updating the NPE via this method is slower as it requires
+ * a memory copy from SDRAM, hence it is recommended that the
+ * ixEthDBVlanTableEntryUpdate function is used where possible.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+IxEthDBStatus ixEthDBVlanTableRangeUpdate(IxEthDBPortId portID)
+{
+    PortInfo *portInfo    = &ixEthDBPortInfo[portID];
+    UINT8 *vlanUpdateZone = (UINT8 *) portInfo->updateMethod.vlanUpdateZone;
+    IxNpeMhMessage message;
+    UINT32 setIndex;
+    IX_STATUS result;
+
+    /* copy membership info and transmit tagging into into exchange area */
+    for (setIndex = 0 ; setIndex < sizeof (portInfo->vlanMembership) ; setIndex++)
+    {
+        /* membership and TTI data are interleaved */
+        vlanUpdateZone[setIndex * 2]     = portInfo->vlanMembership[setIndex];
+        vlanUpdateZone[setIndex * 2 + 1] = portInfo->transmitTaggingInfo[setIndex];
+    }
+
+    IX_OSAL_CACHE_FLUSH(vlanUpdateZone, FULL_VLAN_BYTE_SIZE);
+    
+    /* build NPE message */
+    FILL_SETPORTVLANTABLERANGE_MSG(message, IX_ETH_DB_PORT_ID_TO_NPE_LOGICAL_ID(portID), 0, 0, 
+        IX_OSAL_MMU_VIRT_TO_PHYS(vlanUpdateZone));
+
+    /* send message */    
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+
+    return result;
+}
+
+/**
+ * @brief adds or removes a VLAN from a port's VLAN membership table
+ * or Transmit Tagging Information table
+ *
+ * @param portID ID of the port
+ * @param vlanID VLAN ID to add or remove
+ * @param table to add or remove from
+ * @param action ADD_VLAN or REMOVE_VLAN
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+IxEthDBStatus ixEthDBPortVlanMembershipChange(IxEthDBPortId portID, IxEthDBVlanId vlanID, IxEthDBVlanSet table, UINT32 action)
+{
+    /* change VLAN in local membership table */
+    ixEthDBLocalVlanMembershipChange(vlanID, table, action);
+    
+    /* send updated entry to NPE */
+    return ixEthDBVlanTableEntryUpdate(portID, VLAN_SET_OFFSET(vlanID));
+}
+
+/**
+ * @brief sets the default port VLAN tag (the lower 3 bytes are the PVID)
+ *
+ * @param portID ID of the port
+ * @param vlanTag port VLAN tag (802.1Q tag)
+ * 
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPortVlanTagSet(IxEthDBPortId portID, IxEthDBVlanTag vlanTag)
+{
+    IxNpeMhMessage message;
+    IX_STATUS result;
+    
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+ 
+    IX_ETH_DB_CHECK_VLAN_TAG(vlanTag);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+        
+    /* add VLAN ID to local membership table */
+    ixEthDBPortVlanMembershipChange(portID, 
+        vlanTag & IX_ETH_DB_802_1Q_VLAN_MASK, 
+        ixEthDBPortInfo[portID].vlanMembership, 
+        ADD_VLAN);
+        
+    /* set tag in portInfo */
+    ixEthDBPortInfo[portID].vlanTag = vlanTag;
+    
+    /* build VLAN_SetDefaultRxVID message */
+    FILL_SETDEFAULTRXVID_MSG(message, 
+        IX_ETH_DB_PORT_ID_TO_NPE_LOGICAL_ID(portID), 
+        IX_IEEE802_1Q_VLAN_TPID, 
+        vlanTag);
+    
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+    
+    return result;
+}
+
+/**
+ * @brief retrieves the default port VLAN tag (the lower 3 bytes are the PVID)
+ *
+ * @param portID ID of the port
+ * @param vlanTag address to write the port VLAN tag (802.1Q tag) into
+ * 
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPortVlanTagGet(IxEthDBPortId portID, IxEthDBVlanTag *vlanTag)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+    
+    IX_ETH_DB_CHECK_REFERENCE(vlanTag);
+    
+    *vlanTag = ixEthDBPortInfo[portID].vlanTag;
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief sets the VLAN tag (the lower 3 bytes are the PVID) of a 
+ * database filtering record
+ *
+ * @param portID ID of the port
+ * @param vlanTag VLAN tag (802.1Q tag)
+ * 
+ * Important: filtering records are automatically converted to 
+ * IX_ETH_DB_FILTERING_VLAN record when added a VLAN tag.
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBVlanTagSet(IxEthDBMacAddr *macAddr, IxEthDBVlanTag vlanTag)
+{
+    HashNode *searchResult;
+    MacDescriptor *descriptor;
+    
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+    
+    IX_ETH_DB_CHECK_VLAN_TAG(vlanTag);
+    
+    searchResult = ixEthDBSearch(macAddr, IX_ETH_DB_ALL_FILTERING_RECORDS);
+    
+    if (searchResult == NULL)
+    {
+        return IX_ETH_DB_NO_SUCH_ADDR;
+    }
+    
+    descriptor = (MacDescriptor *) searchResult->data;
+    
+    /* set record type to VLAN if not already set */
+    descriptor->type = IX_ETH_DB_FILTERING_VLAN_RECORD;
+    
+    /* add vlan tag */
+    descriptor->recordData.filteringVlanData.ieee802_1qTag = vlanTag;
+    
+    /* transaction completed */
+    ixEthDBReleaseHashNode(searchResult);
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief retrieves the VLAN tag (the lower 3 bytes are the PVID) from a 
+ * database VLAN filtering record
+ *
+ * @param portID ID of the port
+ * @param vlanTag address to write the VLAN tag (802.1Q tag) into
+ * 
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBVlanTagGet(IxEthDBMacAddr *macAddr, IxEthDBVlanTag *vlanTag)
+{
+    HashNode *searchResult;
+    MacDescriptor *descriptor;
+    
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+    
+    IX_ETH_DB_CHECK_REFERENCE(vlanTag);
+    
+    searchResult = ixEthDBSearch(macAddr, IX_ETH_DB_FILTERING_VLAN_RECORD);
+    
+    if (searchResult == NULL)
+    {
+        return IX_ETH_DB_NO_SUCH_ADDR;
+    }
+    
+    descriptor = (MacDescriptor *) searchResult->data;
+        
+    /* get vlan tag */
+    *vlanTag = descriptor->recordData.filteringVlanData.ieee802_1qTag;
+    
+    /* transaction completed */
+    ixEthDBReleaseHashNode(searchResult);
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief adds a VLAN to a port's VLAN membership table
+ *
+ * @param portID ID of the port
+ * @param vlanID VLAN ID to add
+ * 
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPortVlanMembershipAdd(IxEthDBPortId portID, IxEthDBVlanId vlanID)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    IX_ETH_DB_CHECK_VLAN_ID(vlanID);
+
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+
+    return ixEthDBPortVlanMembershipChange(portID, vlanID, ixEthDBPortInfo[portID].vlanMembership, ADD_VLAN);
+}
+
+/**
+ * @brief removes a VLAN from a port's VLAN membership table
+ *
+ * @param portID ID of the port
+ * @param vlanID VLAN ID to remove
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPortVlanMembershipRemove(IxEthDBPortId portID, IxEthDBVlanId vlanID)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+
+    IX_ETH_DB_CHECK_VLAN_ID(vlanID);
+
+    /* for safety isolate only the VLAN ID in the tag (the lower 12 bits) */
+    vlanID = vlanID & IX_ETH_DB_802_1Q_VLAN_MASK;
+    
+    /* check we're not asked to remove the default port VID */
+    if (vlanID == IX_ETH_DB_GET_VLAN_ID(ixEthDBPortInfo[portID].vlanTag))
+    {
+        return IX_ETH_DB_NO_PERMISSION;
+    }
+    
+    return ixEthDBPortVlanMembershipChange(portID, vlanID, ixEthDBPortInfo[portID].vlanMembership, REMOVE_VLAN);
+}
+
+/**
+ * @brief adds or removes a VLAN range from a port's 
+ * VLAN membership table or TTI table
+ *
+ * @param portID ID of the port
+ * @param vlanIDMin start of the VLAN range
+ * @param vlanIDMax end of the VLAN range
+ * @param table VLAN set to add or remove from
+ * @param action ADD_VLAN or REMOVE_VLAN
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+IxEthDBStatus ixEthDBPortVlanMembershipRangeChange(IxEthDBPortId portID, IxEthDBVlanId vlanIDMin, IxEthDBVlanId vlanIDMax, IxEthDBVlanSet table, UINT32 action)
+{
+    UINT32 setOffsetMin, setOffsetMax;
+    
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+    
+    IX_ETH_DB_CHECK_VLAN_ID(vlanIDMin);
+    
+    IX_ETH_DB_CHECK_VLAN_ID(vlanIDMax);
+    
+    /* for safety isolate only the VLAN ID in the tags (the lower 12 bits) */
+    vlanIDMin = vlanIDMin & IX_ETH_DB_802_1Q_VLAN_MASK;
+    vlanIDMax = vlanIDMax & IX_ETH_DB_802_1Q_VLAN_MASK;
+    
+    /* is this a range? */
+    if (vlanIDMax < vlanIDMin)
+    {
+        return IX_ETH_DB_INVALID_VLAN;
+    }
+    
+    /* check that we're not specifically asked to remove the default port VID */
+    if (action == REMOVE_VLAN && vlanIDMax == vlanIDMin && IX_ETH_DB_GET_VLAN_ID(ixEthDBPortInfo[portID].vlanTag) == vlanIDMin)
+    {
+        return IX_ETH_DB_NO_PERMISSION;
+    }
+    
+    /* compute set offsets */
+    setOffsetMin = VLAN_SET_OFFSET(vlanIDMin);
+    setOffsetMax = VLAN_SET_OFFSET(vlanIDMax);
+
+    /* change VLAN range */
+    for (; vlanIDMin <= vlanIDMax ; vlanIDMin++)
+    {
+        /* change vlan in local membership table */
+        ixEthDBLocalVlanMembershipChange(vlanIDMin, table, action);
+    }
+
+    /* if the range is within one set (max 8 VLANs in one table byte) we can just update that entry in the NPE */
+    if (setOffsetMin == setOffsetMax)
+    {
+        /* send updated entry to NPE */
+        return ixEthDBVlanTableEntryUpdate(portID, setOffsetMin);
+    }
+    else
+    {
+        /* update a zone of the membership/transmit tag info table */
+        return ixEthDBVlanTableRangeUpdate(portID);
+    }
+}
+
+/**
+ * @brief adds a VLAN range to a port's VLAN membership table
+ *
+ * @param portID ID of the port
+ * @param vlanIDMin start of the VLAN range
+ * @param vlanIDMax end of the VLAN range
+ * 
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPortVlanMembershipRangeAdd(IxEthDBPortId portID, IxEthDBVlanId vlanIDMin, IxEthDBVlanId vlanIDMax)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    return ixEthDBPortVlanMembershipRangeChange(portID, vlanIDMin, vlanIDMax, ixEthDBPortInfo[portID].vlanMembership, ADD_VLAN);
+}
+
+/**
+ * @brief removes a VLAN range from a port's VLAN membership table
+ *
+ * @param portID ID of the port
+ * @param vlanIDMin start of the VLAN range
+ * @param vlanIDMax end of the VLAN range
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPortVlanMembershipRangeRemove(IxEthDBPortId portID, IxEthDBVlanId vlanIDMin, IxEthDBVlanId vlanIDMax)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    return ixEthDBPortVlanMembershipRangeChange(portID, vlanIDMin, vlanIDMax, ixEthDBPortInfo[portID].vlanMembership, REMOVE_VLAN);
+}
+
+/**
+ * @brief sets a port's VLAN membership table or TTI table and
+ * updates the NPE VLAN configuration
+ *
+ * @param portID ID of the port
+ * @param portVlanTable port VLAN table to set
+ * @param vlanSet new set contents
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBPortVlanTableSet(IxEthDBPortId portID, IxEthDBVlanSet portVlanTable, IxEthDBVlanSet vlanSet)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+    
+    IX_ETH_DB_CHECK_REFERENCE(vlanSet);
+
+    memcpy(portVlanTable, vlanSet, sizeof (IxEthDBVlanSet));
+    
+    return ixEthDBVlanTableRangeUpdate(portID);
+}
+
+/**
+ * @brief retireves a port's VLAN membership table or TTI table
+ *
+ * @param portID ID of the port
+ * @param portVlanTable port VLAN table to retrieve
+ * @param vlanSet address to 
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBVlanTableGet(IxEthDBPortId portID, IxEthDBVlanSet portVlanTable, IxEthDBVlanSet vlanSet)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+    
+    IX_ETH_DB_CHECK_REFERENCE(vlanSet);
+    
+    memcpy(vlanSet, portVlanTable, sizeof (IxEthDBVlanSet));
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief sets a port's VLAN membership table
+ *
+ * @param portID ID of the port
+ * @param vlanSet new VLAN membership table
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPortVlanMembershipSet(IxEthDBPortId portID, IxEthDBVlanSet vlanSet)
+{
+    IxEthDBVlanId vlanID;
+
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    IX_ETH_DB_CHECK_REFERENCE(vlanSet);
+
+    /* set the bit corresponding to the PVID just in case */
+    vlanID = IX_ETH_DB_GET_VLAN_ID(ixEthDBPortInfo[portID].vlanTag);
+    vlanSet[VLAN_SET_OFFSET(vlanID)] |= 1 << VLAN_SET_MASK(vlanID);
+    
+    return ixEthDBPortVlanTableSet(portID, ixEthDBPortInfo[portID].vlanMembership, vlanSet);
+}
+
+/**
+ * @brief retrieves a port's VLAN membership table
+ *
+ * @param portID ID of the port
+ * @param vlanSet location to store the port's VLAN membership table
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPortVlanMembershipGet(IxEthDBPortId portID, IxEthDBVlanSet vlanSet)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    return ixEthDBVlanTableGet(portID, ixEthDBPortInfo[portID].vlanMembership, vlanSet);
+}
+
+/**
+ * @brief enables or disables Egress tagging for one VLAN ID
+ *
+ * @param portID ID of the port
+ * @param vlanID VLAN ID to enable or disable Egress tagging on
+ * @param enabled TRUE to enable and FALSE to disable tagging
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBEgressVlanEntryTaggingEnabledSet(IxEthDBPortId portID, IxEthDBVlanId vlanID, BOOL enabled)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    IX_ETH_DB_CHECK_VLAN_ID(vlanID);
+
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+
+    return ixEthDBPortVlanMembershipChange(portID, vlanID, ixEthDBPortInfo[portID].transmitTaggingInfo, enabled? ADD_VLAN : REMOVE_VLAN);
+}
+
+/**
+ * @brief retrieves the Egress tagging status for one VLAN ID
+ *
+ * @param portID ID of the port
+ * @param vlanID VLAN ID to retrieve the tagging status for
+ * @param enabled location to store the tagging status
+ * (TRUE - tagging enabled, FALSE - tagging disabled)
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBEgressVlanEntryTaggingEnabledGet(IxEthDBPortId portID, IxEthDBVlanId vlanID, BOOL *enabled)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+    
+    IX_ETH_DB_CHECK_REFERENCE(enabled);
+
+    IX_ETH_DB_CHECK_VLAN_ID(vlanID);
+    
+    *enabled = ((ixEthDBPortInfo[portID].transmitTaggingInfo[VLAN_SET_OFFSET(vlanID)] & (1 << VLAN_SET_MASK(vlanID))) != 0);
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief enables or disables Egress VLAN tagging for a VLAN range
+ *
+ * @param portID ID of the port
+ * @param vlanIDMin start of VLAN range
+ * @param vlanIDMax end of VLAN range
+ * @param enabled TRUE to enable or FALSE to disable VLAN tagging
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBEgressVlanRangeTaggingEnabledSet(IxEthDBPortId portID, IxEthDBVlanId vlanIDMin, IxEthDBVlanId vlanIDMax, BOOL enabled)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    return ixEthDBPortVlanMembershipRangeChange(portID, vlanIDMin, vlanIDMax, ixEthDBPortInfo[portID].transmitTaggingInfo, enabled? ADD_VLAN : REMOVE_VLAN);
+}
+
+/**
+ * @brief sets the Egress VLAN tagging table (the Transmit Tagging
+ * Information table)
+ *
+ * @param portID ID of the port
+ * @param vlanSet new TTI table
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBEgressVlanTaggingEnabledSet(IxEthDBPortId portID, IxEthDBVlanSet vlanSet)
+{
+    IxEthDBVlanId vlanID;
+
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    IX_ETH_DB_CHECK_REFERENCE(vlanSet);
+
+    /* set the PVID bit just in case */
+    vlanID = IX_ETH_DB_GET_VLAN_ID(ixEthDBPortInfo[portID].vlanTag);
+    vlanSet[VLAN_SET_OFFSET(vlanID)] |= 1 << VLAN_SET_MASK(vlanID);
+    
+    return ixEthDBPortVlanTableSet(portID, ixEthDBPortInfo[portID].transmitTaggingInfo, vlanSet);
+}
+
+/**
+ * @brief retrieves the Egress VLAN tagging table (the Transmit 
+ * Tagging Information table)
+ *
+ * @param portID ID of the port
+ * @param vlanSet location to store the port's TTI table
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBEgressVlanTaggingEnabledGet(IxEthDBPortId portID, IxEthDBVlanSet vlanSet)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    return ixEthDBVlanTableGet(portID, ixEthDBPortInfo[portID].transmitTaggingInfo, vlanSet);
+}
+
+/**
+ * @brief sends the NPE the updated frame filter and default
+ * Ingress tagging
+ *
+ * @param portID ID of the port
+ * 
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+IxEthDBStatus ixEthDBIngressVlanModeUpdate(IxEthDBPortId portID)
+{
+    PortInfo *portInfo = &ixEthDBPortInfo[portID];
+    IxNpeMhMessage message;
+    IX_STATUS result;
+
+    FILL_SETRXTAGMODE_MSG(message, portID, portInfo->npeFrameFilter, portInfo->npeTaggingAction);
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+
+    return result;
+}
+
+/**
+ * @brief sets the default Ingress tagging behavior
+ *
+ * @param portID ID of the port
+ * @param taggingAction default tagging behavior
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBIngressVlanTaggingEnabledSet(IxEthDBPortId portID, IxEthDBTaggingAction taggingAction)
+{
+    PortInfo *portInfo;
+    
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+    
+    portInfo = &ixEthDBPortInfo[portID];
+    
+    if (taggingAction == IX_ETH_DB_PASS_THROUGH)
+    {
+        portInfo->npeTaggingAction = 0x00;
+    }
+    else if (taggingAction == IX_ETH_DB_ADD_TAG)
+    {
+        portInfo->npeTaggingAction = 0x02;
+    }
+    else if (taggingAction == IX_ETH_DB_REMOVE_TAG)
+    {
+        portInfo->npeTaggingAction = 0x01;
+    }
+    else
+    {
+        return IX_ETH_DB_INVALID_ARG;
+    }
+    
+    portInfo->taggingAction = taggingAction;
+    
+    return ixEthDBIngressVlanModeUpdate(portID);
+}
+
+/**
+ * @brief retrieves the default Ingress tagging behavior of a port
+ *
+ * @param portID ID of the port
+ * @param taggingAction location to save the default tagging behavior
+ * 
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBIngressVlanTaggingEnabledGet(IxEthDBPortId portID, IxEthDBTaggingAction *taggingAction)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+
+    IX_ETH_DB_CHECK_REFERENCE(taggingAction);
+    
+    *taggingAction = ixEthDBPortInfo[portID].taggingAction;
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief sets the Ingress acceptable frame type filter
+ *
+ * @param portID ID of the port
+ * @param frameFilter acceptable frame type filter
+ * 
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBAcceptableFrameTypeSet(IxEthDBPortId portID, IxEthDBFrameFilter frameFilter)
+{
+    PortInfo *portInfo;
+    IxEthDBStatus result = IX_ETH_DB_SUCCESS;
+        
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+    
+    /* check parameter range 
+       the ORed value of the valid values is 0x7
+       a value having extra bits is invalid */
+    if ((frameFilter | 0x7) != 0x7 || frameFilter == 0)
+    {
+        return IX_ETH_DB_INVALID_ARG;
+    }
+    
+    portInfo = &ixEthDBPortInfo[portID];
+    
+    portInfo->frameFilter    = frameFilter;
+    portInfo->npeFrameFilter = 0; /* allow all by default */
+    
+    /* if accepting priority tagged but not all VLAN tagged
+       set the membership table to contain only VLAN ID 0 
+       hence remove vlans 1-4094 and add VLAN ID 0 */
+    if (((frameFilter & IX_ETH_DB_PRIORITY_TAGGED_FRAMES) != 0)
+        && ((frameFilter & IX_ETH_DB_VLAN_TAGGED_FRAMES) == 0))
+    {
+        result = ixEthDBPortVlanMembershipRangeChange(portID, 
+            1, IX_ETH_DB_802_1Q_MAX_VLAN_ID, portInfo->vlanMembership, REMOVE_VLAN);
+
+        if (result == IX_ETH_DB_SUCCESS)
+        {
+            ixEthDBLocalVlanMembershipChange(0, portInfo->vlanMembership, ADD_VLAN);
+            result = ixEthDBVlanTableRangeUpdate(portID);
+        }
+    }
+    
+    /* untagged only? */
+    if (frameFilter == IX_ETH_DB_UNTAGGED_FRAMES)
+    {
+        portInfo->npeFrameFilter = 0x01;
+    }
+    
+    /* tagged only? */
+    if ((frameFilter & IX_ETH_DB_UNTAGGED_FRAMES) == 0)
+    {
+        portInfo->npeFrameFilter = 0x02;
+    }
+
+    if (result == IX_ETH_DB_SUCCESS)
+    {
+        result = ixEthDBIngressVlanModeUpdate(portID);
+    }
+
+    return result;
+}
+
+/**
+ * @brief retrieves the acceptable frame type filter for a port
+ *
+ * @param portID ID of the port
+ * @param frameFilter location to store the frame filter
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBAcceptableFrameTypeGet(IxEthDBPortId portID, IxEthDBFrameFilter *frameFilter)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+    
+    IX_ETH_DB_CHECK_REFERENCE(frameFilter);
+    
+    *frameFilter = ixEthDBPortInfo[portID].frameFilter;
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief sends an NPE the updated configuration related
+ * to one QoS priority (associated traffic class and AQM mapping)
+ *
+ * @param portID ID of the port
+ * @param classIndex QoS priority (traffic class index)
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBUpdateTrafficClass(IxEthDBPortId portID, UINT32 classIndex)
+{
+    IxNpeMhMessage message;
+    IX_STATUS result;
+
+    UINT32 trafficClass = ixEthDBPortInfo[portID].priorityTable[classIndex];
+    UINT32 aqmQueue     = ixEthDBPortInfo[portID].ixEthDBTrafficClassAQMAssignments[trafficClass];
+    
+    FILL_SETRXQOSENTRY(message, IX_ETH_DB_PORT_ID_TO_NPE_LOGICAL_ID(portID), classIndex, trafficClass, aqmQueue);
+    
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+    
+    return result;
+}
+
+/**
+ * @brief sets the priority mapping table
+ *
+ * @param portID ID of the port
+ * @param priorityTable new priority mapping table
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPriorityMappingTableSet(IxEthDBPortId portID, IxEthDBPriorityTable priorityTable)
+{
+    UINT32 classIndex;
+    
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+    
+    IX_ETH_DB_CHECK_REFERENCE(priorityTable);
+           
+    for (classIndex = 0 ; classIndex < IX_IEEE802_1Q_QOS_PRIORITY_COUNT ; classIndex++)
+    {
+        /* check range */
+        if (priorityTable[classIndex] >= ixEthDBPortInfo[portID].ixEthDBTrafficClassCount)
+        {
+            return IX_ETH_DB_INVALID_PRIORITY;
+        }
+    }
+    
+    /* set new traffic classes */
+    for (classIndex = 0 ; classIndex < IX_IEEE802_1Q_QOS_PRIORITY_COUNT ; classIndex++)
+    {
+        ixEthDBPortInfo[portID].priorityTable[classIndex] = priorityTable[classIndex];
+        
+        if (ixEthDBUpdateTrafficClass(portID, classIndex) != IX_ETH_DB_SUCCESS)
+        {
+            return IX_ETH_DB_FAIL;
+        }
+    }
+    
+    return IX_ETH_DB_SUCCESS;
+ }
+
+/**
+ * @brief retrieves a port's priority mapping table
+ *
+ * @param portID ID of the port
+ * @param priorityTable location to store the priority table
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPriorityMappingTableGet(IxEthDBPortId portID, IxEthDBPriorityTable priorityTable)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+    
+    IX_ETH_DB_CHECK_REFERENCE(priorityTable);
+    
+    memcpy(priorityTable, ixEthDBPortInfo[portID].priorityTable, sizeof (IxEthDBPriorityTable));
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief sets one QoS priority => traffic class mapping
+ *
+ * @param portID ID of the port
+ * @param userPriority QoS (user) priority
+ * @param trafficClass associated traffic class
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPriorityMappingClassSet(IxEthDBPortId portID, IxEthDBPriority userPriority, IxEthDBPriority trafficClass)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+
+    /* check ranges for userPriority and trafficClass */
+    if (userPriority >= IX_IEEE802_1Q_QOS_PRIORITY_COUNT || trafficClass >= ixEthDBPortInfo[portID].ixEthDBTrafficClassCount)
+    {
+        return IX_ETH_DB_INVALID_PRIORITY;
+    }
+    
+    ixEthDBPortInfo[portID].priorityTable[userPriority] = trafficClass;
+    
+    return ixEthDBUpdateTrafficClass(portID, userPriority);
+}
+
+/**
+ * @brief retrieves one QoS priority => traffic class mapping
+ *
+ * @param portID ID of the port
+ * @param userPriority QoS (user) priority
+ * @param trafficClass location to store the associated traffic class
+ * 
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBPriorityMappingClassGet(IxEthDBPortId portID, IxEthDBPriority userPriority, IxEthDBPriority *trafficClass)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+    
+    IX_ETH_DB_CHECK_REFERENCE(trafficClass);
+    
+    /* check userPriority range */
+    if (userPriority >= IX_IEEE802_1Q_QOS_PRIORITY_COUNT)
+    {
+        return IX_ETH_DB_INVALID_PRIORITY;
+    }
+    
+    *trafficClass = ixEthDBPortInfo[portID].priorityTable[userPriority];
+    
+    return IX_ETH_DB_SUCCESS;
+}
+
+/**
+ * @brief enables or disables the source port extraction
+ * from the VLAN TPID field
+ *
+ * @param portID ID of the port
+ * @param enable TRUE to enable or FALSE to disable
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC 
+IxEthDBStatus ixEthDBVlanPortExtractionEnable(IxEthDBPortId portID, BOOL enable)
+{
+    IxNpeMhMessage message;
+    IX_STATUS result;
+
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_VLAN_QOS);
+
+    FILL_SETPORTIDEXTRACTIONMODE(message, portID, enable);
+
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+    
+    return result;
+}

cpu/ixp/npe/IxEthDBWiFi.c

@@ -0,0 +1,480 @@
+/**
+ * @file IxEthDBAPI.c
+ *
+ * @brief Implementation of the public API
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxEthDB_p.h"
+
+/* forward prototypes */
+IX_ETH_DB_PUBLIC
+MacTreeNode *ixEthDBGatewaySelect(MacTreeNode *stations);
+
+/**
+ * @brief sets the BBSID value for the WiFi header conversion feature
+ *
+ * @param portID ID of the port
+ * @param bbsid pointer to the 6-byte BBSID value
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBWiFiBBSIDSet(IxEthDBPortId portID, IxEthDBMacAddr *bbsid)
+{
+    IxNpeMhMessage message;
+    IX_STATUS result;
+    
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_WIFI_HEADER_CONVERSION);
+ 
+    IX_ETH_DB_CHECK_REFERENCE(bbsid);
+    
+    memcpy(ixEthDBPortInfo[portID].bbsid, bbsid, sizeof (IxEthDBMacAddr));
+
+    FILL_SETBBSID_MSG(message, portID, bbsid);
+
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+    
+    return result;
+}
+
+/**
+ * @brief updates the Frame Control and Duration/ID WiFi header
+ * conversion parameters in an NPE
+ *
+ * @param portID ID of the port
+ *
+ * This function will send a message to the NPE updating the 
+ * frame conversion parameters for 802.3 => 802.11 header conversion.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or IX_ETH_DB_FAIL otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+IxEthDBStatus ixEthDBWiFiFrameControlDurationIDUpdate(IxEthDBPortId portID)
+{
+    IxNpeMhMessage message;
+    IX_STATUS result;
+
+    FILL_SETFRAMECONTROLDURATIONID(message, portID, ixEthDBPortInfo[portID].frameControlDurationID);
+    
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+    
+    return result;
+}
+
+/**
+ * @brief sets the Duration/ID WiFi frame header conversion parameter
+ *
+ * @param portID ID of the port
+ * @param durationID 16-bit value containing the new Duration/ID parameter
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBWiFiDurationIDSet(IxEthDBPortId portID, UINT16 durationID)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_WIFI_HEADER_CONVERSION);
+
+    ixEthDBPortInfo[portID].frameControlDurationID = (ixEthDBPortInfo[portID].frameControlDurationID & 0xFFFF0000) | durationID;
+    
+    return ixEthDBWiFiFrameControlDurationIDUpdate(portID);
+}
+
+/**
+ * @brief sets the Frame Control WiFi frame header conversion parameter
+ *
+ * @param portID ID of the port
+ * @param durationID 16-bit value containing the new Frame Control parameter
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBWiFiFrameControlSet(IxEthDBPortId portID, UINT16 frameControl)
+{
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_WIFI_HEADER_CONVERSION);
+
+    ixEthDBPortInfo[portID].frameControlDurationID = (ixEthDBPortInfo[portID].frameControlDurationID & 0xFFFF) | (frameControl << 16); 
+    
+    return ixEthDBWiFiFrameControlDurationIDUpdate(portID);
+}
+
+/**
+ * @brief removes a WiFi header conversion record
+ *
+ * @param portID ID of the port
+ * @param macAddr MAC address of the record to remove
+ *
+ * Note that this function is documented in the main
+ * component header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBWiFiEntryRemove(IxEthDBPortId portID, IxEthDBMacAddr *macAddr)
+{
+    MacDescriptor recordTemplate;
+    
+    IX_ETH_DB_CHECK_PORT(portID);
+    
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_WIFI_HEADER_CONVERSION);
+    
+    memcpy(recordTemplate.macAddress, macAddr, sizeof (IxEthDBMacAddr));
+    
+    recordTemplate.type   = IX_ETH_DB_WIFI_RECORD;
+    recordTemplate.portID = portID;
+    
+    return ixEthDBRemove(&recordTemplate, NULL);
+}
+
+/**
+ * @brief adds a WiFi header conversion record
+ *
+ * @param portID ID of the port
+ * @param macAddr MAC address of the record to add
+ * @param gatewayMacAddr address of the gateway (or
+ * NULL if this is a station record)
+ *
+ * This function adds a record of type AP_TO_AP (gateway is not NULL)
+ * or AP_TO_STA (gateway is NULL) in the main database as a 
+ * WiFi header conversion record.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ *
+ * @internal
+ */
+IX_ETH_DB_PRIVATE
+IxEthDBStatus ixEthDBWiFiEntryAdd(IxEthDBPortId portID, IxEthDBMacAddr *macAddr, IxEthDBMacAddr *gatewayMacAddr)
+{
+    MacDescriptor recordTemplate;
+
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+
+    IX_ETH_DB_CHECK_REFERENCE(macAddr);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_WIFI_HEADER_CONVERSION);
+
+    memcpy(recordTemplate.macAddress, macAddr, sizeof (IxEthDBMacAddr));
+    
+    recordTemplate.type   = IX_ETH_DB_WIFI_RECORD;
+    recordTemplate.portID = portID;
+    
+    if (gatewayMacAddr != NULL)
+    {
+        memcpy(recordTemplate.recordData.wifiData.gwMacAddress, gatewayMacAddr, sizeof (IxEthDBMacAddr));
+        
+        recordTemplate.recordData.wifiData.type = IX_ETH_DB_WIFI_AP_TO_AP;
+    }
+    else
+    {
+        memset(recordTemplate.recordData.wifiData.gwMacAddress, 0, sizeof (IxEthDBMacAddr));
+
+        recordTemplate.recordData.wifiData.type = IX_ETH_DB_WIFI_AP_TO_STA;
+    }
+    
+    return ixEthDBAdd(&recordTemplate, NULL);
+}
+
+/**
+ * @brief adds a WiFi header conversion record
+ *
+ * @param portID ID of the port
+ * @param macAddr MAC address of the record to add
+ * @param gatewayMacAddr address of the gateway 
+ *
+ * This function adds a record of type AP_TO_AP
+ * in the main database as a WiFi header conversion record.
+ *
+ * This is simply a wrapper over @ref ixEthDBWiFiEntryAdd().
+ *
+ * Note that this function is documented in the main
+ * component header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBWiFiAccessPointEntryAdd(IxEthDBPortId portID, IxEthDBMacAddr *macAddr, IxEthDBMacAddr *gatewayMacAddr)
+{
+    IX_ETH_DB_CHECK_REFERENCE(gatewayMacAddr);
+
+    return ixEthDBWiFiEntryAdd(portID, macAddr, gatewayMacAddr);
+}
+
+/**
+ * @brief adds a WiFi header conversion record
+ *
+ * @param portID ID of the port
+ * @param macAddr MAC address of the record to add
+ *
+ * This function adds a record of type AP_TO_STA
+ * in the main database as a WiFi header conversion record.
+ *
+ * This is simply a wrapper over @ref ixEthDBWiFiEntryAdd().
+ *
+ * Note that this function is documented in the main
+ * component header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed
+ * successfully or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBWiFiStationEntryAdd(IxEthDBPortId portID, IxEthDBMacAddr *macAddr)
+{
+    return ixEthDBWiFiEntryAdd(portID, macAddr, NULL);
+}
+
+/**
+ * @brief selects a set of gateways from a tree of 
+ * WiFi header conversion records
+ *
+ * @param stations binary tree containing pointers to WiFi header
+ * conversion records
+ *
+ * This function browses through the input binary tree, identifies
+ * records of type AP_TO_AP, clones these records and appends them
+ * to a vine (a single right-branch binary tree) which is returned
+ * as result. A maximum of MAX_GW_SIZE entries containing gateways
+ * will be cloned from the original tree.
+ *
+ * @return vine (linear binary tree) containing record
+ * clones of AP_TO_AP type, which have a gateway field
+ *
+ * @internal
+ */
+IX_ETH_DB_PUBLIC
+MacTreeNode *ixEthDBGatewaySelect(MacTreeNode *stations)
+{
+    MacTreeNodeStack *stack;
+    MacTreeNode *gateways, *insertionPlace;
+    UINT32 gwIndex = 1; /* skip the empty root */
+    
+    if (stations == NULL)
+    {
+        return NULL;
+    }
+
+    stack = ixOsalCacheDmaMalloc(sizeof (MacTreeNodeStack));
+
+    if (stack == NULL)
+    {
+        ERROR_LOG("DB: (WiFi) failed to allocate the node stack for gateway tree linearization, out of memory?\n");
+        return NULL;
+    }
+    
+    /* initialize root node */
+    gateways = insertionPlace = NULL;
+        
+    /* start browsing the station tree */
+    NODE_STACK_INIT(stack);
+    
+    /* initialize stack by pushing the tree root at offset 0 */
+    NODE_STACK_PUSH(stack, stations, 0);
+    
+    while (NODE_STACK_NONEMPTY(stack))
+    {
+        MacTreeNode *node;
+        UINT32 offset;
+       
+        NODE_STACK_POP(stack, node, offset);
+
+        /* we can store maximum 31 (32 total, 1 empty root) entries in the gateway tree */
+        if (offset > (MAX_GW_SIZE - 1)) break;
+        
+        /* check if this record has a gateway address */
+        if (node->descriptor != NULL && node->descriptor->recordData.wifiData.type == IX_ETH_DB_WIFI_AP_TO_AP)
+        {
+            /* found a record, create an insertion place */
+            if (insertionPlace != NULL)
+            {
+                insertionPlace->right = ixEthDBAllocMacTreeNode();
+                insertionPlace        = insertionPlace->right;
+            }
+            else
+            {
+                gateways       = ixEthDBAllocMacTreeNode();
+                insertionPlace = gateways;
+            }
+
+            if (insertionPlace == NULL)
+            {
+                /* no nodes left, bail out with what we have */
+                ixOsalCacheDmaFree(stack);
+                return gateways;
+            }
+            
+            /* clone the original record for the gateway tree */
+            insertionPlace->descriptor = ixEthDBCloneMacDescriptor(node->descriptor);
+            
+            /* insert and update the offset in the original record */
+            node->descriptor->recordData.wifiData.gwAddressIndex = gwIndex++;
+        }
+        
+        /* browse the tree */
+        if (node->left != NULL)
+        {
+            NODE_STACK_PUSH(stack, node->left, LEFT_CHILD_OFFSET(offset));
+        }
+
+        if (node->right != NULL)
+        {
+            NODE_STACK_PUSH(stack, node->right, RIGHT_CHILD_OFFSET(offset));
+        }
+    }
+    
+    ixOsalCacheDmaFree(stack);
+    return gateways;    
+}
+
+/**
+ * @brief downloads the WiFi header conversion table to an NPE
+ *
+ * @param portID ID of the port
+ *
+ * This function prepares the WiFi header conversion tables and
+ * downloads them to the specified NPE port.
+ *
+ * The header conversion tables consist in the main table of
+ * addresses and the secondary table of gateways. AP_TO_AP records
+ * from the first table contain index fields into the second table
+ * for gateway selection.
+ *
+ * Note that this function is documented in the main component
+ * header file, IxEthDB.h.
+ *
+ * @return IX_ETH_DB_SUCCESS if the operation completed successfully
+ * or an appropriate error message otherwise
+ */
+IX_ETH_DB_PUBLIC
+IxEthDBStatus ixEthDBWiFiConversionTableDownload(IxEthDBPortId portID)
+{
+    IxEthDBPortMap query;
+    MacTreeNode *stations = NULL, *gateways = NULL, *gateway = NULL;
+    IxNpeMhMessage message;
+    PortInfo *portInfo;
+    IX_STATUS result;
+
+    IX_ETH_DB_CHECK_PORT(portID);
+
+    IX_ETH_DB_CHECK_SINGLE_NPE(portID);
+    
+    IX_ETH_DB_CHECK_FEATURE(portID, IX_ETH_DB_WIFI_HEADER_CONVERSION);
+
+    portInfo = &ixEthDBPortInfo[portID];
+   
+    SET_DEPENDENCY_MAP(query, portID);
+    
+    ixEthDBUpdateLock();
+
+    stations = ixEthDBQuery(NULL, query, IX_ETH_DB_WIFI_RECORD, MAX_ELT_SIZE);
+    gateways = ixEthDBGatewaySelect(stations);
+    
+    /* clean up gw area */
+    memset((void *) portInfo->updateMethod.npeGwUpdateZone, FULL_GW_BYTE_SIZE, 0);
+
+    /* write all gateways */
+    gateway = gateways;
+
+    while (gateway != NULL)
+    {
+        ixEthDBNPEGatewayNodeWrite((void *) (((UINT32) portInfo->updateMethod.npeGwUpdateZone) 
+            + gateway->descriptor->recordData.wifiData.gwAddressIndex * ELT_ENTRY_SIZE), 
+            gateway);
+
+        gateway = gateway->right;
+    }
+
+    /* free the gateway tree */
+    if (gateways != NULL)
+    {
+        ixEthDBFreeMacTreeNode(gateways);
+    }
+
+    FILL_SETAPMACTABLE_MSG(message, 
+        IX_OSAL_MMU_VIRT_TO_PHYS(portInfo->updateMethod.npeGwUpdateZone));
+
+    IX_ETHDB_SEND_NPE_MSG(IX_ETH_DB_PORT_ID_TO_NPE(portID), message, result);
+
+    if (result == IX_SUCCESS)
+    {
+        /* update the main tree (the stations tree) */
+        portInfo->updateMethod.searchTree = stations;
+        
+        result = ixEthDBNPEUpdateHandler(portID, IX_ETH_DB_WIFI_RECORD);
+    }
+
+    ixEthDBUpdateUnlock();
+
+    return result;
+}

cpu/ixp/npe/IxEthMii.c

@@ -0,0 +1,497 @@
+/**
+ * @file IxEthMii.c
+ *
+ * @author Intel Corporation
+ * @date
+ *
+ * @brief  MII control functions
+ *
+ * Design Notes:
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxOsal.h"
+
+#include "IxEthAcc.h"
+#include "IxEthMii_p.h"
+
+#ifdef __wince
+#include "IxOsPrintf.h"
+#endif
+
+/* Array to store the phy IDs of the discovered phys */
+PRIVATE UINT32 ixEthMiiPhyId[IXP425_ETH_ACC_MII_MAX_ADDR];
+
+/*********************************************************
+ *
+ * Scan for PHYs on the MII bus. This function returns
+ * an array of booleans, one for each PHY address.
+ * If a PHY is found at a particular address, the
+ * corresponding entry in the array is set to TRUE.
+ *
+ */
+
+PUBLIC IX_STATUS
+ixEthMiiPhyScan(BOOL phyPresent[], UINT32 maxPhyCount)
+{
+    UINT32 i;
+    UINT16 regval, regvalId1, regvalId2;
+
+    /*Search for PHYs on the MII*/
+    /*Search for existant phys on the MDIO bus*/
+
+    if ((phyPresent == NULL) || 
+	(maxPhyCount > IXP425_ETH_ACC_MII_MAX_ADDR))
+    {
+	return IX_FAIL;
+    }
+
+    /* fill the array */
+    for(i=0;
+        i<IXP425_ETH_ACC_MII_MAX_ADDR;
+	i++)
+    {
+	phyPresent[i] = FALSE;
+    }
+
+    /* iterate through the PHY addresses */
+    for(i=0;
+	maxPhyCount > 0 && i<IXP425_ETH_ACC_MII_MAX_ADDR;
+	i++)
+    {
+	ixEthMiiPhyId[i] = IX_ETH_MII_INVALID_PHY_ID;
+	if(ixEthAccMiiReadRtn(i,
+			      IX_ETH_MII_CTRL_REG,
+			      &regval) == IX_ETH_ACC_SUCCESS)
+	{
+	    if((regval & 0xffff) != 0xffff)
+	    {
+		maxPhyCount--;
+		/*Need to read the register twice here to flush PHY*/
+		ixEthAccMiiReadRtn(i,  IX_ETH_MII_PHY_ID1_REG, &regvalId1);
+		ixEthAccMiiReadRtn(i,  IX_ETH_MII_PHY_ID1_REG, &regvalId1);
+		ixEthAccMiiReadRtn(i,  IX_ETH_MII_PHY_ID2_REG, &regvalId2);
+		ixEthMiiPhyId[i] = (regvalId1 << IX_ETH_MII_REG_SHL) | regvalId2;
+		if ((ixEthMiiPhyId[i] == IX_ETH_MII_KS8995_PHY_ID)
+		    || (ixEthMiiPhyId[i] == IX_ETH_MII_LXT971_PHY_ID)
+		    || (ixEthMiiPhyId[i] == IX_ETH_MII_LXT972_PHY_ID)
+		    || (ixEthMiiPhyId[i] == IX_ETH_MII_LXT973_PHY_ID)
+		    || (ixEthMiiPhyId[i] == IX_ETH_MII_LXT973A3_PHY_ID)
+		    || (ixEthMiiPhyId[i] == IX_ETH_MII_LXT9785_PHY_ID)
+		    )
+		{
+		    /* supported phy */
+		    phyPresent[i] = TRUE;
+		} /* end of if(ixEthMiiPhyId) */
+		else
+		{
+		    if (ixEthMiiPhyId[i] != IX_ETH_MII_INVALID_PHY_ID)
+		    {
+			/* unsupported phy */
+                        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+                                   IX_OSAL_LOG_DEV_STDOUT,
+				    "ixEthMiiPhyScan : unexpected Mii PHY ID %8.8x\n", 
+				    ixEthMiiPhyId[i], 2, 3, 4, 5, 6);
+			ixEthMiiPhyId[i] = IX_ETH_MII_UNKNOWN_PHY_ID;
+			phyPresent[i] = TRUE;
+		    }
+		} 
+	    }
+	}
+    }
+    return IX_SUCCESS;
+}
+
+/************************************************************
+ *
+ * Configure the PHY at the specified address
+ *
+ */
+PUBLIC IX_STATUS
+ixEthMiiPhyConfig(UINT32 phyAddr,
+		  BOOL speed100,
+		  BOOL fullDuplex,
+		  BOOL autonegotiate)
+{
+    UINT16 regval=0;
+
+    /* parameter check */
+    if ((phyAddr < IXP425_ETH_ACC_MII_MAX_ADDR) &&
+	(ixEthMiiPhyId[phyAddr] != IX_ETH_MII_INVALID_PHY_ID))
+    {
+    /*
+     * set the control register
+     */
+	if(autonegotiate)
+	{
+	    regval |= IX_ETH_MII_CR_AUTO_EN | IX_ETH_MII_CR_RESTART;
+	}
+	else
+	{
+	    if(speed100)
+	    {
+		regval |= IX_ETH_MII_CR_100;
+	    }
+	    if(fullDuplex)
+	    {
+		regval |= IX_ETH_MII_CR_FDX;
+	    }
+	} /* end of if-else() */
+	if (ixEthAccMiiWriteRtn(phyAddr, 
+				IX_ETH_MII_CTRL_REG, 
+				regval) == IX_ETH_ACC_SUCCESS)
+	{
+	    return IX_SUCCESS;
+	}
+    } /* end of if(phyAddr) */
+    return IX_FAIL;
+}
+
+/******************************************************************
+ *
+ *  Enable the PHY Loopback at the specified address
+ */
+PUBLIC IX_STATUS
+ixEthMiiPhyLoopbackEnable (UINT32 phyAddr)
+{
+  UINT16 regval ;  
+
+  if ((phyAddr < IXP425_ETH_ACC_MII_MAX_ADDR) && 
+      (IX_ETH_MII_INVALID_PHY_ID != ixEthMiiPhyId[phyAddr]))
+  {
+      /* read/write the control register */
+      if(ixEthAccMiiReadRtn (phyAddr,
+			     IX_ETH_MII_CTRL_REG, 
+			     &regval) 
+	 == IX_ETH_ACC_SUCCESS)
+      {
+	  if(ixEthAccMiiWriteRtn (phyAddr, 
+				  IX_ETH_MII_CTRL_REG, 
+				  regval | IX_ETH_MII_CR_LOOPBACK)
+	     == IX_ETH_ACC_SUCCESS)
+	  {
+	      return IX_SUCCESS;
+	  }
+      }
+  }
+  return IX_FAIL;
+}
+
+/******************************************************************
+ *
+ *  Disable the PHY Loopback at the specified address
+ */
+PUBLIC IX_STATUS
+ixEthMiiPhyLoopbackDisable (UINT32 phyAddr)
+{
+  UINT16 regval ;  
+
+  if ((phyAddr < IXP425_ETH_ACC_MII_MAX_ADDR) && 
+      (IX_ETH_MII_INVALID_PHY_ID != ixEthMiiPhyId[phyAddr]))
+  {
+      /* read/write the control register */
+      if(ixEthAccMiiReadRtn (phyAddr,
+			     IX_ETH_MII_CTRL_REG, 
+			     &regval) 
+	 == IX_ETH_ACC_SUCCESS)
+      {
+	  if(ixEthAccMiiWriteRtn (phyAddr, 
+				  IX_ETH_MII_CTRL_REG, 
+				  regval & (~IX_ETH_MII_CR_LOOPBACK))
+	     == IX_ETH_ACC_SUCCESS)
+	  {
+	      return IX_SUCCESS;
+	  }
+      }
+  }
+  return IX_FAIL;
+}
+
+/******************************************************************
+ *
+ *  Reset the PHY at the specified address
+ */
+PUBLIC IX_STATUS
+ixEthMiiPhyReset(UINT32 phyAddr)
+{
+    UINT32 timeout;
+    UINT16 regval;
+
+    if ((phyAddr < IXP425_ETH_ACC_MII_MAX_ADDR) &&
+	(ixEthMiiPhyId[phyAddr] != IX_ETH_MII_INVALID_PHY_ID))
+    {
+	if ((ixEthMiiPhyId[phyAddr] == IX_ETH_MII_LXT971_PHY_ID)	||
+	    (ixEthMiiPhyId[phyAddr] == IX_ETH_MII_LXT972_PHY_ID)	||
+	    (ixEthMiiPhyId[phyAddr] == IX_ETH_MII_LXT973_PHY_ID)	||
+	    (ixEthMiiPhyId[phyAddr] == IX_ETH_MII_LXT973A3_PHY_ID)	||
+		(ixEthMiiPhyId[phyAddr] == IX_ETH_MII_LXT9785_PHY_ID)
+	    )
+	{
+	    /* use the control register to reset the phy */
+	    ixEthAccMiiWriteRtn(phyAddr, 
+				IX_ETH_MII_CTRL_REG,
+				IX_ETH_MII_CR_RESET);
+	 
+	    /* poll until the reset bit is cleared */
+	    timeout = 0;
+	    do
+	    {
+		ixOsalSleep (IX_ETH_MII_RESET_POLL_MS);
+
+		/* read the control register and check for timeout */
+		ixEthAccMiiReadRtn(phyAddr, 
+				   IX_ETH_MII_CTRL_REG,
+				   &regval);
+		if ((regval & IX_ETH_MII_CR_RESET) == 0)
+		{
+		    /* timeout bit is self-cleared */
+		    break;
+		}
+		timeout += IX_ETH_MII_RESET_POLL_MS;
+	    }
+	    while (timeout < IX_ETH_MII_RESET_DELAY_MS);
+
+	    /* check for timeout */
+	    if (timeout >= IX_ETH_MII_RESET_DELAY_MS)
+	    {
+		ixEthAccMiiWriteRtn(phyAddr, IX_ETH_MII_CTRL_REG,
+				    IX_ETH_MII_CR_NORM_EN);
+		return IX_FAIL;
+	    }
+
+	    return IX_SUCCESS;
+	} /* end of if(ixEthMiiPhyId) */
+	else if (ixEthMiiPhyId[phyAddr] == IX_ETH_MII_KS8995_PHY_ID)
+	{
+	    /* reset bit is reserved, just reset the control register */
+	    ixEthAccMiiWriteRtn(phyAddr, IX_ETH_MII_CTRL_REG,
+				IX_ETH_MII_CR_NORM_EN);
+	    return IX_SUCCESS;
+	}
+	else
+	{
+	    /* unknown PHY, set the control register reset bit,
+	     * wait 2 s. and clear the control register.
+	     */
+	    ixEthAccMiiWriteRtn(phyAddr, IX_ETH_MII_CTRL_REG,
+				IX_ETH_MII_CR_RESET);
+	    
+	    ixOsalSleep (IX_ETH_MII_RESET_DELAY_MS);
+	    
+	    ixEthAccMiiWriteRtn(phyAddr, IX_ETH_MII_CTRL_REG,
+				IX_ETH_MII_CR_NORM_EN);
+	    return IX_SUCCESS;
+	} /* end of if-else(ixEthMiiPhyId) */
+    } /* end of if(phyAddr) */
+    return IX_FAIL;
+}
+
+/*****************************************************************
+ *
+ *  Link state query functions
+ */
+
+PUBLIC IX_STATUS
+ixEthMiiLinkStatus(UINT32 phyAddr,
+           BOOL *linkUp,
+           BOOL *speed100,
+           BOOL *fullDuplex,
+           BOOL *autoneg)
+{
+    UINT16 ctrlRegval, statRegval, regval, regval4, regval5;
+
+    /* check the parameters */
+    if ((linkUp == NULL) || 
+	(speed100 == NULL) || 
+	(fullDuplex == NULL) ||
+	(autoneg == NULL))
+    {
+	return IX_FAIL;
+    }
+
+    *linkUp = FALSE;
+    *speed100 = FALSE;
+    *fullDuplex = FALSE;
+    *autoneg = FALSE;
+
+    if ((phyAddr < IXP425_ETH_ACC_MII_MAX_ADDR) &&
+	(ixEthMiiPhyId[phyAddr] != IX_ETH_MII_INVALID_PHY_ID))
+    {
+	if ((ixEthMiiPhyId[phyAddr] == IX_ETH_MII_LXT971_PHY_ID)	||
+	    (ixEthMiiPhyId[phyAddr] == IX_ETH_MII_LXT972_PHY_ID)	||
+	    (ixEthMiiPhyId[phyAddr] == IX_ETH_MII_LXT9785_PHY_ID)
+		)
+	{
+	    /* --------------------------------------------------*/
+	    /* Retrieve information from PHY specific register   */
+	    /* --------------------------------------------------*/
+	    if (ixEthAccMiiReadRtn(phyAddr, 
+				   IX_ETH_MII_STAT2_REG, 
+				   &regval) != IX_ETH_ACC_SUCCESS)
+	    {
+		return IX_FAIL;
+	    }
+	    *linkUp = ((regval & IX_ETH_MII_SR2_LINK) != 0);
+	    *speed100 = ((regval & IX_ETH_MII_SR2_100) != 0);
+	    *fullDuplex = ((regval & IX_ETH_MII_SR2_FD) != 0);
+	    *autoneg = ((regval & IX_ETH_MII_SR2_AUTO) != 0);
+	    return IX_SUCCESS;
+	} /* end of if(ixEthMiiPhyId) */
+	else
+	{    
+	    /* ----------------------------------------------------*/
+	    /* Retrieve information from status and ctrl registers */
+	    /* ----------------------------------------------------*/
+	    if (ixEthAccMiiReadRtn(phyAddr,  
+				   IX_ETH_MII_CTRL_REG, 
+				   &ctrlRegval) != IX_ETH_ACC_SUCCESS)
+	    {
+		return IX_FAIL;
+	    }
+	    ixEthAccMiiReadRtn(phyAddr,  IX_ETH_MII_STAT_REG, &statRegval);
+	    
+	    *linkUp = ((statRegval & IX_ETH_MII_SR_LINK_STATUS) != 0);
+	    if (*linkUp)
+	    {
+		*autoneg = ((ctrlRegval & IX_ETH_MII_CR_AUTO_EN) != 0) &&
+		    ((statRegval &  IX_ETH_MII_SR_AUTO_SEL) != 0) &&
+		    ((statRegval & IX_ETH_MII_SR_AUTO_NEG) != 0);
+		
+		if (*autoneg)
+		{
+		    /* mask the current stat values with the capabilities */
+		    ixEthAccMiiReadRtn(phyAddr, IX_ETH_MII_AN_ADS_REG, &regval4);
+		    ixEthAccMiiReadRtn(phyAddr, IX_ETH_MII_AN_PRTN_REG, &regval5);
+		    /* merge the flags from the 3 registers */
+		    regval = (statRegval & ((regval4 & regval5) << 6));
+		    /* initialise from status register values */
+		    if ((regval & IX_ETH_MII_SR_TX_FULL_DPX) != 0)
+		    {
+			/* 100 Base X full dplx */
+			*speed100 = TRUE;
+			*fullDuplex = TRUE;
+			return IX_SUCCESS;
+		    }
+		    if ((regval & IX_ETH_MII_SR_TX_HALF_DPX) != 0)
+		    {
+			/* 100 Base X half dplx */
+			*speed100 = TRUE;
+			return IX_SUCCESS;
+		    }
+		    if ((regval & IX_ETH_MII_SR_10T_FULL_DPX) != 0)
+		    {
+			/* 10 mb full dplx */
+			*fullDuplex = TRUE;
+			return IX_SUCCESS;
+		    }
+		    if ((regval & IX_ETH_MII_SR_10T_HALF_DPX) != 0)
+		    {
+			/* 10 mb half dplx */
+			return IX_SUCCESS;
+		    }
+		} /* end of if(autoneg) */
+		else
+		{
+		    /* autonegotiate not complete, return setup parameters */
+		    *speed100 = ((ctrlRegval & IX_ETH_MII_CR_100) != 0);
+		    *fullDuplex = ((ctrlRegval & IX_ETH_MII_CR_FDX) != 0);
+		}
+	    } /* end of if(linkUp) */
+	} /* end of if-else(ixEthMiiPhyId) */
+    } /* end of if(phyAddr) */
+    else
+    {
+	return IX_FAIL;
+    } /* end of if-else(phyAddr) */
+    return IX_SUCCESS;
+}
+
+/*****************************************************************
+ *
+ *  Link state display functions
+ */
+
+PUBLIC IX_STATUS
+ixEthMiiPhyShow (UINT32 phyAddr)
+{
+    BOOL linkUp, speed100, fullDuplex, autoneg;
+    UINT16 cregval;
+    UINT16 sregval;
+    
+
+    ixEthAccMiiReadRtn(phyAddr,  IX_ETH_MII_STAT_REG, &sregval);
+    ixEthAccMiiReadRtn(phyAddr,  IX_ETH_MII_CTRL_REG, &cregval);
+
+    /* get link information */
+    if (ixEthMiiLinkStatus(phyAddr,
+			   &linkUp,
+			   &speed100,
+			   &fullDuplex,
+			   &autoneg) != IX_ETH_ACC_SUCCESS)
+    {
+	printf("PHY Status unknown\n");
+	return IX_FAIL;
+    }
+
+    printf("PHY ID [phyAddr]: %8.8x\n",ixEthMiiPhyId[phyAddr]);
+    printf( " Status reg:  %4.4x\n",sregval);
+    printf( " control reg: %4.4x\n",cregval);
+    /* display link information */
+    printf("PHY Status:\n");
+    printf("    Link is %s\n",
+	   (linkUp ? "Up" : "Down"));
+    if((sregval & IX_ETH_MII_SR_REMOTE_FAULT) != 0)
+    {
+	printf("    Remote fault detected\n");
+    }
+    printf("    Auto Negotiation %s\n",
+	   (autoneg ? "Completed" : "Not Completed"));
+
+    printf("PHY Configuration:\n");
+    printf("    Speed %sMb/s\n",
+	   (speed100 ? "100" : "10"));
+    printf("    %s Duplex\n",
+	   (fullDuplex ? "Full" : "Half"));
+    printf("    Auto Negotiation %s\n",
+	   (autoneg ? "Enabled" : "Disabled"));
+    return IX_SUCCESS;
+}
+

cpu/ixp/npe/IxFeatureCtrl.c

@@ -0,0 +1,422 @@
+/**
+ * @file IxFeatureCtrl.c
+ *
+ * @author Intel Corporation
+ * @date 29-Jan-2003
+ *
+ * @brief Feature Control Public API Implementation
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+#include "IxOsal.h"
+#include "IxVersionId.h"
+#include "IxFeatureCtrl.h"
+
+/* Macro to read from the Feature Control Register */
+#define IX_FEATURE_CTRL_READ(result) \
+do { \
+ixFeatureCtrlExpMap(); \
+(result) = IX_OSAL_READ_LONG(ixFeatureCtrlRegister); \
+} while (0)
+
+/* Macro to write to the Feature Control Register */
+#define IX_FEATURE_CTRL_WRITE(value) \
+do { \
+ixFeatureCtrlExpMap(); \
+IX_OSAL_WRITE_LONG(ixFeatureCtrlRegister, (value)); \
+} while (0)
+
+/*
+ * This is the offset of the feature register relative to the base of the
+ * Expansion Bus Controller MMR.
+ */
+#define IX_FEATURE_CTRL_REG_OFFSET (0x00000028)
+
+
+/* Boolean to mark the fact that the EXP_CONFIG address space was mapped */
+PRIVATE BOOL ixFeatureCtrlExpCfgRegionMapped = FALSE;
+
+/* Pointer holding the virtual address of the Feature Control Register */
+PRIVATE VUINT32 *ixFeatureCtrlRegister = NULL;
+
+/* Place holder to store the software configuration */
+PRIVATE BOOL swConfiguration[IX_FEATURECTRL_SWCONFIG_MAX];
+
+/* Flag to control swConfiguration[] is initialized once */
+PRIVATE BOOL swConfigurationFlag = FALSE ;
+
+/* Array containing component mask values */
+#ifdef __ixp42X
+UINT32 componentMask[IX_FEATURECTRL_MAX_COMPONENTS] = {
+    (0x1<<IX_FEATURECTRL_RCOMP),
+    (0x1<<IX_FEATURECTRL_USB),
+    (0x1<<IX_FEATURECTRL_HASH),
+    (0x1<<IX_FEATURECTRL_AES),
+    (0x1<<IX_FEATURECTRL_DES),
+    (0x1<<IX_FEATURECTRL_HDLC),
+    (0x1<<IX_FEATURECTRL_AAL),
+    (0x1<<IX_FEATURECTRL_HSS),
+    (0x1<<IX_FEATURECTRL_UTOPIA),
+    (0x1<<IX_FEATURECTRL_ETH0),
+    (0x1<<IX_FEATURECTRL_ETH1),
+    (0x1<<IX_FEATURECTRL_NPEA),
+    (0x1<<IX_FEATURECTRL_NPEB),
+    (0x1<<IX_FEATURECTRL_NPEC),
+    (0x1<<IX_FEATURECTRL_PCI),
+    IX_FEATURECTRL_COMPONENT_NOT_AVAILABLE,
+    (0x3<<IX_FEATURECTRL_UTOPIA_PHY_LIMIT),
+    (0x1<<IX_FEATURECTRL_UTOPIA_PHY_LIMIT_BIT2),
+    IX_FEATURECTRL_COMPONENT_NOT_AVAILABLE,
+    IX_FEATURECTRL_COMPONENT_NOT_AVAILABLE,
+    IX_FEATURECTRL_COMPONENT_NOT_AVAILABLE,
+    IX_FEATURECTRL_COMPONENT_NOT_AVAILABLE,
+    IX_FEATURECTRL_COMPONENT_NOT_AVAILABLE
+};
+#elif defined (__ixp46X)
+UINT32 componentMask[IX_FEATURECTRL_MAX_COMPONENTS] = {
+    (0x1<<IX_FEATURECTRL_RCOMP),
+    (0x1<<IX_FEATURECTRL_USB),
+    (0x1<<IX_FEATURECTRL_HASH),
+    (0x1<<IX_FEATURECTRL_AES),
+    (0x1<<IX_FEATURECTRL_DES),
+    (0x1<<IX_FEATURECTRL_HDLC),
+    IX_FEATURECTRL_COMPONENT_ALWAYS_AVAILABLE,  /* AAL component is always on */
+    (0x1<<IX_FEATURECTRL_HSS),
+    (0x1<<IX_FEATURECTRL_UTOPIA),
+    (0x1<<IX_FEATURECTRL_ETH0),
+    (0x1<<IX_FEATURECTRL_ETH1),
+    (0x1<<IX_FEATURECTRL_NPEA),
+    (0x1<<IX_FEATURECTRL_NPEB),
+    (0x1<<IX_FEATURECTRL_NPEC),
+    (0x1<<IX_FEATURECTRL_PCI),
+    (0x1<<IX_FEATURECTRL_ECC_TIMESYNC),
+    (0x3<<IX_FEATURECTRL_UTOPIA_PHY_LIMIT),
+    (0x1<<IX_FEATURECTRL_UTOPIA_PHY_LIMIT_BIT2), /* NOT TO BE USED */
+    (0x1<<IX_FEATURECTRL_USB_HOST_CONTROLLER),
+    (0x1<<IX_FEATURECTRL_NPEA_ETH),
+    (0x1<<IX_FEATURECTRL_NPEB_ETH),
+    (0x1<<IX_FEATURECTRL_RSA),
+    (0x3<<IX_FEATURECTRL_XSCALE_MAX_FREQ),
+    (0x1<<IX_FEATURECTRL_XSCALE_MAX_FREQ_BIT2)
+};
+#endif /* __ixp42X */
+
+/**
+ * Forward declaration
+ */
+PRIVATE
+void ixFeatureCtrlExpMap(void);
+
+PRIVATE 
+void ixFeatureCtrlSwConfigurationInit(void);
+
+/**
+ * Function to map EXP_CONFIG space
+ */
+PRIVATE
+void ixFeatureCtrlExpMap(void)
+{
+    UINT32 expCfgBaseAddress = 0;
+
+    /* If the EXP Configuration space has already been mapped then
+     * return */
+    if (ixFeatureCtrlExpCfgRegionMapped == TRUE)
+    {
+	return;
+    }
+
+    /* Map (get virtual address) for the EXP_CONFIG space */
+    expCfgBaseAddress = (UINT32)
+	(IX_OSAL_MEM_MAP(IX_OSAL_IXP400_EXP_BUS_REGS_PHYS_BASE,
+			   IX_OSAL_IXP400_EXP_REG_MAP_SIZE));
+
+    /* Assert that the mapping operation succeeded */
+    IX_OSAL_ASSERT(expCfgBaseAddress);
+
+    /* Set the address of the Feature register */
+    ixFeatureCtrlRegister =
+	(VUINT32 *) (expCfgBaseAddress + IX_FEATURE_CTRL_REG_OFFSET);
+
+    /* Mark the fact that the EXP_CONFIG space has already been mapped */
+    ixFeatureCtrlExpCfgRegionMapped = TRUE;
+}
+
+/**
+ * Function definition: ixFeatureCtrlSwConfigurationInit
+ * This function will only initialize software configuration once.
+ */
+PRIVATE void ixFeatureCtrlSwConfigurationInit(void)
+{
+  UINT32 i;
+  if (FALSE == swConfigurationFlag)
+  {
+    for (i=0; i<IX_FEATURECTRL_SWCONFIG_MAX ; i++)
+    {
+        /* By default, all software configuration are enabled */
+        swConfiguration[i]= TRUE ;
+    }
+    /*Make sure this function only initializes swConfiguration[] once*/
+    swConfigurationFlag = TRUE ;
+  }  
+}
+
+/**
+ * Function definition: ixFeatureCtrlRead
+ */
+IxFeatureCtrlReg 
+ixFeatureCtrlRead (void)
+{
+    IxFeatureCtrlReg result;
+
+#if CPU!=SIMSPARCSOLARIS
+    /* Read the feature control register */
+    IX_FEATURE_CTRL_READ(result);
+    return result;
+#else
+    /* Return an invalid value for VxWorks simulation */
+    result = 0xFFFFFFFF;
+    return result;
+#endif
+}
+
+/**
+ * Function definition: ixFeatureCtrlWrite
+ */
+void
+ixFeatureCtrlWrite (IxFeatureCtrlReg expUnitReg)
+{
+#if CPU!=SIMSPARCSOLARIS
+    /* Write value to feature control register */
+    IX_FEATURE_CTRL_WRITE(expUnitReg);
+#endif
+}
+
+
+/**
+ * Function definition: ixFeatureCtrlHwCapabilityRead
+ */
+IxFeatureCtrlReg
+ixFeatureCtrlHwCapabilityRead (void)
+{ 
+  IxFeatureCtrlReg currentReg, hwCapability;
+  
+  /* Capture a copy of feature control register */
+  currentReg = ixFeatureCtrlRead(); 
+
+  /* Try to enable all hardware components. 
+   * Only software disable hardware can be enabled again */
+  ixFeatureCtrlWrite(0);
+  
+  /* Read feature control register to know the hardware capability. */ 
+  hwCapability = ixFeatureCtrlRead();
+     
+  /* Restore initial feature control value */
+  ixFeatureCtrlWrite(currentReg);
+
+  /* return Hardware Capability */
+  return hwCapability;  
+}
+
+
+/**
+ * Function definition: ixFeatureCtrlComponentCheck
+ */
+IX_STATUS 
+ixFeatureCtrlComponentCheck (IxFeatureCtrlComponentType componentType)
+{
+  IxFeatureCtrlReg expUnitReg; 
+  UINT32 mask = 0;
+
+  /* Lookup mask of component */
+  mask=componentMask[componentType];
+
+  /* Check if mask is available or not */
+  if(IX_FEATURECTRL_COMPONENT_NOT_AVAILABLE == mask)
+  {
+      return IX_FEATURE_CTRL_COMPONENT_DISABLED;
+  }
+
+  if(IX_FEATURECTRL_COMPONENT_ALWAYS_AVAILABLE == mask)
+  {
+      return IX_FEATURE_CTRL_COMPONENT_ENABLED;
+  }
+
+  /* Read feature control register to know current hardware capability. */ 
+  expUnitReg = ixFeatureCtrlRead();
+
+  /* For example: To check for Hashing Coprocessor (bit-2) 
+   *                   expUniteg    = 0x0010
+   *                  ~expUnitReg   = 0x1101 
+   *                  componentType = 0x0100
+   *    ~expUnitReg & componentType = 0x0100 (Not zero)                      
+   */
+ 
+  /* 
+   * Inverse the bit value because available component is 0 in value 
+   */
+  expUnitReg = ~expUnitReg ;
+
+  if (expUnitReg & mask)
+  {
+     return (IX_FEATURE_CTRL_COMPONENT_ENABLED);
+  }   
+  else
+  {  
+     return (IX_FEATURE_CTRL_COMPONENT_DISABLED);
+  } 
+}
+
+
+/**
+ * Function definition: ixFeatureCtrlProductIdRead
+ */
+IxFeatureCtrlProductId
+ixFeatureCtrlProductIdRead ()
+{
+#if CPU!=SIMSPARCSOLARIS
+  IxFeatureCtrlProductId  pdId = 0 ;
+   
+  /* Use ARM instruction to move register0 from coprocessor to ARM register */ 
+    
+#ifndef __wince
+    __asm("mrc p15, 0, %0, cr0, cr0, 0;" : "=r"(pdId) :); 
+#else
+      
+#ifndef IN_KERNEL
+        BOOL  mode;
+#endif
+    extern  IxFeatureCtrlProductId AsmixFeatureCtrlProductIdRead();
+    
+#ifndef IN_KERNEL
+    mode = SetKMode(TRUE);
+#endif
+    pdId = AsmixFeatureCtrlProductIdRead();
+#ifndef IN_KERNEL
+    SetKMode(mode);
+#endif
+
+#endif
+  return (pdId);
+#else
+  /* Return an invalid value for VxWorks simulation */
+  return 0xffffffff;
+#endif
+}
+
+/**
+ * Function definition: ixFeatureCtrlDeviceRead
+ */
+IxFeatureCtrlDeviceId
+ixFeatureCtrlDeviceRead ()
+{
+  return ((ixFeatureCtrlProductIdRead() >> IX_FEATURE_CTRL_DEVICE_TYPE_OFFSET) 
+             & IX_FEATURE_CTRL_DEVICE_TYPE_MASK);
+} /* End function ixFeatureCtrlDeviceRead */
+
+
+/**
+ * Function definition: ixFeatureCtrlSwConfigurationCheck
+ */
+IX_STATUS
+ixFeatureCtrlSwConfigurationCheck (IxFeatureCtrlSwConfig swConfigType)
+{
+  if (swConfigType >= IX_FEATURECTRL_SWCONFIG_MAX)  
+  {
+     ixOsalLog(IX_OSAL_LOG_LVL_WARNING, 
+               IX_OSAL_LOG_DEV_STDOUT,
+               "FeatureCtrl: Invalid software configuraiton input.\n",
+               0, 0, 0, 0, 0, 0);  
+
+     return IX_FEATURE_CTRL_SWCONFIG_DISABLED;
+  }
+
+  /* The function will only initialize once. */
+  ixFeatureCtrlSwConfigurationInit();
+  
+  /* Check and return software configuration */
+  return  ((swConfiguration[(UINT32)swConfigType] == TRUE) ? IX_FEATURE_CTRL_SWCONFIG_ENABLED: IX_FEATURE_CTRL_SWCONFIG_DISABLED);
+}
+
+/**
+ * Function definition: ixFeatureCtrlSwConfigurationWrite
+ */
+void
+ixFeatureCtrlSwConfigurationWrite (IxFeatureCtrlSwConfig swConfigType, BOOL enabled)
+{
+  if (swConfigType >= IX_FEATURECTRL_SWCONFIG_MAX)  
+  {
+     ixOsalLog(IX_OSAL_LOG_LVL_WARNING, 
+               IX_OSAL_LOG_DEV_STDOUT,
+               "FeatureCtrl: Invalid software configuraiton input.\n",
+               0, 0, 0, 0, 0, 0);  
+
+     return;
+  }
+
+  /* The function will only initialize once. */
+  ixFeatureCtrlSwConfigurationInit();
+  
+  /* Write software configuration */
+  swConfiguration[(UINT32)swConfigType]=enabled ;
+}
+
+/**
+ * Function definition: ixFeatureCtrlIxp400SwVersionShow
+ */
+void
+ixFeatureCtrlIxp400SwVersionShow (void)
+{
+    printf ("\nIXP400 Software Release %s %s\n\n", IX_VERSION_ID, IX_VERSION_INTERNAL_ID);
+
+}
+
+/**
+ * Function definition: ixFeatureCtrlSoftwareBuildGet
+ */
+IxFeatureCtrlBuildDevice
+ixFeatureCtrlSoftwareBuildGet (void)
+{
+    #ifdef __ixp42X
+    return IX_FEATURE_CTRL_SW_BUILD_IXP42X;
+    #else
+    return IX_FEATURE_CTRL_SW_BUILD_IXP46X;
+    #endif
+}

cpu/ixp/npe/IxNpeDl.c

@@ -0,0 +1,972 @@
+/**
+ * @file IxNpeDl.c
+ *
+ * @author Intel Corporation
+ * @date 08 January 2002
+ *
+ * @brief This file contains the implementation of the public API for the
+ *        IXP425 NPE Downloader component
+ *
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+/*
+ * Put the system defined include files required
+ */
+
+/*
+ * Put the user defined include files required
+ */
+#include "IxNpeDl.h"
+#include "IxNpeDlImageMgr_p.h"
+#include "IxNpeDlNpeMgr_p.h"
+#include "IxNpeDlMacros_p.h"
+#include "IxFeatureCtrl.h"
+#include "IxOsal.h"
+/*
+ * #defines used in this file
+ */
+ #define IMAGEID_MAJOR_NUMBER_DEFAULT 0
+ #define IMAGEID_MINOR_NUMBER_DEFAULT 0
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+typedef struct
+{
+    BOOL validImage;
+    IxNpeDlImageId imageId;
+} IxNpeDlNpeState;
+
+/* module statistics counters */
+typedef struct
+{
+    UINT32 attemptedDownloads;
+    UINT32 successfulDownloads;
+    UINT32 criticalFailDownloads;
+} IxNpeDlStats;
+
+/*
+ * Variable declarations global to this file only.  Externs are followed
+ * by static variables.
+ */
+static IxNpeDlNpeState ixNpeDlNpeState[IX_NPEDL_NPEID_MAX] =
+{
+    {FALSE, {IX_NPEDL_NPEID_MAX, 0, 0, 0}},
+    {FALSE, {IX_NPEDL_NPEID_MAX, 0, 0, 0}},
+    {FALSE, {IX_NPEDL_NPEID_MAX, 0, 0, 0}}
+};
+
+static IxNpeDlStats ixNpeDlStats;
+
+/*
+ * Software guard to prevent NPE from being started multiple times.
+ */
+static BOOL ixNpeDlNpeStarted[IX_NPEDL_NPEID_MAX] ={FALSE, FALSE, FALSE} ;
+
+
+/*
+ * static function prototypes.
+ */
+PRIVATE IX_STATUS
+ixNpeDlNpeInitAndStartInternal (UINT32 *imageLibrary, UINT32 imageId);
+
+/*
+ * Function definition: ixNpeDlMicrocodeImageLibraryOverride
+ */
+PUBLIC IX_STATUS
+ixNpeDlMicrocodeImageLibraryOverride (UINT32 *clientImageLibrary)
+{
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Entering ixNpeDlMicrocodeImageLibraryOverride\n");
+
+    if (clientImageLibrary == NULL)
+    {
+        status = IX_NPEDL_PARAM_ERR;
+        IX_NPEDL_ERROR_REPORT ("ixNpeDlMicrocodeImageLibraryOverride - "
+                               "invalid parameter\n");
+    }
+    else
+    {
+        status = ixNpeDlImageMgrMicrocodeImageLibraryOverride (clientImageLibrary);
+        if (status != IX_SUCCESS)
+        {
+            status = IX_FAIL;
+        }
+    } /* end of if-else(clientImageLibrary) */
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Exiting ixNpeDlMicrocodeImageLibraryOverride : "
+                     "status = %d\n", status);
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlImageDownload
+ */
+PUBLIC IX_STATUS
+ixNpeDlImageDownload (IxNpeDlImageId *imageIdPtr,
+                      BOOL verify)
+{
+    UINT32        imageSize;
+    UINT32       *imageCodePtr  = NULL;
+    IX_STATUS     status;
+    IxNpeDlNpeId  npeId           = imageIdPtr->npeId;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Entering ixNpeDlImageDownload\n");
+
+    ixNpeDlStats.attemptedDownloads++;
+
+    /* Check input parameters */
+    if ((npeId >= IX_NPEDL_NPEID_MAX) || (npeId < 0))
+    {
+        status = IX_NPEDL_PARAM_ERR;
+        IX_NPEDL_ERROR_REPORT ("ixNpeDlImageDownload - invalid parameter\n");
+    }
+    else
+    {
+        /* Ensure initialisation has been completed */
+        ixNpeDlNpeMgrInit();
+
+	/* If not IXP42X A0 stepping, proceed to check for existence of npe's */
+	if ((IX_FEATURE_CTRL_SILICON_TYPE_A0 !=
+	     (ixFeatureCtrlProductIdRead() & IX_FEATURE_CTRL_SILICON_STEPPING_MASK))
+	    || (IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X != ixFeatureCtrlDeviceRead ()))
+	{
+            if (npeId == IX_NPEDL_NPEID_NPEA)
+            {
+                if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEA) ==
+                    IX_FEATURE_CTRL_COMPONENT_DISABLED)
+                {
+                    IX_NPEDL_WARNING_REPORT("Warning: the NPE A component you specified does"
+                                            " not exist\n");
+                    return IX_SUCCESS;
+                }
+            } /* end of if(npeId) */
+            else if (npeId == IX_NPEDL_NPEID_NPEB)
+            {
+                if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEB)==
+                    IX_FEATURE_CTRL_COMPONENT_DISABLED)
+                {
+                    IX_NPEDL_WARNING_REPORT("Warning: the NPE B component you specified"
+                                            " does not exist\n");
+                    return IX_SUCCESS;
+                }
+            } /* end of elseif(npeId) */
+            else if (npeId == IX_NPEDL_NPEID_NPEC)
+            {
+                if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEC)==
+                    IX_FEATURE_CTRL_COMPONENT_DISABLED)
+                {
+                    IX_NPEDL_WARNING_REPORT("Warning: the NPE C component you specified"
+                                            " does not exist\n");
+                    return IX_SUCCESS;
+                }
+            } /* end of elseif(npeId) */
+        } /* end of if(IX_FEATURE_CTRL_SILICON_TYPE_B0) */ /*End of Silicon Type Check*/
+
+        /* stop and reset the NPE */
+        if (IX_SUCCESS != ixNpeDlNpeStopAndReset (npeId))
+        {
+            IX_NPEDL_ERROR_REPORT ("Failed to stop and reset NPE\n");
+            return IX_FAIL;
+        }
+
+        /* Locate image */
+        status = ixNpeDlImageMgrImageLocate (imageIdPtr, &imageCodePtr,
+                                             &imageSize);
+        if (IX_SUCCESS == status)
+        {
+            /*
+             * If download was successful, store image Id in list of
+             * currently loaded images. If a critical error occured
+             * during download, record that the NPE has an invalid image
+             */
+            status = ixNpeDlNpeMgrImageLoad (npeId, imageCodePtr,
+                                             verify);
+            if (IX_SUCCESS == status)
+            {
+                ixNpeDlNpeState[npeId].imageId = *imageIdPtr;
+                ixNpeDlNpeState[npeId].validImage = TRUE;
+                ixNpeDlStats.successfulDownloads++;
+
+                status =  ixNpeDlNpeExecutionStart (npeId);
+            }
+            else if ((status == IX_NPEDL_CRITICAL_NPE_ERR) ||
+                     (status == IX_NPEDL_CRITICAL_MICROCODE_ERR))
+            {
+                ixNpeDlNpeState[npeId].imageId = *imageIdPtr;
+                ixNpeDlNpeState[npeId].validImage = FALSE;
+                ixNpeDlStats.criticalFailDownloads++;
+            }
+        } /* end of if(IX_SUCCESS) */ /* condition: image located successfully in microcode image */
+    } /* end of if-else(npeId) */ /* condition: parameter checks ok */
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Exiting ixNpeDlImageDownload : status = %d\n", status);
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlAvailableImagesCountGet
+ */
+PUBLIC IX_STATUS
+ixNpeDlAvailableImagesCountGet (UINT32 *numImagesPtr)
+{
+    IX_STATUS status;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Entering ixNpeDlAvailableImagesCountGet\n");
+
+    /* Check input parameters */
+    if (numImagesPtr == NULL)
+    {
+        status = IX_NPEDL_PARAM_ERR;
+        IX_NPEDL_ERROR_REPORT ("ixNpeDlAvailableImagesCountGet - "
+                               "invalid parameter\n");
+    }
+    else
+    {
+        /*
+         * Use ImageMgr module to get no. of images listed in Image Library Header.
+         * If NULL is passed as imageListPtr parameter to following function,
+         * it will only fill number of images into numImagesPtr
+         */
+        status = ixNpeDlImageMgrImageListExtract (NULL, numImagesPtr);
+    } /* end of if-else(numImagesPtr) */
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Exiting ixNpeDlAvailableImagesCountGet : "
+                     "status = %d\n", status);
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlAvailableImagesListGet
+ */
+PUBLIC IX_STATUS
+ixNpeDlAvailableImagesListGet (IxNpeDlImageId *imageIdListPtr,
+                               UINT32 *listSizePtr)
+{
+    IX_STATUS status;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Entering ixNpeDlAvailableImagesListGet\n");
+
+    /* Check input parameters */
+    if ((imageIdListPtr == NULL) || (listSizePtr == NULL))
+    {
+        status = IX_NPEDL_PARAM_ERR;
+        IX_NPEDL_ERROR_REPORT ("ixNpeDlAvailableImagesListGet - "
+                               "invalid parameter\n");
+    }
+    else
+    {
+        /* Call ImageMgr to get list of images listed in Image Library Header */
+        status = ixNpeDlImageMgrImageListExtract (imageIdListPtr,
+                                                  listSizePtr);
+    } /* end of if-else(imageIdListPtr) */
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Exiting ixNpeDlAvailableImagesListGet : status = %d\n",
+                     status);
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlLoadedImageGet
+ */
+PUBLIC IX_STATUS
+ixNpeDlLoadedImageGet (IxNpeDlNpeId npeId,
+                       IxNpeDlImageId *imageIdPtr)
+{
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Entering ixNpeDlLoadedImageGet\n");
+
+    /* Check input parameters */
+    if ((npeId >= IX_NPEDL_NPEID_MAX) || (npeId < 0) || (imageIdPtr == NULL))
+    {
+        status = IX_NPEDL_PARAM_ERR;
+        IX_NPEDL_ERROR_REPORT ("ixNpeDlLoadedImageGet - invalid parameter\n");
+    }
+    else
+    {
+
+         /* If not IXP42X A0 stepping, proceed to check for existence of npe's */
+         if ((IX_FEATURE_CTRL_SILICON_TYPE_A0 !=
+	      (ixFeatureCtrlProductIdRead() & IX_FEATURE_CTRL_SILICON_STEPPING_MASK))
+	     || (IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X != ixFeatureCtrlDeviceRead ()))
+        {
+            if (npeId == IX_NPEDL_NPEID_NPEA &&
+                (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEA) ==
+                 IX_FEATURE_CTRL_COMPONENT_DISABLED))
+            {
+                IX_NPEDL_WARNING_REPORT("Warning: the NPE A component you specified does"
+                                        " not exist\n");
+                return IX_SUCCESS;
+            } /* end of if(npeId) */
+
+            if (npeId == IX_NPEDL_NPEID_NPEB &&
+                (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEB) ==
+                 IX_FEATURE_CTRL_COMPONENT_DISABLED))
+            {
+                IX_NPEDL_WARNING_REPORT("Warning: the NPE B component you specified does"
+                                        " not exist\n");
+                return IX_SUCCESS;
+            } /* end of if(npeId) */
+
+            if (npeId == IX_NPEDL_NPEID_NPEC &&
+                (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEC) ==
+                 IX_FEATURE_CTRL_COMPONENT_DISABLED))
+            {
+                IX_NPEDL_WARNING_REPORT("Warning: the NPE C component you specified does"
+                                        " not exist\n");
+                return IX_SUCCESS;
+            } /* end of if(npeId) */
+        } /* end of if not IXP42x-A0 silicon */
+
+        if (ixNpeDlNpeState[npeId].validImage)
+        {
+            /* use npeId to get imageId from list of currently loaded
+               images */
+            *imageIdPtr = ixNpeDlNpeState[npeId].imageId;
+        }
+        else
+        {
+            status = IX_FAIL;
+        } /* end of if-else(ixNpeDlNpeState) */
+    } /* end of if-else(npeId) */
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Exiting ixNpeDlLoadedImageGet : status = %d\n",
+                     status);
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlLatestImageGet
+ */
+PUBLIC IX_STATUS
+ixNpeDlLatestImageGet (
+    IxNpeDlNpeId npeId,
+    IxNpeDlFunctionalityId functionalityId,
+    IxNpeDlImageId *imageIdPtr)
+{
+    IX_STATUS status;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Entering ixNpeDlLatestImageGet\n");
+
+    /* Check input parameters */
+    if ((npeId >= IX_NPEDL_NPEID_MAX) ||
+        (npeId < 0) ||
+        (imageIdPtr == NULL))
+    {
+        status = IX_NPEDL_PARAM_ERR;
+        IX_NPEDL_ERROR_REPORT ("ixNpeDlLatestImageGet - "
+                               "invalid parameter\n");
+    } /* end of if(npeId) */
+    else
+    {
+
+	/* If not IXP42X A0 stepping, proceed to check for existence of npe's */
+	if ((IX_FEATURE_CTRL_SILICON_TYPE_A0 !=
+	     (ixFeatureCtrlProductIdRead() & IX_FEATURE_CTRL_SILICON_STEPPING_MASK))
+	    || (IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X != ixFeatureCtrlDeviceRead ()))
+        {
+            if (npeId == IX_NPEDL_NPEID_NPEA &&
+                (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEA) ==
+                 IX_FEATURE_CTRL_COMPONENT_DISABLED))
+            {
+                IX_NPEDL_WARNING_REPORT("Warning: the NPE A component you specified does"
+                                        " not exist\n");
+                return IX_SUCCESS;
+            } /* end of if(npeId) */
+
+            if (npeId == IX_NPEDL_NPEID_NPEB &&
+                (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEB) ==
+                 IX_FEATURE_CTRL_COMPONENT_DISABLED))
+            {
+                IX_NPEDL_WARNING_REPORT("Warning: the NPE B component you specified does"
+                                        " not exist\n");
+                return IX_SUCCESS;
+            } /* end of if(npeId) */
+
+            if (npeId == IX_NPEDL_NPEID_NPEC &&
+                (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEC) ==
+                 IX_FEATURE_CTRL_COMPONENT_DISABLED))
+            {
+                IX_NPEDL_WARNING_REPORT("Warning: the NPE C component you specified does"
+                                        " not exist\n");
+                return IX_SUCCESS;
+            } /* end of if(npeId) */
+        } /* end of if not IXP42x-A0 silicon */
+
+        imageIdPtr->npeId = npeId;
+        imageIdPtr->functionalityId = functionalityId;
+        imageIdPtr->major = IMAGEID_MAJOR_NUMBER_DEFAULT;
+        imageIdPtr->minor = IMAGEID_MINOR_NUMBER_DEFAULT;
+        /* Call ImageMgr to get list of images listed in Image Library Header */
+        status = ixNpeDlImageMgrLatestImageExtract(imageIdPtr);
+    } /* end of if-else(npeId) */
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Exiting ixNpeDlLatestImageGet : status = %d\n",
+                     status);
+
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlNpeStopAndReset
+ */
+PUBLIC IX_STATUS
+ixNpeDlNpeStopAndReset (IxNpeDlNpeId npeId)
+{
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Entering ixNpeDlNpeStopAndReset\n");
+
+    /* Ensure initialisation has been completed */
+    ixNpeDlNpeMgrInit();
+
+    /* If not IXP42X A0 stepping, proceed to check for existence of npe's */
+    if ((IX_FEATURE_CTRL_SILICON_TYPE_A0 !=
+	 (ixFeatureCtrlProductIdRead() & IX_FEATURE_CTRL_SILICON_STEPPING_MASK))
+	|| (IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X != ixFeatureCtrlDeviceRead ()))
+    {
+        /*
+         * Check whether NPE is present
+         */
+        if (IX_NPEDL_NPEID_NPEA == npeId)
+        {
+            /* Check whether NPE A is present */
+            if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEA)==
+                IX_FEATURE_CTRL_COMPONENT_DISABLED)
+            {
+                /* NPE A does not present */
+                IX_NPEDL_WARNING_REPORT ("ixNpeDlNpeStopAndReset - Warning:NPEA does not present.\n");
+                return IX_SUCCESS;
+            }
+        } /* end of if(IX_NPEDL_NPEID_NPEA) */
+        else if (IX_NPEDL_NPEID_NPEB == npeId)
+        {
+            /* Check whether NPE B is present */
+            if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEB)==
+                IX_FEATURE_CTRL_COMPONENT_DISABLED)
+            {
+                /* NPE B does not present */
+                IX_NPEDL_WARNING_REPORT ("ixNpeDlNpeStopAndReset - Warning:NPEB does not present.\n");
+                return IX_SUCCESS;
+            }
+        } /* end of elseif(IX_NPEDL_NPEID_NPEB) */
+        else if (IX_NPEDL_NPEID_NPEC == npeId)
+        {
+            /* Check whether NPE C is present */
+            if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEC)==
+                IX_FEATURE_CTRL_COMPONENT_DISABLED)
+            {
+                /* NPE C does not present */
+                IX_NPEDL_WARNING_REPORT ("ixNpeDlNpeStopAndReset - Warning:NPEC does not present.\n");
+                return IX_SUCCESS;
+            }
+        } /* end of elseif(IX_NPEDL_NPEID_NPEC) */
+        else
+        {
+            /* Invalid NPE ID */
+            IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeStopAndReset - invalid Npe ID\n");
+            status = IX_NPEDL_PARAM_ERR;
+        } /* end of if-else(IX_NPEDL_NPEID_NPEC) */
+    } /* end of if not IXP42x-A0 Silicon */
+
+    if (status == IX_SUCCESS)
+    {
+        /* call NpeMgr function to stop the NPE */
+        status = ixNpeDlNpeMgrNpeStop (npeId);
+        if (status == IX_SUCCESS)
+        {
+            /* call NpeMgr function to reset the NPE */
+            status = ixNpeDlNpeMgrNpeReset (npeId);
+        }
+    } /* end of if(status) */
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Exiting ixNpeDlNpeStopAndReset : status = %d\n", status);
+
+    if (IX_SUCCESS == status)
+    {
+        /* Indicate NPE has been stopped */
+        ixNpeDlNpeStarted[npeId] = FALSE ;
+    }
+
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlNpeExecutionStart
+ */
+PUBLIC IX_STATUS
+ixNpeDlNpeExecutionStart (IxNpeDlNpeId npeId)
+{
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Entering ixNpeDlNpeExecutionStart\n");
+
+    /* If not IXP42X A0 stepping, proceed to check for existence of npe's */
+    if ((IX_FEATURE_CTRL_SILICON_TYPE_A0 !=
+	 (ixFeatureCtrlProductIdRead() & IX_FEATURE_CTRL_SILICON_STEPPING_MASK))
+	|| (IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X != ixFeatureCtrlDeviceRead ()))
+    {
+        /*
+         * Check whether NPE is present
+         */
+        if (IX_NPEDL_NPEID_NPEA == npeId)
+        {
+            /* Check whether NPE A is present */
+            if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEA)==
+                IX_FEATURE_CTRL_COMPONENT_DISABLED)
+            {
+                /* NPE A does not present */
+                IX_NPEDL_WARNING_REPORT ("ixNpeDlNpeExecutionStart - Warning:NPEA does not present.\n");
+                return IX_SUCCESS;
+            }
+        } /* end of if(IX_NPEDL_NPEID_NPEA) */
+        else if (IX_NPEDL_NPEID_NPEB == npeId)
+        {
+            /* Check whether NPE B is present */
+            if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEB)==
+                IX_FEATURE_CTRL_COMPONENT_DISABLED)
+            {
+                /* NPE B does not present */
+                IX_NPEDL_WARNING_REPORT ("ixNpeDlNpeExecutionStart - Warning:NPEB does not present.\n");
+                return IX_SUCCESS;
+            }
+        } /* end of elseif(IX_NPEDL_NPEID_NPEB) */
+        else if (IX_NPEDL_NPEID_NPEC ==  npeId)
+        {
+            /* Check whether NPE C is present */
+            if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEC)==
+                IX_FEATURE_CTRL_COMPONENT_DISABLED)
+            {
+                /* NPE C does not present */
+                IX_NPEDL_WARNING_REPORT ("ixNpeDlNpeExecutionStart - Warning:NPEC does not present.\n");
+                return IX_SUCCESS;
+            }
+        } /* end of elseif(IX_NPEDL_NPEID_NPEC) */
+        else
+        {
+            /* Invalid NPE ID */
+            IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeExecutionStart - invalid Npe ID\n");
+            return IX_NPEDL_PARAM_ERR;
+        } /* end of if-else(IX_NPEDL_NPEID_NPEC) */
+    } /* end of if not IXP42x-A0 Silicon */
+
+    if (TRUE == ixNpeDlNpeStarted[npeId])
+    {
+        /* NPE has been started. */
+        return IX_SUCCESS ;
+    }
+
+    /* Ensure initialisation has been completed */
+    ixNpeDlNpeMgrInit();
+
+    /* call NpeMgr function to start the NPE */
+    status = ixNpeDlNpeMgrNpeStart (npeId);
+
+    if (IX_SUCCESS == status)
+    {
+        /* Indicate NPE has started */
+        ixNpeDlNpeStarted[npeId] = TRUE ;
+    }
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Exiting ixNpeDlNpeExecutionStart : status = %d\n",
+                     status);
+
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlNpeExecutionStop
+ */
+PUBLIC IX_STATUS
+ixNpeDlNpeExecutionStop (IxNpeDlNpeId npeId)
+{
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Entering ixNpeDlNpeExecutionStop\n");
+
+    /* Ensure initialisation has been completed */
+    ixNpeDlNpeMgrInit();
+
+    /* If not IXP42X A0 stepping, proceed to check for existence of npe's */
+    if ((IX_FEATURE_CTRL_SILICON_TYPE_A0 !=
+	 (ixFeatureCtrlProductIdRead() & IX_FEATURE_CTRL_SILICON_STEPPING_MASK))
+	|| (IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X != ixFeatureCtrlDeviceRead ()))
+    {
+        /*
+         * Check whether NPE is present
+         */
+        if (IX_NPEDL_NPEID_NPEA == npeId)
+        {
+            /* Check whether NPE A is present */
+            if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEA)==
+                IX_FEATURE_CTRL_COMPONENT_DISABLED)
+            {
+                /* NPE A does not present */
+                IX_NPEDL_WARNING_REPORT ("ixNpeDlNpeExecutionStop - Warning:NPEA does not present.\n");
+                return IX_SUCCESS;
+            }
+        } /* end of if(IX_NPEDL_NPEID_NPEA) */
+        else if (IX_NPEDL_NPEID_NPEB == npeId)
+        {
+            /* Check whether NPE B is present */
+            if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEB)==
+                IX_FEATURE_CTRL_COMPONENT_DISABLED)
+            {
+                /* NPE B does not present */
+                IX_NPEDL_WARNING_REPORT ("ixNpeDlNpeExecutionStop - Warning:NPEB does not present.\n");
+                return IX_SUCCESS;
+            }
+        } /* end of elseif(IX_NPEDL_NPEID_NPEB) */
+        else if (IX_NPEDL_NPEID_NPEC == npeId)
+        {
+            /* Check whether NPE C is present */
+            if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEC)==
+                IX_FEATURE_CTRL_COMPONENT_DISABLED)
+            {
+                /* NPE C does not present */
+                IX_NPEDL_WARNING_REPORT ("ixNpeDlNpeExecutionStop - Warning:NPEC does not present.\n");
+                return IX_SUCCESS;
+            }
+        } /* end of elseif(IX_NPEDL_NPEID_NPEC) */
+        else
+        {
+            /* Invalid NPE ID */
+            IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeExecutionStop - invalid Npe ID\n");
+            status = IX_NPEDL_PARAM_ERR;
+        } /* end of if-else(IX_NPEDL_NPEID_NPEC) */
+    } /* end of if not IXP42X-AO Silicon */
+
+    if (status == IX_SUCCESS)
+    {
+        /* call NpeMgr function to stop the NPE */
+        status = ixNpeDlNpeMgrNpeStop (npeId);
+    }
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Exiting ixNpeDlNpeExecutionStop : status = %d\n",
+                     status);
+
+    if (IX_SUCCESS == status)
+    {
+        /* Indicate NPE has been stopped */
+        ixNpeDlNpeStarted[npeId] = FALSE ;
+    }
+
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlUnload
+ */
+PUBLIC IX_STATUS
+ixNpeDlUnload (void)
+{
+    IX_STATUS status;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Entering ixNpeDlUnload\n");
+
+    status = ixNpeDlNpeMgrUninit();
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Exiting ixNpeDlUnload : status = %d\n",
+                     status);
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlStatsShow
+ */
+PUBLIC void
+ixNpeDlStatsShow (void)
+{
+    ixOsalLog (IX_OSAL_LOG_LVL_USER,
+               IX_OSAL_LOG_DEV_STDOUT,
+               "\nixNpeDlStatsShow:\n"
+               "\tDownloads Attempted by user: %u\n"
+               "\tSuccessful Downloads: %u\n"
+               "\tFailed Downloads (due to Critical Error): %u\n\n",
+               ixNpeDlStats.attemptedDownloads,
+               ixNpeDlStats.successfulDownloads,
+               ixNpeDlStats.criticalFailDownloads,
+               0,0,0);
+
+    ixNpeDlImageMgrStatsShow ();
+    ixNpeDlNpeMgrStatsShow ();
+}
+
+/*
+ * Function definition: ixNpeDlStatsReset
+ */
+PUBLIC void
+ixNpeDlStatsReset (void)
+{
+    ixNpeDlStats.attemptedDownloads = 0;
+    ixNpeDlStats.successfulDownloads = 0;
+    ixNpeDlStats.criticalFailDownloads = 0;
+
+    ixNpeDlImageMgrStatsReset ();
+    ixNpeDlNpeMgrStatsReset ();
+}
+
+/*
+ * Function definition: ixNpeDlNpeInitAndStartInternal
+ */
+PRIVATE IX_STATUS
+ixNpeDlNpeInitAndStartInternal (UINT32 *imageLibrary,
+                                UINT32 imageId)
+{
+    UINT32        imageSize;
+    UINT32       *imageCodePtr  = NULL;
+    IX_STATUS     status;
+    IxNpeDlNpeId  npeId = IX_NPEDL_NPEID_FROM_IMAGEID_GET(imageId);
+    IxFeatureCtrlDeviceId deviceId = IX_NPEDL_DEVICEID_FROM_IMAGEID_GET(imageId);
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Entering ixNpeDlNpeInitAndStartInternal\n");
+
+    ixNpeDlStats.attemptedDownloads++;
+
+    /* Check input parameter device correctness */
+    if ((deviceId >= IX_FEATURE_CTRL_DEVICE_TYPE_MAX) ||
+        (deviceId < IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X))
+    {
+        status = IX_NPEDL_PARAM_ERR;
+        IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeInitAndStartInternal - "
+                               "invalid parameter\n");
+    } /* End valid device id checking */
+
+    /* Check input parameters */
+    else if ((npeId >= IX_NPEDL_NPEID_MAX) || (npeId < 0))
+    {
+        status = IX_NPEDL_PARAM_ERR;
+        IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeInitAndStartInternal - "
+                               "invalid parameter\n");
+    }
+
+    else
+    {
+        /* Ensure initialisation has been completed */
+        ixNpeDlNpeMgrInit();
+
+        /* Checking if image being loaded is meant for device that is running.
+         * Image is forward compatible. i.e Image built for IXP42X should run
+         * on IXP46X but not vice versa.*/
+        if (deviceId > (ixFeatureCtrlDeviceRead() & IX_FEATURE_CTRL_DEVICE_TYPE_MASK))
+        {
+            IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeInitAndStartInternal - "
+                                   "Device type mismatch. NPE Image not "
+                                   "meant for device in use \n");
+            return IX_NPEDL_DEVICE_ERR;
+        }/* if statement - matching image device and current device */
+
+	/* If not IXP42X A0 stepping, proceed to check for existence of npe's */
+	if ((IX_FEATURE_CTRL_SILICON_TYPE_A0 !=
+	     (ixFeatureCtrlProductIdRead() & IX_FEATURE_CTRL_SILICON_STEPPING_MASK))
+	    || (IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X != ixFeatureCtrlDeviceRead ()))
+        {
+            if (npeId == IX_NPEDL_NPEID_NPEA)
+            {
+                if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEA) ==
+                    IX_FEATURE_CTRL_COMPONENT_DISABLED)
+                {
+                    IX_NPEDL_WARNING_REPORT("Warning: the NPE A component you specified does"
+                                            " not exist\n");
+                    return IX_SUCCESS;
+                }
+            } /* end of if(npeId) */
+            else if (npeId == IX_NPEDL_NPEID_NPEB)
+            {
+                if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEB)==
+                    IX_FEATURE_CTRL_COMPONENT_DISABLED)
+                {
+                    IX_NPEDL_WARNING_REPORT("Warning: the NPE B component you specified"
+                                            " does not exist\n");
+                    return IX_SUCCESS;
+                }
+            } /* end of elseif(npeId) */
+            else if (npeId == IX_NPEDL_NPEID_NPEC)
+            {
+                if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_NPEC)==
+                    IX_FEATURE_CTRL_COMPONENT_DISABLED)
+                {
+                    IX_NPEDL_WARNING_REPORT("Warning: the NPE C component you specified"
+                                            " does not exist\n");
+                    return IX_SUCCESS;
+                }
+            } /* end of elseif(npeId) */
+        } /* end of if not IXP42X-A0 Silicon */
+
+        /* stop and reset the NPE */
+        status = ixNpeDlNpeStopAndReset (npeId);
+        if (IX_SUCCESS != status)
+        {
+            IX_NPEDL_ERROR_REPORT ("Failed to stop and reset NPE\n");
+            return status;
+        }
+
+        /* Locate image */
+        status = ixNpeDlImageMgrImageFind (imageLibrary, imageId,
+                                           &imageCodePtr, &imageSize);
+        if (IX_SUCCESS == status)
+        {
+            /*
+             * If download was successful, store image Id in list of
+             * currently loaded images. If a critical error occured
+             * during download, record that the NPE has an invalid image
+             */
+            status = ixNpeDlNpeMgrImageLoad (npeId, imageCodePtr, TRUE);
+            if (IX_SUCCESS == status)
+            {
+                ixNpeDlNpeState[npeId].validImage = TRUE;
+                ixNpeDlStats.successfulDownloads++;
+
+                status = ixNpeDlNpeExecutionStart (npeId);
+            }
+            else if ((status == IX_NPEDL_CRITICAL_NPE_ERR) ||
+                     (status == IX_NPEDL_CRITICAL_MICROCODE_ERR))
+            {
+                ixNpeDlNpeState[npeId].validImage = FALSE;
+                ixNpeDlStats.criticalFailDownloads++;
+            }
+
+            /* NOTE - The following section of code is here to support
+             * a deprecated function ixNpeDlLoadedImageGet().  When that
+             * function is removed from the API, this code should be revised.
+             */
+            ixNpeDlNpeState[npeId].imageId.npeId = npeId;
+            ixNpeDlNpeState[npeId].imageId.functionalityId =
+                IX_NPEDL_FUNCTIONID_FROM_IMAGEID_GET(imageId);
+            ixNpeDlNpeState[npeId].imageId.major =
+                IX_NPEDL_MAJOR_FROM_IMAGEID_GET(imageId);
+            ixNpeDlNpeState[npeId].imageId.minor =
+                IX_NPEDL_MINOR_FROM_IMAGEID_GET(imageId);
+        } /* end of if(IX_SUCCESS) */ /* condition: image located successfully in microcode image */
+    } /* end of if-else(npeId-deviceId) */ /* condition: parameter checks ok */
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+                     "Exiting ixNpeDlNpeInitAndStartInternal : "
+                     "status = %d\n", status);
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlCustomImageNpeInitAndStart
+ */
+PUBLIC IX_STATUS
+ixNpeDlCustomImageNpeInitAndStart (UINT32 *imageLibrary,
+                                   UINT32 imageId)
+{
+    IX_STATUS status;
+
+    if (imageLibrary == NULL)
+    {
+        status = IX_NPEDL_PARAM_ERR;
+        IX_NPEDL_ERROR_REPORT ("ixNpeDlCustomImageNpeInitAndStart "
+                               "- invalid parameter\n");
+    }
+    else
+    {
+        status = ixNpeDlNpeInitAndStartInternal (imageLibrary, imageId);
+    } /* end of if-else(imageLibrary) */
+
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlNpeInitAndStart
+ */
+PUBLIC IX_STATUS
+ixNpeDlNpeInitAndStart (UINT32 imageId)
+{
+    return ixNpeDlNpeInitAndStartInternal (NULL, imageId);
+}
+
+/*
+ * Function definition: ixNpeDlLoadedImageFunctionalityGet
+ */
+PUBLIC IX_STATUS
+ixNpeDlLoadedImageFunctionalityGet (IxNpeDlNpeId npeId,
+                                    UINT8 *functionalityId)
+{
+    /* Check input parameters */
+    if ((npeId >= IX_NPEDL_NPEID_MAX) || (npeId < 0))
+    {
+        IX_NPEDL_ERROR_REPORT ("ixNpeDlLoadedImageFunctionalityGet "
+                               "- invalid parameter\n");
+        return IX_NPEDL_PARAM_ERR;
+    }
+    if (functionalityId == NULL)
+    {
+        IX_NPEDL_ERROR_REPORT ("ixNpeDlLoadedImageFunctionalityGet "
+                               "- invalid parameter\n");
+        return IX_NPEDL_PARAM_ERR;
+    }
+
+    if (ixNpeDlNpeState[npeId].validImage)
+    {
+        *functionalityId = ixNpeDlNpeState[npeId].imageId.functionalityId;
+        return IX_SUCCESS;
+    }
+    else
+    {
+        return IX_FAIL;
+    }
+}

cpu/ixp/npe/IxNpeDlImageMgr.c

@@ -0,0 +1,675 @@
+/**
+ * @file IxNpeDlImageMgr.c
+ *
+ * @author Intel Corporation
+ * @date 09 January 2002
+ *
+ * @brief This file contains the implementation of the private API for the 
+ *        IXP425 NPE Downloader ImageMgr module
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+
+/*
+ * Put the system defined include files required.
+ */
+#include "IxOsal.h"
+
+/*
+ * Put the user defined include files required.
+ */
+#include "IxNpeDlImageMgr_p.h"
+#include "IxNpeDlMacros_p.h"
+
+/*
+ * define the flag which toggles the firmare inclusion
+ */
+#define IX_NPE_MICROCODE_FIRMWARE_INCLUDED 1
+#include "IxNpeMicrocode.h"
+
+/*
+ * Indicates the start of an NPE Image, in new NPE Image Library format.
+ * 2 consecutive occurances indicates the end of the NPE Image Library
+ */
+#define NPE_IMAGE_MARKER 0xfeedf00d
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+
+/*
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+typedef struct
+{
+    UINT32 size;
+    UINT32 offset;
+    UINT32 id;
+} IxNpeDlImageMgrImageEntry;
+
+/*
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+typedef union
+{
+    IxNpeDlImageMgrImageEntry image;
+    UINT32 eohMarker;
+} IxNpeDlImageMgrHeaderEntry;
+
+/*
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+typedef struct
+{
+    UINT32 signature;
+    /* 1st entry in the header (there may be more than one) */
+    IxNpeDlImageMgrHeaderEntry entry[1];
+} IxNpeDlImageMgrImageLibraryHeader;
+
+
+/*
+ * NPE Image Header definition, used in new NPE Image Library format
+ */
+typedef struct
+{
+    UINT32 marker;
+    UINT32 id;
+    UINT32 size;
+} IxNpeDlImageMgrImageHeader;
+
+/* module statistics counters */
+typedef struct
+{
+    UINT32 invalidSignature;
+    UINT32 imageIdListOverflow;
+    UINT32 imageIdNotFound;
+} IxNpeDlImageMgrStats;
+
+
+/*
+ * Variable declarations global to this file only.  Externs are followed by
+ * static variables.
+ */
+static IxNpeDlImageMgrStats ixNpeDlImageMgrStats;
+
+/* default image */
+#ifdef IX_NPEDL_READ_MICROCODE_FROM_FILE
+static UINT32 *IxNpeMicroCodeImageLibrary = NULL;  /* Gets set to proper value at runtime */
+#else
+static UINT32 *IxNpeMicroCodeImageLibrary = (UINT32 *)IxNpeMicrocode_array;
+#endif
+
+
+/*
+ * static function prototypes.
+ */
+PRIVATE BOOL
+ixNpeDlImageMgrSignatureCheck (UINT32 *microCodeImageLibrary);
+
+PRIVATE void  
+ixNpeDlImageMgrImageIdFormat (UINT32 rawImageId, IxNpeDlImageId *imageId);
+
+PRIVATE BOOL
+ixNpeDlImageMgrImageIdCompare (IxNpeDlImageId *imageIdA, 
+				 IxNpeDlImageId *imageIdB);
+				 
+PRIVATE BOOL
+ixNpeDlImageMgrNpeFunctionIdCompare (IxNpeDlImageId *imageIdA,
+    				       IxNpeDlImageId *imageIdB);
+
+PRIVATE IX_STATUS
+ixNpeDlImageMgrImageFind_legacy (UINT32 *imageLibrary,
+                                 UINT32 imageId,
+                                 UINT32 **imagePtr,
+                                 UINT32 *imageSize);
+
+/*
+ * Function definition: ixNpeDlImageMgrMicrocodeImageLibraryOverride
+ *
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * AND/OR LEGACY API FUNCTIONS. TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+IX_STATUS
+ixNpeDlImageMgrMicrocodeImageLibraryOverride (
+    UINT32 *clientImageLibrary)
+{
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT, 
+		     "Entering ixNpeDlImageMgrMicrocodeImageLibraryOverride\n");
+
+    if (ixNpeDlImageMgrSignatureCheck (clientImageLibrary))
+    {
+	IxNpeMicroCodeImageLibrary = clientImageLibrary;
+    }
+    else
+    {
+	IX_NPEDL_ERROR_REPORT ("ixNpeDlImageMgrMicrocodeImageLibraryOverride: "
+			       "Client-supplied image has invalid signature\n");
+	status = IX_FAIL;
+    }
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT, 
+		     "Exiting ixNpeDlImageMgrMicrocodeImageLibraryOverride: status = %d\n",
+		     status);
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlImageMgrImageListExtract
+ *
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * AND/OR LEGACY API FUNCTIONS. TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+IX_STATUS
+ixNpeDlImageMgrImageListExtract (
+    IxNpeDlImageId *imageListPtr,
+    UINT32 *numImages)
+{
+    UINT32 rawImageId;
+    IxNpeDlImageId formattedImageId;
+    IX_STATUS status = IX_SUCCESS;
+    UINT32 imageCount = 0;
+    IxNpeDlImageMgrImageLibraryHeader *header;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT, 
+		     "Entering ixNpeDlImageMgrImageListExtract\n");
+
+    header = (IxNpeDlImageMgrImageLibraryHeader *) IxNpeMicroCodeImageLibrary;
+
+    if (ixNpeDlImageMgrSignatureCheck (IxNpeMicroCodeImageLibrary))
+    {
+	/* for each image entry in the image header ... */
+	while (header->entry[imageCount].eohMarker !=
+	       IX_NPEDL_IMAGEMGR_END_OF_HEADER)
+	{
+	    /*
+	     * if the image list container from calling function has capacity,
+	     * add the image id to the list 
+	     */
+	    if ((imageListPtr != NULL) && (imageCount < *numImages))
+	    {
+		rawImageId = header->entry[imageCount].image.id;
+	        ixNpeDlImageMgrImageIdFormat (rawImageId, &formattedImageId);
+		imageListPtr[imageCount] = formattedImageId;
+	    }
+	    /* imageCount reflects no. of image entries in image library header */
+	    imageCount++;  
+	}
+	
+	/*
+	 * if image list container from calling function was too small to
+	 * contain all image ids in the header, set return status to FAIL
+	 */
+	if ((imageListPtr != NULL) && (imageCount > *numImages))
+	{
+	    status = IX_FAIL;
+	    IX_NPEDL_ERROR_REPORT ("ixNpeDlImageMgrImageListExtract: "
+				   "number of Ids found exceeds list capacity\n");
+	    ixNpeDlImageMgrStats.imageIdListOverflow++;
+	}
+	/* return number of image ids found in image library header */
+	*numImages = imageCount;  
+    }
+    else
+    {
+	status = IX_FAIL;
+	IX_NPEDL_ERROR_REPORT ("ixNpeDlImageMgrImageListExtract: "
+			       "invalid signature in image\n");
+    }
+    
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT, 
+		     "Exiting ixNpeDlImageMgrImageListExtract: status = %d\n",
+		     status);
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlImageMgrImageLocate
+ *
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * AND/OR LEGACY API FUNCTIONS. TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+IX_STATUS
+ixNpeDlImageMgrImageLocate (
+    IxNpeDlImageId *imageId,
+    UINT32 **imagePtr,
+    UINT32 *imageSize)
+{
+    UINT32 imageOffset;
+    UINT32 rawImageId;
+    IxNpeDlImageId formattedImageId;
+    /* used to index image entries in image library header */
+    UINT32 imageCount = 0;   
+    IX_STATUS status = IX_FAIL;
+    IxNpeDlImageMgrImageLibraryHeader *header;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlImageMgrImageLocate\n");
+
+    header = (IxNpeDlImageMgrImageLibraryHeader *) IxNpeMicroCodeImageLibrary;
+
+    if (ixNpeDlImageMgrSignatureCheck (IxNpeMicroCodeImageLibrary))
+    {
+	/* for each image entry in the image library header ... */
+	while (header->entry[imageCount].eohMarker !=
+	       IX_NPEDL_IMAGEMGR_END_OF_HEADER)
+	{
+	    rawImageId = header->entry[imageCount].image.id;
+	    ixNpeDlImageMgrImageIdFormat (rawImageId, &formattedImageId);
+	    /* if a match for imageId is found in the image library header... */
+	    if (ixNpeDlImageMgrImageIdCompare (imageId, &formattedImageId))
+	    {
+		/*
+		 * get pointer to the image in the image library using offset from
+		 * 1st word in image library
+		 */
+		imageOffset = header->entry[imageCount].image.offset;
+		*imagePtr = &IxNpeMicroCodeImageLibrary[imageOffset];
+		/* get the image size */
+		*imageSize = header->entry[imageCount].image.size;
+		status = IX_SUCCESS;
+		break;
+	    }
+	    imageCount++;
+	}
+	if (status != IX_SUCCESS)
+	{
+	    IX_NPEDL_ERROR_REPORT ("ixNpeDlImageMgrImageLocate: "
+				   "imageId not found in image library header\n");
+	    ixNpeDlImageMgrStats.imageIdNotFound++;
+	}
+    }
+    else
+    {
+	IX_NPEDL_ERROR_REPORT ("ixNpeDlImageMgrImageLocate: "
+			       "invalid signature in image library\n");
+    }
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlImageMgrImageLocate: status = %d\n", status);
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlImageMgrLatestImageExtract
+ *
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * AND/OR LEGACY API FUNCTIONS. TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+IX_STATUS
+ixNpeDlImageMgrLatestImageExtract (IxNpeDlImageId *imageId)
+{
+    UINT32 imageCount = 0; 
+    UINT32 rawImageId;
+    IxNpeDlImageId formattedImageId;
+    IX_STATUS status = IX_FAIL;
+    IxNpeDlImageMgrImageLibraryHeader *header;
+    
+    
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlImageMgrLatestImageExtract\n");
+		     
+    header = (IxNpeDlImageMgrImageLibraryHeader *) IxNpeMicroCodeImageLibrary;
+    
+    if (ixNpeDlImageMgrSignatureCheck (IxNpeMicroCodeImageLibrary))
+    {
+	/* for each image entry in the image library header ... */
+	while (header->entry[imageCount].eohMarker !=
+	       IX_NPEDL_IMAGEMGR_END_OF_HEADER)
+	{
+	    rawImageId = header->entry[imageCount].image.id;
+	    ixNpeDlImageMgrImageIdFormat (rawImageId, &formattedImageId);
+	    /* 
+	     * if a match for the npe Id and functionality Id of the imageId is
+	     *  found in the image library header... 
+	     */
+            if(ixNpeDlImageMgrNpeFunctionIdCompare(imageId, &formattedImageId))
+            {
+                if(imageId->major <= formattedImageId.major)
+                {
+                    if(imageId->minor < formattedImageId.minor)
+                    {
+                        imageId->minor = formattedImageId.minor;
+                    }
+                    imageId->major = formattedImageId.major;
+                }
+                status = IX_SUCCESS;
+            }
+            imageCount++;
+	}
+	if (status != IX_SUCCESS)
+	{
+	    IX_NPEDL_ERROR_REPORT ("ixNpeDlImageMgrLatestImageExtract: "
+				   "imageId not found in image library header\n");
+	    ixNpeDlImageMgrStats.imageIdNotFound++;
+	}
+    }
+    else
+    {
+	IX_NPEDL_ERROR_REPORT ("ixNpeDlImageMgrLatestImageGet: "
+			       "invalid signature in image library\n");
+    }
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlImageMgrLatestImageGet: status = %d\n", status);
+    return status;
+}
+
+/*
+ * Function definition: ixNpeDlImageMgrSignatureCheck
+ *
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * AND/OR LEGACY API FUNCTIONS. TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+PRIVATE BOOL
+ixNpeDlImageMgrSignatureCheck (UINT32 *microCodeImageLibrary)
+{
+    IxNpeDlImageMgrImageLibraryHeader *header =
+	(IxNpeDlImageMgrImageLibraryHeader *) microCodeImageLibrary;
+    BOOL result = TRUE;
+
+    if (header->signature != IX_NPEDL_IMAGEMGR_SIGNATURE)
+    {
+	result = FALSE;
+	ixNpeDlImageMgrStats.invalidSignature++;
+    }
+
+    return result;
+}
+
+
+/*
+ * Function definition: ixNpeDlImageMgrImageIdFormat
+ *
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * AND/OR LEGACY API FUNCTIONS. TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+PRIVATE void
+ixNpeDlImageMgrImageIdFormat (
+    UINT32 rawImageId, 
+    IxNpeDlImageId *imageId)
+{  
+    imageId->npeId = (rawImageId >>
+				IX_NPEDL_IMAGEID_NPEID_OFFSET) &
+	IX_NPEDL_NPEIMAGE_FIELD_MASK;
+    imageId->functionalityId = (rawImageId >> 
+				  IX_NPEDL_IMAGEID_FUNCTIONID_OFFSET) &
+	IX_NPEDL_NPEIMAGE_FIELD_MASK;
+    imageId->major = (rawImageId >>
+				IX_NPEDL_IMAGEID_MAJOR_OFFSET) &
+	IX_NPEDL_NPEIMAGE_FIELD_MASK;
+    imageId->minor = (rawImageId >>
+				IX_NPEDL_IMAGEID_MINOR_OFFSET) &
+	IX_NPEDL_NPEIMAGE_FIELD_MASK;
+
+}
+
+
+/*
+ * Function definition: ixNpeDlImageMgrImageIdCompare
+ *
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * AND/OR LEGACY API FUNCTIONS. TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+PRIVATE BOOL
+ixNpeDlImageMgrImageIdCompare (
+    IxNpeDlImageId *imageIdA,
+    IxNpeDlImageId *imageIdB)
+{
+    if ((imageIdA->npeId   == imageIdB->npeId)   &&
+	(imageIdA->functionalityId == imageIdB->functionalityId) &&
+	(imageIdA->major   == imageIdB->major)   &&
+	(imageIdA->minor   == imageIdB->minor))
+    {
+	return TRUE;
+    }
+    else
+    {
+	return FALSE;
+    }
+}
+
+/*
+ * Function definition: ixNpeDlImageMgrNpeFunctionIdCompare
+ *
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * AND/OR LEGACY API FUNCTIONS. TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+PRIVATE BOOL
+ixNpeDlImageMgrNpeFunctionIdCompare (
+    IxNpeDlImageId *imageIdA,
+    IxNpeDlImageId *imageIdB)
+{
+    if ((imageIdA->npeId   == imageIdB->npeId)   &&
+	(imageIdA->functionalityId == imageIdB->functionalityId))
+    {
+	return TRUE;
+    }
+    else
+    {
+	return FALSE;
+    }
+}
+
+
+/*
+ * Function definition: ixNpeDlImageMgrStatsShow
+ */
+void
+ixNpeDlImageMgrStatsShow (void)
+{
+    ixOsalLog (IX_OSAL_LOG_LVL_USER,
+               IX_OSAL_LOG_DEV_STDOUT,
+               "\nixNpeDlImageMgrStatsShow:\n"
+               "\tInvalid Image Signatures: %u\n"
+               "\tImage Id List capacity too small: %u\n"
+               "\tImage Id not found: %u\n\n",
+               ixNpeDlImageMgrStats.invalidSignature,
+               ixNpeDlImageMgrStats.imageIdListOverflow,
+               ixNpeDlImageMgrStats.imageIdNotFound,
+               0,0,0);
+}
+
+
+/*
+ * Function definition: ixNpeDlImageMgrStatsReset
+ */
+void
+ixNpeDlImageMgrStatsReset (void)
+{
+    ixNpeDlImageMgrStats.invalidSignature = 0;
+    ixNpeDlImageMgrStats.imageIdListOverflow = 0;
+    ixNpeDlImageMgrStats.imageIdNotFound = 0;
+}
+
+
+/*
+ * Function definition: ixNpeDlImageMgrImageFind_legacy
+ *
+ * FOR BACKWARD-COMPATIBILITY WITH OLD NPE IMAGE LIBRARY FORMAT
+ * AND/OR LEGACY API FUNCTIONS. TO BE DEPRECATED IN A FUTURE RELEASE
+ */
+PRIVATE IX_STATUS
+ixNpeDlImageMgrImageFind_legacy (
+    UINT32 *imageLibrary,
+    UINT32 imageId,
+    UINT32 **imagePtr,
+    UINT32 *imageSize)
+{
+    UINT32 imageOffset;
+    /* used to index image entries in image library header */
+    UINT32 imageCount = 0;   
+    IX_STATUS status = IX_FAIL;
+    IxNpeDlImageMgrImageLibraryHeader *header;
+    BOOL imageFound = FALSE;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlImageMgrImageFind\n");
+
+
+    /* If user didn't specify a library to use, use the default
+     * one from IxNpeMicrocode.h
+     */
+    if (imageLibrary == NULL)
+    {
+	imageLibrary = IxNpeMicroCodeImageLibrary;
+    }
+    
+    if (ixNpeDlImageMgrSignatureCheck (imageLibrary))
+    {
+	header = (IxNpeDlImageMgrImageLibraryHeader *) imageLibrary;
+    
+	/* for each image entry in the image library header ... */
+	while ((header->entry[imageCount].eohMarker !=
+               IX_NPEDL_IMAGEMGR_END_OF_HEADER) && !(imageFound))
+	{
+	    /* if a match for imageId is found in the image library header... */
+	    if (imageId == header->entry[imageCount].image.id)
+	    {
+		/*
+		 * get pointer to the image in the image library using offset from
+		 * 1st word in image library
+		 */
+		imageOffset = header->entry[imageCount].image.offset;
+		*imagePtr = &imageLibrary[imageOffset];
+		/* get the image size */
+		*imageSize = header->entry[imageCount].image.size;
+		status = IX_SUCCESS;
+		imageFound = TRUE;
+	    }
+	    imageCount++;
+	}
+	if (status != IX_SUCCESS)
+	{
+	    IX_NPEDL_ERROR_REPORT ("ixNpeDlImageMgrImageFind: "
+				   "imageId not found in image library header\n");
+	    ixNpeDlImageMgrStats.imageIdNotFound++;
+	}
+    }
+    else
+    {
+	IX_NPEDL_ERROR_REPORT ("ixNpeDlImageMgrImageFind: "
+			       "invalid signature in image library\n");
+    }
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlImageMgrImageFind: status = %d\n", status);
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlImageMgrImageFind
+ */
+IX_STATUS
+ixNpeDlImageMgrImageFind (
+    UINT32 *imageLibrary,
+    UINT32 imageId,
+    UINT32 **imagePtr,
+    UINT32 *imageSize)
+{
+    IxNpeDlImageMgrImageHeader *image;
+    UINT32 offset = 0;
+
+    /* If user didn't specify a library to use, use the default
+     * one from IxNpeMicrocode.h
+     */
+    if (imageLibrary == NULL)
+    {
+#ifdef IX_NPEDL_READ_MICROCODE_FROM_FILE
+	if (ixNpeMicrocode_binaryArray == NULL)
+        {
+	    printk (KERN_ERR "ixp400.o:  ERROR, no Microcode found in memory\n");
+	    return IX_FAIL;
+	}
+	else
+	{
+	    imageLibrary = ixNpeMicrocode_binaryArray;
+	}
+#else
+	imageLibrary = IxNpeMicroCodeImageLibrary;
+#endif /* IX_NPEDL_READ_MICROCODE_FROM_FILE */
+    }
+
+    /* For backward's compatibility with previous image format */
+    if (ixNpeDlImageMgrSignatureCheck(imageLibrary))
+    {
+        return ixNpeDlImageMgrImageFind_legacy(imageLibrary,
+                                               imageId,
+                                               imagePtr,
+                                               imageSize);
+    }
+
+    while (*(imageLibrary+offset) == NPE_IMAGE_MARKER)
+    {
+        image = (IxNpeDlImageMgrImageHeader *)(imageLibrary+offset);
+        offset += sizeof(IxNpeDlImageMgrImageHeader)/sizeof(UINT32);
+        
+        if (image->id == imageId)
+        {
+            *imagePtr = imageLibrary + offset;
+            *imageSize = image->size;
+            return IX_SUCCESS;
+        }
+        /* 2 consecutive NPE_IMAGE_MARKER's indicates end of library */
+        else if (image->id == NPE_IMAGE_MARKER)
+        {
+	    IX_NPEDL_ERROR_REPORT ("ixNpeDlImageMgrImageFind: "
+				   "imageId not found in image library header\n");
+	    ixNpeDlImageMgrStats.imageIdNotFound++;
+            /* reached end of library, image not found */
+            return IX_FAIL;
+        }
+        offset += image->size;
+    }
+
+    /* If we get here, our image library may be corrupted */
+    IX_NPEDL_ERROR_REPORT ("ixNpeDlImageMgrImageFind: "
+                           "image library format may be invalid or corrupted\n");
+    return IX_FAIL;
+}
+

cpu/ixp/npe/IxNpeDlNpeMgr.c

@@ -0,0 +1,936 @@
+/**
+ * @file IxNpeDlNpeMgr.c
+ *
+ * @author Intel Corporation
+ * @date 09 January 2002
+ *
+ * @brief This file contains the implementation of the private API for the
+ *        IXP425 NPE Downloader NpeMgr module
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+
+/*
+ * Put the user defined include files required.
+ */
+
+
+/*
+ * Put the user defined include files required.
+ */
+#include "IxOsal.h"
+#include "IxNpeDl.h"
+#include "IxNpeDlNpeMgr_p.h"
+#include "IxNpeDlNpeMgrUtils_p.h"
+#include "IxNpeDlNpeMgrEcRegisters_p.h"
+#include "IxNpeDlMacros_p.h"
+#include "IxFeatureCtrl.h"
+
+/*
+ * #defines and macros used in this file.
+ */
+#define IX_NPEDL_BYTES_PER_WORD                   4
+
+/* used to read download map from version in microcode image */
+#define IX_NPEDL_BLOCK_TYPE_INSTRUCTION           0x00000000
+#define IX_NPEDL_BLOCK_TYPE_DATA                  0x00000001
+#define IX_NPEDL_BLOCK_TYPE_STATE                 0x00000002
+#define IX_NPEDL_END_OF_DOWNLOAD_MAP              0x0000000F
+
+/*
+ * masks used to extract address info from State information context
+ * register addresses as read from microcode image 
+ */
+#define IX_NPEDL_MASK_STATE_ADDR_CTXT_REG         0x0000000F
+#define IX_NPEDL_MASK_STATE_ADDR_CTXT_NUM         0x000000F0
+
+/* LSB offset of Context Number field in State-Info Context Address */
+#define IX_NPEDL_OFFSET_STATE_ADDR_CTXT_NUM       4
+
+/* size (in words) of single State Information entry (ctxt reg address|data) */
+#define IX_NPEDL_STATE_INFO_ENTRY_SIZE            2
+
+
+ #define IX_NPEDL_RESET_NPE_PARITY  0x0800
+ #define IX_NPEDL_PARITY_BIT_MASK   0x3F00FFFF
+ #define IX_NPEDL_CONFIG_CTRL_REG_MASK  0x3F3FFFFF
+
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+
+typedef struct
+{
+    UINT32 type;
+    UINT32 offset;
+} IxNpeDlNpeMgrDownloadMapBlockEntry;
+
+typedef union
+{
+    IxNpeDlNpeMgrDownloadMapBlockEntry block;
+    UINT32 eodmMarker;
+} IxNpeDlNpeMgrDownloadMapEntry;
+
+typedef struct
+{
+    /* 1st entry in the download map (there may be more than one) */
+    IxNpeDlNpeMgrDownloadMapEntry entry[1];
+} IxNpeDlNpeMgrDownloadMap;
+
+
+/* used to access an instruction or data block in a microcode image */
+typedef struct
+{
+    UINT32 npeMemAddress;
+    UINT32 size;
+    UINT32 data[1];
+} IxNpeDlNpeMgrCodeBlock;
+
+/* used to access each Context Reg entry state-information block */
+typedef struct
+{
+    UINT32 addressInfo;
+    UINT32 value;
+} IxNpeDlNpeMgrStateInfoCtxtRegEntry;
+
+/* used to access a state-information block in a microcode image */
+typedef struct
+{
+    UINT32 size;
+    IxNpeDlNpeMgrStateInfoCtxtRegEntry ctxtRegEntry[1];
+} IxNpeDlNpeMgrStateInfoBlock;
+ 
+/* used to store some useful NPE information for easy access */
+typedef struct
+{
+    UINT32 baseAddress;
+    UINT32 insMemSize;
+    UINT32 dataMemSize;
+} IxNpeDlNpeInfo;
+
+/* used to distinguish instruction and data memory operations */
+typedef enum 
+{
+  IX_NPEDL_MEM_TYPE_INSTRUCTION = 0,
+  IX_NPEDL_MEM_TYPE_DATA
+} IxNpeDlNpeMemType;
+
+/* used to hold a reset value for a particular ECS register */
+typedef struct
+{
+    UINT32 regAddr;
+    UINT32 regResetVal;
+} IxNpeDlEcsRegResetValue;
+
+/* prototype of function to write either Instruction or Data memory */
+typedef IX_STATUS (*IxNpeDlNpeMgrMemWrite) (UINT32 npeBaseAddress,
+					    UINT32 npeMemAddress,
+					    UINT32 npeMemData,
+					    BOOL verify);
+
+/* module statistics counters */
+typedef struct
+{
+    UINT32 instructionBlocksLoaded;
+    UINT32 dataBlocksLoaded;
+    UINT32 stateInfoBlocksLoaded;
+    UINT32 criticalNpeErrors;
+    UINT32 criticalMicrocodeErrors;
+    UINT32 npeStarts;
+    UINT32 npeStops;
+    UINT32 npeResets;
+} IxNpeDlNpeMgrStats;
+
+
+/*
+ * Variable declarations global to this file only.  Externs are followed by
+ * static variables.
+ */
+static IxNpeDlNpeInfo ixNpeDlNpeInfo[] =
+{
+    {
+	0,
+	IX_NPEDL_INS_MEMSIZE_WORDS_NPEA,
+	IX_NPEDL_DATA_MEMSIZE_WORDS_NPEA
+    },
+    {
+	0,
+	IX_NPEDL_INS_MEMSIZE_WORDS_NPEB,
+	IX_NPEDL_DATA_MEMSIZE_WORDS_NPEB
+    },
+    {
+	0,
+	IX_NPEDL_INS_MEMSIZE_WORDS_NPEC,
+	IX_NPEDL_DATA_MEMSIZE_WORDS_NPEC
+    }
+};
+
+/* contains Reset values for Context Store Registers  */
+static UINT32 ixNpeDlCtxtRegResetValues[] =
+{
+    IX_NPEDL_CTXT_REG_RESET_STEVT,
+    IX_NPEDL_CTXT_REG_RESET_STARTPC,
+    IX_NPEDL_CTXT_REG_RESET_REGMAP,
+    IX_NPEDL_CTXT_REG_RESET_CINDEX,
+};
+
+/* contains Reset values for Context Store Registers  */
+static IxNpeDlEcsRegResetValue ixNpeDlEcsRegResetValues[] =
+{
+    {IX_NPEDL_ECS_BG_CTXT_REG_0,    IX_NPEDL_ECS_BG_CTXT_REG_0_RESET},
+    {IX_NPEDL_ECS_BG_CTXT_REG_1,    IX_NPEDL_ECS_BG_CTXT_REG_1_RESET},
+    {IX_NPEDL_ECS_BG_CTXT_REG_2,    IX_NPEDL_ECS_BG_CTXT_REG_2_RESET},
+    {IX_NPEDL_ECS_PRI_1_CTXT_REG_0, IX_NPEDL_ECS_PRI_1_CTXT_REG_0_RESET},
+    {IX_NPEDL_ECS_PRI_1_CTXT_REG_1, IX_NPEDL_ECS_PRI_1_CTXT_REG_1_RESET},
+    {IX_NPEDL_ECS_PRI_1_CTXT_REG_2, IX_NPEDL_ECS_PRI_1_CTXT_REG_2_RESET},
+    {IX_NPEDL_ECS_PRI_2_CTXT_REG_0, IX_NPEDL_ECS_PRI_2_CTXT_REG_0_RESET},
+    {IX_NPEDL_ECS_PRI_2_CTXT_REG_1, IX_NPEDL_ECS_PRI_2_CTXT_REG_1_RESET},
+    {IX_NPEDL_ECS_PRI_2_CTXT_REG_2, IX_NPEDL_ECS_PRI_2_CTXT_REG_2_RESET},
+    {IX_NPEDL_ECS_DBG_CTXT_REG_0,   IX_NPEDL_ECS_DBG_CTXT_REG_0_RESET},
+    {IX_NPEDL_ECS_DBG_CTXT_REG_1,   IX_NPEDL_ECS_DBG_CTXT_REG_1_RESET},
+    {IX_NPEDL_ECS_DBG_CTXT_REG_2,   IX_NPEDL_ECS_DBG_CTXT_REG_2_RESET},
+    {IX_NPEDL_ECS_INSTRUCT_REG,     IX_NPEDL_ECS_INSTRUCT_REG_RESET}
+};
+
+static IxNpeDlNpeMgrStats ixNpeDlNpeMgrStats;
+
+/* Set when NPE register memory has been mapped */
+static BOOL ixNpeDlMemInitialised = FALSE;
+
+
+/*
+ * static function prototypes.
+ */
+PRIVATE IX_STATUS
+ixNpeDlNpeMgrMemLoad (IxNpeDlNpeId npeId, UINT32 npeBaseAddress,
+		      IxNpeDlNpeMgrCodeBlock *codeBlockPtr,
+		      BOOL verify, IxNpeDlNpeMemType npeMemType);
+PRIVATE IX_STATUS
+ixNpeDlNpeMgrStateInfoLoad (UINT32 npeBaseAddress,
+			    IxNpeDlNpeMgrStateInfoBlock *codeBlockPtr,
+			    BOOL verify);
+PRIVATE BOOL
+ixNpeDlNpeMgrBitsSetCheck (UINT32 npeBaseAddress, UINT32 regOffset,
+			   UINT32 expectedBitsSet);
+
+PRIVATE UINT32
+ixNpeDlNpeMgrBaseAddressGet (IxNpeDlNpeId npeId);
+
+/*
+ * Function definition: ixNpeDlNpeMgrBaseAddressGet
+ */
+PRIVATE UINT32
+ixNpeDlNpeMgrBaseAddressGet (IxNpeDlNpeId npeId)
+{
+    IX_OSAL_ASSERT (ixNpeDlMemInitialised);
+    return ixNpeDlNpeInfo[npeId].baseAddress;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrInit
+ */
+void
+ixNpeDlNpeMgrInit (void)
+{
+    /* Only map the memory once */
+    if (!ixNpeDlMemInitialised)
+    {
+	UINT32 virtAddr;
+
+	/* map the register memory for NPE-A */
+	virtAddr = (UINT32) IX_OSAL_MEM_MAP (IX_NPEDL_NPEBASEADDRESS_NPEA,
+					    IX_OSAL_IXP400_NPEA_MAP_SIZE); 
+	IX_OSAL_ASSERT(virtAddr);
+	ixNpeDlNpeInfo[IX_NPEDL_NPEID_NPEA].baseAddress = virtAddr;
+
+	/* map the register memory for NPE-B */
+	virtAddr = (UINT32) IX_OSAL_MEM_MAP (IX_NPEDL_NPEBASEADDRESS_NPEB,
+					    IX_OSAL_IXP400_NPEB_MAP_SIZE); 
+	IX_OSAL_ASSERT(virtAddr);
+	ixNpeDlNpeInfo[IX_NPEDL_NPEID_NPEB].baseAddress = virtAddr;
+
+	/* map the register memory for NPE-C */
+	virtAddr = (UINT32) IX_OSAL_MEM_MAP (IX_NPEDL_NPEBASEADDRESS_NPEC,
+					    IX_OSAL_IXP400_NPEC_MAP_SIZE); 
+	IX_OSAL_ASSERT(virtAddr);
+	ixNpeDlNpeInfo[IX_NPEDL_NPEID_NPEC].baseAddress = virtAddr;
+
+	ixNpeDlMemInitialised = TRUE;
+    }
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrUninit
+ */
+IX_STATUS
+ixNpeDlNpeMgrUninit (void)
+{
+    if (!ixNpeDlMemInitialised)
+    {
+	return IX_FAIL;
+    }
+
+    IX_OSAL_MEM_UNMAP (ixNpeDlNpeInfo[IX_NPEDL_NPEID_NPEA].baseAddress);
+    IX_OSAL_MEM_UNMAP (ixNpeDlNpeInfo[IX_NPEDL_NPEID_NPEB].baseAddress);
+    IX_OSAL_MEM_UNMAP (ixNpeDlNpeInfo[IX_NPEDL_NPEID_NPEC].baseAddress);
+
+    ixNpeDlNpeInfo[IX_NPEDL_NPEID_NPEA].baseAddress = 0;
+    ixNpeDlNpeInfo[IX_NPEDL_NPEID_NPEB].baseAddress = 0;
+    ixNpeDlNpeInfo[IX_NPEDL_NPEID_NPEC].baseAddress = 0;
+
+    ixNpeDlMemInitialised = FALSE;
+
+    return IX_SUCCESS;
+}
+
+/*
+ * Function definition: ixNpeDlNpeMgrImageLoad
+ */
+IX_STATUS
+ixNpeDlNpeMgrImageLoad (
+    IxNpeDlNpeId npeId,
+    UINT32 *imageCodePtr,
+    BOOL verify)
+{
+    UINT32 npeBaseAddress;
+    IxNpeDlNpeMgrDownloadMap *downloadMap;
+    UINT32 *blockPtr;
+    UINT32 mapIndex = 0;
+    IX_STATUS status = IX_SUCCESS;
+    
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlNpeMgrImageLoad\n");
+
+    /* get base memory address of NPE from npeId */
+    npeBaseAddress = ixNpeDlNpeMgrBaseAddressGet (npeId);
+
+    /* check execution status of NPE to verify NPE Stop was successful */
+    if (!ixNpeDlNpeMgrBitsSetCheck (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXCTL,
+				    IX_NPEDL_EXCTL_STATUS_STOP))
+    {
+	IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeMgrImageDownload - "
+			       "NPE was not stopped before download\n");
+	status = IX_FAIL;
+    }
+    else
+    {
+	/*
+	 * Read Download Map, checking each block type and calling
+	 * appropriate function to perform download 
+	 */
+	downloadMap = (IxNpeDlNpeMgrDownloadMap *) imageCodePtr;
+	while ((downloadMap->entry[mapIndex].eodmMarker != 
+		IX_NPEDL_END_OF_DOWNLOAD_MAP)
+	       && (status == IX_SUCCESS))
+	{
+	    /* calculate pointer to block to be downloaded */
+	    blockPtr = imageCodePtr +
+		downloadMap->entry[mapIndex].block.offset;
+
+	    switch (downloadMap->entry[mapIndex].block.type)
+	    {
+	    case IX_NPEDL_BLOCK_TYPE_INSTRUCTION:
+		status = ixNpeDlNpeMgrMemLoad (npeId, npeBaseAddress, 
+					     (IxNpeDlNpeMgrCodeBlock *)blockPtr,
+					       verify,
+					       IX_NPEDL_MEM_TYPE_INSTRUCTION);
+		break;
+	    case IX_NPEDL_BLOCK_TYPE_DATA:
+		status = ixNpeDlNpeMgrMemLoad (npeId, npeBaseAddress,
+                                             (IxNpeDlNpeMgrCodeBlock *)blockPtr,
+					       verify, IX_NPEDL_MEM_TYPE_DATA);
+		break;
+	    case IX_NPEDL_BLOCK_TYPE_STATE:
+		status = ixNpeDlNpeMgrStateInfoLoad (npeBaseAddress,
+				       (IxNpeDlNpeMgrStateInfoBlock *) blockPtr,
+						     verify);
+		break;
+	    default:
+		IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeMgrImageLoad: "
+				       "unknown block type in download map\n");
+		status = IX_NPEDL_CRITICAL_MICROCODE_ERR;
+		ixNpeDlNpeMgrStats.criticalMicrocodeErrors++;
+		break;
+	    }
+	    mapIndex++;
+	}/* loop: for each entry in download map, while status == SUCCESS */
+    }/* condition: NPE stopped before attempting download */
+    
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT, 
+		     "Exiting ixNpeDlNpeMgrImageLoad : status = %d\n",
+		     status);
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrMemLoad
+ */
+PRIVATE IX_STATUS
+ixNpeDlNpeMgrMemLoad (
+    IxNpeDlNpeId npeId,
+    UINT32 npeBaseAddress,
+    IxNpeDlNpeMgrCodeBlock *blockPtr,
+    BOOL verify,
+    IxNpeDlNpeMemType npeMemType)
+{
+    UINT32 npeMemAddress;
+    UINT32 blockSize;
+    UINT32 memSize = 0;
+    IxNpeDlNpeMgrMemWrite memWriteFunc = NULL;
+    UINT32 localIndex = 0;
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlNpeMgrMemLoad\n");
+    
+    /*
+     * select NPE EXCTL reg read/write commands depending on memory
+     * type (instruction/data) to be accessed
+     */
+    if (npeMemType == IX_NPEDL_MEM_TYPE_INSTRUCTION)
+    {
+	memSize = ixNpeDlNpeInfo[npeId].insMemSize;
+	memWriteFunc = (IxNpeDlNpeMgrMemWrite) ixNpeDlNpeMgrInsMemWrite;
+    }
+    else if (npeMemType == IX_NPEDL_MEM_TYPE_DATA)
+    {
+	memSize = ixNpeDlNpeInfo[npeId].dataMemSize;
+	memWriteFunc = (IxNpeDlNpeMgrMemWrite) ixNpeDlNpeMgrDataMemWrite;
+    }
+
+    /*
+     * NPE memory is loaded contiguously from each block, so only address
+     * of 1st word in block is needed
+     */
+    npeMemAddress = blockPtr->npeMemAddress;
+    /* number of words of instruction/data microcode in block to download */
+    blockSize = blockPtr->size;
+    if ((npeMemAddress + blockSize) > memSize)
+    {
+	IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeMgrMemLoad: "
+			       "Block size too big for NPE memory\n");
+	status = IX_NPEDL_CRITICAL_MICROCODE_ERR;
+	ixNpeDlNpeMgrStats.criticalMicrocodeErrors++;
+    }
+    else
+    {
+	for (localIndex = 0; localIndex < blockSize; localIndex++)
+	{
+	    status = memWriteFunc (npeBaseAddress, npeMemAddress,
+				   blockPtr->data[localIndex], verify);
+
+	    if (status != IX_SUCCESS)
+	    {
+		IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeMgrMemLoad: "
+				       "write to NPE memory failed\n");
+		status = IX_NPEDL_CRITICAL_NPE_ERR;
+		ixNpeDlNpeMgrStats.criticalNpeErrors++;
+		break;   /* abort download */
+	    }
+	    /* increment target (word)address in NPE memory */
+	    npeMemAddress++;   
+	}
+    }/* condition: block size will fit in NPE memory */
+
+    if (status == IX_SUCCESS)
+    {
+	if (npeMemType == IX_NPEDL_MEM_TYPE_INSTRUCTION)
+	{
+	    ixNpeDlNpeMgrStats.instructionBlocksLoaded++;
+	}
+	else if (npeMemType == IX_NPEDL_MEM_TYPE_DATA)
+	{
+	    ixNpeDlNpeMgrStats.dataBlocksLoaded++;
+	}
+    }
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlNpeMgrMemLoad : status = %d\n", status);
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrStateInfoLoad
+ */
+PRIVATE IX_STATUS
+ixNpeDlNpeMgrStateInfoLoad (
+    UINT32 npeBaseAddress,
+    IxNpeDlNpeMgrStateInfoBlock *blockPtr,
+    BOOL verify)
+{
+    UINT32 blockSize;
+    UINT32 ctxtRegAddrInfo; 
+    UINT32 ctxtRegVal;
+    IxNpeDlCtxtRegNum ctxtReg; /* identifies Context Store reg (0-3) */
+    UINT32 ctxtNum;            /* identifies Context number (0-16)   */
+    UINT32 i;
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlNpeMgrStateInfoLoad\n");
+
+    /* block size contains number of words of state-info in block */
+    blockSize = blockPtr->size;
+    
+    ixNpeDlNpeMgrDebugInstructionPreExec (npeBaseAddress);
+
+    /* for each state-info context register entry in block */
+    for (i = 0; i < (blockSize/IX_NPEDL_STATE_INFO_ENTRY_SIZE); i++)
+    {
+	/* each state-info entry is 2 words (address, value) in length */
+	ctxtRegAddrInfo = (blockPtr->ctxtRegEntry[i]).addressInfo;
+	ctxtRegVal      = (blockPtr->ctxtRegEntry[i]).value;
+	
+	ctxtReg = (ctxtRegAddrInfo & IX_NPEDL_MASK_STATE_ADDR_CTXT_REG);
+	ctxtNum = (ctxtRegAddrInfo & IX_NPEDL_MASK_STATE_ADDR_CTXT_NUM) >> 
+	    IX_NPEDL_OFFSET_STATE_ADDR_CTXT_NUM;
+	
+	/* error-check Context Register No. and Context Number values  */
+	/* NOTE that there is no STEVT register for Context 0 */
+	if ((ctxtReg < 0) ||
+	    (ctxtReg >= IX_NPEDL_CTXT_REG_MAX) ||
+	    (ctxtNum > IX_NPEDL_CTXT_NUM_MAX) ||
+	    ((ctxtNum == 0) && (ctxtReg == IX_NPEDL_CTXT_REG_STEVT)))
+	{
+	    IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeMgrStateInfoLoad: "
+				   "invalid Context Register Address\n");
+	    status = IX_NPEDL_CRITICAL_MICROCODE_ERR;
+	    ixNpeDlNpeMgrStats.criticalMicrocodeErrors++;
+	    break;   /* abort download */
+	}    
+	
+	status = ixNpeDlNpeMgrCtxtRegWrite (npeBaseAddress, ctxtNum, ctxtReg,
+					    ctxtRegVal, verify);
+	if (status != IX_SUCCESS)
+	{
+	    IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeMgrStateInfoLoad: "
+				   "write of state-info to NPE failed\n");
+	    status = IX_NPEDL_CRITICAL_NPE_ERR;
+	    ixNpeDlNpeMgrStats.criticalNpeErrors++;
+	    break;   /* abort download */
+	}
+    }/* loop: for each context reg entry in State Info block */
+    
+    ixNpeDlNpeMgrDebugInstructionPostExec (npeBaseAddress);
+
+    if (status == IX_SUCCESS)
+    {
+	ixNpeDlNpeMgrStats.stateInfoBlocksLoaded++;
+    }
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlNpeMgrStateInfoLoad : status = %d\n",
+		     status);
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrNpeReset
+ */
+IX_STATUS
+ixNpeDlNpeMgrNpeReset (
+    IxNpeDlNpeId npeId)
+{
+    UINT32 npeBaseAddress;
+    IxNpeDlCtxtRegNum ctxtReg; /* identifies Context Store reg (0-3) */
+    UINT32 ctxtNum;            /* identifies Context number (0-16)   */
+    UINT32 regAddr;
+    UINT32 regVal;
+    UINT32 localIndex;
+    UINT32 indexMax;
+    IX_STATUS status = IX_SUCCESS;
+    IxFeatureCtrlReg unitFuseReg;
+    UINT32 ixNpeConfigCtrlRegVal;
+    
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT, 
+		     "Entering ixNpeDlNpeMgrNpeReset\n");
+    
+    /* get base memory address of NPE from npeId */
+    npeBaseAddress = ixNpeDlNpeMgrBaseAddressGet (npeId);
+
+    /* pre-store the NPE Config Control Register Value */
+    IX_NPEDL_REG_READ (npeBaseAddress, IX_NPEDL_REG_OFFSET_CTL, &ixNpeConfigCtrlRegVal);
+    
+    ixNpeConfigCtrlRegVal |= 0x3F000000;
+    
+    /* disable the parity interrupt */
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_CTL, (ixNpeConfigCtrlRegVal & IX_NPEDL_PARITY_BIT_MASK));
+    
+    ixNpeDlNpeMgrDebugInstructionPreExec (npeBaseAddress);
+
+    /*
+     * clear the FIFOs
+     */
+    while (ixNpeDlNpeMgrBitsSetCheck (npeBaseAddress,
+				      IX_NPEDL_REG_OFFSET_WFIFO,
+				      IX_NPEDL_MASK_WFIFO_VALID))
+    {
+	/* read from the Watch-point FIFO until empty */
+	IX_NPEDL_REG_READ (npeBaseAddress, IX_NPEDL_REG_OFFSET_WFIFO,
+			   &regVal);
+    }
+    
+    while (ixNpeDlNpeMgrBitsSetCheck (npeBaseAddress,
+					  IX_NPEDL_REG_OFFSET_STAT,
+				      IX_NPEDL_MASK_STAT_OFNE))
+    {
+	/* read from the outFIFO until empty */
+	IX_NPEDL_REG_READ (npeBaseAddress, IX_NPEDL_REG_OFFSET_FIFO,
+			   &regVal);
+    }
+    
+    while (ixNpeDlNpeMgrBitsSetCheck (npeBaseAddress,
+				      IX_NPEDL_REG_OFFSET_STAT,
+				      IX_NPEDL_MASK_STAT_IFNE))
+    {
+	/*
+	 * step execution of the NPE intruction to read inFIFO using
+	 * the Debug Executing Context stack
+	 */
+	status = ixNpeDlNpeMgrDebugInstructionExec (npeBaseAddress,
+					   IX_NPEDL_INSTR_RD_FIFO, 0, 0);
+
+    if (IX_SUCCESS != status)
+    {
+        return status;   
+    }
+    
+    }
+    
+    /*
+     * Reset the mailbox reg
+     */
+    /* ...from XScale side */
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_MBST,
+			IX_NPEDL_REG_RESET_MBST);
+    /* ...from NPE side */
+    status = ixNpeDlNpeMgrDebugInstructionExec (npeBaseAddress,
+				       IX_NPEDL_INSTR_RESET_MBOX, 0, 0);
+
+    if (IX_SUCCESS != status)
+    {
+        return status;   
+    }
+
+    /* 
+     *   Reset the physical registers in the NPE register file:
+     *   Note: no need to save/restore REGMAP for Context 0 here
+     *   since all Context Store regs are reset in subsequent code
+     */
+    for (regAddr = 0;
+	 (regAddr < IX_NPEDL_TOTAL_NUM_PHYS_REG) && (status != IX_FAIL);
+	 regAddr++)
+    {
+	/* for each physical register in the NPE reg file, write 0 : */
+	status = ixNpeDlNpeMgrPhysicalRegWrite (npeBaseAddress, regAddr,
+						0, TRUE);
+	if (status != IX_SUCCESS)
+	{
+	    return status;  /* abort reset */
+	}
+    }
+    
+
+    /*
+     * Reset the context store:
+     */
+    for (ctxtNum = IX_NPEDL_CTXT_NUM_MIN;
+	 ctxtNum <= IX_NPEDL_CTXT_NUM_MAX; ctxtNum++)
+    {	
+	/* set each context's Context Store registers to reset values: */
+	for (ctxtReg = 0; ctxtReg < IX_NPEDL_CTXT_REG_MAX; ctxtReg++)
+	{
+	    /* NOTE that there is no STEVT register for Context 0 */
+	    if (!((ctxtNum == 0) && (ctxtReg == IX_NPEDL_CTXT_REG_STEVT)))
+	    { 
+		regVal = ixNpeDlCtxtRegResetValues[ctxtReg];
+		status = ixNpeDlNpeMgrCtxtRegWrite (npeBaseAddress, ctxtNum,
+						    ctxtReg, regVal, TRUE);
+		if (status != IX_SUCCESS)
+		{
+		    return status;  /* abort reset */
+		}
+	    }
+	}
+    }
+
+    ixNpeDlNpeMgrDebugInstructionPostExec (npeBaseAddress);
+
+    /* write Reset values to Execution Context Stack registers */
+    indexMax = sizeof (ixNpeDlEcsRegResetValues) /
+	sizeof (IxNpeDlEcsRegResetValue);
+    for (localIndex = 0; localIndex < indexMax; localIndex++)
+    {
+	regAddr = ixNpeDlEcsRegResetValues[localIndex].regAddr;
+	regVal = ixNpeDlEcsRegResetValues[localIndex].regResetVal;
+	ixNpeDlNpeMgrExecAccRegWrite (npeBaseAddress, regAddr, regVal);
+    }
+    
+    /* clear the profile counter */
+    ixNpeDlNpeMgrCommandIssue (npeBaseAddress, 
+			       IX_NPEDL_EXCTL_CMD_CLR_PROFILE_CNT);
+    
+    /* clear registers EXCT, AP0, AP1, AP2 and AP3 */
+    for (regAddr = IX_NPEDL_REG_OFFSET_EXCT;
+	     regAddr <= IX_NPEDL_REG_OFFSET_AP3;
+	 regAddr += IX_NPEDL_BYTES_PER_WORD)
+    {
+	IX_NPEDL_REG_WRITE (npeBaseAddress, regAddr, 0);
+    }
+    
+    /* Reset the Watch-count register */
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_WC, 0);
+    
+    /*
+     * WR IXA00055043 - Remove IMEM Parity Introduced by NPE Reset Operation
+     */
+
+    /*
+     * Call the feature control API to fused out and reset the NPE and its
+     * coprocessor - to reset internal states and remove parity error
+     */
+    unitFuseReg = ixFeatureCtrlRead ();
+    unitFuseReg |= (IX_NPEDL_RESET_NPE_PARITY << npeId);
+    ixFeatureCtrlWrite (unitFuseReg);
+
+    /* call the feature control API to un-fused and un-reset the NPE & COP */
+    unitFuseReg &= (~(IX_NPEDL_RESET_NPE_PARITY << npeId));
+    ixFeatureCtrlWrite (unitFuseReg);
+
+    /*
+     * Call NpeMgr function to stop the NPE again after the Feature Control
+     * has unfused and Un-Reset the NPE and its associated Coprocessors
+     */
+    status = ixNpeDlNpeMgrNpeStop (npeId);
+
+    /* restore NPE configuration bus Control Register - Parity Settings  */
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_CTL, 
+        (ixNpeConfigCtrlRegVal & IX_NPEDL_CONFIG_CTRL_REG_MASK));
+
+    ixNpeDlNpeMgrStats.npeResets++;
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlNpeMgrNpeReset : status = %d\n", status);
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrNpeStart
+ */
+IX_STATUS
+ixNpeDlNpeMgrNpeStart (
+    IxNpeDlNpeId npeId)
+{
+    UINT32    npeBaseAddress;
+    UINT32    ecsRegVal;
+    BOOL      npeRunning;
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT, 
+		     "Entering ixNpeDlNpeMgrNpeStart\n");
+
+    /* get base memory address of NPE from npeId */
+    npeBaseAddress = ixNpeDlNpeMgrBaseAddressGet (npeId);
+
+    /*
+     * ensure only Background Context Stack Level is Active by turning off
+     * the Active bit in each of the other Executing Context Stack levels
+     */
+    ecsRegVal = ixNpeDlNpeMgrExecAccRegRead (npeBaseAddress,
+					     IX_NPEDL_ECS_PRI_1_CTXT_REG_0);
+    ecsRegVal &= ~IX_NPEDL_MASK_ECS_REG_0_ACTIVE;
+    ixNpeDlNpeMgrExecAccRegWrite (npeBaseAddress, IX_NPEDL_ECS_PRI_1_CTXT_REG_0,
+				  ecsRegVal);
+
+    ecsRegVal = ixNpeDlNpeMgrExecAccRegRead (npeBaseAddress,
+					     IX_NPEDL_ECS_PRI_2_CTXT_REG_0);
+    ecsRegVal &= ~IX_NPEDL_MASK_ECS_REG_0_ACTIVE;
+    ixNpeDlNpeMgrExecAccRegWrite (npeBaseAddress, IX_NPEDL_ECS_PRI_2_CTXT_REG_0,
+				  ecsRegVal);
+
+    ecsRegVal = ixNpeDlNpeMgrExecAccRegRead (npeBaseAddress,
+					     IX_NPEDL_ECS_DBG_CTXT_REG_0);
+    ecsRegVal &= ~IX_NPEDL_MASK_ECS_REG_0_ACTIVE;
+    ixNpeDlNpeMgrExecAccRegWrite (npeBaseAddress, IX_NPEDL_ECS_DBG_CTXT_REG_0,
+				  ecsRegVal);
+    
+    /* clear the pipeline */
+    ixNpeDlNpeMgrCommandIssue (npeBaseAddress, IX_NPEDL_EXCTL_CMD_NPE_CLR_PIPE);
+    
+    /* start NPE execution by issuing command through EXCTL register on NPE */
+    ixNpeDlNpeMgrCommandIssue (npeBaseAddress, IX_NPEDL_EXCTL_CMD_NPE_START);
+
+    /*
+     * check execution status of NPE to verify NPE Start operation was
+     * successful
+     */
+    npeRunning = ixNpeDlNpeMgrBitsSetCheck (npeBaseAddress,
+					    IX_NPEDL_REG_OFFSET_EXCTL,
+					    IX_NPEDL_EXCTL_STATUS_RUN);
+    if (npeRunning)
+    {
+	ixNpeDlNpeMgrStats.npeStarts++;
+    }
+    else
+    {
+	IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeMgrNpeStart: "
+			       "failed to start NPE execution\n");
+	status = IX_FAIL;
+    }
+
+    
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlNpeMgrNpeStart : status = %d\n", status);
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrNpeStop
+ */
+IX_STATUS
+ixNpeDlNpeMgrNpeStop (
+    IxNpeDlNpeId npeId)
+{
+    UINT32    npeBaseAddress;
+    IX_STATUS status = IX_SUCCESS;
+    
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlNpeMgrNpeStop\n");
+    
+    /* get base memory address of NPE from npeId */
+    npeBaseAddress = ixNpeDlNpeMgrBaseAddressGet (npeId);
+
+    /* stop NPE execution by issuing command through EXCTL register on NPE */
+    ixNpeDlNpeMgrCommandIssue (npeBaseAddress, IX_NPEDL_EXCTL_CMD_NPE_STOP);
+
+    /* verify that NPE Stop was successful */
+    if (! ixNpeDlNpeMgrBitsSetCheck (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXCTL,
+				     IX_NPEDL_EXCTL_STATUS_STOP))
+    {
+	IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeMgrNpeStop: "
+			       "failed to stop NPE execution\n");
+	status = IX_FAIL;
+    }
+
+    ixNpeDlNpeMgrStats.npeStops++;
+    
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlNpeMgrNpeStop : status = %d\n", status);
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrBitsSetCheck
+ */
+PRIVATE BOOL
+ixNpeDlNpeMgrBitsSetCheck (
+    UINT32 npeBaseAddress,
+    UINT32 regOffset,
+    UINT32 expectedBitsSet)
+{
+    UINT32 regVal;
+    IX_NPEDL_REG_READ (npeBaseAddress, regOffset, &regVal);
+
+    return expectedBitsSet == (expectedBitsSet & regVal);
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrStatsShow
+ */
+void
+ixNpeDlNpeMgrStatsShow (void)
+{
+    ixOsalLog (IX_OSAL_LOG_LVL_USER,
+               IX_OSAL_LOG_DEV_STDOUT,
+               "\nixNpeDlNpeMgrStatsShow:\n"
+               "\tInstruction Blocks loaded: %u\n"
+               "\tData Blocks loaded: %u\n"
+               "\tState Information Blocks loaded: %u\n"
+               "\tCritical NPE errors: %u\n"
+               "\tCritical Microcode errors: %u\n",
+               ixNpeDlNpeMgrStats.instructionBlocksLoaded,
+               ixNpeDlNpeMgrStats.dataBlocksLoaded,
+               ixNpeDlNpeMgrStats.stateInfoBlocksLoaded,
+               ixNpeDlNpeMgrStats.criticalNpeErrors,
+               ixNpeDlNpeMgrStats.criticalMicrocodeErrors,
+               0);
+
+    ixOsalLog (IX_OSAL_LOG_LVL_USER,
+               IX_OSAL_LOG_DEV_STDOUT,
+               "\tSuccessful NPE Starts: %u\n"
+               "\tSuccessful NPE Stops: %u\n"
+               "\tSuccessful NPE Resets: %u\n\n",
+               ixNpeDlNpeMgrStats.npeStarts,
+               ixNpeDlNpeMgrStats.npeStops,
+               ixNpeDlNpeMgrStats.npeResets,
+               0,0,0);
+
+    ixNpeDlNpeMgrUtilsStatsShow ();
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrStatsReset
+ */
+void
+ixNpeDlNpeMgrStatsReset (void)
+{
+    ixNpeDlNpeMgrStats.instructionBlocksLoaded = 0;
+    ixNpeDlNpeMgrStats.dataBlocksLoaded = 0;
+    ixNpeDlNpeMgrStats.stateInfoBlocksLoaded = 0;
+    ixNpeDlNpeMgrStats.criticalNpeErrors = 0;
+    ixNpeDlNpeMgrStats.criticalMicrocodeErrors = 0;
+    ixNpeDlNpeMgrStats.npeStarts = 0;
+    ixNpeDlNpeMgrStats.npeStops = 0;
+    ixNpeDlNpeMgrStats.npeResets = 0;
+
+    ixNpeDlNpeMgrUtilsStatsReset ();
+}

cpu/ixp/npe/IxNpeDlNpeMgrUtils.c

@@ -0,0 +1,806 @@
+/**
+ * @file IxNpeDlNpeMgrUtils.c
+ *
+ * @author Intel Corporation
+ * @date 18 February 2002
+ *
+ * @brief This file contains the implementation of the private API for the 
+ *        IXP425 NPE Downloader NpeMgr Utils module
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+
+/*
+ * Put the system defined include files required.
+ */
+#define IX_NPE_DL_MAX_NUM_OF_RETRIES 1000000 /**< Maximum number of
+                                              * retries before
+                                              * timeout
+					                          */  
+
+/*
+ * Put the user defined include files required.
+ */
+#include "IxOsal.h"
+#include "IxNpeDl.h"
+#include "IxNpeDlNpeMgrUtils_p.h"
+#include "IxNpeDlNpeMgrEcRegisters_p.h"
+#include "IxNpeDlMacros_p.h"
+
+/*
+ * #defines and macros used in this file.
+ */
+
+/* used to bit-mask a number of bytes */
+#define IX_NPEDL_MASK_LOWER_BYTE_OF_WORD  0x000000FF
+#define IX_NPEDL_MASK_LOWER_SHORT_OF_WORD 0x0000FFFF
+#define IX_NPEDL_MASK_FULL_WORD           0xFFFFFFFF
+
+#define IX_NPEDL_BYTES_PER_WORD           4
+#define IX_NPEDL_BYTES_PER_SHORT          2
+
+#define IX_NPEDL_REG_SIZE_BYTE            8
+#define IX_NPEDL_REG_SIZE_SHORT           16
+#define IX_NPEDL_REG_SIZE_WORD            32
+
+/*
+ * Introduce extra read cycles after issuing read command to NPE
+ * so that we read the register after the NPE has updated it
+ * This is to overcome race condition between XScale and NPE
+ */
+#define IX_NPEDL_DELAY_READ_CYCLES        2
+/*
+ * To mask top three MSBs of 32bit word to download into NPE IMEM
+ */
+#define IX_NPEDL_MASK_UNUSED_IMEM_BITS    0x1FFFFFFF;
+
+
+/*
+ * typedefs
+ */
+typedef struct
+{
+    UINT32 regAddress;
+    UINT32 regSize;
+} IxNpeDlCtxtRegAccessInfo;
+
+/* module statistics counters */
+typedef struct
+{
+    UINT32 insMemWrites;
+    UINT32 insMemWriteFails;
+    UINT32 dataMemWrites;
+    UINT32 dataMemWriteFails;
+    UINT32 ecsRegWrites;
+    UINT32 ecsRegReads;
+    UINT32 dbgInstructionExecs;
+    UINT32 contextRegWrites;
+    UINT32 physicalRegWrites;
+    UINT32 nextPcWrites;
+} IxNpeDlNpeMgrUtilsStats;
+
+
+/*
+ * Variable declarations global to this file only.  Externs are followed by
+ * static variables.
+ */
+
+/* 
+ * contains useful address and function pointers to read/write Context Regs, 
+ * eliminating some switch or if-else statements in places
+ */
+static IxNpeDlCtxtRegAccessInfo ixNpeDlCtxtRegAccInfo[IX_NPEDL_CTXT_REG_MAX] =
+{
+    {
+	IX_NPEDL_CTXT_REG_ADDR_STEVT,
+	IX_NPEDL_REG_SIZE_BYTE
+    },
+    {
+	IX_NPEDL_CTXT_REG_ADDR_STARTPC,
+	IX_NPEDL_REG_SIZE_SHORT
+    },
+    {
+	IX_NPEDL_CTXT_REG_ADDR_REGMAP,
+	IX_NPEDL_REG_SIZE_SHORT
+    },
+    {
+	IX_NPEDL_CTXT_REG_ADDR_CINDEX,
+	IX_NPEDL_REG_SIZE_BYTE
+    }
+};
+
+static UINT32 ixNpeDlSavedExecCount = 0;
+static UINT32 ixNpeDlSavedEcsDbgCtxtReg2 = 0;
+
+static IxNpeDlNpeMgrUtilsStats ixNpeDlNpeMgrUtilsStats;
+
+
+/*
+ * static function prototypes.
+ */
+PRIVATE __inline__ void
+ixNpeDlNpeMgrWriteCommandIssue (UINT32 npeBaseAddress, UINT32 cmd,
+				UINT32 addr, UINT32 data);
+
+PRIVATE __inline__ UINT32
+ixNpeDlNpeMgrReadCommandIssue (UINT32 npeBaseAddress, UINT32 cmd, UINT32 addr);
+
+PRIVATE IX_STATUS
+ixNpeDlNpeMgrLogicalRegRead (UINT32 npeBaseAddress, UINT32 regAddr,
+			     UINT32 regSize, UINT32 ctxtNum, UINT32 *regVal);
+
+PRIVATE IX_STATUS
+ixNpeDlNpeMgrLogicalRegWrite (UINT32 npeBaseAddress, UINT32 regAddr,
+			      UINT32 regVal, UINT32 regSize,
+			      UINT32 ctxtNum, BOOL verify);
+
+/*
+ * Function definition: ixNpeDlNpeMgrWriteCommandIssue
+ */
+PRIVATE __inline__ void
+ixNpeDlNpeMgrWriteCommandIssue (
+    UINT32 npeBaseAddress,
+    UINT32 cmd,
+    UINT32 addr,
+    UINT32 data)
+{
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXDATA, data);
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXAD, addr);
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXCTL, cmd);
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrReadCommandIssue
+ */
+PRIVATE __inline__ UINT32
+ixNpeDlNpeMgrReadCommandIssue (
+    UINT32 npeBaseAddress,
+    UINT32 cmd,
+    UINT32 addr)
+{
+    UINT32 data = 0;
+    int i;
+
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXAD, addr);
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXCTL, cmd);
+    for (i = 0; i <= IX_NPEDL_DELAY_READ_CYCLES; i++)
+    {
+	IX_NPEDL_REG_READ (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXDATA, &data);
+    }
+
+    return data;
+}
+
+/*
+ * Function definition: ixNpeDlNpeMgrInsMemWrite
+ */
+IX_STATUS
+ixNpeDlNpeMgrInsMemWrite (
+    UINT32 npeBaseAddress,
+    UINT32 insMemAddress,
+    UINT32 insMemData,
+    BOOL verify)
+{
+    UINT32 insMemDataRtn;
+
+    ixNpeDlNpeMgrWriteCommandIssue (npeBaseAddress,
+				    IX_NPEDL_EXCTL_CMD_WR_INS_MEM,
+				    insMemAddress, insMemData);
+    if (verify)
+    {
+        /* write invalid data to this reg, so we can see if we're reading 
+	   the EXDATA register too early */
+	IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXDATA,
+			    ~insMemData);
+
+        /*Disabled since top 3 MSB are not used for Azusa hardware Refer WR:IXA00053900*/
+        insMemData&=IX_NPEDL_MASK_UNUSED_IMEM_BITS;
+
+        insMemDataRtn=ixNpeDlNpeMgrReadCommandIssue (npeBaseAddress,
+                                           IX_NPEDL_EXCTL_CMD_RD_INS_MEM,
+                                           insMemAddress);
+
+        insMemDataRtn&=IX_NPEDL_MASK_UNUSED_IMEM_BITS;
+
+	if (insMemData != insMemDataRtn)
+	{
+	    ixNpeDlNpeMgrUtilsStats.insMemWriteFails++;
+	    return IX_FAIL;
+	}
+    }
+
+    ixNpeDlNpeMgrUtilsStats.insMemWrites++;
+    return IX_SUCCESS;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrDataMemWrite
+ */
+IX_STATUS
+ixNpeDlNpeMgrDataMemWrite (
+    UINT32 npeBaseAddress,
+    UINT32 dataMemAddress,
+    UINT32 dataMemData,
+    BOOL verify)
+{
+    ixNpeDlNpeMgrWriteCommandIssue (npeBaseAddress,
+				    IX_NPEDL_EXCTL_CMD_WR_DATA_MEM,
+				    dataMemAddress, dataMemData);
+    if (verify)
+    {
+        /* write invalid data to this reg, so we can see if we're reading 
+	   the EXDATA register too early */
+	IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXDATA, ~dataMemData);
+
+	if (dataMemData !=
+	    ixNpeDlNpeMgrReadCommandIssue (npeBaseAddress,
+					   IX_NPEDL_EXCTL_CMD_RD_DATA_MEM,
+					   dataMemAddress))
+	{
+	    ixNpeDlNpeMgrUtilsStats.dataMemWriteFails++;
+	    return IX_FAIL;
+	}
+    }
+
+    ixNpeDlNpeMgrUtilsStats.dataMemWrites++;
+    return IX_SUCCESS;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrExecAccRegWrite
+ */
+void
+ixNpeDlNpeMgrExecAccRegWrite (
+    UINT32 npeBaseAddress,
+    UINT32 regAddress,
+    UINT32 regData)
+{
+    ixNpeDlNpeMgrWriteCommandIssue (npeBaseAddress,
+				    IX_NPEDL_EXCTL_CMD_WR_ECS_REG,
+				    regAddress, regData);
+    ixNpeDlNpeMgrUtilsStats.ecsRegWrites++;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrExecAccRegRead
+ */
+UINT32
+ixNpeDlNpeMgrExecAccRegRead (
+    UINT32 npeBaseAddress,
+    UINT32 regAddress)
+{
+    ixNpeDlNpeMgrUtilsStats.ecsRegReads++;
+    return ixNpeDlNpeMgrReadCommandIssue (npeBaseAddress,
+					  IX_NPEDL_EXCTL_CMD_RD_ECS_REG,
+					  regAddress);
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrCommandIssue
+ */
+void
+ixNpeDlNpeMgrCommandIssue (
+    UINT32 npeBaseAddress,
+    UINT32 command)     
+{
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlNpeMgrCommandIssue\n");
+
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXCTL, command);
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlNpeMgrCommandIssue\n");
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrDebugInstructionPreExec
+ */
+void
+ixNpeDlNpeMgrDebugInstructionPreExec(
+    UINT32 npeBaseAddress)
+{
+    /* turn off the halt bit by clearing Execution Count register. */
+    /* save reg contents 1st and restore later */
+    IX_NPEDL_REG_READ (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXCT,
+		       &ixNpeDlSavedExecCount);
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXCT, 0);
+
+    /* ensure that IF and IE are on (temporarily), so that we don't end up
+     * stepping forever */
+    ixNpeDlSavedEcsDbgCtxtReg2 = ixNpeDlNpeMgrExecAccRegRead (npeBaseAddress,
+				                   IX_NPEDL_ECS_DBG_CTXT_REG_2);
+
+    ixNpeDlNpeMgrExecAccRegWrite (npeBaseAddress, IX_NPEDL_ECS_DBG_CTXT_REG_2,
+				  (ixNpeDlSavedEcsDbgCtxtReg2 |
+				   IX_NPEDL_MASK_ECS_DBG_REG_2_IF |
+				   IX_NPEDL_MASK_ECS_DBG_REG_2_IE));
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrDebugInstructionExec
+ */
+IX_STATUS
+ixNpeDlNpeMgrDebugInstructionExec(
+    UINT32 npeBaseAddress,
+    UINT32 npeInstruction,
+    UINT32 ctxtNum,
+    UINT32 ldur)
+{
+    UINT32 ecsDbgRegVal;
+    UINT32 oldWatchcount, newWatchcount;
+    UINT32 retriesCount = 0;
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlNpeMgrDebugInstructionExec\n");
+
+    /* set the Active bit, and the LDUR, in the debug level */
+    ecsDbgRegVal = IX_NPEDL_MASK_ECS_REG_0_ACTIVE |
+	(ldur << IX_NPEDL_OFFSET_ECS_REG_0_LDUR);
+
+    ixNpeDlNpeMgrExecAccRegWrite (npeBaseAddress, IX_NPEDL_ECS_DBG_CTXT_REG_0,
+				  ecsDbgRegVal);
+
+    /*
+     * set CCTXT at ECS DEBUG L3 to specify in which context to execute the
+     * instruction, and set SELCTXT at ECS DEBUG Level to specify which context
+     * store to access.
+     * Debug ECS Level Reg 1 has form  0x000n000n, where n = context number
+     */
+    ecsDbgRegVal = (ctxtNum << IX_NPEDL_OFFSET_ECS_REG_1_CCTXT) |
+	(ctxtNum << IX_NPEDL_OFFSET_ECS_REG_1_SELCTXT);
+
+    ixNpeDlNpeMgrExecAccRegWrite (npeBaseAddress, IX_NPEDL_ECS_DBG_CTXT_REG_1,
+				  ecsDbgRegVal);
+
+    /* clear the pipeline */
+    ixNpeDlNpeMgrCommandIssue (npeBaseAddress, IX_NPEDL_EXCTL_CMD_NPE_CLR_PIPE);
+
+    /* load NPE instruction into the instruction register */
+    ixNpeDlNpeMgrExecAccRegWrite (npeBaseAddress, IX_NPEDL_ECS_INSTRUCT_REG,
+				  npeInstruction);
+
+    /* we need this value later to wait for completion of NPE execution step */
+    IX_NPEDL_REG_READ (npeBaseAddress, IX_NPEDL_REG_OFFSET_WC, &oldWatchcount);
+
+    /* issue a Step One command via the Execution Control register */
+    ixNpeDlNpeMgrCommandIssue (npeBaseAddress, IX_NPEDL_EXCTL_CMD_NPE_STEP);
+
+	/* Watch Count register increments when NPE completes an instruction */
+	IX_NPEDL_REG_READ (npeBaseAddress, IX_NPEDL_REG_OFFSET_WC,
+        &newWatchcount);
+        
+    /*
+     * force the XScale to wait until the NPE has finished execution step
+     * NOTE that this delay will be very small, just long enough to allow a
+     * single NPE instruction to complete execution; if instruction execution
+     * is not completed before timeout retries, exit the while loop
+     */
+    while ((IX_NPE_DL_MAX_NUM_OF_RETRIES > retriesCount) 
+        && (newWatchcount == oldWatchcount))
+    {
+	    /* Watch Count register increments when NPE completes an instruction */
+	    IX_NPEDL_REG_READ (npeBaseAddress, IX_NPEDL_REG_OFFSET_WC,
+		    &newWatchcount);
+			   
+        retriesCount++;
+    }    
+
+    if (IX_NPE_DL_MAX_NUM_OF_RETRIES > retriesCount)
+    {
+        ixNpeDlNpeMgrUtilsStats.dbgInstructionExecs++;
+    }
+    else
+    {
+        /* Return timeout status as the instruction has not been executed
+         * after maximum retries */
+        status = IX_NPEDL_CRITICAL_NPE_ERR;
+    }
+    
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlNpeMgrDebugInstructionExec\n");
+		     
+    return status;
+}    
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrDebugInstructionPostExec
+ */
+void
+ixNpeDlNpeMgrDebugInstructionPostExec(
+    UINT32 npeBaseAddress)
+{
+    /* clear active bit in debug level */
+    ixNpeDlNpeMgrExecAccRegWrite (npeBaseAddress, IX_NPEDL_ECS_DBG_CTXT_REG_0,
+				  0);
+
+    /* clear the pipeline */
+    ixNpeDlNpeMgrCommandIssue (npeBaseAddress, IX_NPEDL_EXCTL_CMD_NPE_CLR_PIPE);
+    
+    /* restore Execution Count register contents. */
+    IX_NPEDL_REG_WRITE (npeBaseAddress, IX_NPEDL_REG_OFFSET_EXCT,
+			ixNpeDlSavedExecCount);
+
+    /* restore IF and IE bits to original values */
+    ixNpeDlNpeMgrExecAccRegWrite (npeBaseAddress, IX_NPEDL_ECS_DBG_CTXT_REG_2,
+				  ixNpeDlSavedEcsDbgCtxtReg2);
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrLogicalRegRead
+ */
+PRIVATE IX_STATUS
+ixNpeDlNpeMgrLogicalRegRead (
+    UINT32 npeBaseAddress, 
+    UINT32 regAddr,
+    UINT32 regSize,
+    UINT32 ctxtNum,
+    UINT32 *regVal)
+{
+    IX_STATUS status = IX_SUCCESS;
+    UINT32 npeInstruction = 0;
+    UINT32 mask = 0;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlNpeMgrLogicalRegRead\n");
+
+    switch (regSize)
+    {
+    case IX_NPEDL_REG_SIZE_BYTE:
+      npeInstruction = IX_NPEDL_INSTR_RD_REG_BYTE;
+      mask = IX_NPEDL_MASK_LOWER_BYTE_OF_WORD;  break;
+    case IX_NPEDL_REG_SIZE_SHORT:
+      npeInstruction = IX_NPEDL_INSTR_RD_REG_SHORT;
+      mask = IX_NPEDL_MASK_LOWER_SHORT_OF_WORD;  break;
+    case IX_NPEDL_REG_SIZE_WORD:
+      npeInstruction = IX_NPEDL_INSTR_RD_REG_WORD;
+      mask = IX_NPEDL_MASK_FULL_WORD;  break;
+    }
+
+    /* make regAddr be the SRC and DEST operands (e.g. movX d0, d0) */
+    npeInstruction |= (regAddr << IX_NPEDL_OFFSET_INSTR_SRC) |
+	(regAddr << IX_NPEDL_OFFSET_INSTR_DEST);
+
+    /* step execution of NPE intruction using Debug Executing Context stack */
+    status = ixNpeDlNpeMgrDebugInstructionExec (npeBaseAddress, npeInstruction,
+				       ctxtNum, IX_NPEDL_RD_INSTR_LDUR);
+
+    if (IX_SUCCESS != status)
+    {
+        return status;
+    }
+    
+    /* read value of register from Execution Data register */
+    IX_NPEDL_REG_READ (npeBaseAddress,	IX_NPEDL_REG_OFFSET_EXDATA, regVal);
+
+   /* align value from left to right */
+    *regVal = (*regVal >> (IX_NPEDL_REG_SIZE_WORD - regSize)) & mask;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlNpeMgrLogicalRegRead\n");
+    
+    return IX_SUCCESS;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrLogicalRegWrite
+ */
+PRIVATE IX_STATUS
+ixNpeDlNpeMgrLogicalRegWrite (
+    UINT32 npeBaseAddress, 
+    UINT32 regAddr,
+    UINT32 regVal,
+    UINT32 regSize,
+    UINT32 ctxtNum,
+    BOOL verify)
+{
+    UINT32 npeInstruction = 0;
+    UINT32 mask = 0;
+    IX_STATUS status = IX_SUCCESS;
+    UINT32 retRegVal;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlNpeMgrLogicalRegWrite\n");
+
+    if (regSize == IX_NPEDL_REG_SIZE_WORD)
+    {
+	/* NPE register addressing is left-to-right: e.g. |d0|d1|d2|d3| */
+	/* Write upper half-word (short) to |d0|d1| */
+	status = ixNpeDlNpeMgrLogicalRegWrite (npeBaseAddress, regAddr,
+				      regVal >> IX_NPEDL_REG_SIZE_SHORT,
+				      IX_NPEDL_REG_SIZE_SHORT,
+				      ctxtNum, verify);
+				      
+	if (IX_SUCCESS != status)
+	{
+	    return status;
+	}
+	
+	/* Write lower half-word (short) to |d2|d3| */
+	status = ixNpeDlNpeMgrLogicalRegWrite (npeBaseAddress,
+				      regAddr + IX_NPEDL_BYTES_PER_SHORT,
+                                    regVal & IX_NPEDL_MASK_LOWER_SHORT_OF_WORD,
+				      IX_NPEDL_REG_SIZE_SHORT,
+				      ctxtNum, verify);
+    
+    if (IX_SUCCESS != status)
+	{
+	    return status;
+	}
+	}
+    else
+    {
+        switch (regSize)
+	{ 
+	case IX_NPEDL_REG_SIZE_BYTE:
+	    npeInstruction = IX_NPEDL_INSTR_WR_REG_BYTE;
+	    mask = IX_NPEDL_MASK_LOWER_BYTE_OF_WORD;  break;
+	case IX_NPEDL_REG_SIZE_SHORT:
+            npeInstruction = IX_NPEDL_INSTR_WR_REG_SHORT;
+	    mask = IX_NPEDL_MASK_LOWER_SHORT_OF_WORD;  break;
+	}
+	/* mask out any redundant bits, so verify will work later */
+	regVal &= mask;
+
+	/* fill dest operand field of  instruction with destination reg addr */
+	npeInstruction |= (regAddr << IX_NPEDL_OFFSET_INSTR_DEST);
+
+	/* fill src operand field of instruction with least-sig 5 bits of val*/
+	npeInstruction |= ((regVal & IX_NPEDL_MASK_IMMED_INSTR_SRC_DATA) <<
+			   IX_NPEDL_OFFSET_INSTR_SRC);
+
+	/* fill coprocessor field of instruction with most-sig 11 bits of val*/
+	npeInstruction |= ((regVal & IX_NPEDL_MASK_IMMED_INSTR_COPROC_DATA) <<
+			   IX_NPEDL_DISPLACE_IMMED_INSTR_COPROC_DATA);
+
+	/* step execution of NPE intruction using Debug ECS */
+	status = ixNpeDlNpeMgrDebugInstructionExec(npeBaseAddress, npeInstruction,
+					  ctxtNum, IX_NPEDL_WR_INSTR_LDUR);
+					  
+	if (IX_SUCCESS != status)
+	{
+	    return status;  
+	} 
+    }/* condition: if reg to be written is 8-bit or 16-bit (not 32-bit) */
+
+    if (verify)
+    {
+    	status = ixNpeDlNpeMgrLogicalRegRead (npeBaseAddress, regAddr,
+    						   regSize, ctxtNum, &retRegVal);
+    						   
+        if (IX_SUCCESS == status)
+        {
+            if (regVal != retRegVal)
+            {
+                status = IX_FAIL;
+            }
+        }        
+    }
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlNpeMgrLogicalRegWrite : status = %d\n",
+		     status);
+    
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrPhysicalRegWrite
+ */
+IX_STATUS
+ixNpeDlNpeMgrPhysicalRegWrite (
+    UINT32 npeBaseAddress,
+    UINT32 regAddr,
+    UINT32 regValue,
+    BOOL verify)
+{
+    IX_STATUS status;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Entering ixNpeDlNpeMgrPhysicalRegWrite\n");
+
+/*
+ * There are 32 physical registers used in an NPE.  These are
+ * treated as 16 pairs of 32-bit registers.  To write one of the pair,
+ * write the pair number (0-16) to the REGMAP for Context 0.  Then write
+ * the value to register  0 or 4 in the regfile, depending on which
+ * register of the pair is to be written
+ */
+
+    /*
+     * set REGMAP for context 0 to (regAddr >> 1) to choose which pair (0-16)
+     * of physical registers to write 
+     */
+    status = ixNpeDlNpeMgrLogicalRegWrite (npeBaseAddress,
+					   IX_NPEDL_CTXT_REG_ADDR_REGMAP,
+					   (regAddr >>
+					  IX_NPEDL_OFFSET_PHYS_REG_ADDR_REGMAP),
+					   IX_NPEDL_REG_SIZE_SHORT, 0, verify);
+    if (status == IX_SUCCESS)
+    {
+	/* regAddr = 0 or 4  */
+	regAddr = (regAddr & IX_NPEDL_MASK_PHYS_REG_ADDR_LOGICAL_ADDR) *
+	    IX_NPEDL_BYTES_PER_WORD;
+    
+    status = ixNpeDlNpeMgrLogicalRegWrite (npeBaseAddress, regAddr, regValue, 
+					   IX_NPEDL_REG_SIZE_WORD, 0, verify);
+    }
+    
+    if (status != IX_SUCCESS)
+    {
+	IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeMgrPhysicalRegWrite: "
+			       "error writing to physical register\n");
+    }
+
+    ixNpeDlNpeMgrUtilsStats.physicalRegWrites++;
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT,
+		     "Exiting ixNpeDlNpeMgrPhysicalRegWrite : status = %d\n",
+		     status);
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrCtxtRegWrite
+ */
+IX_STATUS
+ixNpeDlNpeMgrCtxtRegWrite (
+    UINT32 npeBaseAddress,
+    UINT32 ctxtNum,
+    IxNpeDlCtxtRegNum ctxtReg,
+    UINT32 ctxtRegVal,
+    BOOL verify)
+{
+    UINT32 tempRegVal;
+    UINT32 ctxtRegAddr;
+    UINT32 ctxtRegSize;
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEDL_TRACE0 (IX_NPEDL_FN_ENTRY_EXIT, 
+		     "Entering ixNpeDlNpeMgrCtxtRegWrite\n");
+
+    /*
+     * Context 0 has no STARTPC. Instead, this value is used to set
+     * NextPC for Background ECS, to set where NPE starts executing code
+     */
+    if ((ctxtNum == 0) && (ctxtReg == IX_NPEDL_CTXT_REG_STARTPC))
+    {
+	/* read BG_CTXT_REG_0, update NEXTPC bits, and write back to reg */
+	tempRegVal = ixNpeDlNpeMgrExecAccRegRead (npeBaseAddress,
+						  IX_NPEDL_ECS_BG_CTXT_REG_0);
+	tempRegVal &= ~IX_NPEDL_MASK_ECS_REG_0_NEXTPC;
+	tempRegVal |= (ctxtRegVal << IX_NPEDL_OFFSET_ECS_REG_0_NEXTPC) &
+	    IX_NPEDL_MASK_ECS_REG_0_NEXTPC;
+	
+	ixNpeDlNpeMgrExecAccRegWrite (npeBaseAddress,
+				      IX_NPEDL_ECS_BG_CTXT_REG_0, tempRegVal);
+
+	ixNpeDlNpeMgrUtilsStats.nextPcWrites++;
+    }
+    else
+    {
+	ctxtRegAddr = ixNpeDlCtxtRegAccInfo[ctxtReg].regAddress;
+	ctxtRegSize = ixNpeDlCtxtRegAccInfo[ctxtReg].regSize;
+	status = ixNpeDlNpeMgrLogicalRegWrite (npeBaseAddress, ctxtRegAddr,
+					       ctxtRegVal, ctxtRegSize,
+					       ctxtNum, verify);
+	if (status != IX_SUCCESS)
+	{
+	    IX_NPEDL_ERROR_REPORT ("ixNpeDlNpeMgrCtxtRegWrite: "
+				   "error writing to context store register\n");
+	}
+	
+	ixNpeDlNpeMgrUtilsStats.contextRegWrites++;
+    }
+
+    IX_NPEDL_TRACE1 (IX_NPEDL_FN_ENTRY_EXIT, 
+		     "Exiting ixNpeDlNpeMgrCtxtRegWrite : status = %d\n",
+		     status);
+
+    return status;
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrUtilsStatsShow
+ */
+void
+ixNpeDlNpeMgrUtilsStatsShow (void)
+{
+    ixOsalLog (IX_OSAL_LOG_LVL_USER,
+               IX_OSAL_LOG_DEV_STDOUT,
+               "\nixNpeDlNpeMgrUtilsStatsShow:\n"
+               "\tInstruction Memory writes: %u\n"
+               "\tInstruction Memory writes failed: %u\n"
+               "\tData Memory writes: %u\n"
+               "\tData Memory writes failed: %u\n",
+               ixNpeDlNpeMgrUtilsStats.insMemWrites,
+               ixNpeDlNpeMgrUtilsStats.insMemWriteFails,
+               ixNpeDlNpeMgrUtilsStats.dataMemWrites,
+               ixNpeDlNpeMgrUtilsStats.dataMemWriteFails,
+               0,0);
+
+    ixOsalLog (IX_OSAL_LOG_LVL_USER,
+               IX_OSAL_LOG_DEV_STDOUT,
+               "\tExecuting Context Stack Register writes: %u\n"
+               "\tExecuting Context Stack Register reads: %u\n"
+               "\tPhysical Register writes: %u\n"
+               "\tContext Store Register writes: %u\n"
+               "\tExecution Backgound Context NextPC writes: %u\n"
+               "\tDebug Instructions Executed: %u\n\n",
+               ixNpeDlNpeMgrUtilsStats.ecsRegWrites,
+               ixNpeDlNpeMgrUtilsStats.ecsRegReads,
+               ixNpeDlNpeMgrUtilsStats.physicalRegWrites,
+               ixNpeDlNpeMgrUtilsStats.contextRegWrites,
+               ixNpeDlNpeMgrUtilsStats.nextPcWrites,
+               ixNpeDlNpeMgrUtilsStats.dbgInstructionExecs);
+}
+
+
+/*
+ * Function definition: ixNpeDlNpeMgrUtilsStatsReset
+ */
+void
+ixNpeDlNpeMgrUtilsStatsReset (void)
+{
+    ixNpeDlNpeMgrUtilsStats.insMemWrites = 0;
+    ixNpeDlNpeMgrUtilsStats.insMemWriteFails = 0;
+    ixNpeDlNpeMgrUtilsStats.dataMemWrites = 0;
+    ixNpeDlNpeMgrUtilsStats.dataMemWriteFails = 0;
+    ixNpeDlNpeMgrUtilsStats.ecsRegWrites = 0;
+    ixNpeDlNpeMgrUtilsStats.ecsRegReads = 0;
+    ixNpeDlNpeMgrUtilsStats.physicalRegWrites = 0;
+    ixNpeDlNpeMgrUtilsStats.contextRegWrites = 0;
+    ixNpeDlNpeMgrUtilsStats.nextPcWrites = 0;
+    ixNpeDlNpeMgrUtilsStats.dbgInstructionExecs = 0;
+}

cpu/ixp/npe/IxNpeMh.c

@@ -0,0 +1,582 @@
+/**
+ * @file IxNpeMh.c
+ *
+ * @author Intel Corporation
+ * @date 18 Jan 2002
+ *
+ * @brief This file contains the implementation of the public API for the
+ * IXP425 NPE Message Handler component.
+ *
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+/*
+ * Put the system defined include files required.
+ */
+
+/*
+ * Put the user defined include files required.
+ */
+
+#include "IxOsal.h"
+#include "IxNpeMhMacros_p.h"
+
+#include "IxNpeMh.h"
+
+#include "IxNpeMhConfig_p.h"
+#include "IxNpeMhReceive_p.h"
+#include "IxNpeMhSend_p.h"
+#include "IxNpeMhSolicitedCbMgr_p.h"
+#include "IxNpeMhUnsolicitedCbMgr_p.h"
+
+/*
+ * #defines and macros used in this file.
+ */
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+
+/*
+ * Variable declarations global to this file only.  Externs are followed by
+ * static variables.
+ */
+
+PRIVATE BOOL ixNpeMhInitialized = FALSE;
+
+/*
+ * Extern function prototypes.
+ */
+
+/*
+ * Static function prototypes.
+ */
+
+/*
+ * Function definition: ixNpeMhInitialize
+ */
+
+PUBLIC IX_STATUS ixNpeMhInitialize (
+    IxNpeMhNpeInterrupts npeInterrupts)
+{
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhInitialize\n");
+
+    /* check the npeInterrupts parameter */
+    if ((npeInterrupts != IX_NPEMH_NPEINTERRUPTS_NO) &&
+        (npeInterrupts != IX_NPEMH_NPEINTERRUPTS_YES))
+    {
+        IX_NPEMH_ERROR_REPORT ("Illegal npeInterrupts parameter value\n");
+        return IX_FAIL;
+    }
+
+    /* parameters are ok ... */
+
+    /* initialize the Receive module */
+    ixNpeMhReceiveInitialize ();
+
+    /* initialize the Solicited Callback Manager module */
+    ixNpeMhSolicitedCbMgrInitialize ();
+
+    /* initialize the Unsolicited Callback Manager module */
+    ixNpeMhUnsolicitedCbMgrInitialize ();
+
+    /* initialize the Configuration module
+     *
+     * NOTE: This module was originally configured before the
+     * others, but the sequence was changed so that interrupts
+     * would only be enabled after the handler functions were
+     * set up.  The above modules need to be initialised to
+     * handle the NPE interrupts.  See SCR #2231.
+     */
+    ixNpeMhConfigInitialize (npeInterrupts);
+
+    ixNpeMhInitialized = TRUE;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhInitialize\n");
+
+    return IX_SUCCESS;
+}
+
+/*
+ * Function definition: ixNpeMhUnload
+ */
+
+PUBLIC IX_STATUS ixNpeMhUnload (void)
+{
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhUnload\n");
+
+    if (!ixNpeMhInitialized)
+    {
+	return IX_FAIL;
+    }
+
+    /* Uninitialize the Configuration module */
+    ixNpeMhConfigUninit ();
+
+    ixNpeMhInitialized = FALSE;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhUnload\n");
+
+    return IX_SUCCESS;
+}
+
+
+/*
+ * Function definition: ixNpeMhUnsolicitedCallbackRegister
+ */
+
+PUBLIC IX_STATUS ixNpeMhUnsolicitedCallbackRegister (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessageId messageId,
+    IxNpeMhCallback unsolicitedCallback)
+{
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhUnsolicitedCallbackRegister\n");
+
+    /* check that we are initialized */
+    if (!ixNpeMhInitialized)
+    {
+        IX_NPEMH_ERROR_REPORT ("IxNpeMh component is not initialized\n");
+        return IX_FAIL;
+    }
+
+    /* check the npeId parameter */
+    if (!ixNpeMhConfigNpeIdIsValid (npeId))
+    {
+        IX_NPEMH_ERROR_REPORT ("NPE ID invalid\n");
+        return IX_FAIL;
+    }
+
+    /* check the messageId parameter */
+    if ((messageId < IX_NPEMH_MIN_MESSAGE_ID)
+        || (messageId > IX_NPEMH_MAX_MESSAGE_ID))
+    {
+        IX_NPEMH_ERROR_REPORT ("Message ID is out of range\n");
+        return IX_FAIL;
+    }
+
+    /* the unsolicitedCallback parameter is allowed to be NULL */
+
+    /* parameters are ok ... */
+
+    /* get the lock to prevent other clients from entering */
+    ixNpeMhConfigLockGet (npeId);
+
+    /* save the unsolicited callback for the message ID */
+    ixNpeMhUnsolicitedCbMgrCallbackSave (
+        npeId, messageId, unsolicitedCallback);
+
+    /* release the lock to allow other clients back in */
+    ixNpeMhConfigLockRelease (npeId);
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhUnsolicitedCallbackRegister\n");
+
+    return IX_SUCCESS;
+}
+
+/*
+ * Function definition: ixNpeMhUnsolicitedCallbackForRangeRegister
+ */
+
+PUBLIC IX_STATUS ixNpeMhUnsolicitedCallbackForRangeRegister (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessageId minMessageId,
+    IxNpeMhMessageId maxMessageId,
+    IxNpeMhCallback unsolicitedCallback)
+{
+    IxNpeMhMessageId messageId;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhUnsolicitedCallbackForRangeRegister\n");
+
+    /* check that we are initialized */
+    if (!ixNpeMhInitialized)
+    {
+        IX_NPEMH_ERROR_REPORT ("IxNpeMh component is not initialized\n");
+        return IX_FAIL;
+    }
+
+    /* check the npeId parameter */
+    if (!ixNpeMhConfigNpeIdIsValid (npeId))
+    {
+        IX_NPEMH_ERROR_REPORT ("NPE ID invalid\n");
+        return IX_FAIL;
+    }
+
+    /* check the minMessageId parameter */
+    if ((minMessageId < IX_NPEMH_MIN_MESSAGE_ID)
+        || (minMessageId > IX_NPEMH_MAX_MESSAGE_ID))
+    {
+        IX_NPEMH_ERROR_REPORT ("Min message ID is out of range\n");
+        return IX_FAIL;
+    }
+
+    /* check the maxMessageId parameter */
+    if ((maxMessageId < IX_NPEMH_MIN_MESSAGE_ID)
+        || (maxMessageId > IX_NPEMH_MAX_MESSAGE_ID))
+    {
+        IX_NPEMH_ERROR_REPORT ("Max message ID is out of range\n");
+        return IX_FAIL;
+    }
+
+    /* check the semantics of the message range parameters */
+    if (minMessageId > maxMessageId)
+    {
+        IX_NPEMH_ERROR_REPORT ("Min message ID greater than max message "
+                               "ID\n");
+        return IX_FAIL;
+    }
+
+    /* the unsolicitedCallback parameter is allowed to be NULL */
+
+    /* parameters are ok ... */
+
+    /* get the lock to prevent other clients from entering */
+    ixNpeMhConfigLockGet (npeId);
+
+    /* for each message ID in the range ... */
+    for (messageId = minMessageId; messageId <= maxMessageId; messageId++)
+    {
+        /* save the unsolicited callback for the message ID */
+        ixNpeMhUnsolicitedCbMgrCallbackSave (
+            npeId, messageId, unsolicitedCallback);
+    }
+
+    /* release the lock to allow other clients back in */
+    ixNpeMhConfigLockRelease (npeId);
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhUnsolicitedCallbackForRangeRegister\n");
+
+    return IX_SUCCESS;
+}
+
+/*
+ * Function definition: ixNpeMhMessageSend
+ */
+
+PUBLIC IX_STATUS ixNpeMhMessageSend (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessage message,
+    UINT32 maxSendRetries)
+{
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhMessageSend\n");
+
+    /* check that we are initialized */
+    if (!ixNpeMhInitialized)
+    {
+        IX_NPEMH_ERROR_REPORT ("IxNpeMh component is not initialized\n");
+        return IX_FAIL;
+    }
+
+    /* check the npeId parameter */
+    if (!ixNpeMhConfigNpeIdIsValid (npeId))
+    {
+        IX_NPEMH_ERROR_REPORT ("NPE ID invalid\n");
+        return IX_FAIL;
+    }
+
+    /* parameters are ok ... */
+
+    /* get the lock to prevent other clients from entering */
+    ixNpeMhConfigLockGet (npeId);
+
+    /* send the message */
+    status = ixNpeMhSendMessageSend (npeId, message, maxSendRetries);
+    if (status != IX_SUCCESS)
+    {
+        IX_NPEMH_ERROR_REPORT ("Failed to send message\n");
+    }
+
+    /* release the lock to allow other clients back in */
+    ixNpeMhConfigLockRelease (npeId);
+
+    IX_NPEMH_TRACE1 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhMessageSend"
+                     " : status = %d\n", status);
+
+    return status;
+}
+
+/*
+ * Function definition: ixNpeMhMessageWithResponseSend
+ */
+
+PUBLIC IX_STATUS ixNpeMhMessageWithResponseSend (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessage message,
+    IxNpeMhMessageId solicitedMessageId,
+    IxNpeMhCallback solicitedCallback,
+    UINT32 maxSendRetries)
+{
+    IX_STATUS status = IX_SUCCESS;
+    IxNpeMhCallback unsolicitedCallback = NULL;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhMessageWithResponseSend\n");
+
+    /* check that we are initialized */
+    if (!ixNpeMhInitialized)
+    {
+        IX_NPEMH_ERROR_REPORT ("IxNpeMh component is not initialized\n");
+        return IX_FAIL;
+    }
+
+    /* the solicitecCallback parameter is allowed to be NULL.  this */
+    /* signifies the client is not interested in the response message */
+
+    /* check the npeId parameter */
+    if (!ixNpeMhConfigNpeIdIsValid (npeId))
+    {
+        IX_NPEMH_ERROR_REPORT ("NPE ID invalid\n");
+        return IX_FAIL;
+    }
+
+    /* check the solicitedMessageId parameter */
+    if ((solicitedMessageId < IX_NPEMH_MIN_MESSAGE_ID)
+        || (solicitedMessageId > IX_NPEMH_MAX_MESSAGE_ID))
+    {
+        IX_NPEMH_ERROR_REPORT ("Solicited message ID is out of range\n");
+        return IX_FAIL;
+    }
+
+    /* check the solicitedMessageId parameter.  if an unsolicited */
+    /* callback has been registered for the specified message ID then */
+    /* report an error and return failure */
+    ixNpeMhUnsolicitedCbMgrCallbackRetrieve (
+        npeId, solicitedMessageId, &unsolicitedCallback);
+    if (unsolicitedCallback != NULL)
+    {
+        IX_NPEMH_ERROR_REPORT ("Solicited message ID conflicts with "
+                               "unsolicited message ID\n");
+        return IX_FAIL;
+    }
+
+    /* parameters are ok ... */
+
+    /* get the lock to prevent other clients from entering */
+    ixNpeMhConfigLockGet (npeId);
+
+    /* send the message */
+    status = ixNpeMhSendMessageWithResponseSend (
+        npeId, message, solicitedMessageId, solicitedCallback,
+        maxSendRetries);
+    if (status != IX_SUCCESS)
+    {
+        IX_NPEMH_ERROR_REPORT ("Failed to send message\n");
+    }
+
+    /* release the lock to allow other clients back in */
+    ixNpeMhConfigLockRelease (npeId);
+
+    IX_NPEMH_TRACE1 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhMessageWithResponseSend"
+                     " : status = %d\n", status);
+
+    return status;
+}
+
+/*
+ * Function definition: ixNpeMhMessagesReceive
+ */
+
+PUBLIC IX_STATUS ixNpeMhMessagesReceive (
+    IxNpeMhNpeId npeId)
+{
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhMessagesReceive\n");
+
+    /* check that we are initialized */
+    if (!ixNpeMhInitialized)
+    {
+        IX_NPEMH_ERROR_REPORT ("IxNpeMh component is not initialized\n");
+        return IX_FAIL;
+    }
+
+    /* check the npeId parameter */
+    if (!ixNpeMhConfigNpeIdIsValid (npeId))
+    {
+        IX_NPEMH_ERROR_REPORT ("NPE ID invalid\n");
+        return IX_FAIL;
+    }
+
+    /* parameters are ok ... */
+
+    /* get the lock to prevent other clients from entering */
+    ixNpeMhConfigLockGet (npeId);
+
+    /* receive messages from the NPE */
+    status = ixNpeMhReceiveMessagesReceive (npeId);
+
+    if (status != IX_SUCCESS)
+    {
+        IX_NPEMH_ERROR_REPORT ("Failed to receive message\n");
+    }
+
+    /* release the lock to allow other clients back in */
+    ixNpeMhConfigLockRelease (npeId);
+
+    IX_NPEMH_TRACE1 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhMessagesReceive"
+                     " : status = %d\n", status);
+
+    return status;
+}
+
+/*
+ * Function definition: ixNpeMhShow
+ */
+
+PUBLIC IX_STATUS ixNpeMhShow (
+    IxNpeMhNpeId npeId)
+{
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhShow\n");
+
+    /* check that we are initialized */
+    if (!ixNpeMhInitialized)
+    {
+        IX_NPEMH_ERROR_REPORT ("IxNpeMh component is not initialized\n");
+        return IX_FAIL;
+    }
+
+    /* check the npeId parameter */
+    if (!ixNpeMhConfigNpeIdIsValid (npeId))
+    {
+        IX_NPEMH_ERROR_REPORT ("NPE ID invalid\n");
+        return IX_FAIL;
+    }
+
+    /* parameters are ok ... */
+
+    /* note we don't get the lock here as printing the statistics */
+    /* to a console may take some time and we don't want to impact */
+    /* system performance.  this means that the statistics displayed */
+    /* may be in a state of flux and make not represent a consistent */
+    /* snapshot. */
+
+    /* display a header */
+    ixOsalLog (IX_OSAL_LOG_LVL_USER, IX_OSAL_LOG_DEV_STDOUT,
+	       "Current state of NPE ID %d:\n\n", npeId, 0, 0, 0, 0, 0);
+
+    /* show the current state of each module */
+
+    /* show the current state of the Configuration module */
+    ixNpeMhConfigShow (npeId);
+
+    /* show the current state of the Receive module */
+    ixNpeMhReceiveShow (npeId);
+
+    /* show the current state of the Send module */
+    ixNpeMhSendShow (npeId);
+
+    /* show the current state of the Solicited Callback Manager module */
+    ixNpeMhSolicitedCbMgrShow (npeId);
+
+    /* show the current state of the Unsolicited Callback Manager module */
+    ixNpeMhUnsolicitedCbMgrShow (npeId);
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhShow\n");
+
+    return IX_SUCCESS;
+}
+
+/*
+ * Function definition: ixNpeMhShowReset
+ */
+
+PUBLIC IX_STATUS ixNpeMhShowReset (
+    IxNpeMhNpeId npeId)
+{
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhShowReset\n");
+
+    /* check that we are initialized */
+    if (!ixNpeMhInitialized)
+    {
+        IX_NPEMH_ERROR_REPORT ("IxNpeMh component is not initialized\n");
+        return IX_FAIL;
+    }
+
+    /* check the npeId parameter */
+    if (!ixNpeMhConfigNpeIdIsValid (npeId))
+    {
+        IX_NPEMH_ERROR_REPORT ("NPE ID invalid\n");
+        return IX_FAIL;
+    }
+
+    /* parameters are ok ... */
+
+    /* note we don't get the lock here as resetting the statistics */
+    /* shouldn't impact system performance. */
+
+    /* reset the current state of each module */
+
+    /* reset the current state of the Configuration module */
+    ixNpeMhConfigShowReset (npeId);
+
+    /* reset the current state of the Receive module */
+    ixNpeMhReceiveShowReset (npeId);
+
+    /* reset the current state of the Send module */
+    ixNpeMhSendShowReset (npeId);
+
+    /* reset the current state of the Solicited Callback Manager module */
+    ixNpeMhSolicitedCbMgrShowReset (npeId);
+
+    /* reset the current state of the Unsolicited Callback Manager module */
+    ixNpeMhUnsolicitedCbMgrShowReset (npeId);
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhShowReset\n");
+
+    return IX_SUCCESS;
+}

cpu/ixp/npe/IxNpeMhConfig.c

@@ -0,0 +1,608 @@
+/**
+ * @file IxNpeMhConfig.c
+ *
+ * @author Intel Corporation
+ * @date 18 Jan 2002
+ *
+ * @brief This file contains the implementation of the private API for the
+ * Configuration module.
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+/*
+ * Put the system defined include files required.
+ */
+
+
+/*
+ * Put the user defined include files required.
+ */
+
+#include "IxOsal.h"
+
+#include "IxNpeMhMacros_p.h"
+
+#include "IxNpeMhConfig_p.h"
+
+/*
+ * #defines and macros used in this file.
+ */
+#define IX_NPE_MH_MAX_NUM_OF_RETRIES 1000000 /**< Maximum number of
+                                              * retries before
+                                              * timeout
+					                          */  
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+
+/**
+ * @struct IxNpeMhConfigStats
+ *
+ * @brief This structure is used to maintain statistics for the
+ * Configuration module.
+ */
+
+typedef struct
+{
+    UINT32 outFifoReads;        /**< outFifo reads */
+    UINT32 inFifoWrites;        /**< inFifo writes */
+    UINT32 maxInFifoFullRetries;   /**< max retries if inFIFO full   */
+    UINT32 maxOutFifoEmptyRetries; /**< max retries if outFIFO empty */
+} IxNpeMhConfigStats;
+
+/*
+ * Variable declarations global to this file only.  Externs are followed by
+ * static variables.
+ */
+
+IxNpeMhConfigNpeInfo ixNpeMhConfigNpeInfo[IX_NPEMH_NUM_NPES] =
+{
+    {
+        0,
+        IX_NPEMH_NPEA_INT,
+	0,
+        0,
+        0,
+        0,
+        0,
+        NULL,
+        FALSE
+    },
+    {
+        0,
+        IX_NPEMH_NPEB_INT,
+	0,
+	0,
+        0,
+        0,
+        0,
+        NULL,
+        FALSE
+    },
+    {
+        0,
+        IX_NPEMH_NPEC_INT,
+        0,
+        0,
+        0,
+        0,
+        0,
+        NULL,
+        FALSE
+    }
+};
+
+PRIVATE IxNpeMhConfigStats ixNpeMhConfigStats[IX_NPEMH_NUM_NPES];
+
+/*
+ * Extern function prototypes.
+ */
+
+/*
+ * Static function prototypes.
+ */
+PRIVATE
+void ixNpeMhConfigIsr (void *parameter);
+
+/*
+ * Function definition: ixNpeMhConfigIsr
+ */
+
+PRIVATE
+void ixNpeMhConfigIsr (void *parameter)
+{
+    IxNpeMhNpeId npeId = (IxNpeMhNpeId)parameter;
+    UINT32 ofint;
+    volatile UINT32 *statusReg =
+        (UINT32 *)ixNpeMhConfigNpeInfo[npeId].statusRegister;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhConfigIsr\n");
+
+    /* get the OFINT (OutFifo interrupt) bit of the status register */
+    IX_NPEMH_REGISTER_READ_BITS (statusReg, &ofint, IX_NPEMH_NPE_STAT_OFINT);
+
+    /* if the OFINT status bit is set */
+    if (ofint)
+    {
+        /* if there is an ISR registered for this NPE */
+        if (ixNpeMhConfigNpeInfo[npeId].isr != NULL)
+        {
+            /* invoke the ISR routine */
+            ixNpeMhConfigNpeInfo[npeId].isr (npeId);
+        }
+        else
+        {
+            /* if we don't service the interrupt the NPE will continue */
+            /* to trigger the interrupt indefinitely */
+            IX_NPEMH_ERROR_REPORT ("No ISR registered to service "
+                                   "interrupt\n");
+        }
+    }
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhConfigIsr\n");
+}
+
+/*
+ * Function definition: ixNpeMhConfigInitialize
+ */
+
+void ixNpeMhConfigInitialize (
+    IxNpeMhNpeInterrupts npeInterrupts)
+{
+    IxNpeMhNpeId npeId;
+    UINT32 virtualAddr[IX_NPEMH_NUM_NPES];
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhConfigInitialize\n");
+
+    /* Request a mapping for the NPE-A config register address space */
+    virtualAddr[IX_NPEMH_NPEID_NPEA] =
+	(UINT32) IX_OSAL_MEM_MAP (IX_NPEMH_NPEA_BASE,
+				     IX_OSAL_IXP400_NPEA_MAP_SIZE);
+    IX_OSAL_ASSERT (virtualAddr[IX_NPEMH_NPEID_NPEA]);
+
+    /* Request a mapping for the NPE-B config register address space */
+    virtualAddr[IX_NPEMH_NPEID_NPEB] =
+	(UINT32) IX_OSAL_MEM_MAP (IX_NPEMH_NPEB_BASE,
+				     IX_OSAL_IXP400_NPEB_MAP_SIZE);
+    IX_OSAL_ASSERT (virtualAddr[IX_NPEMH_NPEID_NPEB]);
+    
+    /* Request a mapping for the NPE-C config register address space */
+    virtualAddr[IX_NPEMH_NPEID_NPEC] =
+	(UINT32) IX_OSAL_MEM_MAP (IX_NPEMH_NPEC_BASE,
+				     IX_OSAL_IXP400_NPEC_MAP_SIZE);
+    IX_OSAL_ASSERT (virtualAddr[IX_NPEMH_NPEID_NPEC]);
+
+    /* for each NPE ... */
+    for (npeId = 0; npeId < IX_NPEMH_NUM_NPES; npeId++)
+    {
+        /* declare a convenience pointer */
+        IxNpeMhConfigNpeInfo *npeInfo = &ixNpeMhConfigNpeInfo[npeId];
+	
+	/* store the virtual addresses of the NPE registers for later use */
+	npeInfo->virtualRegisterBase  = virtualAddr[npeId];
+	npeInfo->statusRegister  = virtualAddr[npeId] + IX_NPEMH_NPESTAT_OFFSET;
+	npeInfo->controlRegister = virtualAddr[npeId] + IX_NPEMH_NPECTL_OFFSET;
+	npeInfo->inFifoRegister  = virtualAddr[npeId] + IX_NPEMH_NPEFIFO_OFFSET;
+	npeInfo->outFifoRegister = virtualAddr[npeId] + IX_NPEMH_NPEFIFO_OFFSET;
+
+        /* for test purposes - to verify the register addresses */
+        IX_NPEMH_TRACE2 (IX_NPEMH_DEBUG, "NPE %d status register  = "
+                         "0x%08X\n", npeId, npeInfo->statusRegister);
+        IX_NPEMH_TRACE2 (IX_NPEMH_DEBUG, "NPE %d control register = "
+                         "0x%08X\n", npeId, npeInfo->controlRegister);
+        IX_NPEMH_TRACE2 (IX_NPEMH_DEBUG, "NPE %d inFifo register  = "
+                         "0x%08X\n", npeId, npeInfo->inFifoRegister);
+        IX_NPEMH_TRACE2 (IX_NPEMH_DEBUG, "NPE %d outFifo register = "
+                         "0x%08X\n", npeId, npeInfo->outFifoRegister);
+
+        /* connect our ISR to the NPE interrupt */
+        (void) ixOsalIrqBind (
+            npeInfo->interruptId, ixNpeMhConfigIsr, (void *)npeId);
+
+        /* initialise a mutex for this NPE */
+        (void) ixOsalMutexInit (&npeInfo->mutex);
+
+        /* if we should service the NPE's "outFIFO not empty" interrupt */
+        if (npeInterrupts == IX_NPEMH_NPEINTERRUPTS_YES)
+        {
+            /* enable the NPE's "outFIFO not empty" interrupt */
+            ixNpeMhConfigNpeInterruptEnable (npeId);
+        }
+        else
+        {
+            /* disable the NPE's "outFIFO not empty" interrupt */
+            ixNpeMhConfigNpeInterruptDisable (npeId);
+        }
+    }
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhConfigInitialize\n");
+}
+
+/*
+ * Function definition: ixNpeMhConfigUninit
+ */
+
+void ixNpeMhConfigUninit (void)
+{
+    IxNpeMhNpeId npeId;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhConfigUninit\n");
+
+    /* for each NPE ... */
+    for (npeId = 0; npeId < IX_NPEMH_NUM_NPES; npeId++)
+    {
+        /* declare a convenience pointer */
+        IxNpeMhConfigNpeInfo *npeInfo = &ixNpeMhConfigNpeInfo[npeId];
+        
+        /* disconnect ISR */
+        ixOsalIrqUnbind(npeInfo->interruptId);
+
+        /* destroy mutex associated with this NPE */
+        ixOsalMutexDestroy(&npeInfo->mutex);
+	
+	IX_OSAL_MEM_UNMAP (npeInfo->virtualRegisterBase);
+
+	npeInfo->virtualRegisterBase  = 0;
+	npeInfo->statusRegister  = 0;
+	npeInfo->controlRegister = 0;
+	npeInfo->inFifoRegister  = 0;
+	npeInfo->outFifoRegister = 0;
+    }
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhConfigUninit\n");
+}
+
+/*
+ * Function definition: ixNpeMhConfigIsrRegister
+ */
+
+void ixNpeMhConfigIsrRegister (
+    IxNpeMhNpeId npeId,
+    IxNpeMhConfigIsr isr)
+{
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhConfigIsrRegister\n");
+
+    /* check if there is already an ISR registered for this NPE */
+    if (ixNpeMhConfigNpeInfo[npeId].isr != NULL)
+    {
+        IX_NPEMH_TRACE0 (IX_NPEMH_DEBUG, "Over-writing registered NPE ISR\n");
+    }
+
+    /* save the ISR routine with the NPE info */
+    ixNpeMhConfigNpeInfo[npeId].isr = isr;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhConfigIsrRegister\n");
+}
+
+/*
+ * Function definition: ixNpeMhConfigNpeInterruptEnable
+ */
+
+BOOL ixNpeMhConfigNpeInterruptEnable (
+    IxNpeMhNpeId npeId)
+{
+    UINT32 ofe;
+    volatile UINT32 *controlReg =
+        (UINT32 *)ixNpeMhConfigNpeInfo[npeId].controlRegister;
+
+    /* get the OFE (OutFifoEnable) bit of the control register */
+    IX_NPEMH_REGISTER_READ_BITS (controlReg, &ofe, IX_NPEMH_NPE_CTL_OFE);
+
+    /* if the interrupt is disabled then we must enable it */
+    if (!ofe)
+    {
+        /* set the OFE (OutFifoEnable) bit of the control register */
+        /* we must set the OFEWE (OutFifoEnableWriteEnable) at the same */
+        /* time for the write to have effect */
+        IX_NPEMH_REGISTER_WRITE_BITS (controlReg,
+                                      (IX_NPEMH_NPE_CTL_OFE |
+                                       IX_NPEMH_NPE_CTL_OFEWE),
+                                      (IX_NPEMH_NPE_CTL_OFE |
+                                       IX_NPEMH_NPE_CTL_OFEWE));
+    }
+
+    /* return the previous state of the interrupt */
+    return (ofe != 0);
+}
+
+/*
+ * Function definition: ixNpeMhConfigNpeInterruptDisable
+ */
+
+BOOL ixNpeMhConfigNpeInterruptDisable (
+    IxNpeMhNpeId npeId)
+{
+    UINT32 ofe;
+    volatile UINT32 *controlReg =
+        (UINT32 *)ixNpeMhConfigNpeInfo[npeId].controlRegister;
+
+    /* get the OFE (OutFifoEnable) bit of the control register */
+    IX_NPEMH_REGISTER_READ_BITS (controlReg, &ofe, IX_NPEMH_NPE_CTL_OFE);
+
+    /* if the interrupt is enabled then we must disable it */
+    if (ofe)
+    {
+        /* unset the OFE (OutFifoEnable) bit of the control register */
+        /* we must set the OFEWE (OutFifoEnableWriteEnable) at the same */
+        /* time for the write to have effect */
+        IX_NPEMH_REGISTER_WRITE_BITS (controlReg,
+                                      (0                    |
+                                       IX_NPEMH_NPE_CTL_OFEWE),
+                                      (IX_NPEMH_NPE_CTL_OFE |
+                                       IX_NPEMH_NPE_CTL_OFEWE));
+    }
+
+    /* return the previous state of the interrupt */
+    return (ofe != 0);
+}
+
+/*
+ * Function definition: ixNpeMhConfigMessageIdGet
+ */
+
+IxNpeMhMessageId ixNpeMhConfigMessageIdGet (
+    IxNpeMhMessage message)
+{
+    /* return the most-significant byte of the first word of the */
+    /* message */
+    return ((IxNpeMhMessageId) ((message.data[0] >> 24) & 0xFF));
+}
+
+/*
+ * Function definition: ixNpeMhConfigNpeIdIsValid
+ */
+
+BOOL ixNpeMhConfigNpeIdIsValid (
+    IxNpeMhNpeId npeId)
+{
+    /* check that the npeId parameter is within the range of valid IDs */
+    return (npeId >= 0 && npeId < IX_NPEMH_NUM_NPES);
+}
+
+/*
+ * Function definition: ixNpeMhConfigLockGet
+ */
+
+void ixNpeMhConfigLockGet (
+    IxNpeMhNpeId npeId)
+{
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhConfigLockGet\n");
+
+    /* lock the mutex for this NPE */
+    (void) ixOsalMutexLock (&ixNpeMhConfigNpeInfo[npeId].mutex, 
+			    IX_OSAL_WAIT_FOREVER);
+
+    /* disable the NPE's "outFIFO not empty" interrupt */
+    ixNpeMhConfigNpeInfo[npeId].oldInterruptState =
+        ixNpeMhConfigNpeInterruptDisable (npeId);
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhConfigLockGet\n");
+}
+
+/*
+ * Function definition: ixNpeMhConfigLockRelease
+ */
+
+void ixNpeMhConfigLockRelease (
+    IxNpeMhNpeId npeId)
+{
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhConfigLockRelease\n");
+
+    /* if the interrupt was previously enabled */
+    if (ixNpeMhConfigNpeInfo[npeId].oldInterruptState)
+    {
+        /* enable the NPE's "outFIFO not empty" interrupt */
+        ixNpeMhConfigNpeInfo[npeId].oldInterruptState =
+            ixNpeMhConfigNpeInterruptEnable (npeId);
+    }
+
+    /* unlock the mutex for this NPE */
+    (void) ixOsalMutexUnlock (&ixNpeMhConfigNpeInfo[npeId].mutex);
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhConfigLockRelease\n");
+}
+
+/*
+ * Function definition: ixNpeMhConfigInFifoWrite
+ */
+
+IX_STATUS ixNpeMhConfigInFifoWrite (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessage message)
+{
+    volatile UINT32 *npeInFifo =
+        (UINT32 *)ixNpeMhConfigNpeInfo[npeId].inFifoRegister;
+    UINT32 retriesCount = 0;
+
+    /* write the first word of the message to the NPE's inFIFO */
+    IX_NPEMH_REGISTER_WRITE (npeInFifo, message.data[0]);
+
+    /* need to wait for room to write second word - see SCR #493,
+       poll for maximum number of retries, if exceed maximum
+       retries, exit from while loop */
+    while ((IX_NPE_MH_MAX_NUM_OF_RETRIES > retriesCount)
+        && ixNpeMhConfigInFifoIsFull (npeId))
+    {
+        retriesCount++;
+    }
+
+    /* Return TIMEOUT status to caller, indicate that NPE Hang / Halt */
+    if (IX_NPE_MH_MAX_NUM_OF_RETRIES == retriesCount)
+    {
+        return IX_NPEMH_CRITICAL_NPE_ERR;   
+    }    
+    
+    /* write the second word of the message to the NPE's inFIFO */
+    IX_NPEMH_REGISTER_WRITE (npeInFifo, message.data[1]);
+
+    /* record in the stats the maximum number of retries needed */
+    if (ixNpeMhConfigStats[npeId].maxInFifoFullRetries < retriesCount)
+    {
+	ixNpeMhConfigStats[npeId].maxInFifoFullRetries = retriesCount;
+    }
+
+    /* update statistical info */
+    ixNpeMhConfigStats[npeId].inFifoWrites++;
+    
+    return IX_SUCCESS;
+}
+
+/*
+ * Function definition: ixNpeMhConfigOutFifoRead
+ */
+
+IX_STATUS ixNpeMhConfigOutFifoRead (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessage *message)
+{
+    volatile UINT32 *npeOutFifo =
+        (UINT32 *)ixNpeMhConfigNpeInfo[npeId].outFifoRegister;
+    UINT32 retriesCount = 0;
+
+    /* read the first word of the message from the NPE's outFIFO */
+    IX_NPEMH_REGISTER_READ (npeOutFifo, &message->data[0]);
+
+    /* need to wait for NPE to write second word - see SCR #493 
+       poll for maximum number of retries, if exceed maximum
+       retries, exit from while loop */
+    while ((IX_NPE_MH_MAX_NUM_OF_RETRIES > retriesCount)
+        && ixNpeMhConfigOutFifoIsEmpty (npeId))
+    {
+        retriesCount++;
+    }
+
+    /* Return TIMEOUT status to caller, indicate that NPE Hang / Halt */
+    if (IX_NPE_MH_MAX_NUM_OF_RETRIES == retriesCount)
+    {
+        return IX_NPEMH_CRITICAL_NPE_ERR;   
+    } 
+    
+    /* read the second word of the message from the NPE's outFIFO */
+    IX_NPEMH_REGISTER_READ (npeOutFifo, &message->data[1]);
+
+    /* record in the stats the maximum number of retries needed */
+    if (ixNpeMhConfigStats[npeId].maxOutFifoEmptyRetries < retriesCount)
+    {
+	ixNpeMhConfigStats[npeId].maxOutFifoEmptyRetries = retriesCount;
+    }
+
+    /* update statistical info */
+    ixNpeMhConfigStats[npeId].outFifoReads++;
+    
+    return IX_SUCCESS;
+}
+
+/*
+ * Function definition: ixNpeMhConfigShow
+ */
+
+void ixNpeMhConfigShow (
+    IxNpeMhNpeId npeId)
+{
+    /* show the message fifo read counter */
+    IX_NPEMH_SHOW ("Message FIFO reads",
+                   ixNpeMhConfigStats[npeId].outFifoReads);
+
+    /* show the message fifo write counter */
+    IX_NPEMH_SHOW ("Message FIFO writes",
+                   ixNpeMhConfigStats[npeId].inFifoWrites);
+
+    /* show the max retries performed when inFIFO full */
+    IX_NPEMH_SHOW ("Max inFIFO Full retries",
+		   ixNpeMhConfigStats[npeId].maxInFifoFullRetries);
+
+    /* show the max retries performed when outFIFO empty */
+    IX_NPEMH_SHOW ("Max outFIFO Empty retries",
+		   ixNpeMhConfigStats[npeId].maxOutFifoEmptyRetries);
+
+    /* show the current status of the inFifo */
+    ixOsalLog (IX_OSAL_LOG_LVL_USER, IX_OSAL_LOG_DEV_STDOUT,
+	       "InFifo is %s and %s\n",
+	       (ixNpeMhConfigInFifoIsEmpty (npeId) ? 
+		(int) "EMPTY" : (int) "NOT EMPTY"),
+	       (ixNpeMhConfigInFifoIsFull (npeId) ? 
+		(int) "FULL" : (int) "NOT FULL"),
+	       0, 0, 0, 0);
+
+    /* show the current status of the outFifo */
+    ixOsalLog (IX_OSAL_LOG_LVL_USER, IX_OSAL_LOG_DEV_STDOUT,
+	       "OutFifo is %s and %s\n",
+	       (ixNpeMhConfigOutFifoIsEmpty (npeId) ? 
+		(int) "EMPTY" : (int) "NOT EMPTY"),
+	       (ixNpeMhConfigOutFifoIsFull (npeId) ? 
+		(int) "FULL" : (int) "NOT FULL"),
+	       0, 0, 0, 0);
+}
+
+/*
+ * Function definition: ixNpeMhConfigShowReset
+ */
+
+void ixNpeMhConfigShowReset (
+    IxNpeMhNpeId npeId)
+{
+    /* reset the message fifo read counter */
+    ixNpeMhConfigStats[npeId].outFifoReads = 0;
+
+    /* reset the message fifo write counter */
+    ixNpeMhConfigStats[npeId].inFifoWrites = 0;
+
+    /* reset the max inFIFO Full retries counter */
+    ixNpeMhConfigStats[npeId].maxInFifoFullRetries = 0;
+
+    /* reset the max outFIFO empty retries counter */
+    ixNpeMhConfigStats[npeId].maxOutFifoEmptyRetries = 0;
+}
+
+

cpu/ixp/npe/IxNpeMhReceive.c

@@ -0,0 +1,320 @@
+/**
+ * @file IxNpeMhReceive.c
+ *
+ * @author Intel Corporation
+ * @date 18 Jan 2002
+ *
+ * @brief This file contains the implementation of the private API for the
+ * Receive module.
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+/*
+ * Put the system defined include files required.
+ */
+
+
+/*
+ * Put the user defined include files required.
+ */
+#include "IxOsal.h"
+#include "IxNpeMhMacros_p.h"
+#include "IxNpeMhConfig_p.h"
+#include "IxNpeMhReceive_p.h"
+#include "IxNpeMhSolicitedCbMgr_p.h"
+#include "IxNpeMhUnsolicitedCbMgr_p.h"
+
+/*
+ * #defines and macros used in this file.
+ */
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+
+/**
+ * @struct IxNpeMhReceiveStats
+ *
+ * @brief This structure is used to maintain statistics for the Receive
+ * module.
+ */
+
+typedef struct
+{
+    UINT32 isrs;        /**< receive ISR invocations */
+    UINT32 receives;    /**< receive messages invocations */
+    UINT32 messages;    /**< messages received */
+    UINT32 solicited;   /**< solicited messages received */
+    UINT32 unsolicited; /**< unsolicited messages received */
+    UINT32 callbacks;   /**< callbacks invoked */
+} IxNpeMhReceiveStats;
+
+/*
+ * Variable declarations global to this file only.  Externs are followed by
+ * static variables.
+ */
+
+PRIVATE IxNpeMhReceiveStats ixNpeMhReceiveStats[IX_NPEMH_NUM_NPES];
+
+/*
+ * Extern function prototypes.
+ */
+
+/*
+ * Static function prototypes.
+ */
+PRIVATE
+void ixNpeMhReceiveIsr (int npeId);
+
+PRIVATE
+void ixNpeMhReceiveIsr (int npeId)
+{
+    int lockKey;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhReceiveIsr\n");
+
+    lockKey = ixOsalIrqLock ();
+
+    /* invoke the message receive routine to get messages from the NPE */
+    ixNpeMhReceiveMessagesReceive (npeId);
+
+    /* update statistical info */
+    ixNpeMhReceiveStats[npeId].isrs++;
+
+    ixOsalIrqUnlock (lockKey);
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhReceiveIsr\n");
+}
+
+/*
+ * Function definition: ixNpeMhReceiveInitialize
+ */
+
+void ixNpeMhReceiveInitialize (void)
+{
+    IxNpeMhNpeId npeId = 0;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhReceiveInitialize\n");
+
+    /* for each NPE ... */
+    for (npeId = 0; npeId < IX_NPEMH_NUM_NPES; npeId++)
+    {
+        /* register our internal ISR for the NPE to handle "outFIFO not */
+        /* empty" interrupts */
+        ixNpeMhConfigIsrRegister (npeId, ixNpeMhReceiveIsr);
+    }
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhReceiveInitialize\n");
+}
+
+/*
+ * Function definition: ixNpeMhReceiveMessagesReceive
+ */
+
+IX_STATUS ixNpeMhReceiveMessagesReceive (
+    IxNpeMhNpeId npeId)
+{
+    IxNpeMhMessage message = { { 0, 0 } };
+    IxNpeMhMessageId messageId = 0;
+    IxNpeMhCallback callback = NULL;
+    IX_STATUS status;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhReceiveMessagesReceive\n");
+
+    /* update statistical info */
+    ixNpeMhReceiveStats[npeId].receives++;
+
+    /* while the NPE has messages in its outFIFO */
+    while (!ixNpeMhConfigOutFifoIsEmpty (npeId))
+    {
+        /* read a message from the NPE's outFIFO */
+        status = ixNpeMhConfigOutFifoRead (npeId, &message);
+
+        if (IX_SUCCESS != status)
+        {
+            return status;
+        }
+        
+        /* get the ID of the message */
+        messageId = ixNpeMhConfigMessageIdGet (message);
+
+	    IX_NPEMH_TRACE2 (IX_NPEMH_DEBUG,
+			 "Received message from NPE %d with ID 0x%02X\n",
+			 npeId, messageId);
+
+        /* update statistical info */
+        ixNpeMhReceiveStats[npeId].messages++;
+
+        /* try to find a matching unsolicited callback for this message. */
+
+        /* we assume the message is unsolicited.  only if there is no */
+        /* unsolicited callback for this message type do we assume the */
+        /* message is solicited.  it is much faster to check for an */
+        /* unsolicited callback, so doing this check first should result */
+        /* in better performance. */
+
+        ixNpeMhUnsolicitedCbMgrCallbackRetrieve (
+            npeId, messageId, &callback);
+
+        if (callback != NULL)
+        {
+	    IX_NPEMH_TRACE0 (IX_NPEMH_DEBUG,
+			     "Found matching unsolicited callback\n");
+
+            /* update statistical info */
+            ixNpeMhReceiveStats[npeId].unsolicited++;
+        }
+
+        /* if no unsolicited callback was found try to find a matching */
+        /* solicited callback for this message */
+        if (callback == NULL)
+        {
+            ixNpeMhSolicitedCbMgrCallbackRetrieve (
+                npeId, messageId, &callback);
+
+            if (callback != NULL)
+            {
+		IX_NPEMH_TRACE0 (IX_NPEMH_DEBUG,
+				 "Found matching solicited callback\n");
+
+                /* update statistical info */
+                ixNpeMhReceiveStats[npeId].solicited++;
+            }
+        }
+
+        /* if a callback (either unsolicited or solicited) was found */
+        if (callback != NULL)
+        {
+            /* invoke the callback to pass the message back to the client */
+            callback (npeId, message);
+
+            /* update statistical info */
+            ixNpeMhReceiveStats[npeId].callbacks++;
+        }
+        else /* no callback (neither unsolicited nor solicited) was found */
+        {
+	    IX_NPEMH_TRACE2 (IX_NPEMH_WARNING,
+			     "No matching callback for NPE %d"
+			     " and ID 0x%02X, discarding message\n",
+			     npeId, messageId);
+
+            /* the message will be discarded.  this is normal behaviour */
+            /* if the client passes a NULL solicited callback when */
+            /* sending a message.  this indicates that the client is not */
+            /* interested in receiving the response.  alternatively a */
+            /* NULL callback here may signify an unsolicited message */
+            /* with no appropriate registered callback. */
+        }
+    }
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhReceiveMessagesReceive\n");
+    
+    return IX_SUCCESS;
+}
+
+/*
+ * Function definition: ixNpeMhReceiveShow
+ */
+
+void ixNpeMhReceiveShow (
+    IxNpeMhNpeId npeId)
+{
+    /* show the ISR invocation counter */
+    IX_NPEMH_SHOW ("Receive ISR invocations",
+                   ixNpeMhReceiveStats[npeId].isrs);
+
+    /* show the receive message invocation counter */
+    IX_NPEMH_SHOW ("Receive messages invocations",
+                   ixNpeMhReceiveStats[npeId].receives);
+
+    /* show the message received counter */
+    IX_NPEMH_SHOW ("Messages received",
+                   ixNpeMhReceiveStats[npeId].messages);
+
+    /* show the solicited message counter */
+    IX_NPEMH_SHOW ("Solicited messages received",
+                   ixNpeMhReceiveStats[npeId].solicited);
+
+    /* show the unsolicited message counter */
+    IX_NPEMH_SHOW ("Unsolicited messages received",
+                   ixNpeMhReceiveStats[npeId].unsolicited);
+
+    /* show the callback invoked counter */
+    IX_NPEMH_SHOW ("Callbacks invoked",
+                   ixNpeMhReceiveStats[npeId].callbacks);
+
+    /* show the message discarded counter */
+    IX_NPEMH_SHOW ("Received messages discarded",
+                   (ixNpeMhReceiveStats[npeId].messages -
+                    ixNpeMhReceiveStats[npeId].callbacks));
+}
+
+/*
+ * Function definition: ixNpeMhReceiveShowReset
+ */
+
+void ixNpeMhReceiveShowReset (
+    IxNpeMhNpeId npeId)
+{
+    /* reset the ISR invocation counter */
+    ixNpeMhReceiveStats[npeId].isrs = 0;
+
+    /* reset the receive message invocation counter */
+    ixNpeMhReceiveStats[npeId].receives = 0;
+
+    /* reset the message received counter */
+    ixNpeMhReceiveStats[npeId].messages = 0;
+
+    /* reset the solicited message counter */
+    ixNpeMhReceiveStats[npeId].solicited = 0;
+
+    /* reset the unsolicited message counter */
+    ixNpeMhReceiveStats[npeId].unsolicited = 0;
+
+    /* reset the callback invoked counter */
+    ixNpeMhReceiveStats[npeId].callbacks = 0;
+}

cpu/ixp/npe/IxNpeMhSend.c

@@ -0,0 +1,307 @@
+/**
+ * @file IxNpeMhSend.c
+ *
+ * @author Intel Corporation
+ * @date 18 Jan 2002
+ *
+ * @brief This file contains the implementation of the private API for the
+ * Send module.
+ *
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+/*
+ * Put the system defined include files required.
+ */
+
+
+/*
+ * Put the user defined include files required.
+ */
+
+#include "IxNpeMhMacros_p.h"
+
+#include "IxNpeMhConfig_p.h"
+#include "IxNpeMhSend_p.h"
+#include "IxNpeMhSolicitedCbMgr_p.h"
+
+/*
+ * #defines and macros used in this file.
+ */
+
+/**
+ * @def IX_NPEMH_INFIFO_RETRY_DELAY_US
+ *
+ * @brief Amount of time (uSecs) to delay between retries
+ * while inFIFO is Full when attempting to send a message
+ */
+#define IX_NPEMH_INFIFO_RETRY_DELAY_US (1)
+
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+
+/**
+ * @struct IxNpeMhSendStats
+ *
+ * @brief This structure is used to maintain statistics for the Send
+ * module.
+ */
+
+typedef struct
+{
+    UINT32 sends;             /**< send invocations */
+    UINT32 sendWithResponses; /**< send with response invocations */
+    UINT32 queueFulls;        /**< fifo queue full occurrences */
+    UINT32 queueFullRetries;  /**< fifo queue full retry occurrences */
+    UINT32 maxQueueFullRetries; /**< max fifo queue full retries */
+    UINT32 callbackFulls;     /**< callback list full occurrences */
+} IxNpeMhSendStats;
+
+/*
+ * Variable declarations global to this file only.  Externs are followed by
+ * static variables.
+ */
+
+PRIVATE IxNpeMhSendStats ixNpeMhSendStats[IX_NPEMH_NUM_NPES];
+
+/*
+ * Extern function prototypes.
+ */
+
+/*
+ * Static function prototypes.
+ */
+PRIVATE
+BOOL ixNpeMhSendInFifoIsFull(
+    IxNpeMhNpeId npeId,
+    UINT32 maxSendRetries);
+
+/*
+ * Function definition: ixNpeMhSendInFifoIsFull
+ */
+
+PRIVATE
+BOOL ixNpeMhSendInFifoIsFull(
+    IxNpeMhNpeId npeId,
+    UINT32 maxSendRetries)
+{
+    BOOL isFull = FALSE;
+    UINT32 numRetries = 0;
+
+    /* check the NPE's inFIFO */
+    isFull = ixNpeMhConfigInFifoIsFull (npeId);
+
+    /* we retry a few times, just to give the NPE a chance to read from */
+    /* the FIFO if the FIFO is currently full */
+    while (isFull && (numRetries++ < maxSendRetries))
+    {
+	if (numRetries >= IX_NPEMH_SEND_RETRIES_DEFAULT)
+	{
+	    /* Delay here for as short a time as possible (1 us). */
+	    /* Adding a delay here should ensure we are not hogging */
+	    /* the AHB bus while we are retrying                    */
+	    ixOsalBusySleep (IX_NPEMH_INFIFO_RETRY_DELAY_US);
+	}
+
+        /* re-check the NPE's inFIFO */
+        isFull = ixNpeMhConfigInFifoIsFull (npeId);
+
+        /* update statistical info */
+        ixNpeMhSendStats[npeId].queueFullRetries++;
+    }
+
+    /* record the highest number of retries that occurred */
+    if (ixNpeMhSendStats[npeId].maxQueueFullRetries < numRetries)
+    {
+	ixNpeMhSendStats[npeId].maxQueueFullRetries = numRetries;
+    }
+
+    if (isFull)
+    {
+        /* update statistical info */
+        ixNpeMhSendStats[npeId].queueFulls++;
+    }
+
+    return isFull;
+}
+
+/*
+ * Function definition: ixNpeMhSendMessageSend
+ */
+
+IX_STATUS ixNpeMhSendMessageSend (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessage message,
+    UINT32 maxSendRetries)
+{
+    IX_STATUS status;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhSendMessageSend\n");
+
+    /* update statistical info */
+    ixNpeMhSendStats[npeId].sends++;
+
+    /* check if the NPE's inFIFO is full - if so return an error */
+    if (ixNpeMhSendInFifoIsFull (npeId, maxSendRetries))
+    {
+        IX_NPEMH_TRACE0 (IX_NPEMH_WARNING, "NPE's inFIFO is full\n");
+        return IX_FAIL;
+    }
+
+    /* write the message to the NPE's inFIFO */
+    status = ixNpeMhConfigInFifoWrite (npeId, message);
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhSendMessageSend\n");
+
+    return status;
+}
+
+/*
+ * Function definition: ixNpeMhSendMessageWithResponseSend
+ */
+
+IX_STATUS ixNpeMhSendMessageWithResponseSend (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessage message,
+    IxNpeMhMessageId solicitedMessageId,
+    IxNpeMhCallback solicitedCallback,
+    UINT32 maxSendRetries)
+{
+    IX_STATUS status = IX_SUCCESS;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhSendMessageWithResponseSend\n");
+
+    /* update statistical info */
+    ixNpeMhSendStats[npeId].sendWithResponses++;
+
+    /* sr: this sleep will call the receive routine (no interrupts used!!!) */
+    ixOsalSleep (IX_NPEMH_INFIFO_RETRY_DELAY_US);
+
+    /* check if the NPE's inFIFO is full - if so return an error */
+    if (ixNpeMhSendInFifoIsFull (npeId, maxSendRetries))
+    {
+        IX_NPEMH_TRACE0 (IX_NPEMH_WARNING, "NPE's inFIFO is full\n");
+        return IX_FAIL;
+    }
+
+    /* save the solicited callback */
+    status = ixNpeMhSolicitedCbMgrCallbackSave (
+        npeId, solicitedMessageId, solicitedCallback);
+    if (status != IX_SUCCESS)
+    {
+        IX_NPEMH_ERROR_REPORT ("Failed to save solicited callback\n");
+
+        /* update statistical info */
+        ixNpeMhSendStats[npeId].callbackFulls++;
+
+        return status;
+    }
+
+    /* write the message to the NPE's inFIFO */
+    status = ixNpeMhConfigInFifoWrite (npeId, message);
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhSendMessageWithResponseSend\n");
+
+    return status;
+}
+
+/*
+ * Function definition: ixNpeMhSendShow
+ */
+
+void ixNpeMhSendShow (
+    IxNpeMhNpeId npeId)
+{
+    /* show the message send invocation counter */
+    IX_NPEMH_SHOW ("Send invocations",
+                   ixNpeMhSendStats[npeId].sends);
+
+    /* show the message send with response invocation counter */
+    IX_NPEMH_SHOW ("Send with response invocations",
+                   ixNpeMhSendStats[npeId].sendWithResponses);
+
+    /* show the fifo queue full occurrence counter */
+    IX_NPEMH_SHOW ("Fifo queue full occurrences",
+                   ixNpeMhSendStats[npeId].queueFulls);
+
+    /* show the fifo queue full retry occurrence counter */
+    IX_NPEMH_SHOW ("Fifo queue full retry occurrences",
+                   ixNpeMhSendStats[npeId].queueFullRetries);
+
+    /* show the fifo queue full maximum retries counter */
+    IX_NPEMH_SHOW ("Maximum fifo queue full retries",
+                   ixNpeMhSendStats[npeId].maxQueueFullRetries);
+
+    /* show the callback list full occurrence counter */
+    IX_NPEMH_SHOW ("Solicited callback list full occurrences",
+                   ixNpeMhSendStats[npeId].callbackFulls);
+}
+
+/*
+ * Function definition: ixNpeMhSendShowReset
+ */
+
+void ixNpeMhSendShowReset (
+    IxNpeMhNpeId npeId)
+{
+    /* reset the message send invocation counter */
+    ixNpeMhSendStats[npeId].sends = 0;
+
+    /* reset the message send with response invocation counter */
+    ixNpeMhSendStats[npeId].sendWithResponses = 0;
+
+    /* reset the fifo queue full occurrence counter */
+    ixNpeMhSendStats[npeId].queueFulls = 0;
+
+    /* reset the fifo queue full retry occurrence counter */
+    ixNpeMhSendStats[npeId].queueFullRetries = 0;
+
+    /* reset the max fifo queue full retries counter */
+    ixNpeMhSendStats[npeId].maxQueueFullRetries = 0;
+
+    /* reset the callback list full occurrence counter */
+    ixNpeMhSendStats[npeId].callbackFulls = 0;
+}

cpu/ixp/npe/IxNpeMhSolicitedCbMgr.c

@@ -0,0 +1,358 @@
+/**
+ * @file IxNpeMhSolicitedCbMgr.c
+ *
+ * @author Intel Corporation
+ * @date 18 Jan 2002
+ *
+ * @brief This file contains the implementation of the private API for the
+ * Solicited Callback Manager module.
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+#ifndef IXNPEMHCONFIG_P_H
+#	define IXNPEMHSOLICITEDCBMGR_C
+#else
+#	error "Error, IxNpeMhConfig_p.h should not be included before this definition."
+#endif
+
+/*
+ * Put the system defined include files required.
+ */
+
+
+/*
+ * Put the user defined include files required.
+ */
+
+#include "IxOsal.h"
+
+#include "IxNpeMhMacros_p.h"
+#include "IxNpeMhSolicitedCbMgr_p.h"
+#include "IxNpeMhConfig_p.h"
+/*
+ * #defines and macros used in this file.
+ */
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+
+/**
+ * @struct IxNpeMhSolicitedCallbackListEntry
+ *
+ * @brief This structure is used to store the information associated with
+ * an entry in the callback list.  This consists of the ID of the send
+ * message (which indicates the ID of the corresponding response message)
+ * and the callback function pointer itself.
+ *
+ */
+
+typedef struct IxNpeMhSolicitedCallbackListEntry
+{
+    /** message ID */
+    IxNpeMhMessageId messageId;
+
+    /** callback function pointer */
+    IxNpeMhCallback callback;
+
+    /** pointer to next entry in the list */
+    struct IxNpeMhSolicitedCallbackListEntry *next;
+} IxNpeMhSolicitedCallbackListEntry;
+
+/**
+ * @struct IxNpeMhSolicitedCallbackList
+ *
+ * @brief This structure is used to maintain the list of response
+ * callbacks.  The number of entries in this list will be variable, and
+ * they will be stored in a linked list fashion for ease of addition and
+ * removal.  The entries themselves are statically allocated, and are
+ * organised into a "free" list and a "callback" list.  Adding an entry
+ * means taking an entry from the "free" list and adding it to the
+ * "callback" list.  Removing an entry means removing it from the
+ * "callback" list and returning it to the "free" list.
+ */
+
+typedef struct
+{
+    /** pointer to the head of the free list */
+    IxNpeMhSolicitedCallbackListEntry *freeHead;
+
+    /** pointer to the head of the callback list */
+    IxNpeMhSolicitedCallbackListEntry *callbackHead;
+
+    /** pointer to the tail of the callback list */
+    IxNpeMhSolicitedCallbackListEntry *callbackTail;
+
+    /** array of entries - the first entry is used as a dummy entry to */
+    /* avoid the scenario of having an empty list, hence '+ 1' */
+    IxNpeMhSolicitedCallbackListEntry entries[IX_NPEMH_MAX_CALLBACKS + 1];
+} IxNpeMhSolicitedCallbackList;
+
+/**
+ * @struct IxNpeMhSolicitedCbMgrStats
+ *
+ * @brief This structure is used to maintain statistics for the Solicited
+ * Callback Manager module.
+ */
+
+typedef struct
+{
+    UINT32 saves;     /**< callback list saves */
+    UINT32 retrieves; /**< callback list retrieves */
+} IxNpeMhSolicitedCbMgrStats;
+
+/*
+ * Variable declarations global to this file only.  Externs are followed by
+ * static variables.
+ */
+
+PRIVATE IxNpeMhSolicitedCallbackList
+ixNpeMhSolicitedCbMgrCallbackLists[IX_NPEMH_NUM_NPES];
+
+PRIVATE IxNpeMhSolicitedCbMgrStats
+ixNpeMhSolicitedCbMgrStats[IX_NPEMH_NUM_NPES];
+
+/*
+ * Extern function prototypes.
+ */
+
+/*
+ * Static function prototypes.
+ */
+
+/*
+ * Function definition: ixNpeMhSolicitedCbMgrInitialize
+ */
+
+void ixNpeMhSolicitedCbMgrInitialize (void)
+{
+    IxNpeMhNpeId npeId;
+    UINT32 localIndex;
+    IxNpeMhSolicitedCallbackList *list = NULL;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhSolicitedCbMgrInitialize\n");
+
+    /* for each NPE ... */
+    for (npeId = 0; npeId < IX_NPEMH_NUM_NPES; npeId++)
+    {
+        /* initialise a pointer to the list for convenience */
+        list = &ixNpeMhSolicitedCbMgrCallbackLists[npeId];
+
+        /* for each entry in the list, after the dummy entry ... */
+        for (localIndex = 1; localIndex <= IX_NPEMH_MAX_CALLBACKS; localIndex++)
+        {
+            /* initialise the entry */
+            list->entries[localIndex].messageId = 0x00;
+            list->entries[localIndex].callback = NULL;
+
+            /* if this entry is before the last entry */
+            if (localIndex < IX_NPEMH_MAX_CALLBACKS)
+            {
+                /* chain this entry to the following entry */
+                list->entries[localIndex].next = &(list->entries[localIndex + 1]);
+            }
+            else /* this entry is the last entry */
+            {
+                /* the last entry isn't chained to anything */
+                list->entries[localIndex].next = NULL;
+            }
+        }
+
+        /* set the free list pointer to point to the first real entry */
+        /* (all real entries begin chained together on the free list) */
+        list->freeHead = &(list->entries[1]);
+
+        /* set the callback list pointers to point to the dummy entry */
+        /* (the callback list is initially empty) */
+        list->callbackHead = &(list->entries[0]);
+        list->callbackTail = &(list->entries[0]);
+    }
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhSolicitedCbMgrInitialize\n");
+}
+
+/*
+ * Function definition: ixNpeMhSolicitedCbMgrCallbackSave
+ */
+
+IX_STATUS ixNpeMhSolicitedCbMgrCallbackSave (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessageId solicitedMessageId,
+    IxNpeMhCallback solicitedCallback)
+{
+    IxNpeMhSolicitedCallbackList *list = NULL;
+    IxNpeMhSolicitedCallbackListEntry *callbackEntry = NULL;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhSolicitedCbMgrCallbackSave\n");
+
+    /* initialise a pointer to the list for convenience */
+    list = &ixNpeMhSolicitedCbMgrCallbackLists[npeId];
+
+    /* check to see if there are any entries in the free list */
+    if (list->freeHead == NULL)
+    {
+        IX_NPEMH_ERROR_REPORT ("Solicited callback list is full\n");
+        return IX_FAIL;
+    }
+
+    /* there is an entry in the free list we can use */
+
+    /* update statistical info */
+    ixNpeMhSolicitedCbMgrStats[npeId].saves++;
+
+    /* remove a callback entry from the start of the free list */
+    callbackEntry = list->freeHead;
+    list->freeHead = callbackEntry->next;
+
+    /* fill in the callback entry with the new data */
+    callbackEntry->messageId = solicitedMessageId;
+    callbackEntry->callback = solicitedCallback;
+
+    /* the new callback entry will be added to the tail of the callback */
+    /* list, so it isn't chained to anything */
+    callbackEntry->next = NULL;
+
+    /* chain new callback entry to the last entry of the callback list */
+    list->callbackTail->next = callbackEntry;
+    list->callbackTail = callbackEntry;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhSolicitedCbMgrCallbackSave\n");
+
+    return IX_SUCCESS;
+}
+
+/*
+ * Function definition: ixNpeMhSolicitedCbMgrCallbackRetrieve
+ */
+
+void ixNpeMhSolicitedCbMgrCallbackRetrieve (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessageId solicitedMessageId,
+    IxNpeMhCallback *solicitedCallback)
+{
+    IxNpeMhSolicitedCallbackList *list = NULL;
+    IxNpeMhSolicitedCallbackListEntry *callbackEntry = NULL;
+    IxNpeMhSolicitedCallbackListEntry *previousEntry = NULL;
+
+    /* initialise a pointer to the list for convenience */
+    list = &ixNpeMhSolicitedCbMgrCallbackLists[npeId];
+
+    /* initialise the callback entry to the first entry of the callback */
+    /* list - we must skip over the dummy entry, which is the previous */
+    callbackEntry = list->callbackHead->next;
+    previousEntry = list->callbackHead;
+
+    /* traverse the callback list looking for an entry with a matching */
+    /* message ID.  note we also save the previous entry's pointer to */
+    /* allow us to unchain the matching entry from the callback list */
+    while ((callbackEntry != NULL) &&
+           (callbackEntry->messageId != solicitedMessageId))
+    {
+        previousEntry = callbackEntry;
+        callbackEntry = callbackEntry->next;
+    }
+
+    /* if we didn't find a matching callback entry */
+    if (callbackEntry == NULL)
+    {
+        /* return a NULL callback in the outgoing parameter */
+        *solicitedCallback = NULL;
+    }
+    else /* we found a matching callback entry */
+    {
+        /* update statistical info */
+        ixNpeMhSolicitedCbMgrStats[npeId].retrieves++;
+
+        /* return the callback in the outgoing parameter */
+        *solicitedCallback = callbackEntry->callback;
+
+        /* unchain callback entry by chaining previous entry to next */
+        previousEntry->next = callbackEntry->next;
+
+        /* if the callback entry is at the tail of the list */
+        if (list->callbackTail == callbackEntry)
+        {
+            /* update the tail of the callback list */
+            list->callbackTail = previousEntry;
+        }
+
+        /* re-initialise the callback entry */
+        callbackEntry->messageId = 0x00;
+        callbackEntry->callback = NULL;
+
+        /* add the callback entry to the start of the free list */
+        callbackEntry->next = list->freeHead;
+        list->freeHead = callbackEntry;
+    }
+}
+
+/*
+ * Function definition: ixNpeMhSolicitedCbMgrShow
+ */
+
+void ixNpeMhSolicitedCbMgrShow (
+    IxNpeMhNpeId npeId)
+{
+    /* show the solicited callback list save counter */
+    IX_NPEMH_SHOW ("Solicited callback list saves",
+                   ixNpeMhSolicitedCbMgrStats[npeId].saves);
+
+    /* show the solicited callback list retrieve counter */
+    IX_NPEMH_SHOW ("Solicited callback list retrieves",
+                   ixNpeMhSolicitedCbMgrStats[npeId].retrieves);
+}
+
+/*
+ * Function definition: ixNpeMhSolicitedCbMgrShowReset
+ */
+
+void ixNpeMhSolicitedCbMgrShowReset (
+    IxNpeMhNpeId npeId)
+{
+    /* reset the solicited callback list save counter */
+    ixNpeMhSolicitedCbMgrStats[npeId].saves = 0;
+
+    /* reset the solicited callback list retrieve counter */
+    ixNpeMhSolicitedCbMgrStats[npeId].retrieves = 0;
+}

cpu/ixp/npe/IxNpeMhUnsolicitedCbMgr.c

@@ -0,0 +1,246 @@
+/**
+ * @file IxNpeMhUnsolicitedCbMgr.c
+ *
+ * @author Intel Corporation
+ * @date 18 Jan 2002
+ *
+ * @brief This file contains the implementation of the private API for
+ * the Unsolicited Callback Manager module.
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+/*
+ * Put the system defined include files required.
+ */
+
+
+/*
+ * Put the user defined include files required.
+ */
+#include "IxOsal.h"
+
+#include "IxNpeMhMacros_p.h"
+
+#include "IxNpeMhUnsolicitedCbMgr_p.h"
+
+
+/*
+ * #defines and macros used in this file.
+ */
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+
+/**
+ * @struct IxNpeMhUnsolicitedCallbackTable
+ *
+ * @brief This structure is used to maintain the list of registered
+ * callbacks.  One entry exists for each message ID, and a NULL entry will
+ * signify that no callback has been registered for that ID.
+ */
+
+typedef struct
+{
+    /** array of entries */
+    IxNpeMhCallback entries[IX_NPEMH_MAX_MESSAGE_ID + 1];
+} IxNpeMhUnsolicitedCallbackTable;
+
+/**
+ * @struct IxNpeMhUnsolicitedCbMgrStats
+ *
+ * @brief This structure is used to maintain statistics for the Unsolicited
+ * Callback Manager module.
+ */
+
+typedef struct
+{
+    UINT32 saves;      /**< callback table saves */
+    UINT32 overwrites; /**< callback table overwrites */
+} IxNpeMhUnsolicitedCbMgrStats;
+
+/*
+ * Variable declarations global to this file only.  Externs are followed by
+ * static variables.
+ */
+
+PRIVATE IxNpeMhUnsolicitedCallbackTable
+ixNpeMhUnsolicitedCallbackTables[IX_NPEMH_NUM_NPES];
+
+PRIVATE IxNpeMhUnsolicitedCbMgrStats
+ixNpeMhUnsolicitedCbMgrStats[IX_NPEMH_NUM_NPES];
+
+/*
+ * Extern function prototypes.
+ */
+
+/*
+ * Static function prototypes.
+ */
+
+/*
+ * Function definition: ixNpeMhUnsolicitedCbMgrInitialize
+ */
+
+void ixNpeMhUnsolicitedCbMgrInitialize (void)
+{
+    IxNpeMhNpeId npeId = 0;
+    IxNpeMhUnsolicitedCallbackTable *table = NULL;
+    IxNpeMhMessageId messageId = 0;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhUnsolicitedCbMgrInitialize\n");
+
+    /* for each NPE ... */
+    for (npeId = 0; npeId < IX_NPEMH_NUM_NPES; npeId++)
+    {
+        /* initialise a pointer to the table for convenience */
+        table = &ixNpeMhUnsolicitedCallbackTables[npeId];
+
+        /* for each message ID ... */
+        for (messageId = IX_NPEMH_MIN_MESSAGE_ID;
+             messageId <= IX_NPEMH_MAX_MESSAGE_ID; messageId++)
+        {
+            /* initialise the callback for this message ID to NULL */
+            table->entries[messageId] = NULL;
+        }
+    }
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhUnsolicitedCbMgrInitialize\n");
+}
+
+/*
+ * Function definition: ixNpeMhUnsolicitedCbMgrCallbackSave
+ */
+
+void ixNpeMhUnsolicitedCbMgrCallbackSave (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessageId unsolicitedMessageId,
+    IxNpeMhCallback unsolicitedCallback)
+{
+    IxNpeMhUnsolicitedCallbackTable *table = NULL;
+
+    /* initialise a pointer to the table for convenience */
+    table = &ixNpeMhUnsolicitedCallbackTables[npeId];
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Entering "
+                     "ixNpeMhUnsolicitedCbMgrCallbackSave\n");
+
+    /* update statistical info */
+    ixNpeMhUnsolicitedCbMgrStats[npeId].saves++;
+
+    /* check if there is a callback already registered for this NPE and */
+    /* message ID */
+    if (table->entries[unsolicitedMessageId] != NULL)
+    {
+	/* if we are overwriting an existing callback */
+	if (unsolicitedCallback != NULL)
+	{
+	    IX_NPEMH_TRACE2 (IX_NPEMH_DEBUG, "Unsolicited callback "
+			     "overwriting existing callback for NPE ID %d "
+			     "message ID 0x%02X\n", npeId, unsolicitedMessageId);
+	}
+	else /* if we are clearing an existing callback */
+	{
+	    IX_NPEMH_TRACE2 (IX_NPEMH_DEBUG, "NULL unsolicited callback "
+			     "clearing existing callback for NPE ID %d "
+			     "message ID 0x%02X\n", npeId, unsolicitedMessageId);
+	}
+
+        /* update statistical info */
+        ixNpeMhUnsolicitedCbMgrStats[npeId].overwrites++;
+    }
+
+    /* save the callback into the table */
+    table->entries[unsolicitedMessageId] = unsolicitedCallback;
+
+    IX_NPEMH_TRACE0 (IX_NPEMH_FN_ENTRY_EXIT, "Exiting "
+                     "ixNpeMhUnsolicitedCbMgrCallbackSave\n");
+}
+
+/*
+ * Function definition: ixNpeMhUnsolicitedCbMgrCallbackRetrieve
+ */
+
+void ixNpeMhUnsolicitedCbMgrCallbackRetrieve (
+    IxNpeMhNpeId npeId,
+    IxNpeMhMessageId unsolicitedMessageId,
+    IxNpeMhCallback *unsolicitedCallback)
+{
+    IxNpeMhUnsolicitedCallbackTable *table = NULL;
+
+    /* initialise a pointer to the table for convenience */
+    table = &ixNpeMhUnsolicitedCallbackTables[npeId];
+
+    /* retrieve the callback from the table */
+    *unsolicitedCallback = table->entries[unsolicitedMessageId];
+}
+
+/*
+ * Function definition: ixNpeMhUnsolicitedCbMgrShow
+ */
+
+void ixNpeMhUnsolicitedCbMgrShow (
+    IxNpeMhNpeId npeId)
+{
+    /* show the unsolicited callback table save counter */
+    IX_NPEMH_SHOW ("Unsolicited callback table saves",
+                   ixNpeMhUnsolicitedCbMgrStats[npeId].saves);
+
+    /* show the unsolicited callback table overwrite counter */
+    IX_NPEMH_SHOW ("Unsolicited callback table overwrites",
+                   ixNpeMhUnsolicitedCbMgrStats[npeId].overwrites);
+}
+
+/*
+ * Function definition: ixNpeMhUnsolicitedCbMgrShowReset
+ */
+
+void ixNpeMhUnsolicitedCbMgrShowReset (
+    IxNpeMhNpeId npeId)
+{
+    /* reset the unsolicited callback table save counter */
+    ixNpeMhUnsolicitedCbMgrStats[npeId].saves = 0;
+
+    /* reset the unsolicited callback table overwrite counter */
+    ixNpeMhUnsolicitedCbMgrStats[npeId].overwrites = 0;
+}

cpu/ixp/npe/IxOsalBufferMgt.c

@@ -0,0 +1,800 @@
+/**
+ * @file IxOsalBufferMgt.c
+ *
+ * @brief Default buffer pool management and buffer management
+ *        Implementation.
+ *
+ * Design Notes:
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+/*
+ * OS may choose to use default bufferMgt by defining
+ * IX_OSAL_USE_DEFAULT_BUFFER_MGT in IxOsalOsBufferMgt.h  
+ */
+
+#include "IxOsal.h"
+
+#define IX_OSAL_BUFFER_FREE_PROTECTION  /* Define this to enable Illegal MBuf Freed Protection*/
+
+/*
+ * The implementation is only used when the following
+ * is defined.
+ */
+#ifdef IX_OSAL_USE_DEFAULT_BUFFER_MGT
+
+
+#define IX_OSAL_MBUF_SYS_SIGNATURE				(0x8BADF00D)
+#define IX_OSAL_MBUF_SYS_SIGNATURE_MASK				(0xEFFFFFFF)
+#define IX_OSAL_MBUF_USED_FLAG					(0x10000000)
+#define IX_OSAL_MBUF_SYS_SIGNATURE_INIT(bufPtr)        		IX_OSAL_MBUF_SIGNATURE (bufPtr) = (UINT32)IX_OSAL_MBUF_SYS_SIGNATURE
+
+/* 
+*  This implementation is protect, the buffer pool management's  ixOsalMBufFree 
+*  against an invalid MBUF pointer argument that already has been freed earlier 
+*  or in other words resides in the free pool of MBUFs. This added feature, 
+*  checks the MBUF "USED" FLAG. The Flag tells if the MBUF is still not freed 
+*  back to the Buffer Pool.
+*  Disable this feature for performance reasons by undef 
+*  IX_OSAL_BUFFER_FREE_PROTECTION macro.
+*/
+#ifdef IX_OSAL_BUFFER_FREE_PROTECTION  /*IX_OSAL_BUFFER_FREE_PROTECTION With Buffer Free protection*/
+
+#define IX_OSAL_MBUF_GET_SYS_SIGNATURE(bufPtr)		(IX_OSAL_MBUF_SIGNATURE (bufPtr)&(IX_OSAL_MBUF_SYS_SIGNATURE_MASK) )
+#define IX_OSAL_MBUF_SET_SYS_SIGNATURE(bufPtr)    do {																											\
+																									IX_OSAL_MBUF_SIGNATURE (bufPtr)&(~IX_OSAL_MBUF_SYS_SIGNATURE_MASK);\
+														    									IX_OSAL_MBUF_SIGNATURE (bufPtr)|=IX_OSAL_MBUF_SYS_SIGNATURE;			\
+																									}while(0)
+
+#define IX_OSAL_MBUF_SET_USED_FLAG(bufPtr)   IX_OSAL_MBUF_SIGNATURE (bufPtr)|=IX_OSAL_MBUF_USED_FLAG
+#define IX_OSAL_MBUF_CLEAR_USED_FLAG(bufPtr) IX_OSAL_MBUF_SIGNATURE (bufPtr)&=~IX_OSAL_MBUF_USED_FLAG
+#define IX_OSAL_MBUF_ISSET_USED_FLAG(bufPtr) (IX_OSAL_MBUF_SIGNATURE (bufPtr)&IX_OSAL_MBUF_USED_FLAG)
+
+#else
+
+#define IX_OSAL_MBUF_GET_SYS_SIGNATURE(bufPtr)	 IX_OSAL_MBUF_SIGNATURE (bufPtr)
+#define IX_OSAL_MBUF_SET_SYS_SIGNATURE(bufPtr)   IX_OSAL_MBUF_SIGNATURE (bufPtr) = IX_OSAL_MBUF_SYS_SIGNATURE
+
+#endif /*IX_OSAL_BUFFER_FREE_PROTECTION With Buffer Free protection*/
+/*
+ * Variable declarations global to this file only.  Externs are followed by
+ * static variables.
+ */
+
+/* 
+ * A unit of 32, used to provide bit-shift for pool
+ * management. Needs some work if users want more than 32 pools.
+ */
+#define IX_OSAL_BUFF_FREE_BITS 32
+
+PRIVATE UINT32 ixOsalBuffFreePools[IX_OSAL_MBUF_MAX_POOLS /
+    IX_OSAL_BUFF_FREE_BITS];
+
+PUBLIC IX_OSAL_MBUF_POOL ixOsalBuffPools[IX_OSAL_MBUF_MAX_POOLS];
+
+static int ixOsalBuffPoolsInUse = 0;
+
+#ifdef IX_OSAL_BUFFER_ALLOC_SEPARATELY
+PRIVATE IX_OSAL_MBUF *
+ixOsalBuffPoolMbufInit (UINT32 mbufSizeAligned,
+                      UINT32 dataSizeAligned,
+                      IX_OSAL_MBUF_POOL *poolPtr);
+#endif
+
+PRIVATE IX_OSAL_MBUF_POOL * ixOsalPoolAlloc (void);                      
+
+/*
+ * Function definition: ixOsalPoolAlloc
+ */
+
+/****************************/
+
+PRIVATE IX_OSAL_MBUF_POOL *
+ixOsalPoolAlloc (void)
+{
+    register unsigned int i = 0;
+
+    /*
+     * Scan for the first free buffer. Free buffers are indicated by 0
+     * on the corrsponding bit in ixOsalBuffFreePools. 
+     */
+    if (ixOsalBuffPoolsInUse >= IX_OSAL_MBUF_MAX_POOLS)
+    {
+        /*
+         * Fail to grab a ptr this time 
+         */
+        return NULL;
+    }
+
+    while (ixOsalBuffFreePools[i / IX_OSAL_BUFF_FREE_BITS] &
+        (1 << (i % IX_OSAL_BUFF_FREE_BITS)))
+        i++;
+    /*
+     * Free buffer found. Mark it as busy and initialize. 
+     */
+    ixOsalBuffFreePools[i / IX_OSAL_BUFF_FREE_BITS] |=
+        (1 << (i % IX_OSAL_BUFF_FREE_BITS));
+
+    memset (&ixOsalBuffPools[i], 0, sizeof (IX_OSAL_MBUF_POOL));
+
+    ixOsalBuffPools[i].poolIdx = i;
+    ixOsalBuffPoolsInUse++;
+
+    return &ixOsalBuffPools[i];
+}
+
+
+#ifdef IX_OSAL_BUFFER_ALLOC_SEPARATELY
+PRIVATE IX_OSAL_MBUF *
+ixOsalBuffPoolMbufInit (UINT32 mbufSizeAligned,
+                      UINT32 dataSizeAligned,
+                      IX_OSAL_MBUF_POOL *poolPtr)
+{
+    UINT8 *dataPtr;
+    IX_OSAL_MBUF *realMbufPtr;
+    /* Allocate cache-aligned memory for mbuf header */
+    realMbufPtr = (IX_OSAL_MBUF *) IX_OSAL_CACHE_DMA_MALLOC (mbufSizeAligned);
+    IX_OSAL_ASSERT (realMbufPtr != NULL);
+    memset (realMbufPtr, 0, mbufSizeAligned);
+
+    /* Allocate cache-aligned memory for mbuf data */
+    dataPtr = (UINT8 *) IX_OSAL_CACHE_DMA_MALLOC (dataSizeAligned);
+    IX_OSAL_ASSERT (dataPtr != NULL);
+    memset (dataPtr, 0, dataSizeAligned);
+
+    /* Fill in mbuf header fields */
+    IX_OSAL_MBUF_MDATA (realMbufPtr) = dataPtr;
+    IX_OSAL_MBUF_ALLOCATED_BUFF_DATA (realMbufPtr) = (UINT32)dataPtr;
+
+    IX_OSAL_MBUF_MLEN (realMbufPtr) = dataSizeAligned;
+    IX_OSAL_MBUF_ALLOCATED_BUFF_LEN (realMbufPtr) = dataSizeAligned;
+
+    IX_OSAL_MBUF_NET_POOL (realMbufPtr) = (IX_OSAL_MBUF_POOL *) poolPtr;
+
+    IX_OSAL_MBUF_SYS_SIGNATURE_INIT(realMbufPtr);
+
+    /* update some statistical information */
+    poolPtr->mbufMemSize += mbufSizeAligned;
+    poolPtr->dataMemSize += dataSizeAligned;
+
+    return realMbufPtr;
+}
+#endif /* #ifdef IX_OSAL_BUFFER_ALLOC_SEPARATELY */
+
+/*
+ * Function definition: ixOsalBuffPoolInit
+ */
+
+PUBLIC IX_OSAL_MBUF_POOL *
+ixOsalPoolInit (UINT32 count, UINT32 size, const char *name)
+{
+
+    /* These variables are only used if UX_OSAL_BUFFER_ALLOC_SEPERATELY
+     * is defined .
+     */
+#ifdef IX_OSAL_BUFFER_ALLOC_SEPARATELY
+    UINT32 i, mbufSizeAligned, dataSizeAligned;
+    IX_OSAL_MBUF *currentMbufPtr = NULL;
+#else
+    void *poolBufPtr;
+    void *poolDataPtr;
+    int mbufMemSize;
+    int dataMemSize;
+#endif
+
+    IX_OSAL_MBUF_POOL *poolPtr = NULL;
+    
+    if (count <= 0)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalPoolInit(): " "count = 0 \n", 0, 0, 0, 0, 0, 0);
+        return NULL;        
+    }
+
+    if (name == NULL)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalPoolInit(): " "NULL name \n", 0, 0, 0, 0, 0, 0);
+        return NULL;        
+    }
+    
+    if (strlen (name) > IX_OSAL_MBUF_POOL_NAME_LEN)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalPoolInit(): "
+            "ERROR - name length should be no greater than %d  \n",
+            IX_OSAL_MBUF_POOL_NAME_LEN, 0, 0, 0, 0, 0);
+        return NULL;
+    }
+
+/* OS can choose whether to allocate all buffers all together (if it 
+ * can handle a huge single alloc request), or to allocate buffers 
+ * separately by the defining IX_OSAL_BUFFER_ALLOC_SEPARATELY.
+ */
+#ifdef IX_OSAL_BUFFER_ALLOC_SEPARATELY
+    /* Get a pool Ptr */
+    poolPtr = ixOsalPoolAlloc ();
+
+    if (poolPtr == NULL)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalPoolInit(): " "Fail to Get PoolPtr \n", 0, 0, 0, 0, 0, 0);    
+        return NULL;
+    }
+
+    mbufSizeAligned = IX_OSAL_MBUF_POOL_SIZE_ALIGN (sizeof (IX_OSAL_MBUF));
+    dataSizeAligned = IX_OSAL_MBUF_POOL_SIZE_ALIGN(size);
+
+    poolPtr->nextFreeBuf = NULL;    
+    poolPtr->mbufMemPtr = NULL;    
+    poolPtr->dataMemPtr = NULL;
+    poolPtr->bufDataSize = dataSizeAligned;
+    poolPtr->totalBufsInPool = count;
+    poolPtr->poolAllocType = IX_OSAL_MBUF_POOL_TYPE_SYS_ALLOC;
+    strcpy (poolPtr->name, name);
+
+
+    for (i = 0; i < count; i++)
+    {
+	    /* create an mbuf */
+	    currentMbufPtr = ixOsalBuffPoolMbufInit (mbufSizeAligned,
+					         dataSizeAligned,
+					         poolPtr);
+
+#ifdef IX_OSAL_BUFFER_FREE_PROTECTION 		
+/* Set the Buffer USED Flag. If not, ixOsalMBufFree will fail.
+   ixOsalMbufFree used here is in a special case whereby, it's 
+   used to add MBUF to the Pool. By specification, ixOsalMbufFree 
+   deallocates an allocated MBUF from Pool.
+*/ 			         
+      IX_OSAL_MBUF_SET_USED_FLAG(currentMbufPtr);
+#endif                             
+	    /* Add it to the pool */
+	    ixOsalMbufFree (currentMbufPtr);
+
+	    /* flush the pool information to RAM */
+	    IX_OSAL_CACHE_FLUSH (currentMbufPtr, mbufSizeAligned);
+    }
+    
+    /*
+     * update the number of free buffers in the pool 
+     */
+    poolPtr->freeBufsInPool = count;
+
+#else 
+/* Otherwise allocate buffers in a continuous block fashion */    
+    poolBufPtr = IX_OSAL_MBUF_POOL_MBUF_AREA_ALLOC (count, mbufMemSize);
+    IX_OSAL_ASSERT (poolBufPtr != NULL);
+    poolDataPtr =
+        IX_OSAL_MBUF_POOL_DATA_AREA_ALLOC (count, size, dataMemSize);
+    IX_OSAL_ASSERT (poolDataPtr != NULL);
+
+    poolPtr = ixOsalNoAllocPoolInit (poolBufPtr, poolDataPtr,
+        count, size, name);
+    if (poolPtr == NULL)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalPoolInit(): " "Fail to get pool ptr \n", 0, 0, 0, 0, 0, 0);
+        return NULL;
+    }
+
+    poolPtr->poolAllocType = IX_OSAL_MBUF_POOL_TYPE_SYS_ALLOC;
+
+#endif /* IX_OSAL_BUFFER_ALLOC_SEPARATELY */
+    return poolPtr;
+}
+
+PUBLIC IX_OSAL_MBUF_POOL *
+ixOsalNoAllocPoolInit (void *poolBufPtr,
+    void *poolDataPtr, UINT32 count, UINT32 size, const char *name)
+{
+    UINT32 i,  mbufSizeAligned, sizeAligned;
+    IX_OSAL_MBUF *currentMbufPtr = NULL;
+    IX_OSAL_MBUF *nextMbufPtr = NULL;
+    IX_OSAL_MBUF_POOL *poolPtr = NULL;
+
+    /*
+     * check parameters 
+     */
+    if (poolBufPtr == NULL)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalNoAllocPoolInit(): "
+            "ERROR - NULL poolBufPtr \n", 0, 0, 0, 0, 0, 0);
+        return NULL;
+    }
+
+    if (count <= 0)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalNoAllocPoolInit(): "
+            "ERROR - count must > 0   \n", 0, 0, 0, 0, 0, 0);
+        return NULL;
+    }
+
+    if (name == NULL)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalNoAllocPoolInit(): "
+            "ERROR - NULL name ptr  \n", 0, 0, 0, 0, 0, 0);
+        return NULL;
+    }
+
+    if (strlen (name) > IX_OSAL_MBUF_POOL_NAME_LEN)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalNoAllocPoolInit(): "
+            "ERROR - name length should be no greater than %d  \n",
+            IX_OSAL_MBUF_POOL_NAME_LEN, 0, 0, 0, 0, 0);
+        return NULL;
+    }
+
+    poolPtr = ixOsalPoolAlloc ();
+
+    if (poolPtr == NULL)
+    {
+        return NULL;
+    }
+
+    /*
+     * Adjust sizes to ensure alignment on cache line boundaries 
+     */
+    mbufSizeAligned =
+        IX_OSAL_MBUF_POOL_SIZE_ALIGN (sizeof (IX_OSAL_MBUF));
+    /*
+     * clear the mbuf memory area 
+     */
+    memset (poolBufPtr, 0, mbufSizeAligned * count);
+
+    if (poolDataPtr != NULL)
+    {
+        /*
+         * Adjust sizes to ensure alignment on cache line boundaries 
+         */
+        sizeAligned = IX_OSAL_MBUF_POOL_SIZE_ALIGN (size);
+        /*
+         * clear the data memory area 
+         */
+        memset (poolDataPtr, 0, sizeAligned * count);
+    }
+    else
+    {
+        sizeAligned = 0;
+    }
+
+    /*
+     * initialise pool fields 
+     */
+    strcpy ((poolPtr)->name, name);
+
+    poolPtr->dataMemPtr = poolDataPtr;
+    poolPtr->mbufMemPtr = poolBufPtr;
+    poolPtr->bufDataSize = sizeAligned;
+    poolPtr->totalBufsInPool = count;
+    poolPtr->mbufMemSize = mbufSizeAligned * count;
+    poolPtr->dataMemSize = sizeAligned * count;
+
+    currentMbufPtr = (IX_OSAL_MBUF *) poolBufPtr;
+
+    poolPtr->nextFreeBuf = currentMbufPtr;
+
+    for (i = 0; i < count; i++)
+    {
+        if (i < (count - 1))
+        {
+            nextMbufPtr =
+                (IX_OSAL_MBUF *) ((unsigned) currentMbufPtr +
+                mbufSizeAligned);
+        }
+        else
+        {                       /* last mbuf in chain */
+            nextMbufPtr = NULL;
+        }
+        IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR (currentMbufPtr) = nextMbufPtr;
+        IX_OSAL_MBUF_NET_POOL (currentMbufPtr) = poolPtr;
+
+        IX_OSAL_MBUF_SYS_SIGNATURE_INIT(currentMbufPtr);
+
+        if (poolDataPtr != NULL)
+        {
+            IX_OSAL_MBUF_MDATA (currentMbufPtr) = poolDataPtr;
+            IX_OSAL_MBUF_ALLOCATED_BUFF_DATA(currentMbufPtr) = (UINT32) poolDataPtr;
+
+            IX_OSAL_MBUF_MLEN (currentMbufPtr) = sizeAligned;
+            IX_OSAL_MBUF_ALLOCATED_BUFF_LEN(currentMbufPtr) = sizeAligned;
+
+            poolDataPtr = (void *) ((unsigned) poolDataPtr + sizeAligned);
+        }
+
+        currentMbufPtr = nextMbufPtr;
+    }
+
+    /*
+     * update the number of free buffers in the pool 
+     */
+    poolPtr->freeBufsInPool = count;
+
+    poolPtr->poolAllocType = IX_OSAL_MBUF_POOL_TYPE_USER_ALLOC;
+
+    return poolPtr;
+}
+
+/* 
+ * Get a mbuf ptr from the pool
+ */
+PUBLIC IX_OSAL_MBUF *
+ixOsalMbufAlloc (IX_OSAL_MBUF_POOL * poolPtr)
+{
+    int lock;
+    IX_OSAL_MBUF *newBufPtr = NULL;
+
+    /*
+     * check parameters 
+     */
+    if (poolPtr == NULL)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalMbufAlloc(): "
+            "ERROR - Invalid Parameter\n", 0, 0, 0, 0, 0, 0);
+        return NULL;
+    }
+
+    lock = ixOsalIrqLock ();
+
+    newBufPtr = poolPtr->nextFreeBuf;
+    if (newBufPtr)
+    {
+        poolPtr->nextFreeBuf =
+            IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR (newBufPtr);
+        IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR (newBufPtr) = NULL;
+
+        /*
+         * update the number of free buffers in the pool 
+         */
+        poolPtr->freeBufsInPool--;
+    }
+    else
+    {
+        /* Return NULL to indicate to caller that request is denied. */
+        ixOsalIrqUnlock (lock);
+
+        return NULL;
+    }
+
+#ifdef IX_OSAL_BUFFER_FREE_PROTECTION
+	/* Set Buffer Used Flag to indicate state.*/
+    IX_OSAL_MBUF_SET_USED_FLAG(newBufPtr);
+#endif
+
+    ixOsalIrqUnlock (lock);
+
+    return newBufPtr;
+}
+
+PUBLIC IX_OSAL_MBUF *
+ixOsalMbufFree (IX_OSAL_MBUF * bufPtr)
+{
+    int lock;
+    IX_OSAL_MBUF_POOL *poolPtr;
+
+    IX_OSAL_MBUF *nextBufPtr = NULL;
+
+    /*
+     * check parameters 
+     */
+    if (bufPtr == NULL)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalMbufFree(): "
+            "ERROR - Invalid Parameter\n", 0, 0, 0, 0, 0, 0);
+        return NULL;
+    }
+
+
+
+    lock = ixOsalIrqLock ();
+
+#ifdef IX_OSAL_BUFFER_FREE_PROTECTION
+	
+	/* Prevention for Buffer freed more than once*/
+    if(!IX_OSAL_MBUF_ISSET_USED_FLAG(bufPtr))
+    {
+   	return NULL;
+    }
+    IX_OSAL_MBUF_CLEAR_USED_FLAG(bufPtr);
+#endif
+	
+    poolPtr = IX_OSAL_MBUF_NET_POOL (bufPtr);
+
+    /*
+     * check the mbuf wrapper signature (if mbuf wrapper was used) 
+     */
+    if (poolPtr->poolAllocType == IX_OSAL_MBUF_POOL_TYPE_SYS_ALLOC)
+    {
+        IX_OSAL_ENSURE ( (IX_OSAL_MBUF_GET_SYS_SIGNATURE(bufPtr) == IX_OSAL_MBUF_SYS_SIGNATURE),
+            "ixOsalBuffPoolBufFree: ERROR - Invalid mbuf signature.");
+    }
+
+    nextBufPtr = IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR (bufPtr);
+
+    IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR (bufPtr) = poolPtr->nextFreeBuf;
+    poolPtr->nextFreeBuf = bufPtr;
+
+    /*
+     * update the number of free buffers in the pool 
+     */
+    poolPtr->freeBufsInPool++;
+
+    ixOsalIrqUnlock (lock);
+
+    return nextBufPtr;
+}
+
+PUBLIC void
+ixOsalMbufChainFree (IX_OSAL_MBUF * bufPtr)
+{
+    while ((bufPtr = ixOsalMbufFree (bufPtr)));
+}
+
+/*
+ * Function definition: ixOsalBuffPoolShow
+ */
+PUBLIC void
+ixOsalMbufPoolShow (IX_OSAL_MBUF_POOL * poolPtr)
+{
+    IX_OSAL_MBUF *nextBufPtr;
+    int count = 0;
+    int lock;
+
+    /*
+     * check parameters 
+     */
+    if (poolPtr == NULL)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalBuffPoolShow(): "
+            "ERROR - Invalid Parameter", 0, 0, 0, 0, 0, 0);
+        /*
+         * return IX_FAIL; 
+         */
+        return;
+    }
+
+    lock = ixOsalIrqLock ();
+    count = poolPtr->freeBufsInPool;
+    nextBufPtr = poolPtr->nextFreeBuf;
+    ixOsalIrqUnlock (lock);
+
+    ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE,
+        IX_OSAL_LOG_DEV_STDOUT, "=== POOL INFORMATION ===\n", 0, 0, 0,
+        0, 0, 0);
+    ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE, IX_OSAL_LOG_DEV_STDOUT,
+        "Pool Name:                   %s\n",
+        (unsigned int) poolPtr->name, 0, 0, 0, 0, 0);
+    ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE, IX_OSAL_LOG_DEV_STDOUT,
+        "Pool Allocation Type:        %d\n",
+        (unsigned int) poolPtr->poolAllocType, 0, 0, 0, 0, 0);
+    ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE, IX_OSAL_LOG_DEV_STDOUT,
+        "Pool Mbuf Mem Usage (bytes): %d\n",
+        (unsigned int) poolPtr->mbufMemSize, 0, 0, 0, 0, 0);
+    ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE, IX_OSAL_LOG_DEV_STDOUT,
+        "Pool Data Mem Usage (bytes): %d\n",
+        (unsigned int) poolPtr->dataMemSize, 0, 0, 0, 0, 0);
+    ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE, IX_OSAL_LOG_DEV_STDOUT,
+        "Mbuf Data Capacity  (bytes): %d\n",
+        (unsigned int) poolPtr->bufDataSize, 0, 0, 0, 0, 0);
+    ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE, IX_OSAL_LOG_DEV_STDOUT,
+        "Total Mbufs in Pool:         %d\n",
+        (unsigned int) poolPtr->totalBufsInPool, 0, 0, 0, 0, 0);
+    ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE, IX_OSAL_LOG_DEV_STDOUT,
+        "Available Mbufs:             %d\n", (unsigned int) count, 0,
+        0, 0, 0, 0);
+    ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE, IX_OSAL_LOG_DEV_STDOUT,
+        "Next Available Mbuf:         %p\n", (unsigned int) nextBufPtr,
+        0, 0, 0, 0, 0);
+
+    if (poolPtr->poolAllocType == IX_OSAL_MBUF_POOL_TYPE_USER_ALLOC)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "Mbuf Mem Area Start address: %p\n",
+            (unsigned int) poolPtr->mbufMemPtr, 0, 0, 0, 0, 0);
+        ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE, IX_OSAL_LOG_DEV_STDOUT,
+            "Data Mem Area Start address: %p\n",
+            (unsigned int) poolPtr->dataMemPtr, 0, 0, 0, 0, 0);
+    }
+}
+
+PUBLIC void
+ixOsalMbufDataPtrReset (IX_OSAL_MBUF * bufPtr)
+{
+    IX_OSAL_MBUF_POOL *poolPtr;
+    UINT8 *poolDataPtr;
+
+    if (bufPtr == NULL)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR, IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalBuffPoolBufDataPtrReset"
+            ": ERROR - Invalid Parameter\n", 0, 0, 0, 0, 0, 0);
+        return;
+    }
+
+    poolPtr = (IX_OSAL_MBUF_POOL *) IX_OSAL_MBUF_NET_POOL (bufPtr);
+    poolDataPtr = poolPtr->dataMemPtr;
+
+    if (poolPtr->poolAllocType == IX_OSAL_MBUF_POOL_TYPE_SYS_ALLOC)
+    {
+        if (IX_OSAL_MBUF_GET_SYS_SIGNATURE(bufPtr) != IX_OSAL_MBUF_SYS_SIGNATURE)
+        {
+            ixOsalLog (IX_OSAL_LOG_LVL_ERROR, IX_OSAL_LOG_DEV_STDOUT,
+                "ixOsalBuffPoolBufDataPtrReset"
+                ": invalid mbuf, cannot reset mData pointer\n", 0, 0,
+                0, 0, 0, 0);
+            return;
+        }
+        IX_OSAL_MBUF_MDATA (bufPtr) = (UINT8*)IX_OSAL_MBUF_ALLOCATED_BUFF_DATA (bufPtr);
+    }
+    else
+    {
+        if (poolDataPtr)
+        {
+            unsigned int bufSize = poolPtr->bufDataSize;
+            unsigned int bufDataAddr =
+                (unsigned int) IX_OSAL_MBUF_MDATA (bufPtr);
+            unsigned int poolDataAddr = (unsigned int) poolDataPtr;
+
+            /*
+             * the pointer is still pointing somewhere in the mbuf payload.
+             * This operation moves the pointer to the beginning of the 
+             * mbuf payload
+             */
+            bufDataAddr = ((bufDataAddr - poolDataAddr) / bufSize) * bufSize;
+            IX_OSAL_MBUF_MDATA (bufPtr) = &poolDataPtr[bufDataAddr];
+        }
+        else
+        {
+            ixOsalLog (IX_OSAL_LOG_LVL_WARNING, IX_OSAL_LOG_DEV_STDOUT,
+                "ixOsalBuffPoolBufDataPtrReset"
+                ": cannot be used if user supplied NULL pointer for pool data area "
+                "when pool was created\n", 0, 0, 0, 0, 0, 0);
+            return;
+        }
+    }
+
+}
+
+/*
+ * Function definition: ixOsalBuffPoolUninit
+ */
+PUBLIC IX_STATUS
+ixOsalBuffPoolUninit (IX_OSAL_MBUF_POOL * pool)
+{
+    if (!pool)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR, IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalBuffPoolUninit: NULL ptr \n", 0, 0, 0, 0, 0, 0);
+        return IX_FAIL;
+    }
+
+    if (pool->freeBufsInPool != pool->totalBufsInPool)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR, IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalBuffPoolUninit: need to return all ptrs to the pool first! \n",
+            0, 0, 0, 0, 0, 0);
+        return IX_FAIL;
+    }
+
+    if (pool->poolAllocType == IX_OSAL_MBUF_POOL_TYPE_SYS_ALLOC)
+    {
+#ifdef IX_OSAL_BUFFER_ALLOC_SEPARATELY
+				UINT32 i;
+				IX_OSAL_MBUF* pBuf;
+				
+				pBuf = pool->nextFreeBuf;
+				/* Freed the Buffer one by one till all the Memory is freed*/
+				for (i= pool->freeBufsInPool; i >0 && pBuf!=NULL ;i--){
+						IX_OSAL_MBUF* pBufTemp;
+						pBufTemp = IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(pBuf);
+						/* Freed MBUF Data Memory area*/
+						IX_OSAL_CACHE_DMA_FREE( (void *) (IX_OSAL_MBUF_ALLOCATED_BUFF_DATA(pBuf)) );
+						/* Freed MBUF Struct Memory area*/
+						IX_OSAL_CACHE_DMA_FREE(pBuf);
+						pBuf = pBufTemp;
+				}
+				
+#else    	
+        IX_OSAL_CACHE_DMA_FREE (pool->mbufMemPtr);
+        IX_OSAL_CACHE_DMA_FREE (pool->dataMemPtr);
+#endif        
+    }
+
+    ixOsalBuffFreePools[pool->poolIdx / IX_OSAL_BUFF_FREE_BITS] &=
+        ~(1 << (pool->poolIdx % IX_OSAL_BUFF_FREE_BITS));
+    ixOsalBuffPoolsInUse--;
+    return IX_SUCCESS;
+}
+
+/*
+ * Function definition: ixOsalBuffPoolDataAreaSizeGet
+ */
+PUBLIC UINT32
+ixOsalBuffPoolDataAreaSizeGet (int count, int size)
+{
+    UINT32 memorySize;
+    memorySize = count * IX_OSAL_MBUF_POOL_SIZE_ALIGN (size);
+    return memorySize;
+}
+
+/*
+ * Function definition: ixOsalBuffPoolMbufAreaSizeGet
+ */
+PUBLIC UINT32
+ixOsalBuffPoolMbufAreaSizeGet (int count)
+{
+    UINT32 memorySize;
+    memorySize =
+        count * IX_OSAL_MBUF_POOL_SIZE_ALIGN (sizeof (IX_OSAL_MBUF));
+    return memorySize;
+}
+
+/*
+ * Function definition: ixOsalBuffPoolFreeCountGet
+ */
+PUBLIC UINT32 ixOsalBuffPoolFreeCountGet(IX_OSAL_MBUF_POOL * poolPtr)
+
+{
+
+   return poolPtr->freeBufsInPool;
+
+}
+
+#endif /* IX_OSAL_USE_DEFAULT_BUFFER_MGT */

cpu/ixp/npe/IxOsalIoMem.c

@@ -0,0 +1,332 @@
+/**
+ * @file IxOsalIoMem.c 
+ *
+ * @brief OS-independent IO/Mem implementation 
+ * 
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+/* Access to the global mem map is only allowed in this file */
+#define IxOsalIoMem_C
+
+#include "IxOsal.h"
+
+#define SEARCH_PHYSICAL_ADDRESS (1)
+#define SEARCH_VIRTUAL_ADDRESS  (2)
+
+/*
+ * Searches for map using one of the following criteria:
+ * 
+ * - enough room to include a zone starting with the physical "requestedAddress" of size "size" (for mapping)
+ * - includes the virtual "requestedAddress" in its virtual address space (already mapped, for unmapping)
+ * - correct coherency
+ *
+ * Returns a pointer to the map or NULL if a suitable map is not found.
+ */
+PRIVATE IxOsalMemoryMap *
+ixOsalMemMapFind (UINT32 requestedAddress,
+    UINT32 size, UINT32 searchCriteria, UINT32 requestedEndianType)
+{
+    UINT32 mapIndex;
+
+    UINT32 numMapElements =
+        sizeof (ixOsalGlobalMemoryMap) / sizeof (IxOsalMemoryMap);
+
+    for (mapIndex = 0; mapIndex < numMapElements; mapIndex++)
+    {
+        IxOsalMemoryMap *map = &ixOsalGlobalMemoryMap[mapIndex];
+
+        if (searchCriteria == SEARCH_PHYSICAL_ADDRESS
+            && requestedAddress >= map->physicalAddress
+            && (requestedAddress + size) <= (map->physicalAddress + map->size)
+            && (map->mapEndianType & requestedEndianType) != 0)
+        {
+            return map;
+        }
+        else if (searchCriteria == SEARCH_VIRTUAL_ADDRESS
+            && requestedAddress >= map->virtualAddress
+            && requestedAddress <= (map->virtualAddress + map->size)
+            && (map->mapEndianType & requestedEndianType) != 0)
+        {
+            return map;
+        }
+        else if (searchCriteria == SEARCH_PHYSICAL_ADDRESS)
+        {
+            ixOsalLog (IX_OSAL_LOG_LVL_DEBUG3,
+                IX_OSAL_LOG_DEV_STDOUT,
+                "Osal: Checking [phys addr 0x%x:size 0x%x:endianType %d]\n",
+                map->physicalAddress, map->size, map->mapEndianType, 0, 0, 0);
+        }
+    }
+
+    /*
+     * not found 
+     */
+    return NULL;
+}
+
+/*
+ * This function maps an I/O mapped physical memory zone of the given size
+ * into a virtual memory zone accessible by the caller and returns a cookie - 
+ * the start address of the virtual memory zone. 
+ * IX_OSAL_MMAP_PHYS_TO_VIRT should NOT therefore be used on the returned 
+ * virtual address.
+ * The memory zone is to be unmapped using ixOsalMemUnmap once the caller has
+ * finished using this zone (e.g. on driver unload) using the cookie as 
+ * parameter.
+ * The IX_OSAL_READ/WRITE_LONG/SHORT macros should be used to read and write 
+ * the mapped memory, adding the necessary offsets to the address cookie.
+ *
+ * Note: this function is not to be used directly. Use IX_OSAL_MEM_MAP 
+ * instead.
+ */
+PUBLIC void *
+ixOsalIoMemMap (UINT32 requestedAddress,
+    UINT32 size, IxOsalMapEndianessType requestedEndianType)
+{
+    IxOsalMemoryMap *map;
+
+    ixOsalLog (IX_OSAL_LOG_LVL_DEBUG3,
+        IX_OSAL_LOG_DEV_STDOUT,
+        "OSAL: Mapping [addr 0x%x:size 0x%x:endianType %d]\n",
+        requestedAddress, size, requestedEndianType, 0, 0, 0);
+
+    if (requestedEndianType == IX_OSAL_LE)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalIoMemMap: Please specify component coherency mode to use MEM functions \n",
+            0, 0, 0, 0, 0, 0);
+        return (NULL);
+    }
+    map = ixOsalMemMapFind (requestedAddress,
+        size, SEARCH_PHYSICAL_ADDRESS, requestedEndianType);
+    if (map != NULL)
+    {
+        UINT32 offset = requestedAddress - map->physicalAddress;
+
+        ixOsalLog (IX_OSAL_LOG_LVL_DEBUG3,
+            IX_OSAL_LOG_DEV_STDOUT, "OSAL: Found map [", 0, 0, 0, 0, 0, 0);
+        ixOsalLog (IX_OSAL_LOG_LVL_DEBUG3,
+            IX_OSAL_LOG_DEV_STDOUT, map->name, 0, 0, 0, 0, 0, 0);
+        ixOsalLog (IX_OSAL_LOG_LVL_DEBUG3,
+            IX_OSAL_LOG_DEV_STDOUT,
+            ":addr 0x%x: virt 0x%x:size 0x%x:ref %d:endianType %d]\n",
+            map->physicalAddress, map->virtualAddress,
+            map->size, map->refCount, map->mapEndianType, 0);
+
+        if (map->type == IX_OSAL_DYNAMIC_MAP && map->virtualAddress == 0)
+        {
+            if (map->mapFunction != NULL)
+            {
+                map->mapFunction (map);
+
+                if (map->virtualAddress == 0)
+                {
+                    /*
+                     * failed 
+                     */
+                    ixOsalLog (IX_OSAL_LOG_LVL_FATAL,
+                        IX_OSAL_LOG_DEV_STDERR,
+                        "OSAL: Remap failed - [addr 0x%x:size 0x%x:endianType %d]\n",
+                        requestedAddress, size, requestedEndianType, 0, 0, 0);
+                    return NULL;
+                }
+            }
+            else
+            {
+                /*
+                 * error, no map function for a dynamic map 
+                 */
+                ixOsalLog (IX_OSAL_LOG_LVL_FATAL,
+                    IX_OSAL_LOG_DEV_STDERR,
+                    "OSAL: No map function for a dynamic map - "
+                    "[addr 0x%x:size 0x%x:endianType %d]\n",
+                    requestedAddress, size, requestedEndianType, 0, 0, 0);
+
+                return NULL;
+            }
+        }
+
+        /*
+         * increment reference count 
+         */
+        map->refCount++;
+
+        return (void *) (map->virtualAddress + offset);
+    }
+
+    /*
+     * requested address is not described in the global memory map 
+     */
+    ixOsalLog (IX_OSAL_LOG_LVL_FATAL,
+        IX_OSAL_LOG_DEV_STDERR,
+        "OSAL: No mapping found - [addr 0x%x:size 0x%x:endianType %d]\n",
+        requestedAddress, size, requestedEndianType, 0, 0, 0);
+    return NULL;
+}
+
+/*
+ * This function unmaps a previously mapped I/O memory zone using
+ * the cookie obtained in the mapping operation. The memory zone in question
+ * becomes unavailable to the caller once unmapped and the cookie should be
+ * discarded.
+ *
+ * This function cannot fail if the given parameter is correct and does not
+ * return a value.
+ *
+ * Note: this function is not to be used directly. Use IX_OSAL_MEM_UNMAP
+ * instead.
+ */
+PUBLIC void
+ixOsalIoMemUnmap (UINT32 requestedAddress, UINT32 endianType)
+{
+    IxOsalMemoryMap *map;
+
+    if (endianType == IX_OSAL_LE)
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_ERROR,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalIoMemUnmap: Please specify component coherency mode to use MEM functions \n",
+            0, 0, 0, 0, 0, 0);
+        return;
+    }
+
+    if (requestedAddress == 0)
+    {
+        /*
+         * invalid virtual address 
+         */
+        return;
+    }
+
+    map =
+        ixOsalMemMapFind (requestedAddress, 0, SEARCH_VIRTUAL_ADDRESS,
+        endianType);
+
+    if (map != NULL)
+    {
+        if (map->refCount > 0)
+        {
+            /*
+             * decrement reference count 
+             */
+            map->refCount--;
+
+            if (map->refCount == 0)
+            {
+                /*
+                 * no longer used, deallocate 
+                 */
+                if (map->type == IX_OSAL_DYNAMIC_MAP
+                    && map->unmapFunction != NULL)
+                {
+                    map->unmapFunction (map);
+                }
+            }
+        }
+    }
+    else
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_WARNING,
+            IX_OSAL_LOG_DEV_STDERR,
+            "OSAL: ixOsServMemUnmap didn't find the requested map "
+            "[virt addr 0x%x: endianType %d], ignoring call\n",
+            requestedAddress, endianType, 0, 0, 0, 0);
+    }
+}
+
+/* 
+ * This function Converts a virtual address into a physical 
+ * address, including the dynamically mapped memory.
+ * 
+ * Parameters	virtAddr - virtual address to convert
+ * Return value: corresponding physical address, or NULL 
+ *               if there is no physical address addressable 
+ *               by the given virtual address
+ * OS: 	VxWorks, Linux, WinCE, QNX, eCos
+ * Reentrant: Yes
+ * IRQ safe: Yes
+ */
+PUBLIC UINT32
+ixOsalIoMemVirtToPhys (UINT32 virtualAddress, UINT32 requestedCoherency)
+{
+    IxOsalMemoryMap *map =
+        ixOsalMemMapFind (virtualAddress, 0, SEARCH_VIRTUAL_ADDRESS,
+        requestedCoherency);
+
+    if (map != NULL)
+    {
+        return map->physicalAddress + virtualAddress - map->virtualAddress;
+    }
+    else
+    {
+        return (UINT32) IX_OSAL_MMU_VIRT_TO_PHYS (virtualAddress);
+    }
+}
+
+/* 
+ * This function Converts a virtual address into a physical 
+ * address, including the dynamically mapped memory.
+ * 
+ * Parameters	virtAddr - virtual address to convert
+ * Return value: corresponding physical address, or NULL 
+ *               if there is no physical address addressable 
+ *               by the given virtual address
+ * OS: 	VxWorks, Linux, WinCE, QNX, eCos
+ * Reentrant: Yes
+ * IRQ safe: Yes
+ */
+PUBLIC UINT32
+ixOsalIoMemPhysToVirt (UINT32 physicalAddress, UINT32 requestedCoherency)
+{
+    IxOsalMemoryMap *map =
+        ixOsalMemMapFind (physicalAddress, 0, SEARCH_PHYSICAL_ADDRESS,
+        requestedCoherency);
+
+    if (map != NULL)
+    {
+        return map->virtualAddress + physicalAddress - map->physicalAddress;
+    }
+    else
+    {
+        return (UINT32) IX_OSAL_MMU_PHYS_TO_VIRT (physicalAddress);
+    }
+}

cpu/ixp/npe/IxOsalOsCacheMMU.c

@@ -0,0 +1,67 @@
+/**
+ * @file IxOsalOsCacheMMU.c (linux)
+ *
+ * @brief Cache MemAlloc and MemFree.
+ *
+ *
+ * @par
+ * IXP400 SW Release version 1.5
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxOsal.h"
+
+#include <malloc.h>
+
+/*
+ * Allocate on a cache line boundary (null pointers are
+ * not affected by this operation). This operation is NOT cache safe.
+ */
+void *
+ixOsalCacheDmaMalloc (UINT32 n)
+{
+	return malloc(n);
+}
+
+/*
+ *
+ */
+void
+ixOsalCacheDmaFree (void *ptr)
+{
+	free(ptr);
+}

cpu/ixp/npe/IxOsalOsMsgQ.c

@@ -0,0 +1,79 @@
+/**
+ * @file IxOsalOsMsgQ.c (eCos)
+ *
+ * @brief OS-specific Message Queue implementation.
+ *
+ *
+ * @par
+ * IXP400 SW Release version 1.5
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxOsal.h"
+
+/*******************************
+ * Public functions
+ *******************************/
+PUBLIC IX_STATUS
+ixOsalMessageQueueCreate (IxOsalMessageQueue * queue,
+    UINT32 msgCount, UINT32 msgLen)
+{
+    diag_printf("%s called\n", __FUNCTION__);
+    return IX_FAIL;
+}
+
+PUBLIC IX_STATUS
+ixOsalMessageQueueDelete (IxOsalMessageQueue * queue)
+{
+    diag_printf("%s called\n", __FUNCTION__);
+    return IX_FAIL;
+}
+
+PUBLIC IX_STATUS
+ixOsalMessageQueueSend (IxOsalMessageQueue * queue, UINT8 * message)
+{
+    diag_printf("%s called\n", __FUNCTION__);
+    return IX_FAIL;
+}
+
+PUBLIC IX_STATUS
+ixOsalMessageQueueReceive (IxOsalMessageQueue * queue, UINT8 * message)
+{
+    diag_printf("%s called\n", __FUNCTION__);
+    return IX_FAIL;
+}
+

cpu/ixp/npe/IxOsalOsSemaphore.c

@@ -0,0 +1,233 @@
+/**
+ * @file IxOsalOsSemaphore.c (eCos)
+ *
+ * @brief Implementation for semaphore and mutex.
+ *
+ *
+ * @par
+ * IXP400 SW Release version 1.5
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxOsal.h"
+#include "IxNpeMhReceive_p.h"
+
+/* Define a large number */
+#define IX_OSAL_MAX_LONG (0x7FFFFFFF)
+
+/* Max timeout in MS, used to guard against possible overflow */
+#define IX_OSAL_MAX_TIMEOUT_MS (IX_OSAL_MAX_LONG/HZ)
+
+
+PUBLIC IX_STATUS
+ixOsalSemaphoreInit (IxOsalSemaphore * sid, UINT32 start_value)
+{
+    diag_printf("%s called\n", __FUNCTION__);
+    return IX_SUCCESS;
+}
+
+/**
+ * DESCRIPTION: If the semaphore is 'empty', the calling thread is blocked.
+ *              If the semaphore is 'full', it is taken and control is returned
+ *              to the caller. If the time indicated in 'timeout' is reached,
+ *              the thread will unblock and return an error indication. If the
+ *              timeout is set to 'IX_OSAL_WAIT_NONE', the thread will never block;
+ *              if it is set to 'IX_OSAL_WAIT_FOREVER', the thread will block until
+ *              the semaphore is available.
+ *
+ *
+ */
+
+
+PUBLIC IX_STATUS
+ixOsalSemaphoreWait (IxOsalOsSemaphore * sid, INT32 timeout)
+{
+    diag_printf("%s called\n", __FUNCTION__);
+    return IX_SUCCESS;
+}
+
+/*
+ * Attempt to get semaphore, return immediately,
+ * no error info because users expect some failures
+ * when using this API.
+ */
+PUBLIC IX_STATUS
+ixOsalSemaphoreTryWait (IxOsalSemaphore * sid)
+{
+    diag_printf("%s called\n", __FUNCTION__);
+    return IX_FAIL;
+}
+
+/**
+ *
+ * DESCRIPTION: This function causes the next available thread in the pend queue
+ *              to be unblocked. If no thread is pending on this semaphore, the
+ *              semaphore becomes 'full'.
+ */
+PUBLIC IX_STATUS
+ixOsalSemaphorePost (IxOsalSemaphore * sid)
+{
+    diag_printf("%s called\n", __FUNCTION__);
+    return IX_SUCCESS;
+}
+
+PUBLIC IX_STATUS
+ixOsalSemaphoreGetValue (IxOsalSemaphore * sid, UINT32 * value)
+{
+    diag_printf("%s called\n", __FUNCTION__);
+    return IX_FAIL;
+}
+
+PUBLIC IX_STATUS
+ixOsalSemaphoreDestroy (IxOsalSemaphore * sid)
+{
+    diag_printf("%s called\n", __FUNCTION__);
+    return IX_FAIL;
+}
+
+/****************************
+ *    Mutex
+ ****************************/
+
+static void drv_mutex_init(IxOsalMutex *mutex)
+{
+	*mutex = 0;
+}
+
+static void drv_mutex_destroy(IxOsalMutex *mutex)
+{
+	*mutex = -1;
+}
+
+static int drv_mutex_trylock(IxOsalMutex *mutex)
+{
+	int result = TRUE;
+
+	if (*mutex == 1)
+		result = FALSE;
+
+	return result;
+}
+
+static void drv_mutex_unlock(IxOsalMutex *mutex)
+{
+	if (*mutex == 1)
+		printf("Trying to unlock unlocked mutex!");
+
+	*mutex = 0;
+}
+
+PUBLIC IX_STATUS
+ixOsalMutexInit (IxOsalMutex * mutex)
+{
+    drv_mutex_init(mutex);
+    return IX_SUCCESS;
+}
+
+PUBLIC IX_STATUS
+ixOsalMutexLock (IxOsalMutex * mutex, INT32 timeout)
+{
+    int tries;
+
+    if (timeout == IX_OSAL_WAIT_NONE) {
+	if (drv_mutex_trylock(mutex))
+	    return IX_SUCCESS;
+	else
+	    return IX_FAIL;
+    }
+
+    tries = (timeout * 1000) / 50;
+    while (1) {
+	if (drv_mutex_trylock(mutex))
+	    return IX_SUCCESS;
+	if (timeout != IX_OSAL_WAIT_FOREVER && tries-- <= 0)
+	    break;
+	udelay(50);
+    }
+    return IX_FAIL;
+}
+
+PUBLIC IX_STATUS
+ixOsalMutexUnlock (IxOsalMutex * mutex)
+{
+    drv_mutex_unlock(mutex);
+    return IX_SUCCESS;
+}
+
+/*
+ * Attempt to get mutex, return immediately,
+ * no error info because users expect some failures
+ * when using this API.
+ */
+PUBLIC IX_STATUS
+ixOsalMutexTryLock (IxOsalMutex * mutex)
+{
+    if (drv_mutex_trylock(mutex))
+	return IX_SUCCESS;
+    return IX_FAIL;
+}
+
+PUBLIC IX_STATUS
+ixOsalMutexDestroy (IxOsalMutex * mutex)
+{
+    drv_mutex_destroy(mutex);
+    return IX_SUCCESS;
+}
+
+PUBLIC IX_STATUS
+ixOsalFastMutexInit (IxOsalFastMutex * mutex)
+{
+    return ixOsalMutexInit(mutex);
+}
+
+PUBLIC IX_STATUS ixOsalFastMutexTryLock(IxOsalFastMutex *mutex)
+{
+    return ixOsalMutexTryLock(mutex);
+}
+
+
+PUBLIC IX_STATUS
+ixOsalFastMutexUnlock (IxOsalFastMutex * mutex)
+{
+    return ixOsalMutexUnlock(mutex);
+}
+
+PUBLIC IX_STATUS
+ixOsalFastMutexDestroy (IxOsalFastMutex * mutex)
+{
+    return ixOsalMutexDestroy(mutex);
+}

cpu/ixp/npe/IxOsalOsServices.c

@@ -0,0 +1,251 @@
+/**
+ * @file IxOsalOsServices.c (linux)
+ *
+ * @brief Implementation for Irq, Mem, sleep.
+ *
+ *
+ * @par
+ * IXP400 SW Release version 1.5
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include <config.h>
+#include <common.h>
+#include "IxOsal.h"
+#include <IxEthAcc.h>
+#include <IxEthDB.h>
+#include <IxNpeDl.h>
+#include <IxQMgr.h>
+#include <IxNpeMh.h>
+
+static char *traceHeaders[] = {
+    "",
+    "[fatal] ",
+    "[error] ",
+    "[warning] ",
+    "[message] ",
+    "[debug1] ",
+    "[debug2] ",
+    "[debug3] ",
+    "[all]"
+};
+
+/* by default trace all but debug message */
+PRIVATE int ixOsalCurrLogLevel = IX_OSAL_LOG_LVL_MESSAGE;
+
+/**************************************
+ * Irq services
+ *************************************/
+
+PUBLIC IX_STATUS
+ixOsalIrqBind (UINT32 vector, IxOsalVoidFnVoidPtr routine, void *parameter)
+{
+    return IX_FAIL;
+}
+
+PUBLIC IX_STATUS
+ixOsalIrqUnbind (UINT32 vector)
+{
+    return IX_FAIL;
+}
+
+PUBLIC UINT32
+ixOsalIrqLock ()
+{
+    return 0;
+}
+
+/* Enable interrupts and task scheduling,
+ * input parameter: irqEnable status returned
+ * by ixOsalIrqLock().
+ */
+PUBLIC void
+ixOsalIrqUnlock (UINT32 lockKey)
+{
+}
+
+PUBLIC UINT32
+ixOsalIrqLevelSet (UINT32 level)
+{
+    return IX_FAIL;
+}
+
+PUBLIC void
+ixOsalIrqEnable (UINT32 irqLevel)
+{
+}
+
+PUBLIC void
+ixOsalIrqDisable (UINT32 irqLevel)
+{
+}
+
+/*********************
+ * Log function
+ *********************/
+
+INT32
+ixOsalLog (IxOsalLogLevel level,
+    IxOsalLogDevice device,
+    char *format, int arg1, int arg2, int arg3, int arg4, int arg5, int arg6)
+{
+    /*
+     * Return -1 for custom display devices
+     */
+    if ((device != IX_OSAL_LOG_DEV_STDOUT)
+        && (device != IX_OSAL_LOG_DEV_STDERR))
+    {
+        debug("ixOsalLog: only IX_OSAL_LOG_DEV_STDOUT and IX_OSAL_LOG_DEV_STDERR are supported \n");
+        return (IX_OSAL_LOG_ERROR);
+    }
+
+    if (level <= ixOsalCurrLogLevel && level != IX_OSAL_LOG_LVL_NONE)
+    {
+#if 0 /* sr: U-Boots printf or debug doesn't return a length */
+        int headerByteCount = (level == IX_OSAL_LOG_LVL_USER) ? 0 : diag_printf(traceHeaders[level - 1]);
+
+        return headerByteCount + diag_printf (format, arg1, arg2, arg3, arg4, arg5, arg6);
+#else
+        int headerByteCount = (level == IX_OSAL_LOG_LVL_USER) ? 0 : strlen(traceHeaders[level - 1]);
+
+        return headerByteCount + strlen(format);
+#endif
+    }
+    else
+    {
+        /*
+         * Return error
+         */
+        return (IX_OSAL_LOG_ERROR);
+    }
+}
+
+PUBLIC UINT32
+ixOsalLogLevelSet (UINT32 level)
+{
+    UINT32 oldLevel;
+
+    /*
+     * Check value first
+     */
+    if ((level < IX_OSAL_LOG_LVL_NONE) || (level > IX_OSAL_LOG_LVL_ALL))
+    {
+        ixOsalLog (IX_OSAL_LOG_LVL_MESSAGE,
+            IX_OSAL_LOG_DEV_STDOUT,
+            "ixOsalLogLevelSet: Log Level is between %d and%d \n",
+            IX_OSAL_LOG_LVL_NONE, IX_OSAL_LOG_LVL_ALL, 0, 0, 0, 0);
+        return IX_OSAL_LOG_LVL_NONE;
+    }
+    oldLevel = ixOsalCurrLogLevel;
+
+    ixOsalCurrLogLevel = level;
+
+    return oldLevel;
+}
+
+/**************************************
+ * Task services
+ *************************************/
+
+PUBLIC void
+ixOsalBusySleep (UINT32 microseconds)
+{
+	udelay(microseconds);
+}
+
+PUBLIC void
+ixOsalSleep (UINT32 milliseconds)
+{
+    if (milliseconds != 0) {
+#if 1
+	/*
+	 * sr: We poll while we wait because interrupts are off in U-Boot
+	 * and CSR expects messages, etc to be dispatched while sleeping.
+	 */
+	int i;
+	IxQMgrDispatcherFuncPtr qDispatcherFunc;
+
+	ixQMgrDispatcherLoopGet(&qDispatcherFunc);
+
+	while (milliseconds--) {
+		for (i = 1; i <= 2; i++)
+			ixNpeMhMessagesReceive(i);
+		(*qDispatcherFunc)(IX_QMGR_QUELOW_GROUP);
+
+		udelay(1000);
+	}
+#endif
+    }
+}
+
+/**************************************
+ * Memory functions
+ *************************************/
+
+void *
+ixOsalMemAlloc (UINT32 size)
+{
+    return (void *)0;
+}
+
+void
+ixOsalMemFree (void *ptr)
+{
+}
+
+/*
+ * Copy count bytes from src to dest ,
+ * returns pointer to the dest mem zone.
+ */
+void *
+ixOsalMemCopy (void *dest, void *src, UINT32 count)
+{
+    IX_OSAL_ASSERT (dest != NULL);
+    IX_OSAL_ASSERT (src != NULL);
+    return (memcpy (dest, src, count));
+}
+
+/*
+ * Fills a memory zone with a given constant byte,
+ * returns pointer to the memory zone.
+ */
+void *
+ixOsalMemSet (void *ptr, UINT8 filler, UINT32 count)
+{
+    IX_OSAL_ASSERT (ptr != NULL);
+    return (memset (ptr, filler, count));
+}

cpu/ixp/npe/IxOsalOsThread.c

@@ -0,0 +1,98 @@
+/**
+ * @file IxOsalOsThread.c (eCos)
+ *
+ * @brief OS-specific thread implementation.
+ *
+ *
+ * @par
+ * IXP400 SW Release version 1.5
+ *
+ * -- Copyright Notice --
+ *
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ *
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#include "IxOsal.h"
+
+/* Thread attribute is ignored */
+PUBLIC IX_STATUS
+ixOsalThreadCreate (IxOsalThread * ptrTid,
+    IxOsalThreadAttr * threadAttr, IxOsalVoidFnVoidPtr entryPoint, void *arg)
+{
+    return IX_SUCCESS;
+}
+
+/*
+ * Start thread after given its thread handle
+ */
+PUBLIC IX_STATUS
+ixOsalThreadStart (IxOsalThread * tId)
+{
+    /* Thread already started upon creation */
+    return IX_SUCCESS;
+}
+
+/*
+ * In Linux threadKill does not actually destroy the thread,
+ * it will stop the signal handling.
+ */
+PUBLIC IX_STATUS
+ixOsalThreadKill (IxOsalThread * tid)
+{
+    return IX_SUCCESS;
+}
+
+PUBLIC void
+ixOsalThreadExit (void)
+{
+}
+
+PUBLIC IX_STATUS
+ixOsalThreadPrioritySet (IxOsalOsThread * tid, UINT32 priority)
+{
+    return IX_SUCCESS;
+}
+
+PUBLIC IX_STATUS
+ixOsalThreadSuspend (IxOsalThread * tId)
+{
+    return IX_SUCCESS;
+
+}
+
+PUBLIC IX_STATUS
+ixOsalThreadResume (IxOsalThread * tId)
+{
+    return IX_SUCCESS;
+}

cpu/ixp/npe/IxQMgrAqmIf.c

@@ -0,0 +1,963 @@
+/* 
+ * @file:    IxQMgrAqmIf.c
+ *
+ * @author Intel Corporation
+ * @date     30-Oct-2001
+ *
+ * @brief    This component provides a set of functions for
+ * perfoming I/O on the AQM hardware.
+ * 
+ * Design Notes: 
+ *              These functions are intended to be as fast as possible
+ * and as a result perform NO PARAMETER CHECKING.
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+/*
+ * Inlines are compiled as function when this is defined.
+ * N.B. Must be placed before #include of "IxQMgrAqmIf_p.h
+ */
+#ifndef IXQMGRAQMIF_P_H
+#    define IXQMGRAQMIF_C
+#else
+#    error
+#endif
+
+/*
+ * User defined include files.
+ */
+#include "IxOsal.h"
+#include "IxQMgr.h"
+#include "IxQMgrAqmIf_p.h"
+#include "IxQMgrLog_p.h"
+
+
+/*
+ * #defines and macros used in this file.
+ */
+
+/* These defines are the bit offsets of the various fields of
+ * the queue configuration register
+ */
+#define IX_QMGR_Q_CONFIG_WRPTR_OFFSET       0x00
+#define IX_QMGR_Q_CONFIG_RDPTR_OFFSET       0x07
+#define IX_QMGR_Q_CONFIG_BADDR_OFFSET       0x0E
+#define IX_QMGR_Q_CONFIG_ESIZE_OFFSET       0x16
+#define IX_QMGR_Q_CONFIG_BSIZE_OFFSET       0x18
+#define IX_QMGR_Q_CONFIG_NE_OFFSET          0x1A
+#define IX_QMGR_Q_CONFIG_NF_OFFSET          0x1D
+
+#define IX_QMGR_BASE_ADDR_16_WORD_ALIGN     0x40
+#define IX_QMGR_BASE_ADDR_16_WORD_SHIFT     0x6
+
+#define IX_QMGR_NE_NF_CLEAR_MASK            0x03FFFFFF
+#define IX_QMGR_NE_MASK                     0x7
+#define IX_QMGR_NF_MASK                     0x7
+#define IX_QMGR_SIZE_MASK                   0x3
+#define IX_QMGR_ENTRY_SIZE_MASK             0x3
+#define IX_QMGR_BADDR_MASK                  0x003FC000
+#define IX_QMGR_RDPTR_MASK                  0x7F
+#define IX_QMGR_WRPTR_MASK                  0x7F
+#define IX_QMGR_RDWRPTR_MASK                0x00003FFF
+
+#define IX_QMGR_AQM_ADDRESS_SPACE_SIZE_IN_WORDS 0x1000
+
+/* Base address of AQM SRAM */
+#define IX_QMGR_AQM_SRAM_BASE_ADDRESS_OFFSET \
+((IX_QMGR_QUECONFIG_BASE_OFFSET) + (IX_QMGR_QUECONFIG_SIZE))
+
+/* Min buffer size used for generating buffer size in QUECONFIG */
+#define IX_QMGR_MIN_BUFFER_SIZE 16
+
+/* Reset values of QMgr hardware registers */
+#define IX_QMGR_QUELOWSTAT_RESET_VALUE    0x33333333
+#define IX_QMGR_QUEUOSTAT_RESET_VALUE     0x00000000
+#define IX_QMGR_QUEUPPSTAT0_RESET_VALUE   0xFFFFFFFF
+#define IX_QMGR_QUEUPPSTAT1_RESET_VALUE   0x00000000
+#define IX_QMGR_INT0SRCSELREG_RESET_VALUE 0x00000000
+#define IX_QMGR_QUEIEREG_RESET_VALUE      0x00000000
+#define IX_QMGR_QINTREG_RESET_VALUE       0xFFFFFFFF
+#define IX_QMGR_QUECONFIG_RESET_VALUE     0x00000000
+
+#define IX_QMGR_PHYSICAL_AQM_BASE_ADDRESS IX_OSAL_IXP400_QMGR_PHYS_BASE
+
+#define IX_QMGR_QUELOWSTAT_BITS_PER_Q (BITS_PER_WORD/IX_QMGR_QUELOWSTAT_NUM_QUE_PER_WORD)
+
+#define IX_QMGR_QUELOWSTAT_QID_MASK 0x7
+#define IX_QMGR_Q_CONFIG_ADDR_GET(qId)\
+        (((qId) * IX_QMGR_NUM_BYTES_PER_WORD) +\
+                  IX_QMGR_QUECONFIG_BASE_OFFSET)
+
+#define IX_QMGR_ENTRY1_OFFSET 0
+#define IX_QMGR_ENTRY2_OFFSET 1
+#define IX_QMGR_ENTRY4_OFFSET 3
+
+/*
+ * Variable declarations global to this file. Externs are followed by
+ * statics.
+ */
+UINT32 aqmBaseAddress = 0;
+/* Store addresses and bit-masks for certain queue access and status registers.
+ * This is to facilitate inlining of QRead, QWrite and QStatusGet functions
+ * in IxQMgr,h
+ */
+extern IxQMgrQInlinedReadWriteInfo ixQMgrQInlinedReadWriteInfo[];
+UINT32 * ixQMgrAqmIfQueAccRegAddr[IX_QMGR_MAX_NUM_QUEUES];
+UINT32 ixQMgrAqmIfQueLowStatRegAddr[IX_QMGR_MIN_QUEUPP_QID];
+UINT32 ixQMgrAqmIfQueLowStatBitsOffset[IX_QMGR_MIN_QUEUPP_QID];
+UINT32 ixQMgrAqmIfQueLowStatBitsMask;
+UINT32 ixQMgrAqmIfQueUppStat0RegAddr;
+UINT32 ixQMgrAqmIfQueUppStat1RegAddr;
+UINT32 ixQMgrAqmIfQueUppStat0BitMask[IX_QMGR_MIN_QUEUPP_QID];
+UINT32 ixQMgrAqmIfQueUppStat1BitMask[IX_QMGR_MIN_QUEUPP_QID];
+
+/* 
+ * Fast mutexes, one for each queue, used to protect peek & poke functions
+ */
+IxOsalFastMutex ixQMgrAqmIfPeekPokeFastMutex[IX_QMGR_MAX_NUM_QUEUES];
+
+/*
+ * Function prototypes
+ */
+PRIVATE unsigned
+watermarkToAqmWatermark (IxQMgrWMLevel watermark );
+
+PRIVATE unsigned
+entrySizeToAqmEntrySize (IxQMgrQEntrySizeInWords entrySize);
+
+PRIVATE unsigned
+bufferSizeToAqmBufferSize (unsigned bufferSizeInWords);
+
+PRIVATE void
+ixQMgrAqmIfRegistersReset (void);
+
+PRIVATE void
+ixQMgrAqmIfEntryAddressGet (unsigned int entryIndex,
+			    UINT32 configRegWord,
+			    unsigned int qEntrySizeInwords,
+			    unsigned int qSizeInWords,
+			    UINT32 **address);
+/*
+ * Function definitions
+ */
+void
+ixQMgrAqmIfInit (void)
+{
+    UINT32 aqmVirtualAddr;
+    int i;
+
+    /* The value of aqmBaseAddress depends on the logical address
+     * assigned by the MMU.
+     */
+    aqmVirtualAddr =
+	(UINT32) IX_OSAL_MEM_MAP(IX_QMGR_PHYSICAL_AQM_BASE_ADDRESS,
+				    IX_OSAL_IXP400_QMGR_MAP_SIZE);
+    IX_OSAL_ASSERT (aqmVirtualAddr);
+    
+    ixQMgrAqmIfBaseAddressSet (aqmVirtualAddr);
+
+    ixQMgrAqmIfRegistersReset ();
+
+    for (i = 0; i< IX_QMGR_MAX_NUM_QUEUES; i++)
+    {
+	ixOsalFastMutexInit(&ixQMgrAqmIfPeekPokeFastMutex[i]);
+
+	/********************************************************************
+	 * Register addresses and bit masks are calculated and stored here to
+	 * facilitate inlining of QRead, QWrite and QStatusGet functions in
+	 * IxQMgr.h.
+	 * These calculations are normally performed dynamically in inlined
+	 * functions in IxQMgrAqmIf_p.h, and their semantics are reused here.
+	 */
+
+	/* AQM Queue access reg addresses, per queue */
+   	ixQMgrAqmIfQueAccRegAddr[i] = 
+	    (UINT32 *)(aqmBaseAddress + IX_QMGR_Q_ACCESS_ADDR_GET(i));
+	ixQMgrQInlinedReadWriteInfo[i].qAccRegAddr = 
+	    (volatile UINT32 *)(aqmBaseAddress + IX_QMGR_Q_ACCESS_ADDR_GET(i));
+
+
+	ixQMgrQInlinedReadWriteInfo[i].qConfigRegAddr = 
+	    (volatile UINT32 *)(aqmBaseAddress + IX_QMGR_Q_CONFIG_ADDR_GET(i));
+
+	/* AQM Queue lower-group (0-31), only */
+	if (i < IX_QMGR_MIN_QUEUPP_QID)
+	{
+	    /* AQM Q underflow/overflow status register addresses, per queue */
+	    ixQMgrQInlinedReadWriteInfo[i].qUOStatRegAddr = 
+		(volatile UINT32 *)(aqmBaseAddress +
+		IX_QMGR_QUEUOSTAT0_OFFSET +
+		((i / IX_QMGR_QUEUOSTAT_NUM_QUE_PER_WORD) *
+		 IX_QMGR_NUM_BYTES_PER_WORD));
+
+	    /* AQM Q underflow status bit masks for status register per queue */
+	    ixQMgrQInlinedReadWriteInfo[i].qUflowStatBitMask = 
+		(IX_QMGR_UNDERFLOW_BIT_OFFSET + 1) <<
+		((i & (IX_QMGR_QUEUOSTAT_NUM_QUE_PER_WORD - 1)) *
+		 (BITS_PER_WORD / IX_QMGR_QUEUOSTAT_NUM_QUE_PER_WORD));
+
+	    /* AQM Q overflow status bit masks for status register, per queue */
+	    ixQMgrQInlinedReadWriteInfo[i].qOflowStatBitMask = 
+		(IX_QMGR_OVERFLOW_BIT_OFFSET + 1) <<
+		((i & (IX_QMGR_QUEUOSTAT_NUM_QUE_PER_WORD - 1)) *
+		 (BITS_PER_WORD / IX_QMGR_QUEUOSTAT_NUM_QUE_PER_WORD));
+
+	    /* AQM Q lower-group (0-31) status register addresses, per queue */
+	    ixQMgrAqmIfQueLowStatRegAddr[i] = aqmBaseAddress +
+		IX_QMGR_QUELOWSTAT0_OFFSET +
+		((i / IX_QMGR_QUELOWSTAT_NUM_QUE_PER_WORD) *
+		 IX_QMGR_NUM_BYTES_PER_WORD);
+
+	    /* AQM Q lower-group (0-31) status register bit offset */
+	    ixQMgrAqmIfQueLowStatBitsOffset[i] =
+		(i & (IX_QMGR_QUELOWSTAT_NUM_QUE_PER_WORD - 1)) * 
+		(BITS_PER_WORD / IX_QMGR_QUELOWSTAT_NUM_QUE_PER_WORD);
+	}
+	else /* AQM Q upper-group (32-63), only */
+	{
+	    /* AQM Q upper-group (32-63) Nearly Empty status reg bit masks */
+	    ixQMgrAqmIfQueUppStat0BitMask[i - IX_QMGR_MIN_QUEUPP_QID] =
+		(1 << (i - IX_QMGR_MIN_QUEUPP_QID));
+
+	    /* AQM Q upper-group (32-63) Full status register bit masks */
+	    ixQMgrAqmIfQueUppStat1BitMask[i - IX_QMGR_MIN_QUEUPP_QID] =
+		(1 << (i - IX_QMGR_MIN_QUEUPP_QID));
+	}
+    }
+
+    /* AQM Q lower-group (0-31) status register bit mask */
+    ixQMgrAqmIfQueLowStatBitsMask = (1 <<
+				    (BITS_PER_WORD /
+				     IX_QMGR_QUELOWSTAT_NUM_QUE_PER_WORD)) - 1;
+
+    /* AQM Q upper-group (32-63) Nearly Empty status register address */
+    ixQMgrAqmIfQueUppStat0RegAddr = aqmBaseAddress + IX_QMGR_QUEUPPSTAT0_OFFSET;
+    
+    /* AQM Q upper-group (32-63) Full status register address */
+    ixQMgrAqmIfQueUppStat1RegAddr = aqmBaseAddress + IX_QMGR_QUEUPPSTAT1_OFFSET;
+}
+
+/*
+ * Uninitialise the AqmIf module by unmapping memory, etc
+ */
+void
+ixQMgrAqmIfUninit (void)
+{
+    UINT32 virtAddr;
+
+    ixQMgrAqmIfBaseAddressGet (&virtAddr);
+    IX_OSAL_MEM_UNMAP (virtAddr);
+    ixQMgrAqmIfBaseAddressSet (0);
+}
+
+/*
+ * Set the the logical base address of AQM
+ */
+void
+ixQMgrAqmIfBaseAddressSet (UINT32 address)
+{
+    aqmBaseAddress = address;
+}
+
+/*
+ * Get the logical base address of AQM
+ */
+void
+ixQMgrAqmIfBaseAddressGet (UINT32 *address)
+{
+    *address = aqmBaseAddress;
+}
+
+/*
+ * Get the logical base address of AQM SRAM
+ */
+void
+ixQMgrAqmIfSramBaseAddressGet (UINT32 *address)
+{
+    *address = aqmBaseAddress                +
+	IX_QMGR_AQM_SRAM_BASE_ADDRESS_OFFSET;
+}
+
+/*
+ * This function will write the status bits of a queue
+ * specified by qId.
+ */
+void
+ixQMgrAqmIfQRegisterBitsWrite (IxQMgrQId qId, 
+			       UINT32 registerBaseAddrOffset,
+			       unsigned queuesPerRegWord,
+			       UINT32 value)
+{
+    volatile UINT32 *registerAddress;
+    UINT32 registerWord;
+    UINT32 statusBitsMask;
+    UINT32 bitsPerQueue;
+
+    bitsPerQueue = BITS_PER_WORD / queuesPerRegWord;
+
+    /*
+     * Calculate the registerAddress
+     * multiple queues split accross registers
+     */
+    registerAddress = (UINT32*)(aqmBaseAddress +
+				registerBaseAddrOffset +
+				((qId / queuesPerRegWord) *
+				 IX_QMGR_NUM_BYTES_PER_WORD));    
+
+    /* Read the current data */
+    ixQMgrAqmIfWordRead (registerAddress, &registerWord);
+
+
+    if( (registerBaseAddrOffset == IX_QMGR_INT0SRCSELREG0_OFFSET) &&
+        (qId == IX_QMGR_QUEUE_0) )
+    {
+      statusBitsMask = 0x7 ;   
+
+      /* Queue 0 at INT0SRCSELREG should not corrupt the value bit-3  */
+      value &=  0x7 ;        
+    }
+    else
+    {     
+      /* Calculate the mask for the status bits for this queue. */
+      statusBitsMask = ((1 << bitsPerQueue) - 1);
+      statusBitsMask <<= ((qId & (queuesPerRegWord - 1)) * bitsPerQueue);
+
+      /* Mask out bits in value that would overwrite other q data */
+      value <<= ((qId & (queuesPerRegWord - 1)) * bitsPerQueue);
+      value &= statusBitsMask;
+    }
+
+    /* Mask out bits to write to */
+    registerWord &= ~statusBitsMask;
+    
+
+    /* Set the write bits */
+    registerWord |= value;
+
+    /*
+     * Write the data
+     */
+    ixQMgrAqmIfWordWrite (registerAddress, registerWord);
+}
+
+/*
+ * This function generates the parameters that can be used to
+ * check if a Qs status matches the specified source select.
+ * It calculates which status word to check (statusWordOffset),
+ * the value to check the status against (checkValue) and the
+ * mask (mask) to mask out all but the bits to check in the status word.
+ */
+void
+ixQMgrAqmIfQStatusCheckValsCalc (IxQMgrQId qId,
+				 IxQMgrSourceId srcSel,
+				 unsigned int *statusWordOffset,
+				 UINT32 *checkValue,
+				 UINT32 *mask)
+{
+    UINT32 shiftVal;
+   
+    if (qId < IX_QMGR_MIN_QUEUPP_QID)
+    {
+	switch (srcSel)
+	{
+	    case IX_QMGR_Q_SOURCE_ID_E:
+		*checkValue = IX_QMGR_Q_STATUS_E_BIT_MASK;
+		*mask = IX_QMGR_Q_STATUS_E_BIT_MASK;
+		break;
+	    case IX_QMGR_Q_SOURCE_ID_NE:
+		*checkValue = IX_QMGR_Q_STATUS_NE_BIT_MASK;
+		*mask = IX_QMGR_Q_STATUS_NE_BIT_MASK;
+		break;
+	    case IX_QMGR_Q_SOURCE_ID_NF:
+		*checkValue = IX_QMGR_Q_STATUS_NF_BIT_MASK;
+		*mask = IX_QMGR_Q_STATUS_NF_BIT_MASK;
+		break;
+	    case IX_QMGR_Q_SOURCE_ID_F:
+		*checkValue = IX_QMGR_Q_STATUS_F_BIT_MASK;
+		*mask = IX_QMGR_Q_STATUS_F_BIT_MASK;
+		break;
+	    case IX_QMGR_Q_SOURCE_ID_NOT_E:
+		*checkValue = 0;
+		*mask = IX_QMGR_Q_STATUS_E_BIT_MASK;
+		break;
+	    case IX_QMGR_Q_SOURCE_ID_NOT_NE:
+		*checkValue = 0;
+		*mask = IX_QMGR_Q_STATUS_NE_BIT_MASK;
+		break;
+	    case IX_QMGR_Q_SOURCE_ID_NOT_NF:
+		*checkValue = 0;
+		*mask = IX_QMGR_Q_STATUS_NF_BIT_MASK;
+		break;
+	    case IX_QMGR_Q_SOURCE_ID_NOT_F:
+		*checkValue = 0;
+		*mask = IX_QMGR_Q_STATUS_F_BIT_MASK;
+		break;
+	    default:
+		/* Should never hit */
+		IX_OSAL_ASSERT(0);
+		break;
+	}
+
+	/* One nibble of status per queue so need to shift the
+	 * check value and mask out to the correct position.
+	 */
+	shiftVal = (qId % IX_QMGR_QUELOWSTAT_NUM_QUE_PER_WORD) * 
+	    IX_QMGR_QUELOWSTAT_BITS_PER_Q;
+
+	/* Calculate the which status word to check from the qId,
+	 * 8 Qs status per word
+	 */
+	*statusWordOffset = qId / IX_QMGR_QUELOWSTAT_NUM_QUE_PER_WORD;
+
+	*checkValue <<= shiftVal;
+	*mask <<= shiftVal;
+    }
+    else
+    {
+	/* One status word */
+	*statusWordOffset = 0;
+	/* Single bits per queue and int source bit hardwired  NE,
+	 * Qs start at 32.
+	 */
+	*mask = 1 << (qId - IX_QMGR_MIN_QUEUPP_QID);
+	*checkValue = *mask;
+    }
+}
+
+void
+ixQMgrAqmIfQInterruptEnable (IxQMgrQId qId)
+{
+    volatile UINT32 *registerAddress;
+    UINT32 registerWord;
+    UINT32 actualBitOffset;
+    
+    if (qId < IX_QMGR_MIN_QUEUPP_QID)
+    {    
+	registerAddress = (UINT32*)(aqmBaseAddress + IX_QMGR_QUEIEREG0_OFFSET);
+    }
+    else
+    {
+	registerAddress = (UINT32*)(aqmBaseAddress + IX_QMGR_QUEIEREG1_OFFSET);
+    }
+
+    actualBitOffset = 1 << (qId % IX_QMGR_MIN_QUEUPP_QID);
+
+    ixQMgrAqmIfWordRead (registerAddress, &registerWord);
+    ixQMgrAqmIfWordWrite (registerAddress, (registerWord | actualBitOffset));
+}
+
+void
+ixQMgrAqmIfQInterruptDisable (IxQMgrQId qId)
+{
+    volatile UINT32 *registerAddress;
+    UINT32 registerWord;
+    UINT32 actualBitOffset;
+
+    if (qId < IX_QMGR_MIN_QUEUPP_QID)
+    {    
+	registerAddress = (UINT32*)(aqmBaseAddress + IX_QMGR_QUEIEREG0_OFFSET);
+    }
+    else
+    {
+	registerAddress = (UINT32*)(aqmBaseAddress + IX_QMGR_QUEIEREG1_OFFSET);
+    }
+
+    actualBitOffset = 1 << (qId % IX_QMGR_MIN_QUEUPP_QID);
+
+    ixQMgrAqmIfWordRead (registerAddress, &registerWord);
+    ixQMgrAqmIfWordWrite (registerAddress, registerWord & (~actualBitOffset));
+}
+
+void
+ixQMgrAqmIfQueCfgWrite (IxQMgrQId qId,
+		       IxQMgrQSizeInWords qSizeInWords,
+		       IxQMgrQEntrySizeInWords entrySizeInWords,
+		       UINT32 freeSRAMAddress)
+{
+    volatile UINT32 *cfgAddress = NULL;
+    UINT32 qCfg = 0;
+    UINT32 baseAddress = 0;
+    unsigned aqmEntrySize = 0;
+    unsigned aqmBufferSize = 0;
+
+    /* Build config register */
+    aqmEntrySize = entrySizeToAqmEntrySize (entrySizeInWords);
+    qCfg |= (aqmEntrySize&IX_QMGR_ENTRY_SIZE_MASK) <<
+	IX_QMGR_Q_CONFIG_ESIZE_OFFSET;
+
+    aqmBufferSize = bufferSizeToAqmBufferSize (qSizeInWords);
+    qCfg |= (aqmBufferSize&IX_QMGR_SIZE_MASK) << IX_QMGR_Q_CONFIG_BSIZE_OFFSET;
+
+    /* baseAddress, calculated relative to aqmBaseAddress and start address  */
+    baseAddress = freeSRAMAddress -
+	(aqmBaseAddress + IX_QMGR_QUECONFIG_BASE_OFFSET);
+		   
+    /* Verify base address aligned to a 16 word boundary */
+    if ((baseAddress % IX_QMGR_BASE_ADDR_16_WORD_ALIGN) != 0)
+    {
+	IX_QMGR_LOG_ERROR0("ixQMgrAqmIfQueCfgWrite () address is not on 16 word boundary\n");
+    }
+    /* Now convert it to a 16 word pointer as required by QUECONFIG register */
+    baseAddress >>= IX_QMGR_BASE_ADDR_16_WORD_SHIFT;
+    
+    
+    qCfg |= (baseAddress << IX_QMGR_Q_CONFIG_BADDR_OFFSET);
+
+
+    cfgAddress = (UINT32*)(aqmBaseAddress +
+			IX_QMGR_Q_CONFIG_ADDR_GET(qId));
+
+
+    /* NOTE: High and Low watermarks are set to zero */
+    ixQMgrAqmIfWordWrite (cfgAddress, qCfg);
+}
+
+void
+ixQMgrAqmIfQueCfgRead (IxQMgrQId qId,
+		       unsigned int numEntries,
+		       UINT32 *baseAddress,
+		       unsigned int *ne,
+		       unsigned int *nf,
+		       UINT32 *readPtr,
+		       UINT32 *writePtr)
+{
+    UINT32 qcfg;
+    UINT32 *cfgAddress = (UINT32*)(aqmBaseAddress + IX_QMGR_Q_CONFIG_ADDR_GET(qId));
+    unsigned int qEntrySizeInwords;
+    unsigned int qSizeInWords;
+    UINT32 *readPtr_ = NULL;
+	
+    /* Read the queue configuration register */
+    ixQMgrAqmIfWordRead (cfgAddress, &qcfg);
+    
+    /* Extract the base address */
+    *baseAddress = (UINT32)((qcfg & IX_QMGR_BADDR_MASK) >>
+			    (IX_QMGR_Q_CONFIG_BADDR_OFFSET));
+
+    /* Base address is a 16 word pointer from the start of AQM SRAM.
+     * Convert to absolute word address.
+     */
+    *baseAddress <<= IX_QMGR_BASE_ADDR_16_WORD_SHIFT;
+    *baseAddress += (UINT32)IX_QMGR_QUECONFIG_BASE_OFFSET;
+
+    /*
+     * Extract the watermarks. 0->0 entries, 1->1 entries, 2->2 entries, 3->4 entries......
+     * If ne > 0 ==> neInEntries = 2^(ne - 1)
+     * If ne == 0 ==> neInEntries = 0
+     * The same applies.
+     */
+    *ne = ((qcfg) >> (IX_QMGR_Q_CONFIG_NE_OFFSET)) & IX_QMGR_NE_MASK;
+    *nf = ((qcfg) >> (IX_QMGR_Q_CONFIG_NF_OFFSET)) & IX_QMGR_NF_MASK;
+
+    if (0 != *ne)
+    {
+	*ne = 1 << (*ne - 1);	
+    }
+    if (0 != *nf)
+    {
+	*nf = 1 << (*nf - 1);
+    }
+
+    /* Get the queue entry size in words */
+    qEntrySizeInwords = ixQMgrQEntrySizeInWordsGet (qId);
+
+    /* Get the queue size in words */
+    qSizeInWords = ixQMgrQSizeInWordsGet (qId);
+
+    ixQMgrAqmIfEntryAddressGet (0/* Entry 0. i.e the readPtr*/,
+				qcfg,
+				qEntrySizeInwords,
+				qSizeInWords,
+				&readPtr_);
+    *readPtr = (UINT32)readPtr_;
+    *readPtr -= (UINT32)aqmBaseAddress;/* Offset, not absolute address */
+
+    *writePtr = (qcfg >> IX_QMGR_Q_CONFIG_WRPTR_OFFSET) & IX_QMGR_WRPTR_MASK;
+    *writePtr = *baseAddress + (*writePtr * (IX_QMGR_NUM_BYTES_PER_WORD));
+    return;
+}
+
+unsigned
+ixQMgrAqmIfLog2 (unsigned number)
+{
+    unsigned count = 0;
+
+    /*
+     * N.B. this function will return 0
+     * for ixQMgrAqmIfLog2 (0)
+     */
+    while (number/2)
+    {
+	number /=2;
+	count++;	
+    }
+
+    return count;
+}
+
+void ixQMgrAqmIfIntSrcSelReg0Bit3Set (void)
+{
+
+    volatile UINT32 *registerAddress;
+    UINT32 registerWord; 
+
+    /*
+     * Calculate the registerAddress
+     * multiple queues split accross registers
+     */
+    registerAddress = (UINT32*)(aqmBaseAddress +
+				IX_QMGR_INT0SRCSELREG0_OFFSET);    
+
+    /* Read the current data */
+    ixQMgrAqmIfWordRead (registerAddress, &registerWord);
+
+    /* Set the write bits */
+    registerWord |= (1<<IX_QMGR_INT0SRCSELREG0_BIT3) ;
+
+    /*
+     * Write the data
+     */
+    ixQMgrAqmIfWordWrite (registerAddress, registerWord);
+}  
+
+
+void
+ixQMgrAqmIfIntSrcSelWrite (IxQMgrQId qId,
+			  IxQMgrSourceId sourceId)
+{
+    ixQMgrAqmIfQRegisterBitsWrite (qId,
+				   IX_QMGR_INT0SRCSELREG0_OFFSET,
+				   IX_QMGR_INTSRC_NUM_QUE_PER_WORD,
+				   sourceId);
+}
+
+
+
+void
+ixQMgrAqmIfWatermarkSet (IxQMgrQId qId,
+			unsigned ne,
+			unsigned nf)
+{
+    volatile UINT32 *address = 0;
+    UINT32 value = 0;
+    unsigned aqmNeWatermark = 0;
+    unsigned aqmNfWatermark = 0;
+
+    address = (UINT32*)(aqmBaseAddress +
+		     IX_QMGR_Q_CONFIG_ADDR_GET(qId));
+
+    aqmNeWatermark = watermarkToAqmWatermark (ne);
+    aqmNfWatermark = watermarkToAqmWatermark (nf);
+
+    /* Read the current watermarks */
+    ixQMgrAqmIfWordRead (address, &value);
+
+    /* Clear out the old watermarks */
+    value &=  IX_QMGR_NE_NF_CLEAR_MASK;
+    
+    /* Generate the value to write */
+    value |= (aqmNeWatermark << IX_QMGR_Q_CONFIG_NE_OFFSET) |
+	(aqmNfWatermark << IX_QMGR_Q_CONFIG_NF_OFFSET); 
+
+    ixQMgrAqmIfWordWrite (address, value);
+
+}
+
+PRIVATE void
+ixQMgrAqmIfEntryAddressGet (unsigned int entryIndex,
+			    UINT32 configRegWord,
+			    unsigned int qEntrySizeInwords,
+			    unsigned int qSizeInWords,
+			    UINT32 **address)
+{
+    UINT32 readPtr;
+    UINT32 baseAddress;
+    UINT32 *topOfAqmSram;
+
+    topOfAqmSram = ((UINT32 *)aqmBaseAddress + IX_QMGR_AQM_ADDRESS_SPACE_SIZE_IN_WORDS);
+
+    /* Extract the base address */
+    baseAddress = (UINT32)((configRegWord & IX_QMGR_BADDR_MASK) >>
+			   (IX_QMGR_Q_CONFIG_BADDR_OFFSET));
+
+    /* Base address is a 16 word pointer from the start of AQM SRAM.
+     * Convert to absolute word address.
+     */
+    baseAddress <<= IX_QMGR_BASE_ADDR_16_WORD_SHIFT;
+    baseAddress += ((UINT32)aqmBaseAddress + (UINT32)IX_QMGR_QUECONFIG_BASE_OFFSET);
+
+    /* Extract the read pointer. Read pointer is a word pointer */
+    readPtr = (UINT32)((configRegWord >>
+			IX_QMGR_Q_CONFIG_RDPTR_OFFSET)&IX_QMGR_RDPTR_MASK);
+
+    /* Read/Write pointers(word pointers)  are offsets from the queue buffer space base address.
+     * Calculate the absolute read pointer address. NOTE: Queues are circular buffers.
+     */
+    readPtr  = (readPtr + (entryIndex * qEntrySizeInwords)) & (qSizeInWords - 1); /* Mask by queue size */
+    *address = (UINT32 *)(baseAddress + (readPtr * (IX_QMGR_NUM_BYTES_PER_WORD)));
+
+    switch (qEntrySizeInwords)
+    {
+	case IX_QMGR_Q_ENTRY_SIZE1:
+	    IX_OSAL_ASSERT((*address + IX_QMGR_ENTRY1_OFFSET) < topOfAqmSram);	    
+	    break;
+	case IX_QMGR_Q_ENTRY_SIZE2:
+	    IX_OSAL_ASSERT((*address + IX_QMGR_ENTRY2_OFFSET) < topOfAqmSram);
+	    break;
+	case IX_QMGR_Q_ENTRY_SIZE4:
+	    IX_OSAL_ASSERT((*address + IX_QMGR_ENTRY4_OFFSET) < topOfAqmSram);
+	    break;
+	default:
+	    IX_QMGR_LOG_ERROR0("Invalid Q Entry size passed to ixQMgrAqmIfEntryAddressGet");
+	    break;
+    }
+    
+}
+
+IX_STATUS
+ixQMgrAqmIfQPeek (IxQMgrQId qId,
+		  unsigned int entryIndex,
+		  unsigned int *entry)
+{
+    UINT32 *cfgRegAddress = (UINT32*)(aqmBaseAddress + IX_QMGR_Q_CONFIG_ADDR_GET(qId));
+    UINT32 *entryAddress = NULL;
+    UINT32 configRegWordOnEntry;
+    UINT32 configRegWordOnExit;
+    unsigned int qEntrySizeInwords;
+    unsigned int qSizeInWords;
+
+    /* Get the queue entry size in words */
+    qEntrySizeInwords = ixQMgrQEntrySizeInWordsGet (qId);
+
+    /* Get the queue size in words */
+    qSizeInWords = ixQMgrQSizeInWordsGet (qId);
+
+    /* Read the config register */
+    ixQMgrAqmIfWordRead (cfgRegAddress, &configRegWordOnEntry);
+
+    /* Get the entry address */
+    ixQMgrAqmIfEntryAddressGet (entryIndex,
+				configRegWordOnEntry,
+				qEntrySizeInwords,
+				qSizeInWords,
+				&entryAddress);
+
+    /* Get the lock or return busy */
+    if (IX_SUCCESS != ixOsalFastMutexTryLock(&ixQMgrAqmIfPeekPokeFastMutex[qId]))
+    {
+	return IX_FAIL;
+    }
+
+    while(qEntrySizeInwords--)
+    {
+	ixQMgrAqmIfWordRead (entryAddress++, entry++);
+    }
+
+    /* Release the lock */
+    ixOsalFastMutexUnlock(&ixQMgrAqmIfPeekPokeFastMutex[qId]);
+
+    /* Read the config register */
+    ixQMgrAqmIfWordRead (cfgRegAddress, &configRegWordOnExit);
+
+    /* Check that the read and write pointers have not changed */
+    if (configRegWordOnEntry != configRegWordOnExit)
+    {
+	return IX_FAIL;
+    }
+
+    return IX_SUCCESS;
+}
+
+IX_STATUS
+ixQMgrAqmIfQPoke (IxQMgrQId qId,
+		  unsigned entryIndex,
+		  unsigned int *entry)
+{
+    UINT32 *cfgRegAddress = (UINT32*)(aqmBaseAddress + IX_QMGR_Q_CONFIG_ADDR_GET(qId));
+    UINT32 *entryAddress = NULL;
+    UINT32 configRegWordOnEntry;
+    UINT32 configRegWordOnExit;
+    unsigned int qEntrySizeInwords;
+    unsigned int qSizeInWords;
+    
+    /* Get the queue entry size in words */
+    qEntrySizeInwords = ixQMgrQEntrySizeInWordsGet (qId);
+
+    /* Get the queue size in words */
+    qSizeInWords = ixQMgrQSizeInWordsGet (qId);
+
+    /* Read the config register */
+    ixQMgrAqmIfWordRead (cfgRegAddress, &configRegWordOnEntry);
+
+    /* Get the entry address */
+    ixQMgrAqmIfEntryAddressGet (entryIndex,
+				configRegWordOnEntry,
+				qEntrySizeInwords,
+				qSizeInWords,
+				&entryAddress);
+
+    /* Get the lock or return busy */
+    if (IX_SUCCESS != ixOsalFastMutexTryLock(&ixQMgrAqmIfPeekPokeFastMutex[qId]))
+    {
+	return IX_FAIL;
+    }
+
+    /* Else read the entry directly from SRAM. This will not move the read pointer */
+    while(qEntrySizeInwords--)
+    {
+	ixQMgrAqmIfWordWrite (entryAddress++, *entry++);
+    }
+
+    /* Release the lock */
+    ixOsalFastMutexUnlock(&ixQMgrAqmIfPeekPokeFastMutex[qId]);
+
+    /* Read the config register */
+    ixQMgrAqmIfWordRead (cfgRegAddress, &configRegWordOnExit);
+
+    /* Check that the read and write pointers have not changed */
+    if (configRegWordOnEntry != configRegWordOnExit)
+    {
+	return IX_FAIL;
+    }
+
+    return IX_SUCCESS;
+}
+
+PRIVATE unsigned
+watermarkToAqmWatermark (IxQMgrWMLevel watermark )
+{
+    unsigned aqmWatermark = 0;
+
+    /*
+     * Watermarks 0("000"),1("001"),2("010"),4("011"),
+     * 8("100"),16("101"),32("110"),64("111")
+     */
+    aqmWatermark = ixQMgrAqmIfLog2 (watermark * 2);
+    
+    return aqmWatermark;
+}
+
+PRIVATE unsigned
+entrySizeToAqmEntrySize (IxQMgrQEntrySizeInWords entrySize)
+{
+    /* entrySize  1("00"),2("01"),4("10") */
+    return (ixQMgrAqmIfLog2 (entrySize));
+}
+
+PRIVATE unsigned
+bufferSizeToAqmBufferSize (unsigned bufferSizeInWords)
+{
+    /* bufferSize 16("00"),32("01),64("10"),128("11") */
+    return (ixQMgrAqmIfLog2 (bufferSizeInWords / IX_QMGR_MIN_BUFFER_SIZE));
+}
+
+/*
+ * Reset AQM registers to default values.
+ */
+PRIVATE void
+ixQMgrAqmIfRegistersReset (void)
+{
+    volatile UINT32 *qConfigWordAddress = NULL;
+    unsigned int i;
+
+    /*
+     * Need to initialize AQM hardware registers to an initial
+     * value as init may have been called as a result of a soft
+     * reset. i.e. soft reset does not reset hardware registers.
+     */
+
+    /* Reset queues 0..31 status registers 0..3 */
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QUELOWSTAT0_OFFSET), 
+			 IX_QMGR_QUELOWSTAT_RESET_VALUE);
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QUELOWSTAT1_OFFSET), 
+			 IX_QMGR_QUELOWSTAT_RESET_VALUE);
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QUELOWSTAT2_OFFSET), 
+			 IX_QMGR_QUELOWSTAT_RESET_VALUE);
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QUELOWSTAT3_OFFSET), 
+			 IX_QMGR_QUELOWSTAT_RESET_VALUE);
+
+    /* Reset underflow/overflow status registers 0..1 */
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QUEUOSTAT0_OFFSET), 
+			 IX_QMGR_QUEUOSTAT_RESET_VALUE);
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QUEUOSTAT1_OFFSET), 
+			 IX_QMGR_QUEUOSTAT_RESET_VALUE);
+    
+    /* Reset queues 32..63 nearly empty status registers */
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QUEUPPSTAT0_OFFSET),
+			 IX_QMGR_QUEUPPSTAT0_RESET_VALUE);
+
+    /* Reset queues 32..63 full status registers */
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QUEUPPSTAT1_OFFSET),
+			 IX_QMGR_QUEUPPSTAT1_RESET_VALUE);
+
+    /* Reset int0 status flag source select registers 0..3 */
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_INT0SRCSELREG0_OFFSET),
+			 IX_QMGR_INT0SRCSELREG_RESET_VALUE);
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_INT0SRCSELREG1_OFFSET),
+			 IX_QMGR_INT0SRCSELREG_RESET_VALUE);
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_INT0SRCSELREG2_OFFSET),
+			 IX_QMGR_INT0SRCSELREG_RESET_VALUE);
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_INT0SRCSELREG3_OFFSET),
+			 IX_QMGR_INT0SRCSELREG_RESET_VALUE);
+	 
+    /* Reset queue interrupt enable register 0..1 */
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QUEIEREG0_OFFSET),
+			 IX_QMGR_QUEIEREG_RESET_VALUE);
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QUEIEREG1_OFFSET),
+			 IX_QMGR_QUEIEREG_RESET_VALUE);
+
+    /* Reset queue interrupt register 0..1 */
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QINTREG0_OFFSET),
+			 IX_QMGR_QINTREG_RESET_VALUE);
+    ixQMgrAqmIfWordWrite((UINT32 *)(aqmBaseAddress + IX_QMGR_QINTREG1_OFFSET),
+			 IX_QMGR_QINTREG_RESET_VALUE);
+
+    /* Reset queue configuration words 0..63 */
+    qConfigWordAddress = (UINT32 *)(aqmBaseAddress + IX_QMGR_QUECONFIG_BASE_OFFSET);
+    for (i = 0; i < (IX_QMGR_QUECONFIG_SIZE / sizeof(UINT32)); i++)
+    {
+	ixQMgrAqmIfWordWrite(qConfigWordAddress,
+			     IX_QMGR_QUECONFIG_RESET_VALUE);
+	/* Next word */
+	qConfigWordAddress++;
+    }
+}
+

cpu/ixp/npe/IxQMgrDispatcher.c

@@ -0,0 +1,1347 @@
+/**
+ * @file    IxQMgrDispatcher.c
+ *
+ * @author Intel Corporation
+ * @date    20-Dec-2001
+ *    
+ * @brief   This file contains the implementation of the Dispatcher sub component
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+/*
+ * User defined include files.
+ */
+#include "IxQMgr.h"
+#include "IxQMgrAqmIf_p.h"
+#include "IxQMgrQCfg_p.h"
+#include "IxQMgrDispatcher_p.h"
+#include "IxQMgrLog_p.h"
+#include "IxQMgrDefines_p.h"
+#include "IxFeatureCtrl.h"
+#include "IxOsal.h"
+
+
+
+/*
+ * #defines and macros used in this file.
+ */
+
+
+/*
+ * This constant is used to indicate the number of priority levels supported
+ */
+#define IX_QMGR_NUM_PRIORITY_LEVELS 3
+
+/* 
+ * This constant is used to set the size of the array of status words
+ */
+#define MAX_Q_STATUS_WORDS      4
+
+/*
+ * This macro is used to check if a given priority is valid
+ */
+#define IX_QMGR_DISPATCHER_PRIORITY_CHECK(priority) \
+(((priority) >= IX_QMGR_Q_PRIORITY_0) && ((priority) <= IX_QMGR_Q_PRIORITY_2))
+
+/*
+ * This macto is used to check that a given interrupt source is valid
+ */
+#define IX_QMGR_DISPATCHER_SOURCE_ID_CHECK(srcSel) \
+(((srcSel) >= IX_QMGR_Q_SOURCE_ID_E) && ((srcSel) <= IX_QMGR_Q_SOURCE_ID_NOT_F))
+
+/*
+ * Number of times a dummy callback is called before logging a trace
+ * message
+ */
+#define LOG_THROTTLE_COUNT 1000000
+
+/* Priority tables limits */
+#define IX_QMGR_MIN_LOW_QUE_PRIORITY_TABLE_INDEX (0)
+#define IX_QMGR_MID_LOW_QUE_PRIORITY_TABLE_INDEX (16)
+#define IX_QMGR_MAX_LOW_QUE_PRIORITY_TABLE_INDEX (31)
+#define IX_QMGR_MIN_UPP_QUE_PRIORITY_TABLE_INDEX (32)
+#define IX_QMGR_MID_UPP_QUE_PRIORITY_TABLE_INDEX (48)
+#define IX_QMGR_MAX_UPP_QUE_PRIORITY_TABLE_INDEX (63)
+ 
+/*
+ * This macro is used to check if a given callback type is valid
+ */
+#define IX_QMGR_DISPATCHER_CALLBACK_TYPE_CHECK(type) \
+            (((type) >= IX_QMGR_TYPE_REALTIME_OTHER) && \
+            ((type) <= IX_QMGR_TYPE_REALTIME_SPORADIC))
+
+/* 
+ * define max index in lower queue to use in loops 
+ */
+#define IX_QMGR_MAX_LOW_QUE_TABLE_INDEX (31)
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+
+/*
+ * Information on a queue needed by the Dispatcher
+ */
+typedef struct 
+{
+    IxQMgrCallback callback;       /* Notification callback                  */
+    IxQMgrCallbackId callbackId;   /* Notification callback identifier       */
+    unsigned dummyCallbackCount;   /* Number of times runs of dummy callback */
+    IxQMgrPriority priority;       /* Dispatch priority                      */
+    unsigned int statusWordOffset; /* Offset to the status word to check     */
+    UINT32 statusMask;             /* Status mask                            */    
+    UINT32 statusCheckValue;       /* Status check value                     */
+    UINT32 intRegCheckMask;	   /* Interrupt register check mask          */
+} IxQMgrQInfo;
+
+/*
+ * Variable declarations global to this file. Externs are followed by
+ * statics.
+ */
+
+/* 
+ * Flag to keep record of what dispatcher set in featureCtrl when ixQMgrInit()
+ * is called. This is needed because it is possible that a client might
+ * change whether the live lock prevention dispatcher is used between
+ * calls to ixQMgrInit() and ixQMgrDispatcherLoopGet(). 
+ */
+PRIVATE IX_STATUS ixQMgrOrigB0Dispatcher = IX_FEATURE_CTRL_COMPONENT_ENABLED;
+
+/* 
+ * keep record of Q types - not in IxQMgrQInfo for performance as
+ * it is only used with ixQMgrDispatcherLoopRunB0LLP()
+ */
+PRIVATE IxQMgrType ixQMgrQTypes[IX_QMGR_MAX_NUM_QUEUES];
+
+/*
+ * This array contains a list of queue identifiers ordered by priority. The table
+ * is split logically between queue identifiers 0-31 and 32-63.
+ */
+static IxQMgrQId priorityTable[IX_QMGR_MAX_NUM_QUEUES];
+
+/*
+ * This flag indicates to the dispatcher that the priority table needs to be rebuilt.
+ */
+static BOOL rebuildTable = FALSE;
+
+/* Dispatcher statistics */
+static IxQMgrDispatcherStats dispatcherStats;
+
+/* Table of queue information */
+static IxQMgrQInfo dispatchQInfo[IX_QMGR_MAX_NUM_QUEUES];
+
+/* Masks use to identify the first queues in the priority tables 
+*  when comparing with the interrupt register
+*/
+static unsigned int lowPriorityTableFirstHalfMask;
+static unsigned int uppPriorityTableFirstHalfMask;
+
+/*
+ * Static function prototypes
+ */
+
+/*
+ * This function is the default callback for all queues
+ */
+PRIVATE void
+dummyCallback (IxQMgrQId qId,	      
+	       IxQMgrCallbackId cbId);
+
+PRIVATE void
+ixQMgrDispatcherReBuildPriorityTable (void);
+
+/*
+ * Function definitions.
+ */
+void
+ixQMgrDispatcherInit (void)
+{
+    int i;
+    IxFeatureCtrlProductId productId = 0;
+    IxFeatureCtrlDeviceId deviceId = 0;
+    BOOL stickyIntSilicon = TRUE; 
+
+    /* Set default priorities */
+    for (i=0; i< IX_QMGR_MAX_NUM_QUEUES; i++)
+    {
+	dispatchQInfo[i].callback = dummyCallback;
+	dispatchQInfo[i].callbackId = 0;
+	dispatchQInfo[i].dummyCallbackCount = 0;
+	dispatchQInfo[i].priority = IX_QMGR_Q_PRIORITY_2;
+	dispatchQInfo[i].statusWordOffset = 0;
+	dispatchQInfo[i].statusCheckValue = 0;
+	dispatchQInfo[i].statusMask = 0;  
+        /* 
+	 * There are two interrupt registers, 32 bits each. One for the lower
+	 * queues(0-31) and one for the upper queues(32-63). Therefore need to
+	 * mod by 32 i.e the min upper queue identifier.
+	 */
+	dispatchQInfo[i].intRegCheckMask = (1<<(i%(IX_QMGR_MIN_QUEUPP_QID)));
+
+        /* 
+         * Set the Q types - will only be used with livelock 
+         */
+        ixQMgrQTypes[i] = IX_QMGR_TYPE_REALTIME_OTHER;
+
+	/* Reset queue statistics */
+	dispatcherStats.queueStats[i].callbackCnt = 0;
+	dispatcherStats.queueStats[i].priorityChangeCnt = 0;
+	dispatcherStats.queueStats[i].intNoCallbackCnt = 0;
+	dispatcherStats.queueStats[i].intLostCallbackCnt = 0;
+        dispatcherStats.queueStats[i].notificationEnabled = FALSE;
+        dispatcherStats.queueStats[i].srcSel = 0;
+
+    }
+
+    /* Priority table. Order the table from queue 0 to 63 */
+    ixQMgrDispatcherReBuildPriorityTable();
+
+    /* Reset statistics */
+    dispatcherStats.loopRunCnt = 0;
+
+    /* Get the device ID for the underlying silicon */
+    deviceId = ixFeatureCtrlDeviceRead();
+    
+    /* Get the product ID for the underlying silicon */
+    productId = ixFeatureCtrlProductIdRead();
+
+    /* 
+     * Check featureCtrl to see if Livelock prevention is required 
+     */
+    ixQMgrOrigB0Dispatcher = ixFeatureCtrlSwConfigurationCheck( 
+                                 IX_FEATURECTRL_ORIGB0_DISPATCHER);
+
+    /*
+     * Check if the silicon supports the sticky interrupt feature.
+     * IF (IXP42X AND A0) -> No sticky interrupt feature supported 
+     */
+    if ((IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X == 
+        (IX_FEATURE_CTRL_DEVICE_TYPE_MASK & deviceId)) &&
+        (IX_FEATURE_CTRL_SILICON_TYPE_A0 == 
+        (IX_FEATURE_CTRL_SILICON_STEPPING_MASK & productId))) 
+    {
+       stickyIntSilicon = FALSE;
+    }
+
+    /*
+     * IF user wants livelock prev option AND silicon supports sticky interrupt 
+     * feature -> enable the sticky interrupt bit
+     */
+    if ((IX_FEATURE_CTRL_SWCONFIG_DISABLED == ixQMgrOrigB0Dispatcher) &&
+         stickyIntSilicon)  
+    {
+        ixQMgrStickyInterruptRegEnable();
+    }
+}
+
+IX_STATUS
+ixQMgrDispatcherPrioritySet (IxQMgrQId qId,
+			     IxQMgrPriority priority)
+{   
+    int ixQMgrLockKey;
+
+    if (!ixQMgrQIsConfigured(qId))
+    {
+	return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+    
+    if (!IX_QMGR_DISPATCHER_PRIORITY_CHECK(priority))
+    {
+	return IX_QMGR_Q_INVALID_PRIORITY;
+    }
+
+    ixQMgrLockKey = ixOsalIrqLock();
+    
+    /* Change priority */
+    dispatchQInfo[qId].priority = priority;
+    /* Set flag */
+    rebuildTable = TRUE;
+
+    ixOsalIrqUnlock(ixQMgrLockKey);
+
+#ifndef NDEBUG
+    /* Update statistics */
+    dispatcherStats.queueStats[qId].priorityChangeCnt++;
+#endif
+
+    return IX_SUCCESS;
+}
+
+IX_STATUS
+ixQMgrNotificationCallbackSet (IxQMgrQId qId,
+			       IxQMgrCallback callback,
+			       IxQMgrCallbackId callbackId)
+{
+    if (!ixQMgrQIsConfigured(qId))
+    {
+	return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+
+    if (NULL == callback)
+    {
+	/* Reset to dummy callback */
+	dispatchQInfo[qId].callback = dummyCallback;
+	dispatchQInfo[qId].dummyCallbackCount = 0;
+	dispatchQInfo[qId].callbackId = 0;
+    }
+    else 
+    {
+	dispatchQInfo[qId].callback = callback;
+	dispatchQInfo[qId].callbackId = callbackId;
+    }
+
+    return IX_SUCCESS;
+}
+
+IX_STATUS
+ixQMgrNotificationEnable (IxQMgrQId qId, 
+			  IxQMgrSourceId srcSel)
+{
+    IxQMgrQStatus qStatusOnEntry;/* The queue status on entry/exit */
+    IxQMgrQStatus qStatusOnExit; /* to this function               */
+    int ixQMgrLockKey;
+
+#ifndef NDEBUG
+    if (!ixQMgrQIsConfigured (qId))
+    {
+	return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+
+    if ((qId < IX_QMGR_MIN_QUEUPP_QID) &&
+       !IX_QMGR_DISPATCHER_SOURCE_ID_CHECK(srcSel))
+    {
+	/* QId 0-31 source id invalid */
+	return IX_QMGR_INVALID_INT_SOURCE_ID;
+    }
+
+    if ((IX_QMGR_Q_SOURCE_ID_NE != srcSel) &&
+	(qId >= IX_QMGR_MIN_QUEUPP_QID))
+    {
+	/*
+	 * For queues 32-63 the interrupt source is fixed to the Nearly
+	 * Empty status flag and therefore should have a srcSel of NE.
+	 */
+	return IX_QMGR_INVALID_INT_SOURCE_ID;
+    }
+#endif
+
+#ifndef NDEBUG
+    dispatcherStats.queueStats[qId].notificationEnabled = TRUE;
+    dispatcherStats.queueStats[qId].srcSel = srcSel;
+#endif
+
+    /* Get the current queue status */
+    ixQMgrAqmIfQueStatRead (qId, &qStatusOnEntry);
+  
+    /* 
+     * Enabling interrupts results in Read-Modify-Write
+     * so need critical section
+     */
+
+    ixQMgrLockKey = ixOsalIrqLock();
+
+    /* Calculate the checkMask and checkValue for this q */
+    ixQMgrAqmIfQStatusCheckValsCalc (qId,
+				     srcSel,
+				     &dispatchQInfo[qId].statusWordOffset,
+				     &dispatchQInfo[qId].statusCheckValue,
+				     &dispatchQInfo[qId].statusMask);
+
+
+    /* Set the interupt source is this queue is in the range 0-31 */
+    if (qId < IX_QMGR_MIN_QUEUPP_QID)
+    {
+	ixQMgrAqmIfIntSrcSelWrite (qId, srcSel);
+    }
+
+    /* Enable the interrupt */
+    ixQMgrAqmIfQInterruptEnable (qId);
+
+    ixOsalIrqUnlock(ixQMgrLockKey);
+    
+    /* Get the current queue status */
+    ixQMgrAqmIfQueStatRead (qId, &qStatusOnExit);
+  
+    /* If the status has changed return a warning */
+    if (qStatusOnEntry != qStatusOnExit)
+    {
+	return IX_QMGR_WARNING;
+    }
+    
+    return IX_SUCCESS;
+}
+
+
+IX_STATUS
+ixQMgrNotificationDisable (IxQMgrQId qId)
+{
+    int ixQMgrLockKey;
+
+#ifndef NDEBUG
+    /* Validate parameters */
+    if (!ixQMgrQIsConfigured (qId))
+    {
+	return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+#endif
+  
+    /* 
+     * Enabling interrupts results in Read-Modify-Write
+     * so need critical section
+     */
+#ifndef NDEBUG
+    dispatcherStats.queueStats[qId].notificationEnabled = FALSE;
+#endif
+
+    ixQMgrLockKey = ixOsalIrqLock();
+
+    ixQMgrAqmIfQInterruptDisable (qId);
+    
+    ixOsalIrqUnlock(ixQMgrLockKey);
+
+    return IX_SUCCESS;    
+}
+
+void 
+ixQMgrStickyInterruptRegEnable(void)
+{
+ /* Use Aqm If function to set Interrupt Register0 Bit-3 */ 
+ ixQMgrAqmIfIntSrcSelReg0Bit3Set ();   
+}
+
+#if !defined __XSCALE__ || defined __linux
+
+/* Count the number of leading zero bits in a word,
+ * and return the same value than the CLZ instruction.
+ *
+ * word (in)    return value (out)
+ * 0x80000000   0
+ * 0x40000000   1
+ * ,,,          ,,,
+ * 0x00000002   30
+ * 0x00000001   31
+ * 0x00000000   32
+ *
+ * The C version of this function is used as a replacement 
+ * for system not providing the equivalent of the CLZ 
+ * assembly language instruction.
+ *
+ * Note that this version is big-endian
+ */
+unsigned int
+ixQMgrCountLeadingZeros(UINT32 word)
+{
+  unsigned int leadingZerosCount = 0;
+
+  if (word == 0)
+  {
+      return 32;
+  }
+  /* search the first bit set by testing the MSB and shifting the input word */
+  while ((word & 0x80000000) == 0)
+  {
+      word <<= 1;
+      leadingZerosCount++;
+  }
+  return leadingZerosCount;
+}
+#endif /* not  __XSCALE__ or __linux */
+
+void
+ixQMgrDispatcherLoopGet (IxQMgrDispatcherFuncPtr *qDispatcherFuncPtr)
+{
+  IxFeatureCtrlProductId productId = 0;
+  IxFeatureCtrlDeviceId deviceId = 0;
+  
+  /* Get the device ID for the underlying silicon */
+  deviceId = ixFeatureCtrlDeviceRead();
+
+  /* Get the product ID for the underlying silicon */
+  productId = ixFeatureCtrlProductIdRead ();
+
+  /* IF (IXP42X AND A0 silicon) -> use ixQMgrDispatcherLoopRunA0 */
+  if ((IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X ==
+      (IX_FEATURE_CTRL_DEVICE_TYPE_MASK & deviceId)) &&
+      (IX_FEATURE_CTRL_SILICON_TYPE_A0 ==  
+      (IX_FEATURE_CTRL_SILICON_STEPPING_MASK & productId)))  
+  {
+    /*For IXP42X A0 silicon */
+    *qDispatcherFuncPtr = &ixQMgrDispatcherLoopRunA0 ;
+  } 
+  else /*For IXP42X B0 or IXP46X silicon*/ 
+  { 
+    if (IX_FEATURE_CTRL_SWCONFIG_ENABLED == ixQMgrOrigB0Dispatcher)
+    {
+        /* Default for IXP42X B0 and IXP46X silicon */
+        *qDispatcherFuncPtr = &ixQMgrDispatcherLoopRunB0;
+    }
+    else 
+    {
+        /* FeatureCtrl indicated that livelock dispatcher be used */
+        *qDispatcherFuncPtr = &ixQMgrDispatcherLoopRunB0LLP;
+    }
+  }
+}
+
+void
+ixQMgrDispatcherLoopRunA0 (IxQMgrDispatchGroup group)
+{
+    UINT32 intRegVal;                /* Interrupt reg val */
+    UINT32 intRegValAfterWrite;      /* Interrupt reg val after writing back */
+    UINT32 intRegCheckMask;          /* Mask for checking interrupt bits */
+    UINT32 qStatusWordsB4Write[MAX_Q_STATUS_WORDS];  /* Status b4 interrupt write */
+    UINT32 qStatusWordsAfterWrite[MAX_Q_STATUS_WORDS]; /* Status after interrupt write */
+    IxQMgrQInfo *currDispatchQInfo;
+    BOOL statusChangeFlag;
+
+    int priorityTableIndex;/* Priority table index */
+    int qIndex;            /* Current queue being processed */
+    int endIndex;          /* Index of last queue to process */
+
+#ifndef NDEBUG
+    IX_OSAL_ASSERT((group == IX_QMGR_QUEUPP_GROUP) || 
+	      (group == IX_QMGR_QUELOW_GROUP));
+#endif
+
+    /* Read Q status registers before interrupt status read/write */
+    ixQMgrAqmIfQStatusRegsRead (group, qStatusWordsB4Write);
+
+    /* Read the interrupt register */
+    ixQMgrAqmIfQInterruptRegRead (group, &intRegVal);
+
+    /* No bit set : nothing to process (the reaminder of the algorithm is
+    * based on the fact that the interrupt register value contains at
+    * least one bit set
+    */
+    if (intRegVal == 0) 
+    {
+#ifndef NDEBUG
+	/* Update statistics */
+	dispatcherStats.loopRunCnt++;
+#endif
+
+	/* Rebuild the priority table if needed */
+	if (rebuildTable)
+	{
+	    ixQMgrDispatcherReBuildPriorityTable ();
+	}
+
+	return;
+    }
+   
+    /* Write it back to clear the interrupt */
+    ixQMgrAqmIfQInterruptRegWrite (group, intRegVal);
+
+    /* Read Q status registers after interrupt status read/write */
+    ixQMgrAqmIfQStatusRegsRead (group, qStatusWordsAfterWrite);
+ 
+    /* get the first queue Id from the interrupt register value */
+    qIndex = (BITS_PER_WORD - 1) - ixQMgrCountLeadingZeros(intRegVal);
+
+    /* check if any change occured during hw register modifications */ 
+    if (IX_QMGR_QUELOW_GROUP == group)
+    {
+	statusChangeFlag = 
+	    (qStatusWordsB4Write[0] != qStatusWordsAfterWrite[0]) ||
+	    (qStatusWordsB4Write[1] != qStatusWordsAfterWrite[1]) ||
+	    (qStatusWordsB4Write[2] != qStatusWordsAfterWrite[2]) ||
+	    (qStatusWordsB4Write[3] != qStatusWordsAfterWrite[3]);
+    }
+    else
+    {
+	statusChangeFlag = 
+	    (qStatusWordsB4Write[0] != qStatusWordsAfterWrite[0]);
+	/* Set the queue range based on the queue group to proccess */
+	qIndex += IX_QMGR_MIN_QUEUPP_QID;
+    }
+
+    if (statusChangeFlag == FALSE)
+    {
+	/* check if the interrupt register contains 
+	 * only 1 bit set (happy day scenario)
+	 */
+	currDispatchQInfo = &dispatchQInfo[qIndex];
+	if (intRegVal == currDispatchQInfo->intRegCheckMask)
+	{
+	    /* only 1 queue event triggered a notification *
+	     * Call the callback function for this queue 
+	     */
+	    currDispatchQInfo->callback (qIndex,
+					 currDispatchQInfo->callbackId);  
+#ifndef NDEBUG
+	    /* Update statistics */
+	    dispatcherStats.queueStats[qIndex].callbackCnt++;
+#endif
+	}
+	else 
+	{
+	    /* the event is triggered by more than 1 queue, 
+	     * the queue search will be starting from the beginning
+	     * or the middle of the priority table
+	     *
+	     * the serach will end when all the bits of the interrupt
+	     * register are cleared. There is no need to maintain
+	     * a seperate value and test it at each iteration.
+	     */
+	    if (IX_QMGR_QUELOW_GROUP == group)
+	    {
+		/* check if any bit related to queues in the first
+		 * half of the priority table is set
+		 */
+		if (intRegVal & lowPriorityTableFirstHalfMask)
+		{
+		    priorityTableIndex = IX_QMGR_MIN_LOW_QUE_PRIORITY_TABLE_INDEX;
+		}
+		else
+		{
+		    priorityTableIndex = IX_QMGR_MID_LOW_QUE_PRIORITY_TABLE_INDEX;
+		}
+	    }
+	    else 
+	    {
+		/* check if any bit related to queues in the first
+		 * half of the priority table is set
+		 */
+		if (intRegVal & uppPriorityTableFirstHalfMask)
+		{
+		    priorityTableIndex = IX_QMGR_MIN_UPP_QUE_PRIORITY_TABLE_INDEX;
+		}
+		else
+		{
+		    priorityTableIndex = IX_QMGR_MID_UPP_QUE_PRIORITY_TABLE_INDEX;
+		}
+	    }
+	    
+	    /* iterate following the priority table until all the bits 
+	     * of the interrupt register are cleared.
+	     */
+	    do
+	    {
+		qIndex = priorityTable[priorityTableIndex++];
+		currDispatchQInfo = &dispatchQInfo[qIndex];
+		intRegCheckMask = currDispatchQInfo->intRegCheckMask;
+		
+		/* If this queue caused this interrupt to be raised */
+		if (intRegVal & intRegCheckMask)
+		{
+		    /* Call the callback function for this queue */
+		    currDispatchQInfo->callback (qIndex,
+						 currDispatchQInfo->callbackId);
+#ifndef NDEBUG
+		    /* Update statistics */
+		    dispatcherStats.queueStats[qIndex].callbackCnt++;
+#endif
+		    
+		    /* Clear the interrupt register bit */
+		    intRegVal &= ~intRegCheckMask;
+		}
+	    }
+	    while(intRegVal);
+	}
+    }
+    else
+    {
+    /* A change in queue status occured during the hw interrupt
+     * register update. To maintain the interrupt consistency, it
+     * is necessary to iterate through all queues of the queue group.
+     */
+
+    /* Read interrupt status again */
+    ixQMgrAqmIfQInterruptRegRead (group, &intRegValAfterWrite);
+
+    if (IX_QMGR_QUELOW_GROUP == group)
+    {
+	priorityTableIndex = IX_QMGR_MIN_LOW_QUE_PRIORITY_TABLE_INDEX;
+	endIndex = IX_QMGR_MAX_LOW_QUE_PRIORITY_TABLE_INDEX;
+    }
+    else
+    {
+	priorityTableIndex = IX_QMGR_MIN_UPP_QUE_PRIORITY_TABLE_INDEX;
+	endIndex = IX_QMGR_MAX_UPP_QUE_PRIORITY_TABLE_INDEX;
+    }
+
+    for ( ; priorityTableIndex<=endIndex; priorityTableIndex++)
+    {
+	qIndex = priorityTable[priorityTableIndex];
+	currDispatchQInfo = &dispatchQInfo[qIndex];
+	intRegCheckMask = currDispatchQInfo->intRegCheckMask;
+
+	/* If this queue caused this interrupt to be raised */
+	if (intRegVal & intRegCheckMask)
+	{  
+	    /* Call the callback function for this queue */
+	    currDispatchQInfo->callback (qIndex,
+					 currDispatchQInfo->callbackId);
+#ifndef NDEBUG
+	    /* Update statistics */
+	    dispatcherStats.queueStats[qIndex].callbackCnt++;
+#endif
+	    
+	} /* if (intRegVal .. */
+
+	/* 
+	 * If interrupt bit is set in intRegValAfterWrite don't
+	 * proceed as this will be caught in next interrupt
+	 */
+	else if ((intRegValAfterWrite & intRegCheckMask) == 0)
+	{
+	    /* Check if an interrupt was lost for this Q */
+	    if (ixQMgrAqmIfQStatusCheck(qStatusWordsB4Write,
+					qStatusWordsAfterWrite,
+					currDispatchQInfo->statusWordOffset,
+					currDispatchQInfo->statusCheckValue,
+					currDispatchQInfo->statusMask))
+	    {
+		/* Call the callback function for this queue */
+		currDispatchQInfo->callback (qIndex, 
+					     dispatchQInfo[qIndex].callbackId);                 
+#ifndef NDEBUG
+		/* Update statistics */
+		dispatcherStats.queueStats[qIndex].callbackCnt++;
+		dispatcherStats.queueStats[qIndex].intLostCallbackCnt++;
+#endif
+	    } /* if ixQMgrAqmIfQStatusCheck(.. */
+	} /* else if ((intRegValAfterWrite ... */
+    } /* for (priorityTableIndex=0 ... */
+    }
+
+    /* Rebuild the priority table if needed */
+    if (rebuildTable)
+    {
+	ixQMgrDispatcherReBuildPriorityTable ();
+    }
+
+#ifndef NDEBUG
+    /* Update statistics */
+    dispatcherStats.loopRunCnt++;
+#endif
+}
+
+
+
+void
+ixQMgrDispatcherLoopRunB0 (IxQMgrDispatchGroup group)
+{
+    UINT32 intRegVal;                /* Interrupt reg val */
+    UINT32 intRegCheckMask;          /* Mask for checking interrupt bits */
+    IxQMgrQInfo *currDispatchQInfo;
+
+
+    int priorityTableIndex; /* Priority table index */
+    int qIndex;             /* Current queue being processed */
+
+#ifndef NDEBUG
+    IX_OSAL_ASSERT((group == IX_QMGR_QUEUPP_GROUP) ||
+              (group == IX_QMGR_QUELOW_GROUP));
+    IX_OSAL_ASSERT((group == IX_QMGR_QUEUPP_GROUP) || 
+	      (group == IX_QMGR_QUELOW_GROUP));
+#endif
+
+    /* Read the interrupt register */
+    ixQMgrAqmIfQInterruptRegRead (group, &intRegVal);
+
+
+    /* No queue has interrupt register set */
+    if (intRegVal != 0)
+    {
+
+            /* Write it back to clear the interrupt */
+            ixQMgrAqmIfQInterruptRegWrite (group, intRegVal);
+
+            /* get the first queue Id from the interrupt register value */
+            qIndex = (BITS_PER_WORD - 1) - ixQMgrCountLeadingZeros(intRegVal);
+
+            if (IX_QMGR_QUEUPP_GROUP == group)
+            {
+                /* Set the queue range based on the queue group to proccess */
+                qIndex += IX_QMGR_MIN_QUEUPP_QID;
+            }
+
+            /* check if the interrupt register contains
+             * only 1 bit set
+             * For example:
+             *                                        intRegVal = 0x0010
+             *               currDispatchQInfo->intRegCheckMask = 0x0010
+             *    intRegVal == currDispatchQInfo->intRegCheckMask is TRUE.
+             */
+             currDispatchQInfo = &dispatchQInfo[qIndex];
+             if (intRegVal == currDispatchQInfo->intRegCheckMask)
+             {
+                /* only 1 queue event triggered a notification *
+                 * Call the callback function for this queue
+                 */
+                currDispatchQInfo->callback (qIndex,
+                                     currDispatchQInfo->callbackId);
+#ifndef NDEBUG
+                /* Update statistics */
+                dispatcherStats.queueStats[qIndex].callbackCnt++;
+#endif
+             }
+             else
+             {
+                 /* the event is triggered by more than 1 queue,
+                  * the queue search will be starting from the beginning
+                  * or the middle of the priority table
+                  *
+                  * the serach will end when all the bits of the interrupt
+                  * register are cleared. There is no need to maintain
+                  * a seperate value and test it at each iteration.
+                  */
+                 if (IX_QMGR_QUELOW_GROUP == group)
+                 {
+                     /* check if any bit related to queues in the first
+                      * half of the priority table is set
+                      */
+                     if (intRegVal & lowPriorityTableFirstHalfMask)
+                     {
+                         priorityTableIndex = IX_QMGR_MIN_LOW_QUE_PRIORITY_TABLE_INDEX;
+                     }
+                     else
+                     {
+                         priorityTableIndex = IX_QMGR_MID_LOW_QUE_PRIORITY_TABLE_INDEX;
+                     }
+                 }
+                else
+                 {
+                     /* check if any bit related to queues in the first
+                      * half of the priority table is set
+                      */
+                     if (intRegVal & uppPriorityTableFirstHalfMask)
+                     {
+                         priorityTableIndex = IX_QMGR_MIN_UPP_QUE_PRIORITY_TABLE_INDEX;
+                     }
+                     else
+                     {
+                         priorityTableIndex = IX_QMGR_MID_UPP_QUE_PRIORITY_TABLE_INDEX;
+                     }
+                 }
+
+                 /* iterate following the priority table until all the bits
+                  * of the interrupt register are cleared.
+                  */
+                 do
+                 {
+                     qIndex = priorityTable[priorityTableIndex++];
+                     currDispatchQInfo = &dispatchQInfo[qIndex];
+                     intRegCheckMask = currDispatchQInfo->intRegCheckMask;
+
+                     /* If this queue caused this interrupt to be raised */
+                     if (intRegVal & intRegCheckMask)
+                     {
+                         /* Call the callback function for this queue */
+                         currDispatchQInfo->callback (qIndex,
+                                              currDispatchQInfo->callbackId);
+#ifndef NDEBUG
+                         /* Update statistics */
+                         dispatcherStats.queueStats[qIndex].callbackCnt++;
+#endif
+
+                         /* Clear the interrupt register bit */
+                         intRegVal &= ~intRegCheckMask;
+                     }
+                  }
+                  while(intRegVal);
+             } /*End of intRegVal == currDispatchQInfo->intRegCheckMask */
+     } /* End of intRegVal != 0 */
+
+#ifndef NDEBUG
+    /* Update statistics */
+    dispatcherStats.loopRunCnt++;
+#endif
+
+    /* Rebuild the priority table if needed */
+    if (rebuildTable)
+    {
+        ixQMgrDispatcherReBuildPriorityTable ();
+    }
+}
+
+void
+ixQMgrDispatcherLoopRunB0LLP (IxQMgrDispatchGroup group)
+{
+    UINT32 intRegVal =0;                /* Interrupt reg val */
+    UINT32 intRegCheckMask;          /* Mask for checking interrupt bits */
+    IxQMgrQInfo *currDispatchQInfo;
+
+    int priorityTableIndex; /* Priority table index */
+    int qIndex;             /* Current queue being processed */
+
+    UINT32 intRegValCopy = 0;
+    UINT32 intEnableRegVal = 0;
+    UINT8 i = 0;
+
+#ifndef NDEBUG
+    IX_OSAL_ASSERT((group == IX_QMGR_QUEUPP_GROUP) ||
+              (group == IX_QMGR_QUELOW_GROUP));
+#endif
+
+    /* Read the interrupt register */
+    ixQMgrAqmIfQInterruptRegRead (group, &intRegVal);
+
+    /* 
+     * mask any interrupts that are not enabled 
+     */
+    ixQMgrAqmIfQInterruptEnableRegRead (group, &intEnableRegVal);
+    intRegVal &= intEnableRegVal;
+
+    /* No queue has interrupt register set */
+    if (intRegVal != 0)
+    {
+        if (IX_QMGR_QUELOW_GROUP == group)
+        {
+            /*
+             * As the sticky bit is set, the interrupt register will 
+             * not clear if write back at this point because the condition
+             * has not been cleared. Take a copy and write back later after
+             * the condition has been cleared
+             */
+            intRegValCopy = intRegVal;
+        }
+        else
+        {
+            /* no sticky for upper Q's, so write back now */
+            ixQMgrAqmIfQInterruptRegWrite (group, intRegVal);
+        }
+
+        /* get the first queue Id from the interrupt register value */
+        qIndex = (BITS_PER_WORD - 1) - ixQMgrCountLeadingZeros(intRegVal);
+
+        if (IX_QMGR_QUEUPP_GROUP == group)
+        {
+            /* Set the queue range based on the queue group to proccess */
+            qIndex += IX_QMGR_MIN_QUEUPP_QID;
+        }
+
+        /* check if the interrupt register contains
+        * only 1 bit set
+        * For example:
+        *                                        intRegVal = 0x0010
+        *               currDispatchQInfo->intRegCheckMask = 0x0010
+        *    intRegVal == currDispatchQInfo->intRegCheckMask is TRUE.
+        */
+        currDispatchQInfo = &dispatchQInfo[qIndex];
+        if (intRegVal == currDispatchQInfo->intRegCheckMask)
+        {
+
+            /* 
+             * check if Q type periodic -  only lower queues can
+             * have there type set to periodic 
+             */
+            if (IX_QMGR_TYPE_REALTIME_PERIODIC == ixQMgrQTypes[qIndex])
+            {
+                /* 
+                 * Disable the notifications on any sporadics 
+                 */
+                for (i=0; i <= IX_QMGR_MAX_LOW_QUE_TABLE_INDEX; i++)
+                {
+                    if (IX_QMGR_TYPE_REALTIME_SPORADIC == ixQMgrQTypes[i])
+                    {
+                        ixQMgrNotificationDisable(i);
+#ifndef NDEBUG
+                        /* Update statistics */
+                        dispatcherStats.queueStats[i].disableCount++;
+#endif
+                    }
+                }
+            }
+
+            currDispatchQInfo->callback (qIndex,
+                                         currDispatchQInfo->callbackId);
+#ifndef NDEBUG
+            /* Update statistics */
+            dispatcherStats.queueStats[qIndex].callbackCnt++;
+#endif
+        }
+        else
+        {
+            /* the event is triggered by more than 1 queue,
+            * the queue search will be starting from the beginning
+            * or the middle of the priority table
+            *
+            * the serach will end when all the bits of the interrupt
+            * register are cleared. There is no need to maintain
+            * a seperate value and test it at each iteration.
+            */
+            if (IX_QMGR_QUELOW_GROUP == group)
+            {
+                /* check if any bit related to queues in the first
+                 * half of the priority table is set
+                 */
+                if (intRegVal & lowPriorityTableFirstHalfMask)
+                {
+                    priorityTableIndex =
+                                       IX_QMGR_MIN_LOW_QUE_PRIORITY_TABLE_INDEX;
+                }
+                else
+                {
+                    priorityTableIndex =
+                                       IX_QMGR_MID_LOW_QUE_PRIORITY_TABLE_INDEX;
+                }
+            }
+            else
+            {
+                /* check if any bit related to queues in the first
+                 * half of the priority table is set
+                 */
+                if (intRegVal & uppPriorityTableFirstHalfMask)
+                {
+                    priorityTableIndex =
+                                       IX_QMGR_MIN_UPP_QUE_PRIORITY_TABLE_INDEX;
+                }
+                else
+                {
+                    priorityTableIndex =
+                                       IX_QMGR_MID_UPP_QUE_PRIORITY_TABLE_INDEX;
+                }
+            }
+
+            /* iterate following the priority table until all the bits
+             * of the interrupt register are cleared.
+             */
+            do
+            {
+                qIndex = priorityTable[priorityTableIndex++];
+                currDispatchQInfo = &dispatchQInfo[qIndex];
+                intRegCheckMask = currDispatchQInfo->intRegCheckMask;
+
+                /* If this queue caused this interrupt to be raised */
+                if (intRegVal & intRegCheckMask)
+                {
+                    /* 
+                     * check if Q type periodic - only lower queues can
+                     * have there type set to periodic. There can only be one
+                     * periodic queue, so the sporadics are only disabled once.
+                     */
+                    if (IX_QMGR_TYPE_REALTIME_PERIODIC == ixQMgrQTypes[qIndex])
+                    {
+                        /* 
+                         * Disable the notifications on any sporadics 
+                         */
+                        for (i=0; i <= IX_QMGR_MAX_LOW_QUE_TABLE_INDEX; i++)
+                        {
+                            if (IX_QMGR_TYPE_REALTIME_SPORADIC == 
+                                    ixQMgrQTypes[i])
+                            {
+                                ixQMgrNotificationDisable(i);
+                                /* 
+                                 * remove from intRegVal as we don't want 
+                                 * to service any sporadics now
+                                 */
+                                intRegVal &= ~dispatchQInfo[i].intRegCheckMask;
+#ifndef NDEBUG
+                                /* Update statistics */
+                                dispatcherStats.queueStats[i].disableCount++;
+#endif
+                            }
+                        }
+                    }
+
+                    currDispatchQInfo->callback (qIndex,
+                                                 currDispatchQInfo->callbackId);
+#ifndef NDEBUG
+                    /* Update statistics */
+                    dispatcherStats.queueStats[qIndex].callbackCnt++;
+#endif
+                    /* Clear the interrupt register bit */
+                    intRegVal &= ~intRegCheckMask;
+                }
+            }
+            while(intRegVal);
+        } /*End of intRegVal == currDispatchQInfo->intRegCheckMask */
+    } /* End of intRegVal != 0 */
+
+#ifndef NDEBUG
+    /* Update statistics */
+    dispatcherStats.loopRunCnt++;
+#endif
+
+    if ((intRegValCopy != 0) && (IX_QMGR_QUELOW_GROUP == group))
+    {
+        /* 
+         * lower groups (therefore sticky) AND at least one enabled interrupt
+         * Write back to clear the interrupt 
+         */
+        ixQMgrAqmIfQInterruptRegWrite (IX_QMGR_QUELOW_GROUP, intRegValCopy);
+    }
+
+    /* Rebuild the priority table if needed */
+    if (rebuildTable)
+    {
+        ixQMgrDispatcherReBuildPriorityTable ();
+    }
+}
+
+PRIVATE void
+ixQMgrDispatcherReBuildPriorityTable (void)
+{
+    UINT32 qIndex;
+    UINT32 priority;
+    int lowQuePriorityTableIndex = IX_QMGR_MIN_LOW_QUE_PRIORITY_TABLE_INDEX;
+    int uppQuePriorityTableIndex = IX_QMGR_MIN_UPP_QUE_PRIORITY_TABLE_INDEX;
+
+    /* Reset the rebuild flag */
+    rebuildTable = FALSE;
+
+    /* initialize the mak used to identify the queues in the first half
+     * of the priority table
+     */
+    lowPriorityTableFirstHalfMask = 0;
+    uppPriorityTableFirstHalfMask = 0;
+    
+    /* For each priority level */
+    for(priority=0; priority<IX_QMGR_NUM_PRIORITY_LEVELS; priority++)
+    {
+	/* Foreach low queue in this priority */
+	for(qIndex=0; qIndex<IX_QMGR_MIN_QUEUPP_QID; qIndex++)
+	{
+	    if (dispatchQInfo[qIndex].priority == priority)
+	    { 
+		/* build the priority table bitmask which match the
+		 * queues of the first half of the priority table 
+		 */
+		if (lowQuePriorityTableIndex < IX_QMGR_MID_LOW_QUE_PRIORITY_TABLE_INDEX) 
+		{
+		    lowPriorityTableFirstHalfMask |= dispatchQInfo[qIndex].intRegCheckMask;
+		}
+		/* build the priority table */
+		priorityTable[lowQuePriorityTableIndex++] = qIndex;
+	    }
+	}
+	/* Foreach upp queue */
+	for(qIndex=IX_QMGR_MIN_QUEUPP_QID; qIndex<=IX_QMGR_MAX_QID; qIndex++)
+	{
+	    if (dispatchQInfo[qIndex].priority == priority)
+	    {
+		/* build the priority table bitmask which match the
+		 * queues of the first half of the priority table 
+		 */
+		if (uppQuePriorityTableIndex < IX_QMGR_MID_UPP_QUE_PRIORITY_TABLE_INDEX) 
+		{
+		    uppPriorityTableFirstHalfMask |= dispatchQInfo[qIndex].intRegCheckMask;
+		}
+		/* build the priority table */
+		priorityTable[uppQuePriorityTableIndex++] = qIndex;
+	    }
+	}
+    }
+}
+
+IxQMgrDispatcherStats*
+ixQMgrDispatcherStatsGet (void)
+{
+    return &dispatcherStats;
+}
+
+PRIVATE void
+dummyCallback (IxQMgrQId qId,
+	       IxQMgrCallbackId cbId)
+{
+    /* Throttle the trace message */
+    if ((dispatchQInfo[qId].dummyCallbackCount % LOG_THROTTLE_COUNT) == 0)
+    {
+	IX_QMGR_LOG_WARNING2("--> dummyCallback: qId (%d), callbackId (%d)\n",qId,cbId);
+    }
+    dispatchQInfo[qId].dummyCallbackCount++;
+
+#ifndef NDEBUG
+    /* Update statistcs */
+    dispatcherStats.queueStats[qId].intNoCallbackCnt++;
+#endif
+}
+void
+ixQMgrLLPShow (int resetStats)
+{
+#ifndef NDEBUG
+    UINT8 i = 0;
+    IxQMgrQInfo *q;
+    UINT32 intEnableRegVal = 0;
+
+    printf ("Livelock statistics are printed on the fly.\n");
+    printf ("qId Type     EnableCnt DisableCnt IntEnableState Callbacks\n");
+    printf ("=== ======== ========= ========== ============== =========\n");
+
+    for (i=0; i<= IX_QMGR_MAX_LOW_QUE_TABLE_INDEX; i++)
+    {
+        q = &dispatchQInfo[i];
+
+        if (ixQMgrQTypes[i] != IX_QMGR_TYPE_REALTIME_OTHER)
+        {
+            printf (" %2d ", i);
+
+            if (ixQMgrQTypes[i] == IX_QMGR_TYPE_REALTIME_SPORADIC)
+            {
+                printf ("Sporadic");
+            }
+            else
+            {
+                printf ("Periodic");
+            }
+
+           
+            ixQMgrAqmIfQInterruptEnableRegRead (IX_QMGR_QUELOW_GROUP, 
+                                                    &intEnableRegVal);
+            	
+
+	    intEnableRegVal &= dispatchQInfo[i].intRegCheckMask;
+            intEnableRegVal = intEnableRegVal >> i;
+
+            printf (" %10d %10d %10d %10d\n",
+                    dispatcherStats.queueStats[i].enableCount,
+                    dispatcherStats.queueStats[i].disableCount,
+                    intEnableRegVal,
+                    dispatcherStats.queueStats[i].callbackCnt);
+
+            if (resetStats)
+            {
+                dispatcherStats.queueStats[i].enableCount =
+                dispatcherStats.queueStats[i].disableCount = 
+                dispatcherStats.queueStats[i].callbackCnt = 0;
+            }
+        }
+    }
+#else
+    IX_QMGR_LOG0("Livelock Prevention statistics are only collected in debug mode\n");
+#endif
+}
+
+void
+ixQMgrPeriodicDone (void)
+{
+    UINT32 i = 0;
+    UINT32 ixQMgrLockKey = 0;
+
+    /* 
+     * for the lower queues
+     */
+    for (i=0; i <= IX_QMGR_MAX_LOW_QUE_TABLE_INDEX; i++)
+    {
+        /*
+         * check for sporadics 
+         */
+        if (IX_QMGR_TYPE_REALTIME_SPORADIC == ixQMgrQTypes[i])
+        {
+             /* 
+              * enable any sporadics 
+              */
+             ixQMgrLockKey = ixOsalIrqLock();
+             ixQMgrAqmIfQInterruptEnable(i);
+             ixOsalIrqUnlock(ixQMgrLockKey);
+#ifndef NDEBUG
+             /* 
+              * Update statistics 
+              */
+             dispatcherStats.queueStats[i].enableCount++;
+             dispatcherStats.queueStats[i].notificationEnabled = TRUE;
+#endif
+        }
+    }
+}
+IX_STATUS
+ixQMgrCallbackTypeSet (IxQMgrQId qId, 
+                       IxQMgrType type)
+{
+    UINT32 ixQMgrLockKey = 0;
+    IxQMgrType ixQMgrOldType =0;
+
+#ifndef NDEBUG
+    if (!ixQMgrQIsConfigured(qId))
+    {
+        return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+    if (qId >= IX_QMGR_MIN_QUEUPP_QID)
+    {
+        return IX_QMGR_PARAMETER_ERROR;
+    }
+    if(!IX_QMGR_DISPATCHER_CALLBACK_TYPE_CHECK(type))
+    {
+        return IX_QMGR_PARAMETER_ERROR;
+    }
+#endif
+
+    ixQMgrOldType = ixQMgrQTypes[qId];
+    ixQMgrQTypes[qId] = type;
+
+    /*
+     * check if Q has been changed from type SPORADIC
+     */
+    if (IX_QMGR_TYPE_REALTIME_SPORADIC == ixQMgrOldType)
+    {
+       /* 
+        * previously Q was a SPORADIC, this means that LLP
+        * might have had it disabled. enable it now.
+        */
+       ixQMgrLockKey = ixOsalIrqLock();
+       ixQMgrAqmIfQInterruptEnable(qId);
+       ixOsalIrqUnlock(ixQMgrLockKey);
+
+#ifndef NDEBUG
+       /* 
+        * Update statistics 
+        */
+       dispatcherStats.queueStats[qId].enableCount++;
+#endif
+    }
+
+    return IX_SUCCESS;
+}
+
+IX_STATUS
+ixQMgrCallbackTypeGet (IxQMgrQId qId, 
+                       IxQMgrType *type)
+{
+#ifndef NDEBUG
+    if (!ixQMgrQIsConfigured(qId))
+    {
+        return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+    if (qId >= IX_QMGR_MIN_QUEUPP_QID)
+    {
+        return IX_QMGR_PARAMETER_ERROR;
+    }
+    if(type == NULL)
+    {
+         return IX_QMGR_PARAMETER_ERROR;
+    }
+#endif
+
+    *type = ixQMgrQTypes[qId];
+    return IX_SUCCESS;
+}

cpu/ixp/npe/IxQMgrInit.c

@@ -0,0 +1,233 @@
+/**
+ * @file    IxQMgrInit.c
+ *
+ * @author Intel Corporation
+ * @date    30-Oct-2001
+ *
+ * @brief:  Provided initialization of the QMgr component and its subcomponents.
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+/*
+ * System defined include files.
+ */
+
+/*
+ * User defined include files.
+ */
+#include "IxOsal.h"
+#include "IxQMgr.h"
+#include "IxQMgrQCfg_p.h"
+#include "IxQMgrDispatcher_p.h"
+#include "IxQMgrLog_p.h"
+#include "IxQMgrQAccess_p.h"
+#include "IxQMgrDefines_p.h"
+#include "IxQMgrAqmIf_p.h"
+
+/*
+ * Set to true if initialized
+ * N.B. global so integration/unit tests can reinitialize
+ */
+BOOL qMgrIsInitialized = FALSE;
+
+/*
+ * Function definitions.
+ */
+IX_STATUS
+ixQMgrInit (void)
+{
+    if (qMgrIsInitialized)
+    {
+	IX_QMGR_LOG0("ixQMgrInit: IxQMgr already initialised\n");
+	return IX_FAIL;
+    }
+
+    /* Initialise the QCfg component */
+    ixQMgrQCfgInit ();
+
+    /* Initialise the Dispatcher component */
+    ixQMgrDispatcherInit ();
+
+    /* Initialise the Access component */
+    ixQMgrQAccessInit ();
+
+    /* Initialization complete */
+    qMgrIsInitialized = TRUE;
+
+    return IX_SUCCESS;
+}
+
+IX_STATUS
+ixQMgrUnload (void)
+{
+    if (!qMgrIsInitialized)
+    {
+	return IX_FAIL;
+    }
+
+    /* Uninitialise the QCfg component */
+    ixQMgrQCfgUninit ();
+
+    /* Uninitialization complete */
+    qMgrIsInitialized = FALSE;
+
+    return IX_SUCCESS;
+}
+
+void
+ixQMgrShow (void)
+{
+    IxQMgrQCfgStats *qCfgStats = NULL;
+    IxQMgrDispatcherStats *dispatcherStats = NULL;
+    int i;
+    UINT32 lowIntRegRead, upIntRegRead;
+
+    qCfgStats = ixQMgrQCfgStatsGet ();
+    dispatcherStats = ixQMgrDispatcherStatsGet ();
+    ixQMgrAqmIfQInterruptRegRead (IX_QMGR_QUELOW_GROUP, &lowIntRegRead);
+    ixQMgrAqmIfQInterruptRegRead (IX_QMGR_QUEUPP_GROUP, &upIntRegRead);
+    printf("Generic Stats........\n");
+    printf("=====================\n");
+    printf("Loop Run Count..........%u\n",dispatcherStats->loopRunCnt);
+    printf("Watermark set count.....%d\n", qCfgStats->wmSetCnt);
+    printf("===========================================\n");
+    printf("On the fly Interrupt Register Stats........\n");
+    printf("===========================================\n");
+    printf("Lower Interrupt Register............0x%08x\n",lowIntRegRead);
+    printf("Upper Interrupt Register............0x%08x\n",upIntRegRead);
+    printf("==============================================\n");
+    printf("Queue Specific Stats........\n");
+    printf("============================\n");
+
+    for (i=0; i<IX_QMGR_MAX_NUM_QUEUES; i++)
+    {
+	if (ixQMgrQIsConfigured(i))
+	{
+	    ixQMgrQShow(i);
+	}
+    }
+
+    printf("============================\n");
+}
+
+IX_STATUS
+ixQMgrQShow (IxQMgrQId qId)
+{
+    IxQMgrQCfgStats *qCfgStats = NULL;
+    IxQMgrDispatcherStats *dispatcherStats = NULL; 
+
+    if (!ixQMgrQIsConfigured(qId))
+    {
+	return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+    
+    dispatcherStats = ixQMgrDispatcherStatsGet ();
+    qCfgStats = ixQMgrQCfgQStatsGet (qId);
+
+    printf("QId %d\n", qId);
+    printf("======\n");
+    printf("  IxQMgrQCfg Stats\n");
+    printf("    Name..................... \"%s\"\n", qCfgStats->qStats[qId].qName);
+    printf("    Size in words............ %u\n", qCfgStats->qStats[qId].qSizeInWords);
+    printf("    Entry size in words...... %u\n", qCfgStats->qStats[qId].qEntrySizeInWords);
+    printf("    Nearly empty watermark... %u\n", qCfgStats->qStats[qId].ne);
+    printf("    Nearly full watermark.... %u\n", qCfgStats->qStats[qId].nf);
+    printf("    Number of full entries... %u\n", qCfgStats->qStats[qId].numEntries);
+    printf("    Sram base address........ 0x%X\n", qCfgStats->qStats[qId].baseAddress);
+    printf("    Read pointer............. 0x%X\n", qCfgStats->qStats[qId].readPtr);
+    printf("    Write pointer............ 0x%X\n", qCfgStats->qStats[qId].writePtr);
+
+#ifndef NDEBUG
+    if (dispatcherStats->queueStats[qId].notificationEnabled)
+    {
+        char *localEvent = "none ????";
+        switch (dispatcherStats->queueStats[qId].srcSel)
+        {
+            case IX_QMGR_Q_SOURCE_ID_E:
+                localEvent = "Empty";
+                break;
+            case IX_QMGR_Q_SOURCE_ID_NE:
+                localEvent = "Nearly Empty";
+                break;
+            case IX_QMGR_Q_SOURCE_ID_NF:
+                localEvent = "Nearly Full";
+                break;
+            case IX_QMGR_Q_SOURCE_ID_F:
+                localEvent = "Full";
+                break;
+            case IX_QMGR_Q_SOURCE_ID_NOT_E:
+                localEvent = "Not Empty";
+                break;
+            case IX_QMGR_Q_SOURCE_ID_NOT_NE:
+                localEvent = "Not Nearly Empty";
+                break;
+            case IX_QMGR_Q_SOURCE_ID_NOT_NF:
+                localEvent = "Not Nearly Full";
+                break;
+            case IX_QMGR_Q_SOURCE_ID_NOT_F:
+                localEvent = "Not Full";
+                break;
+            default :
+                break;
+        }
+        printf("    Notifications localEvent...... %s\n", localEvent);
+    }
+    else
+    {
+        printf("    Notifications............ not enabled\n");
+    }
+    printf("  IxQMgrDispatcher Stats\n");
+    printf("    Callback count................%d\n",
+	  dispatcherStats->queueStats[qId].callbackCnt);
+    printf("    Priority change count.........%d\n",
+	  dispatcherStats->queueStats[qId].priorityChangeCnt);
+    printf("    Interrupt no callback count...%d\n",
+	  dispatcherStats->queueStats[qId].intNoCallbackCnt);
+    printf("    Interrupt lost callback count...%d\n",
+	  dispatcherStats->queueStats[qId].intLostCallbackCnt);
+#endif
+
+    return IX_SUCCESS;
+}
+
+
+
+

cpu/ixp/npe/IxQMgrQAccess.c

@@ -0,0 +1,796 @@
+/**
+ * @file    IxQMgrQAccess.c
+ *
+ * @author Intel Corporation
+ * @date    30-Oct-2001
+ *
+ * @brief   This file contains functions for putting entries on a queue and
+ * removing entries from a queue.
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+/*
+ * Inlines are compiled as function when this is defined.
+ * N.B. Must be placed before #include of "IxQMgr.h"
+ */
+#ifndef IXQMGR_H
+#    define IXQMGRQACCESS_C
+#else
+#    error
+#endif
+
+/*
+ * System defined include files.
+ */
+
+/*
+ * User defined include files.
+ */
+#include "IxQMgr.h"
+#include "IxQMgrAqmIf_p.h"
+#include "IxQMgrQAccess_p.h"
+#include "IxQMgrQCfg_p.h"
+#include "IxQMgrDefines_p.h"
+
+/*
+ * Global variables and extern definitions
+ */
+extern IxQMgrQInlinedReadWriteInfo ixQMgrQInlinedReadWriteInfo[];
+
+/*
+ * Function definitions.
+ */
+void
+ixQMgrQAccessInit (void)
+{   
+}
+
+IX_STATUS
+ixQMgrQReadWithChecks (IxQMgrQId qId,
+                       UINT32 *entry)
+{
+    IxQMgrQEntrySizeInWords entrySizeInWords;
+    IxQMgrQInlinedReadWriteInfo *infoPtr;
+
+    if (NULL == entry)
+    {
+	return IX_QMGR_PARAMETER_ERROR;
+    }
+
+    /* Check QId */
+    if (!ixQMgrQIsConfigured(qId))
+    {
+	return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+
+    /* Get the q entry size in words */
+    entrySizeInWords = ixQMgrQEntrySizeInWordsGet (qId);
+
+    ixQMgrAqmIfQPop (qId, entrySizeInWords, entry);	    
+
+    /* reset the current read count if the counter wrapped around 
+    * (unsigned arithmetic)
+    */
+    infoPtr = &ixQMgrQInlinedReadWriteInfo[qId];
+    if (infoPtr->qReadCount-- > infoPtr->qSizeInEntries)
+    {
+	infoPtr->qReadCount = 0;
+    }
+
+    /* Check if underflow occurred on the read */
+    if (ixQMgrAqmIfUnderflowCheck (qId))
+    {
+	return IX_QMGR_Q_UNDERFLOW;
+    }
+    
+    return IX_SUCCESS;
+}
+
+/* this function reads the remaining of the q entry
+ * for queues configured with many words.
+ * (the first word of the entry is already read 
+ * in the inlined function and the entry pointer already
+ * incremented
+ */
+IX_STATUS
+ixQMgrQReadMWordsMinus1 (IxQMgrQId qId,
+			 UINT32 *entry)
+{
+    IxQMgrQInlinedReadWriteInfo *infoPtr = &ixQMgrQInlinedReadWriteInfo[qId];
+    UINT32 entrySize = infoPtr->qEntrySizeInWords;
+    volatile UINT32 *qAccRegAddr = infoPtr->qAccRegAddr;
+    
+    while (--entrySize)
+    {
+	/* read the entry and accumulate the result */
+	*(++entry) = IX_OSAL_READ_LONG(++qAccRegAddr);
+    }
+    /* underflow is available for lower queues only */
+    if (qId < IX_QMGR_MIN_QUEUPP_QID)
+    {
+	/* get the queue status */
+	UINT32 status = IX_OSAL_READ_LONG(infoPtr->qUOStatRegAddr);
+	
+	/* check the underflow status */
+	if (status & infoPtr->qUflowStatBitMask)
+	{
+	    /* the queue is empty 
+	     *  clear the underflow status bit if it was set 
+	     */
+	    IX_OSAL_WRITE_LONG(infoPtr->qUOStatRegAddr,
+				 status & ~infoPtr->qUflowStatBitMask);
+	    return IX_QMGR_Q_UNDERFLOW;
+	}
+    }
+    return IX_SUCCESS;
+}
+
+IX_STATUS
+ixQMgrQWriteWithChecks (IxQMgrQId qId,
+                        UINT32 *entry)
+{
+    IxQMgrQEntrySizeInWords entrySizeInWords;
+    IxQMgrQInlinedReadWriteInfo *infoPtr;
+
+    if (NULL == entry)
+    {
+	return IX_QMGR_PARAMETER_ERROR;
+    }
+
+    /* Check QId */
+    if (!ixQMgrQIsConfigured(qId))
+    {
+	return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+
+    /* Get the q entry size in words */
+    entrySizeInWords = ixQMgrQEntrySizeInWordsGet (qId);
+    
+    ixQMgrAqmIfQPush (qId, entrySizeInWords, entry);
+
+    /* reset the current read count if the counter wrapped around 
+    * (unsigned arithmetic)
+    */
+    infoPtr = &ixQMgrQInlinedReadWriteInfo[qId];
+    if (infoPtr->qWriteCount++ >= infoPtr->qSizeInEntries)
+    {
+	infoPtr->qWriteCount = infoPtr->qSizeInEntries;
+    }
+
+    /* Check if overflow occurred on the write*/
+    if (ixQMgrAqmIfOverflowCheck (qId))
+    {
+	return IX_QMGR_Q_OVERFLOW;
+    }
+         
+    return IX_SUCCESS;
+}
+
+IX_STATUS
+ixQMgrQPeek (IxQMgrQId qId,
+	     unsigned int entryIndex,
+	     UINT32 *entry)
+{
+    unsigned int numEntries;
+
+#ifndef NDEBUG
+    if ((NULL == entry) || (entryIndex >= IX_QMGR_Q_SIZE_INVALID))
+    {
+	return IX_QMGR_PARAMETER_ERROR;
+    }
+
+    if (!ixQMgrQIsConfigured(qId))
+    {
+	return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+#endif
+    
+    if (IX_SUCCESS != ixQMgrQNumEntriesGet (qId, &numEntries))
+    {
+	return IX_FAIL;
+    }
+
+    if (entryIndex >= numEntries) /* entryIndex starts at 0 */
+    {
+	return IX_QMGR_ENTRY_INDEX_OUT_OF_BOUNDS;
+    }
+
+    return ixQMgrAqmIfQPeek (qId, entryIndex, entry);
+}
+
+IX_STATUS
+ixQMgrQPoke (IxQMgrQId qId,
+	     unsigned entryIndex,
+	     UINT32 *entry)
+{
+    unsigned int numEntries;
+
+#ifndef NDEBUG
+    if ((NULL == entry) || (entryIndex > 128))
+    {
+	return IX_QMGR_PARAMETER_ERROR;
+    }
+
+    if (!ixQMgrQIsConfigured(qId))
+    {
+	return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+#endif
+        
+    if (IX_SUCCESS != ixQMgrQNumEntriesGet (qId, &numEntries))
+    {
+	return IX_FAIL;
+    }
+
+    if (numEntries < (entryIndex + 1)) /* entryIndex starts at 0 */
+    {
+	return IX_QMGR_ENTRY_INDEX_OUT_OF_BOUNDS;
+    }
+
+    return ixQMgrAqmIfQPoke (qId, entryIndex, entry);
+}
+
+IX_STATUS
+ixQMgrQStatusGetWithChecks (IxQMgrQId qId,
+                            IxQMgrQStatus *qStatus)
+{
+    if (NULL == qStatus)
+    {
+	return IX_QMGR_PARAMETER_ERROR;
+    }
+   
+    if (!ixQMgrQIsConfigured (qId)) 
+    {
+        return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+
+    ixQMgrAqmIfQueStatRead (qId, qStatus);
+
+    return IX_SUCCESS;
+}
+
+IX_STATUS
+ixQMgrQNumEntriesGet (IxQMgrQId qId,
+		      unsigned *numEntriesPtr)
+{
+    UINT32 qPtrs;
+    UINT32 qStatus;
+    unsigned numEntries;
+    IxQMgrQInlinedReadWriteInfo *infoPtr;
+
+
+#ifndef NDEBUG
+    if (NULL == numEntriesPtr)
+    {
+	return IX_QMGR_PARAMETER_ERROR;
+    }
+
+    /* Check QId */
+    if (!ixQMgrQIsConfigured(qId))
+    {
+	return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+#endif
+
+    /* get fast access data */
+    infoPtr = &ixQMgrQInlinedReadWriteInfo[qId];
+
+    /* get snapshot */
+    qPtrs = IX_OSAL_READ_LONG(infoPtr->qConfigRegAddr);
+
+    /* Mod subtraction of pointers to get number of words in Q. */
+    numEntries = (qPtrs - (qPtrs >> 7)) & 0x7f;
+  
+    if (numEntries == 0)
+    {
+	/* 
+	 * Could mean either full or empty queue
+	 * so look at status
+	 */
+	ixQMgrAqmIfQueStatRead (qId, &qStatus);
+
+	if (qId < IX_QMGR_MIN_QUEUPP_QID)
+	{
+	    if (qStatus & IX_QMGR_Q_STATUS_E_BIT_MASK)
+	    {
+		/* Empty */
+		*numEntriesPtr = 0;
+	    }
+	    else if (qStatus & IX_QMGR_Q_STATUS_F_BIT_MASK)
+	    {
+		/* Full */
+		*numEntriesPtr = infoPtr->qSizeInEntries;
+	    }
+	    else
+	    {	    
+		/* 
+		 * Queue status and read/write pointers are volatile.
+		 * The queue state has changed since we took the
+		 * snapshot of the read and write pointers.
+		 * Client can retry if they wish
+		 */
+		*numEntriesPtr = 0;
+		return IX_QMGR_WARNING;
+	    }
+	}
+	else /* It is an upper queue which does not have an empty status bit maintained */
+	{
+	    if (qStatus & IX_QMGR_Q_STATUS_F_BIT_MASK)
+	    {
+		/* The queue is Full at the time of snapshot. */
+		*numEntriesPtr = infoPtr->qSizeInEntries;
+	    }
+	    else
+	    {
+  	       /* The queue is either empty, either moving,
+	        * Client can retry if they wish
+	        */
+		*numEntriesPtr = 0;
+	        return IX_QMGR_WARNING;
+	    }
+	}
+    }
+    else
+    {
+	*numEntriesPtr = (numEntries / infoPtr->qEntrySizeInWords) & (infoPtr->qSizeInEntries - 1);
+    }
+    
+    return IX_SUCCESS;
+}
+
+#if defined(__wince) && defined(NO_INLINE_APIS)
+
+PUBLIC IX_STATUS
+ixQMgrQRead (IxQMgrQId qId,
+      UINT32 *entryPtr)
+{
+    extern IxQMgrQInlinedReadWriteInfo ixQMgrQInlinedReadWriteInfo[];
+    IxQMgrQInlinedReadWriteInfo *infoPtr = &ixQMgrQInlinedReadWriteInfo[qId];
+    UINT32 entry, entrySize;
+
+    /* get a new entry */
+    entrySize = infoPtr->qEntrySizeInWords;
+    entry = IX_OSAL_READ_LONG(infoPtr->qAccRegAddr);
+
+    if (entrySize != IX_QMGR_Q_ENTRY_SIZE1)
+    { 
+    *entryPtr = entry;
+  /* process the remaining part of the entry */
+   return ixQMgrQReadMWordsMinus1(qId, entryPtr);
+    }
+
+    /* underflow is available for lower queues only */
+    if (qId < IX_QMGR_MIN_QUEUPP_QID)
+    {
+ /* the counter of queue entries is decremented. In happy 
+    * day scenario there are many entries in the queue
+  * and the counter does not reach zero.
+  */
+     if (infoPtr->qReadCount-- == 0)
+ {
+       /* There is maybe no entry in the queue
+      * qReadCount is now negative, but will be corrected before
+      * the function returns.
+         */
+     UINT32 qPtrs; /* queue internal pointers */
+
+     /* when a queue is empty, the hw guarantees to return 
+       * a null value. If the value is not null, the queue is
+      * not empty.
+        */
+     if (entry == 0)
+     {
+       /* get the queue status */
+      UINT32 status = IX_OSAL_READ_LONG(infoPtr->qUOStatRegAddr);
+   
+        /* check the underflow status */
+        if (status & infoPtr->qUflowStatBitMask)
+        {
+           /* the queue is empty 
+          *  clear the underflow status bit if it was set 
+            */
+          IX_OSAL_WRITE_LONG(infoPtr->qUOStatRegAddr,
+                    status & ~infoPtr->qUflowStatBitMask);
+         *entryPtr = 0;
+          infoPtr->qReadCount = 0;
+            return IX_QMGR_Q_UNDERFLOW;
+     }
+       }
+       /* store the result */
+      *entryPtr = entry;
+
+      /* No underflow occured : someone is filling the queue
+       * or the queue contains null entries.
+       * The current counter needs to be
+       * updated from the current number of entries in the queue
+       */
+
+     /* get snapshot of queue pointers */
+        qPtrs = IX_OSAL_READ_LONG(infoPtr->qConfigRegAddr);
+
+       /* Mod subtraction of pointers to get number of words in Q. */
+      qPtrs = (qPtrs - (qPtrs >> 7)) & 0x7f; 
+  
+       if (qPtrs == 0)
+     {
+       /* no entry in the queue */
+     infoPtr->qReadCount = 0;
+        }
+       else
+        {
+       /* convert the number of words inside the queue
+      * to a number of entries 
+       */
+     infoPtr->qReadCount = qPtrs & (infoPtr->qSizeInEntries - 1);
+        }
+       return IX_SUCCESS;
+  }
+    }
+    *entryPtr = entry;
+    return IX_SUCCESS;
+}
+
+PUBLIC IX_STATUS
+ixQMgrQBurstRead (IxQMgrQId qId,
+          UINT32 numEntries,
+          UINT32 *entries)
+{
+    extern IxQMgrQInlinedReadWriteInfo ixQMgrQInlinedReadWriteInfo[];
+    IxQMgrQInlinedReadWriteInfo *infoPtr = &ixQMgrQInlinedReadWriteInfo[qId];
+    UINT32 nullCheckEntry;
+
+    if (infoPtr->qEntrySizeInWords == IX_QMGR_Q_ENTRY_SIZE1)
+    {
+    volatile UINT32 *qAccRegAddr = infoPtr->qAccRegAddr;
+
+    /* the code is optimized to take care of data dependencies:
+  * Durig a read, there are a few cycles needed to get the 
+   * read complete. During these cycles, it is poossible to
+    * do some CPU, e.g. increment pointers and decrement 
+   * counters.
+     */
+
+ /* fetch a queue entry */
+   nullCheckEntry = IX_OSAL_READ_LONG(infoPtr->qAccRegAddr);
+
+ /* iterate the specified number of queue entries */ 
+    while (--numEntries)
+    {
+       /* check the result of the previous read */
+     if (nullCheckEntry == 0)
+        {
+       /* if we read a NULL entry, stop. We have underflowed */
+        break;
+      }
+       else
+        {
+       /* write the entry */
+       *entries = nullCheckEntry;
+      /* fetch next entry */
+      nullCheckEntry = IX_OSAL_READ_LONG(qAccRegAddr);
+      /* increment the write address */
+       entries++;
+      }
+   }
+   /* write the pre-fetched entry */
+   *entries = nullCheckEntry;
+    }
+    else
+    {
+    IxQMgrQEntrySizeInWords entrySizeInWords = infoPtr->qEntrySizeInWords;
+  /* read the specified number of queue entries */
+    nullCheckEntry = 0;
+ while (numEntries--)
+    {
+       int i;
+
+      for (i = 0; i < entrySizeInWords; i++)
+      {
+       *entries = IX_OSAL_READ_LONG(infoPtr->qAccRegAddr + i);
+       nullCheckEntry |= *entries++;
+       }
+
+       /* if we read a NULL entry, stop. We have underflowed */
+        if (nullCheckEntry == 0)
+        {
+       break;
+      }
+       nullCheckEntry = 0;
+ }
+    }
+
+    /* reset the current read count : next access to the read function 
+     * will force a underflow status check 
+     */
+    infoPtr->qWriteCount = 0;
+
+    /* Check if underflow occurred on the read */
+    if (nullCheckEntry == 0 && qId < IX_QMGR_MIN_QUEUPP_QID)
+    {
+  /* get the queue status */
+  UINT32 status = IX_OSAL_READ_LONG(infoPtr->qUOStatRegAddr);
+
+   if (status & infoPtr->qUflowStatBitMask)
+    {
+       /* clear the underflow status bit if it was set */
+      IX_OSAL_WRITE_LONG(infoPtr->qUOStatRegAddr,
+                status & ~infoPtr->qUflowStatBitMask);
+     return IX_QMGR_Q_UNDERFLOW;
+ }
+    }
+
+    return IX_SUCCESS;
+}
+
+PUBLIC IX_STATUS
+ixQMgrQWrite (IxQMgrQId qId,
+         UINT32 *entry)
+{
+    extern IxQMgrQInlinedReadWriteInfo ixQMgrQInlinedReadWriteInfo[];
+    IxQMgrQInlinedReadWriteInfo *infoPtr = &ixQMgrQInlinedReadWriteInfo[qId];
+    UINT32 entrySize;
+
+    /* write the entry */
+    IX_OSAL_WRITE_LONG(infoPtr->qAccRegAddr, *entry);
+    entrySize = infoPtr->qEntrySizeInWords;
+
+    if (entrySize != IX_QMGR_Q_ENTRY_SIZE1)
+    {   
+    /* process the remaining part of the entry */
+   volatile UINT32 *qAccRegAddr = infoPtr->qAccRegAddr;
+    while (--entrySize)
+ {
+       ++entry;
+        IX_OSAL_WRITE_LONG(++qAccRegAddr, *entry);
+    }
+   entrySize = infoPtr->qEntrySizeInWords;
+    }
+
+    /* overflow is available for lower queues only */
+    if (qId < IX_QMGR_MIN_QUEUPP_QID)
+    {   
+  UINT32 qSize = infoPtr->qSizeInEntries;
+ /* increment the current number of entries in the queue
+  * and check for overflow 
+   */
+ if (infoPtr->qWriteCount++ == qSize)
+    {
+       /* the queue may have overflow */
+       UINT32 qPtrs; /* queue internal pointers */
+  
+       /* get the queue status */
+      UINT32 status = IX_OSAL_READ_LONG(infoPtr->qUOStatRegAddr);
+
+       /* read the status twice because the status may 
+         * not be immediately ready after the write operation
+        */
+     if ((status & infoPtr->qOflowStatBitMask) ||
+        ((status = IX_OSAL_READ_LONG(infoPtr->qUOStatRegAddr))
+         & infoPtr->qOflowStatBitMask))
+     {
+       /* the queue is full, clear the overflow status
+      *  bit if it was set 
+       */
+     IX_OSAL_WRITE_LONG(infoPtr->qUOStatRegAddr,
+                    status & ~infoPtr->qOflowStatBitMask);
+     infoPtr->qWriteCount = infoPtr->qSizeInEntries;
+     return IX_QMGR_Q_OVERFLOW;
+      }
+       /* No overflow occured : someone is draining the queue
+       * and the current counter needs to be
+       * updated from the current number of entries in the queue
+       */
+
+     /* get q pointer snapshot */
+        qPtrs = IX_OSAL_READ_LONG(infoPtr->qConfigRegAddr);
+
+       /* Mod subtraction of pointers to get number of words in Q. */
+      qPtrs = (qPtrs - (qPtrs >> 7)) & 0x7f; 
+
+     if (qPtrs == 0)
+     {
+       /* the queue may be full at the time of the 
+         * snapshot. Next access will check 
+         * the overflow status again.
+        */
+     infoPtr->qWriteCount = qSize;
+       }
+       else 
+       {
+       /* convert the number of words to a number of entries */
+        if (entrySize == IX_QMGR_Q_ENTRY_SIZE1)
+     {
+           infoPtr->qWriteCount = qPtrs & (qSize - 1);
+     }
+       else
+        {
+           infoPtr->qWriteCount = (qPtrs / entrySize) & (qSize - 1);
+       }
+       }
+   }
+    }
+    return IX_SUCCESS;
+}
+
+PUBLIC IX_STATUS
+ixQMgrQBurstWrite (IxQMgrQId qId,
+          unsigned numEntries,
+        UINT32 *entries)
+{
+    extern IxQMgrQInlinedReadWriteInfo ixQMgrQInlinedReadWriteInfo[];
+    IxQMgrQInlinedReadWriteInfo *infoPtr = &ixQMgrQInlinedReadWriteInfo[qId];
+    UINT32 status;
+
+    /* update the current write count */
+    infoPtr->qWriteCount += numEntries;
+
+    if (infoPtr->qEntrySizeInWords == IX_QMGR_Q_ENTRY_SIZE1)
+    {
+    volatile UINT32 *qAccRegAddr = infoPtr->qAccRegAddr;
+    while (numEntries--)
+    {
+       IX_OSAL_WRITE_LONG(qAccRegAddr, *entries);
+        entries++;
+  }
+    }
+    else
+    {
+ IxQMgrQEntrySizeInWords entrySizeInWords = infoPtr->qEntrySizeInWords;
+  int i;
+
+  /* write each queue entry */
+    while (numEntries--)
+    {
+       /* write the queueEntrySize number of words for each entry */
+       for (i = 0; i < entrySizeInWords; i++)
+      {
+       IX_OSAL_WRITE_LONG((infoPtr->qAccRegAddr + i), *entries);
+     entries++;
+      }
+   }
+    }
+
+    /* check if the write count overflows */
+    if (infoPtr->qWriteCount > infoPtr->qSizeInEntries)
+    {
+  /* reset the current write count */
+ infoPtr->qWriteCount = infoPtr->qSizeInEntries;
+    }
+
+    /* Check if overflow occurred on the write operation */
+    if (qId < IX_QMGR_MIN_QUEUPP_QID)
+    {
+   /* get the queue status */
+  status = IX_OSAL_READ_LONG(infoPtr->qUOStatRegAddr);
+
+  /* read the status twice because the status may 
+     * not be ready at the time of the write
+     */
+ if ((status & infoPtr->qOflowStatBitMask) ||
+        ((status = IX_OSAL_READ_LONG(infoPtr->qUOStatRegAddr))
+         & infoPtr->qOflowStatBitMask))
+ {
+       /* clear the underflow status bit if it was set */
+      IX_OSAL_WRITE_LONG(infoPtr->qUOStatRegAddr,
+                status & ~infoPtr->qOflowStatBitMask);
+     return IX_QMGR_Q_OVERFLOW;
+  }
+    }
+
+    return IX_SUCCESS;
+}
+
+PUBLIC IX_STATUS
+ixQMgrQStatusGet (IxQMgrQId qId,
+          IxQMgrQStatus *qStatus)
+{
+    /* read the status of a queue in the range 0-31 */
+    if (qId < IX_QMGR_MIN_QUEUPP_QID)
+    {
+  extern UINT32 ixQMgrAqmIfQueLowStatRegAddr[];
+   extern UINT32 ixQMgrAqmIfQueLowStatBitsOffset[];
+    extern UINT32 ixQMgrAqmIfQueLowStatBitsMask;
+    extern IxQMgrQInlinedReadWriteInfo ixQMgrQInlinedReadWriteInfo[];
+    IxQMgrQInlinedReadWriteInfo *infoPtr = &ixQMgrQInlinedReadWriteInfo[qId];
+   volatile UINT32 *lowStatRegAddr = (UINT32*)ixQMgrAqmIfQueLowStatRegAddr[qId];
+   volatile UINT32 *qUOStatRegAddr = infoPtr->qUOStatRegAddr;
+
+  UINT32 lowStatBitsOffset = ixQMgrAqmIfQueLowStatBitsOffset[qId];
+    UINT32 lowStatBitsMask   = ixQMgrAqmIfQueLowStatBitsMask;
+   UINT32 underflowBitMask  = infoPtr->qUflowStatBitMask;
+  UINT32 overflowBitMask   = infoPtr->qOflowStatBitMask;
+
+  /* read the status register for this queue */
+   *qStatus = IX_OSAL_READ_LONG(lowStatRegAddr);
+ /* mask out the status bits relevant only to this queue */
+  *qStatus = (*qStatus >> lowStatBitsOffset) & lowStatBitsMask;
+
+   /* Check if the queue has overflowed */
+ if (IX_OSAL_READ_LONG(qUOStatRegAddr) & overflowBitMask)
+  {
+       /* clear the overflow status bit if it was set */
+       IX_OSAL_WRITE_LONG(qUOStatRegAddr,
+                 (IX_OSAL_READ_LONG(qUOStatRegAddr) &
+               ~overflowBitMask));
+       *qStatus |= IX_QMGR_Q_STATUS_OF_BIT_MASK;
+   }
+
+   /* Check if the queue has underflowed */
+        if (IX_OSAL_READ_LONG(qUOStatRegAddr) & underflowBitMask)
+ {
+       /* clear the underflow status bit if it was set */
+      IX_OSAL_WRITE_LONG(qUOStatRegAddr,
+                 (IX_OSAL_READ_LONG(qUOStatRegAddr) &
+               ~underflowBitMask));
+      *qStatus |= IX_QMGR_Q_STATUS_UF_BIT_MASK;
+   }
+    }
+    else /* read status of a queue in the range 32-63 */
+    {
+ extern UINT32 ixQMgrAqmIfQueUppStat0RegAddr;
+    extern UINT32 ixQMgrAqmIfQueUppStat1RegAddr;
+    extern UINT32 ixQMgrAqmIfQueUppStat0BitMask[];
+  extern UINT32 ixQMgrAqmIfQueUppStat1BitMask[];
+
+  volatile UINT32 *qNearEmptyStatRegAddr = (UINT32*)ixQMgrAqmIfQueUppStat0RegAddr;
+    volatile UINT32 *qFullStatRegAddr      = (UINT32*)ixQMgrAqmIfQueUppStat1RegAddr;
+    int maskIndex = qId - IX_QMGR_MIN_QUEUPP_QID;
+   UINT32 qNearEmptyStatBitMask = ixQMgrAqmIfQueUppStat0BitMask[maskIndex];
+    UINT32 qFullStatBitMask      = ixQMgrAqmIfQueUppStat1BitMask[maskIndex];
+
+    /* Reset the status bits */
+ *qStatus = 0;
+
+   /* Check if the queue is nearly empty */
+    if (IX_OSAL_READ_LONG(qNearEmptyStatRegAddr) & qNearEmptyStatBitMask)
+ {
+       *qStatus |= IX_QMGR_Q_STATUS_NE_BIT_MASK;
+   }
+
+   /* Check if the queue is full */
+    if (IX_OSAL_READ_LONG(qFullStatRegAddr) & qFullStatBitMask)
+   {
+       *qStatus |= IX_QMGR_Q_STATUS_F_BIT_MASK;
+    }
+    }
+    return IX_SUCCESS;
+}
+#endif /* def NO_INLINE_APIS */

cpu/ixp/npe/IxQMgrQCfg.c

@@ -0,0 +1,543 @@
+/**
+ * @file    QMgrQCfg.c
+ *
+ * @author Intel Corporation
+ * @date    30-Oct-2001
+ * 
+ * @brief   This modules provides an interface for setting up the static
+ * configuration of AQM queues.This file contains the following
+ * functions:
+ *
+ * 
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+/*
+ * System defined include files.
+ */
+
+/*
+ * User defined include files.
+ */
+#include "IxOsal.h"
+#include "IxQMgr.h"
+#include "IxQMgrAqmIf_p.h"
+#include "IxQMgrQCfg_p.h"
+#include "IxQMgrDefines_p.h"
+
+/*
+ * #defines and macros used in this file.
+ */
+
+#define IX_QMGR_MIN_ENTRY_SIZE_IN_WORDS 16
+
+/* Total size of SRAM */
+#define IX_QMGR_AQM_SRAM_SIZE_IN_BYTES 0x4000
+
+/*
+ * Check that qId is a valid queue identifier. This is provided to
+ * make the code easier to read.
+ */
+#define IX_QMGR_QID_IS_VALID(qId) \
+(((qId) >= (IX_QMGR_MIN_QID)) && ((qId) <= (IX_QMGR_MAX_QID)))
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+
+/*
+ * This struct describes an AQM queue.
+ * N.b. bufferSizeInWords and qEntrySizeInWords are stored in the queue
+ * as these are requested by Access in the data path. sizeInEntries is
+ * not required by the data path so it can be calculated dynamically.
+ * 
+ */
+typedef struct
+{
+    char qName[IX_QMGR_MAX_QNAME_LEN+1];       /* Textual description of a queue*/
+    IxQMgrQSizeInWords qSizeInWords;           /* The number of words in the queue */
+    IxQMgrQEntrySizeInWords qEntrySizeInWords; /* Number of words per queue entry*/
+    BOOL isConfigured;                         /* This flag is TRUE if the queue has
+                                                *   been configured
+                                                */
+} IxQMgrCfgQ;
+
+/*
+ * Variable declarations global to this file. Externs are followed by
+ * statics.
+ */
+
+extern UINT32 * ixQMgrAqmIfQueAccRegAddr[]; 
+
+/* Store data required to inline read and write access
+ */
+IxQMgrQInlinedReadWriteInfo ixQMgrQInlinedReadWriteInfo[IX_QMGR_MAX_NUM_QUEUES];
+
+static IxQMgrCfgQ cfgQueueInfo[IX_QMGR_MAX_NUM_QUEUES];
+
+/* This pointer holds the starting address of AQM SRAM not used by
+ * the AQM queues.
+ */
+static UINT32 freeSramAddress=0;
+
+/* 4 words of zeroed memory for inline access */
+static UINT32 zeroedPlaceHolder[4] = { 0, 0, 0, 0 };
+
+static BOOL cfgInitialized = FALSE;
+
+static IxOsalMutex ixQMgrQCfgMutex;
+
+/*
+ * Statistics
+ */
+static IxQMgrQCfgStats stats;
+
+/*
+ * Function declarations
+ */
+PRIVATE BOOL
+watermarkLevelIsOk (IxQMgrQId qId, IxQMgrWMLevel level);
+
+PRIVATE BOOL
+qSizeInWordsIsOk (IxQMgrQSizeInWords qSize);
+
+PRIVATE BOOL
+qEntrySizeInWordsIsOk (IxQMgrQEntrySizeInWords entrySize);
+
+/*
+ * Function definitions.
+ */
+void
+ixQMgrQCfgInit (void)
+{
+    int loopIndex;
+    
+    for (loopIndex=0; loopIndex < IX_QMGR_MAX_NUM_QUEUES;loopIndex++)
+    {
+	/* info for code inlining */
+	ixQMgrAqmIfQueAccRegAddr[loopIndex] = zeroedPlaceHolder;
+
+	/* info for code inlining */
+	ixQMgrQInlinedReadWriteInfo[loopIndex].qReadCount = 0;
+	ixQMgrQInlinedReadWriteInfo[loopIndex].qWriteCount = 0;
+	ixQMgrQInlinedReadWriteInfo[loopIndex].qAccRegAddr = zeroedPlaceHolder;
+	ixQMgrQInlinedReadWriteInfo[loopIndex].qUOStatRegAddr = zeroedPlaceHolder;
+	ixQMgrQInlinedReadWriteInfo[loopIndex].qUflowStatBitMask = 0;
+	ixQMgrQInlinedReadWriteInfo[loopIndex].qOflowStatBitMask = 0;
+	ixQMgrQInlinedReadWriteInfo[loopIndex].qEntrySizeInWords = 0;
+	ixQMgrQInlinedReadWriteInfo[loopIndex].qSizeInEntries = 0;
+	ixQMgrQInlinedReadWriteInfo[loopIndex].qConfigRegAddr = zeroedPlaceHolder;
+   }
+
+    /* Initialise the AqmIf component */
+    ixQMgrAqmIfInit ();
+   
+    /* Reset all queues to have queue name = NULL, entry size = 0 and
+     * isConfigured = false
+     */
+    for (loopIndex=0; loopIndex < IX_QMGR_MAX_NUM_QUEUES;loopIndex++)
+    {
+	strcpy (cfgQueueInfo[loopIndex].qName, "");
+	cfgQueueInfo[loopIndex].qSizeInWords = 0;
+	cfgQueueInfo[loopIndex].qEntrySizeInWords = 0;
+	cfgQueueInfo[loopIndex].isConfigured = FALSE;
+
+	/* Statistics */
+	stats.qStats[loopIndex].isConfigured = FALSE;
+	stats.qStats[loopIndex].qName = cfgQueueInfo[loopIndex].qName;
+    }
+
+    /* Statistics */
+    stats.wmSetCnt = 0;
+
+    ixQMgrAqmIfSramBaseAddressGet (&freeSramAddress);
+    
+    ixOsalMutexInit(&ixQMgrQCfgMutex);
+
+    cfgInitialized = TRUE;
+}
+
+void
+ixQMgrQCfgUninit (void)
+{
+    cfgInitialized = FALSE;
+
+    /* Uninitialise the AqmIf component */
+    ixQMgrAqmIfUninit ();
+}
+
+IX_STATUS
+ixQMgrQConfig (char *qName,
+	      IxQMgrQId qId,
+	      IxQMgrQSizeInWords qSizeInWords,
+	      IxQMgrQEntrySizeInWords qEntrySizeInWords)
+{
+    UINT32 aqmLocalBaseAddress;
+
+    if (!cfgInitialized)
+    {
+        return IX_FAIL;
+    }
+    
+    if (!IX_QMGR_QID_IS_VALID(qId))
+    {
+	return IX_QMGR_INVALID_Q_ID;
+    }
+    
+    else if (NULL == qName)
+    {
+	return IX_QMGR_PARAMETER_ERROR;
+    }
+    
+    else if (strlen (qName) > IX_QMGR_MAX_QNAME_LEN)
+    {
+	return IX_QMGR_PARAMETER_ERROR;
+    }
+
+    else if (!qSizeInWordsIsOk (qSizeInWords))
+    {
+	return IX_QMGR_INVALID_QSIZE;
+    }
+
+    else if (!qEntrySizeInWordsIsOk (qEntrySizeInWords))
+    {
+	return IX_QMGR_INVALID_Q_ENTRY_SIZE;
+    }
+    
+    else if (cfgQueueInfo[qId].isConfigured)
+    {
+	return IX_QMGR_Q_ALREADY_CONFIGURED;
+    }
+   
+    ixOsalMutexLock(&ixQMgrQCfgMutex, IX_OSAL_WAIT_FOREVER);
+
+    /* Write the config register */
+    ixQMgrAqmIfQueCfgWrite (qId,
+			   qSizeInWords,
+			   qEntrySizeInWords,
+			   freeSramAddress);
+
+
+    strcpy (cfgQueueInfo[qId].qName, qName);
+    cfgQueueInfo[qId].qSizeInWords = qSizeInWords;
+    cfgQueueInfo[qId].qEntrySizeInWords = qEntrySizeInWords;
+
+    /* store pre-computed information in the same cache line
+     * to facilitate inlining of QRead and QWrite functions 
+     * in IxQMgr.h
+     */
+    ixQMgrQInlinedReadWriteInfo[qId].qReadCount = 0;
+    ixQMgrQInlinedReadWriteInfo[qId].qWriteCount = 0;
+    ixQMgrQInlinedReadWriteInfo[qId].qEntrySizeInWords = qEntrySizeInWords;
+    ixQMgrQInlinedReadWriteInfo[qId].qSizeInEntries = 
+		(UINT32)qSizeInWords / (UINT32)qEntrySizeInWords;
+
+    /* Calculate the new freeSramAddress from the size of the queue
+     * currently being configured.
+     */
+    freeSramAddress += (qSizeInWords * IX_QMGR_NUM_BYTES_PER_WORD);
+
+    /* Get the virtual SRAM address */
+    ixQMgrAqmIfBaseAddressGet (&aqmLocalBaseAddress);
+
+    IX_OSAL_ASSERT((freeSramAddress - (aqmLocalBaseAddress + (IX_QMGR_QUEBUFFER_SPACE_OFFSET))) <=
+	      IX_QMGR_QUE_BUFFER_SPACE_SIZE);
+
+    /* The queue is now configured */
+    cfgQueueInfo[qId].isConfigured = TRUE;
+
+    ixOsalMutexUnlock(&ixQMgrQCfgMutex);
+
+#ifndef NDEBUG
+    /* Update statistics */
+    stats.qStats[qId].isConfigured = TRUE;
+    stats.qStats[qId].qName = cfgQueueInfo[qId].qName;
+#endif
+    return IX_SUCCESS;
+}
+
+IxQMgrQSizeInWords
+ixQMgrQSizeInWordsGet (IxQMgrQId qId)
+{
+    /* No parameter checking as this is used on the data path */
+    return (cfgQueueInfo[qId].qSizeInWords);
+}
+
+IX_STATUS
+ixQMgrQSizeInEntriesGet (IxQMgrQId qId,
+			 unsigned *qSizeInEntries)
+{
+    if (!ixQMgrQIsConfigured(qId))
+    {
+        return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+
+    if(NULL == qSizeInEntries)
+    {
+        return IX_QMGR_PARAMETER_ERROR;
+    }
+
+    *qSizeInEntries = (UINT32)(cfgQueueInfo[qId].qSizeInWords) /
+        (UINT32)cfgQueueInfo[qId].qEntrySizeInWords;
+
+    return IX_SUCCESS;
+}
+
+IxQMgrQEntrySizeInWords
+ixQMgrQEntrySizeInWordsGet (IxQMgrQId qId)
+{
+    /* No parameter checking as this is used on the data path */
+    return (cfgQueueInfo[qId].qEntrySizeInWords);
+}
+
+IX_STATUS
+ixQMgrWatermarkSet (IxQMgrQId qId,
+		    IxQMgrWMLevel ne,
+		    IxQMgrWMLevel nf)
+{    
+    IxQMgrQStatus qStatusOnEntry;/* The queue status on entry/exit */
+    IxQMgrQStatus qStatusOnExit; /* to this function               */
+
+    if (!ixQMgrQIsConfigured(qId))
+    {
+        return IX_QMGR_Q_NOT_CONFIGURED;
+    }
+
+    if (!watermarkLevelIsOk (qId, ne))
+    {
+	return IX_QMGR_INVALID_Q_WM;
+    }
+
+    if (!watermarkLevelIsOk (qId, nf))
+    {
+	return IX_QMGR_INVALID_Q_WM;
+    }
+
+    /* Get the current queue status */
+    ixQMgrAqmIfQueStatRead (qId, &qStatusOnEntry);
+
+#ifndef NDEBUG
+    /* Update statistics */
+    stats.wmSetCnt++;
+#endif
+
+    ixQMgrAqmIfWatermarkSet (qId,
+			    ne,
+			    nf);
+
+    /* Get the current queue status */
+    ixQMgrAqmIfQueStatRead (qId, &qStatusOnExit);
+  
+    /* If the status has changed return a warning */
+    if (qStatusOnEntry != qStatusOnExit)
+    {
+	return IX_QMGR_WARNING;
+    }
+
+    return IX_SUCCESS;
+}
+
+IX_STATUS
+ixQMgrAvailableSramAddressGet (UINT32 *address,
+			      unsigned *sizeOfFreeRam)
+{
+    UINT32 aqmLocalBaseAddress;
+
+    if ((NULL == address)||(NULL == sizeOfFreeRam)) 
+    {
+	return IX_QMGR_PARAMETER_ERROR;
+    }
+    if (!cfgInitialized)
+    {
+	return IX_FAIL;
+    }
+
+    *address = freeSramAddress;
+
+    /* Get the virtual SRAM address */
+    ixQMgrAqmIfBaseAddressGet (&aqmLocalBaseAddress);
+
+    /* 
+     * Calculate the size in bytes of free sram 
+     * i.e. current free SRAM virtual pointer from
+     *      (base + total size)
+     */
+    *sizeOfFreeRam = 
+	(aqmLocalBaseAddress +
+	IX_QMGR_AQM_SRAM_SIZE_IN_BYTES) -
+	freeSramAddress;
+
+    if (0 == *sizeOfFreeRam)
+    {
+	return IX_QMGR_NO_AVAILABLE_SRAM;
+    }
+
+    return IX_SUCCESS;
+}
+
+BOOL
+ixQMgrQIsConfigured (IxQMgrQId qId)
+{
+    if (!IX_QMGR_QID_IS_VALID(qId))
+    {
+	return FALSE;
+    }
+
+    return cfgQueueInfo[qId].isConfigured;
+}
+
+IxQMgrQCfgStats*
+ixQMgrQCfgStatsGet (void)
+{
+    return &stats;
+}
+
+IxQMgrQCfgStats*
+ixQMgrQCfgQStatsGet (IxQMgrQId qId)
+{
+    unsigned int ne;
+    unsigned int nf;
+    UINT32 baseAddress;
+    UINT32 readPtr;
+    UINT32 writePtr;
+
+    stats.qStats[qId].qSizeInWords = cfgQueueInfo[qId].qSizeInWords;
+    stats.qStats[qId].qEntrySizeInWords = cfgQueueInfo[qId].qEntrySizeInWords;
+    
+    if (IX_SUCCESS != ixQMgrQNumEntriesGet (qId, &stats.qStats[qId].numEntries))
+    {
+        if (IX_QMGR_WARNING != ixQMgrQNumEntriesGet (qId, &stats.qStats[qId].numEntries))
+        {
+	   IX_QMGR_LOG_WARNING1("Failed to get the number of entries in queue.... %d\n", qId);
+        }
+    }
+
+    ixQMgrAqmIfQueCfgRead (qId,
+			   stats.qStats[qId].numEntries,
+			   &baseAddress,
+			   &ne,
+			   &nf,
+			   &readPtr,
+			   &writePtr);
+        
+    stats.qStats[qId].baseAddress = baseAddress;
+    stats.qStats[qId].ne = ne;
+    stats.qStats[qId].nf = nf;
+    stats.qStats[qId].readPtr = readPtr;
+    stats.qStats[qId].writePtr = writePtr;
+
+    return &stats;
+}
+
+/* 
+ * Static function definitions
+ */
+
+PRIVATE BOOL
+watermarkLevelIsOk (IxQMgrQId qId, IxQMgrWMLevel level)
+{
+    unsigned qSizeInEntries;
+
+    switch (level)
+    {
+	case IX_QMGR_Q_WM_LEVEL0: 
+	case IX_QMGR_Q_WM_LEVEL1: 
+	case IX_QMGR_Q_WM_LEVEL2: 
+	case IX_QMGR_Q_WM_LEVEL4: 
+	case IX_QMGR_Q_WM_LEVEL8: 
+	case IX_QMGR_Q_WM_LEVEL16:
+	case IX_QMGR_Q_WM_LEVEL32:
+	case IX_QMGR_Q_WM_LEVEL64:
+	    break;
+	default:
+	    return FALSE;
+    }
+
+    /* Check watermark is not bigger than the qSizeInEntries */
+    ixQMgrQSizeInEntriesGet(qId, &qSizeInEntries);
+
+    if ((unsigned)level > qSizeInEntries)
+    {
+	return FALSE;
+    }
+
+    return TRUE;
+}
+
+PRIVATE BOOL
+qSizeInWordsIsOk (IxQMgrQSizeInWords qSize)
+{
+    BOOL status;
+
+    switch (qSize)
+    {	
+	case IX_QMGR_Q_SIZE16:
+	case IX_QMGR_Q_SIZE32:
+	case IX_QMGR_Q_SIZE64:
+	case IX_QMGR_Q_SIZE128:
+	    status = TRUE;
+	    break;
+	default:
+	    status = FALSE;
+	    break;
+    }
+
+    return status;
+}
+
+PRIVATE BOOL
+qEntrySizeInWordsIsOk (IxQMgrQEntrySizeInWords entrySize)
+{
+    BOOL status;
+
+    switch (entrySize)
+    {
+	case IX_QMGR_Q_ENTRY_SIZE1:
+	case IX_QMGR_Q_ENTRY_SIZE2:
+	case IX_QMGR_Q_ENTRY_SIZE4:
+	    status = TRUE;
+	    break;
+	default:
+	    status = FALSE;
+	    break;
+    }
+
+    return status;
+}

cpu/ixp/npe/Makefile

@@ -0,0 +1,91 @@
+#
+# (C) Copyright 2006
+# Wolfgang Denk, DENX Software Engineering, wd@denx.de.
+#
+# See file CREDITS for list of people who contributed to this
+# project.
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation; either version 2 of
+# the License, or (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+# MA 02111-1307 USA
+#
+
+include $(TOPDIR)/config.mk
+
+LIB := libnpe.a
+
+CFLAGS  += -I$(TOPDIR)/cpu/ixp/npe/include -DCONFIG_IXP425_COMPONENT_ETHDB
+
+OBJS := npe.o \
+	miiphy.o \
+	IxOsalBufferMgt.o \
+	IxOsalIoMem.o \
+	IxOsalOsCacheMMU.o \
+	IxOsalOsMsgQ.o \
+	IxOsalOsSemaphore.o \
+	IxOsalOsServices.o \
+	IxOsalOsThread.o \
+	IxEthAcc.o \
+	IxEthAccCommon.o \
+	IxEthAccControlInterface.o \
+	IxEthAccDataPlane.o \
+	IxEthAccMac.o \
+	IxEthAccMii.o \
+	IxEthDBAPI.o \
+	IxEthDBAPISupport.o \
+	IxEthDBCore.o \
+	IxEthDBEvents.o \
+	IxEthDBFeatures.o \
+	IxEthDBFirewall.o \
+	IxEthDBHashtable.o \
+	IxEthDBLearning.o \
+	IxEthDBMem.o \
+	IxEthDBNPEAdaptor.o \
+	IxEthDBPortUpdate.o \
+	IxEthDBReports.o \
+	IxEthDBSearch.o \
+	IxEthDBSpanningTree.o \
+	IxEthDBUtil.o \
+	IxEthDBVlan.o \
+	IxEthDBWiFi.o \
+	IxEthMii.o \
+	IxQMgrAqmIf.o \
+	IxQMgrDispatcher.o \
+	IxQMgrInit.o \
+	IxQMgrQAccess.o \
+	IxQMgrQCfg.o \
+	IxFeatureCtrl.o \
+	IxNpeDl.o \
+	IxNpeDlImageMgr.o \
+	IxNpeDlNpeMgr.o \
+	IxNpeDlNpeMgrUtils.o \
+	IxNpeMicrocode.o \
+	IxNpeMh.o \
+	IxNpeMhConfig.o \
+	IxNpeMhReceive.o \
+	IxNpeMhSend.o \
+	IxNpeMhSolicitedCbMgr.o \
+	IxNpeMhUnsolicitedCbMgr.o
+
+all:	$(LIB)
+
+$(LIB):	$(OBJS)
+	$(AR) crv $@ $(OBJS)
+
+#########################################################################
+
+.depend:	Makefile $(OBJS:.o=.c)
+		$(CC) -M $(CFLAGS) $(OBJS:.o=.c) > $@
+
+sinclude .depend

cpu/ixp/npe/include/IxAssert.h

@@ -0,0 +1,71 @@
+/**
+ * @file IxAssert.h
+ *
+ * @date 21-MAR-2002 (replaced by OSAL)
+ *
+ * @brief This file contains assert and ensure macros used by the IXP400 software
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+/**
+ * @defgroup IxAssert IXP400 Assertion Macros (IxAssert) API
+ *
+ * @brief Assertion support
+ *
+ * @{
+ */
+
+#ifndef IXASSERT_H
+
+#ifndef __doxygen_HIDE
+#define IXASSERT_H
+#endif /* __doxygen_HIDE */
+
+#include "IxOsalBackward.h"
+
+#endif /* IXASSERT_H */
+
+/**
+ * @} addtogroup IxAssert
+ */
+
+
+

cpu/ixp/npe/include/IxAtmSch.h

@@ -0,0 +1,504 @@
+/**
+ * @file    IxAtmSch.h
+ *
+ * @date    23-NOV-2001
+ *
+ * @brief   Header file for the IXP400 ATM Traffic Shaper
+ *
+ * This component demonstrates an ATM Traffic Shaper implementation. It
+ * will perform shaping on upto 12 ports and total of 44 VCs accross all ports,
+ * 32 are intended for AAL0/5 and 12 for OAM (1 per port).
+ * The supported traffic types are;1 rt-VBR VC where PCR = SCR.
+ * (Effectively CBR) and Up-to 44 VBR VCs.
+ *
+ * This component models the ATM ports and VCs and is capable of producing
+ * a schedule of ATM cells per port which can be supplied to IxAtmdAcc
+ * for execution on the data path.
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ *
+ * @sa IxAtmm.h
+ *
+ */
+
+/**
+ * @defgroup IxAtmSch IXP400 ATM Transmit Scheduler (IxAtmSch) API
+ *
+ * @brief IXP400 ATM scheduler component Public API
+ *
+ * @{
+ */
+
+#ifndef IXATMSCH_H
+#define IXATMSCH_H
+
+#include "IxOsalTypes.h"
+#include "IxAtmTypes.h"
+
+/*
+ * #defines and macros used in this file.
+ */
+
+/* Return codes */
+
+/** 
+ * @ingroup IxAtmSch
+ *
+ * @def IX_ATMSCH_RET_NOT_ADMITTED
+ * @brief Indicates that CAC function has rejected VC registration due
+ *         to insufficient line capacity.
+*/
+#define IX_ATMSCH_RET_NOT_ADMITTED 2
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @def IX_ATMSCH_RET_QUEUE_FULL
+ *  @brief Indicates that the VC queue is full, no more demand can be
+ *         queued at this time.
+ */
+#define IX_ATMSCH_RET_QUEUE_FULL 3
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ *  @def IX_ATMSCH_RET_QUEUE_EMPTY
+ *  @brief Indicates that all VC queues on this port are empty and
+ *         therefore there are no cells to be scheduled at this time.
+ */
+#define IX_ATMSCH_RET_QUEUE_EMPTY 4
+
+/*
+ * Function declarations
+ */
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @fn ixAtmSchInit(void)
+ *
+ *  @brief This function is used to initialize the ixAtmSch component. It
+ *         should be called before any other IxAtmSch API function.
+ *
+ * @param None
+ *
+ * @return
+ * - <b>IX_SUCCESS :</b> indicates that
+ *          -# The ATM scheduler component has been successfully initialized.
+ *          -# The scheduler is ready to accept Port modelling requests.
+ * - <b>IX_FAIL :</b> Some internal error has prevented the scheduler component
+ *          from initialising.
+ */
+PUBLIC IX_STATUS
+ixAtmSchInit(void);
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @fn ixAtmSchPortModelInitialize( IxAtmLogicalPort port,
+                                       unsigned int portRate,
+                                       unsigned int minCellsToSchedule)
+ *
+ * @brief This function shall be called first to initialize an ATM port before
+ *         any other ixAtmSch API calls may be made for that port.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - The specific port to initialize.  Valid
+ *          values range from 0 to IX_UTOPIA_MAX_PORTS - 1, representing a 
+ *          maximum of IX_UTOPIA_MAX_PORTS possible ports.
+ *
+ * @param portRate unsigned int [in] - Value indicating the upstream capacity
+ *          of the indicated port.  The value should be supplied in
+ *          units of ATM (53 bytes) cells per second.
+ *          A port rate of 800Kbits/s is the equivalent 
+ *          of 1886 cells per second
+ *
+ * @param minCellsToSchedule unsigned int [in] - This parameter specifies the minimum
+ *          number of cells which the scheduler will put in a schedule
+ *          table for this port. This value sets the worst case CDVT for VCs
+ *          on this port i.e. CDVT = 1*minCellsToSchedule/portRate.
+ * @return
+ *    - <b>IX_SUCCESS :</b> indicates that
+ *          -# The ATM scheduler has been successfully initialized.
+ *          -# The requested port model has been established.
+ *          -# The scheduler is ready to accept VC modelling requests
+ *            on the ATM port.
+ *    - <b>IX_FAIL :</b> indicates the requested port could not be
+ * initialized.  */
+PUBLIC IX_STATUS
+ixAtmSchPortModelInitialize( IxAtmLogicalPort port,
+                                       unsigned int portRate,
+                                       unsigned int minCellsToSchedule);
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @fn ixAtmSchPortRateModify( IxAtmLogicalPort port,
+                        unsigned int portRate)
+ *
+ *  @brief This function is called to modify the portRate on a
+ *         previously initialized port, typically in the event that
+ *         the line condition of the port changes.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Specifies the ATM port which is to be
+ *          modified.
+ *
+ * @param portRate unsigned int [in] - Value indicating the new upstream
+ *          capacity for this port in cells/second.
+ *          A port rate of 800Kbits/s is the equivalent 
+ *          of 1886 cells per second
+ *
+ * @return
+ * - <b>IX_SUCCESS :</b> The port rate has been successfully modified.<br>
+ * - <b>IX_FAIL :</b> The port rate could not be modified, either
+ *      because the input data was invalid, or the new port rate is
+ *      insufficient to support established ATM VC contracts on this
+ *      port.
+ *
+ * @warning The IxAtmSch component will validate the supplied port
+ *          rate is sufficient to support all established VC
+ *          contracts on the port.  If the new port rate is
+ *          insufficient to support all established contracts then
+ *          the request to modify the port rate will be rejected.
+ *          In this event, the user is expected to remove
+ *          established contracts using the ixAtmSchVcModelRemove
+ *          interface and then retry this interface.
+ *
+ * @sa ixAtmSchVcModelRemove() */
+PUBLIC IX_STATUS
+ixAtmSchPortRateModify( IxAtmLogicalPort port,
+                        unsigned int portRate);
+
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @fn ixAtmSchVcModelSetup( IxAtmLogicalPort port,
+                      IxAtmTrafficDescriptor *trafficDesc,
+                      IxAtmSchedulerVcId *vcId)
+ *
+ *  @brief A client calls this interface to set up an upstream
+ *         (transmitting) virtual connection model (VC) on the
+ *         specified ATM port.  This function also provides the
+ *         virtual * connection admission control (CAC) service to the
+ *         client.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Specifies the ATM port on which the upstream
+ *          VC is to be established.
+ *
+ * @param *trafficDesc @ref IxAtmTrafficDescriptor [in] - Pointer to a structure
+ *          describing the requested traffic contract of the VC to be
+ *          established.  This structure contains the typical ATM
+ *          traffic descriptor values (e.g. PCR, SCR, MBS, CDVT, etc.)
+ *          defined by the ATM standard.
+ *
+ * @param *vcId @ref IxAtmSchedulerVcId [out] - This value will be filled with the
+ *              port-unique identifier for this virtual connection.  A
+ *              valid identification is a non-negative number.
+ *
+ * @return
+ * - <b>IX_SUCCESS :</b> The VC has been successfully established on
+ *      this port.  The client may begin to submit demand on this VC.
+ * - <b>IX_ATMSCH_RET_NOT_ADMITTED :</b> The VC cannot be established
+ *      on this port because there is insufficient upstream capacity
+ *      available to support the requested traffic contract descriptor
+ * - <b>IX_FAIL :</b>Input data are invalid.  VC has not been
+ *      established.
+ */
+PUBLIC IX_STATUS
+ixAtmSchVcModelSetup( IxAtmLogicalPort port,
+                      IxAtmTrafficDescriptor *trafficDesc,
+                      IxAtmSchedulerVcId *vcId);
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @fn ixAtmSchVcConnIdSet( IxAtmLogicalPort port,
+                     IxAtmSchedulerVcId vcId,
+                     IxAtmConnId vcUserConnId)
+ *
+ *  @brief A client calls this interface to set the vcUserConnId for a VC on
+ *         the specified ATM port. This vcUserConnId will default to
+ *         IX_ATM_IDLE_CELLS_CONNID if this function is not called for a VC.
+ *         Hence if the client does not call this function for a VC then only idle
+ *         cells will be scheduled for this VC.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Specifies the ATM port on which the upstream
+ *        VC is has been established.
+ *
+ * @param vcId @ref IxAtmSchedulerVcId [in] - This is the unique identifier for this virtual
+ *        connection. A valid identification is a non-negative number and is
+ *        all ports.
+ *
+ * @param vcUserConnId @ref IxAtmConnId [in] - The connId is used to refer to a VC in schedule
+ *        table entries. It is treated as the Id by which the scheduler client
+ *        knows the VC. It is used in any communicatations from the Scheduler
+ *        to the scheduler user e.g. schedule table entries.
+ *
+ * @return
+ * - <b>IX_SUCCESS :</b> The id has successfully been set.
+ * - <b>IX_FAIL :</b>Input data are invalid. connId id is not established.
+ */
+PUBLIC IX_STATUS
+ixAtmSchVcConnIdSet( IxAtmLogicalPort port,
+                     IxAtmSchedulerVcId vcId,
+                     IxAtmConnId vcUserConnId);
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @fn ixAtmSchVcModelRemove( IxAtmLogicalPort port,
+                       IxAtmSchedulerVcId vcId)
+ *
+ *  @brief Interface called by the client to remove a previously
+ *         established VC on a particular port.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Specifies the ATM port on which the VC to be
+ *          removed is established.
+ *
+ * @param vcId @ref IxAtmSchedulerVcId [in] - Identifies the VC to be removed.  This is the
+ *          value returned by the @ref ixAtmSchVcModelSetup call which
+ *          established the relevant VC.
+ *
+ * @return
+ * - <b>IX_SUCCESS :</b> The VC has been successfully removed from
+ *      this port. It is no longer modelled on this port.
+ * - <b>IX_FAIL :</b>Input data are invalid. The VC is still being modeled
+ *      by the traffic shaper.
+ *
+ * @sa ixAtmSchVcModelSetup() 
+ */
+PUBLIC IX_STATUS
+ixAtmSchVcModelRemove( IxAtmLogicalPort port,
+                       IxAtmSchedulerVcId vcId);
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @fn ixAtmSchVcQueueUpdate( IxAtmLogicalPort port,
+                       IxAtmSchedulerVcId vcId,
+                       unsigned int numberOfCells)
+ *
+ *  @brief The client calls this function to notify IxAtmSch that the
+ *         user of a VC has submitted cells for transmission.
+ *
+ *  This information is stored, aggregated from a number of calls to
+ *  ixAtmSchVcQueueUpdate and eventually used in the call to
+ *  ixAtmSchTableUpdate.
+ *
+ *  Normally IxAtmSch will update the VC queue by adding the number of
+ *  cells to the current queue length.  However, if IxAtmSch
+ *  determines that the user has over-submitted for the VC and
+ *  exceeded its transmission quota the queue request can be rejected.
+ *  The user should resubmit the request later when the queue has been
+ *  depleted.
+ *
+ *  This implementation of ixAtmSchVcQueueUpdate uses no operating
+ *  system or external facilities, either directly or indirectly.
+ *  This allows clients to call this function form within an interrupt handler.
+ *
+ *  This interface is structurally compatible with the
+ *  IxAtmdAccSchQueueUpdate callback type definition required for
+ *  IXP400 ATM scheduler interoperability.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Specifies the ATM port on which the VC to be
+ *          updated is established.
+ *
+ * @param vcId @ref IxAtmSchedulerVcId [in] - Identifies the VC to be updated.  This is the
+ *          value returned by the @ref ixAtmSchVcModelSetup call which
+ *          established the relevant VC.
+ *
+ * @param numberOfCells unsigned int [in] - Indicates how many ATM cells should
+ *          be added to the queue for this VC.
+ *
+ * @return
+ *  - <b>IX_SUCCESS :</b> The VC queue has been successfully updated.
+ *  - <b>IX_ATMSCH_RET_QUEUE_FULL :</b> The VC queue has reached a
+ *       preset limit.  This indicates the client has over-submitted
+ *       and exceeded its transmission quota.  The request is
+ *       rejected.  The VC queue is not updated.  The VC user is
+ *       advised to resubmit the request later.
+ *  - <b>IX_FAIL :</b> The input are invalid.  No VC queue is updated.
+ *
+ * @warning IxAtmSch assumes that the calling software ensures that
+ *          calls to ixAtmSchVcQueueUpdate, ixAtmSchVcQueueClear and
+ *          ixAtmSchTableUpdate are both self and mutually exclusive
+ *          for the same port.
+ *
+ * @sa ixAtmSchVcQueueUpdate(), ixAtmSchVcQueueClear(), ixAtmSchTableUpdate().  */
+PUBLIC IX_STATUS
+ixAtmSchVcQueueUpdate( IxAtmLogicalPort port,
+                       IxAtmSchedulerVcId vcId,
+                       unsigned int numberOfCells);
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @fn ixAtmSchVcQueueClear( IxAtmLogicalPort port,
+                      IxAtmSchedulerVcId vcId)
+ *
+ *  @brief The client calls this function to remove all currently
+ *         queued cells from a registered VC.  The pending cell count
+ *         for the specified VC is reset to zero.
+ *
+ *  This interface is structurally compatible with the
+ *  IxAtmdAccSchQueueClear callback type definition required for
+ *  IXP400 ATM scheduler interoperability.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Specifies the ATM port on which the VC to be
+ *          cleared is established.
+ *
+ * @param vcId @ref IxAtmSchedulerVcId [in] - Identifies the VC to be cleared.  This is the
+ *          value returned by the @ref ixAtmSchVcModelSetup call which
+ *          established the relevant VC.
+ *
+ * @return
+ *  - <b>IX_SUCCESS :</b> The VC queue has been successfully cleared.
+ *  - <b>IX_FAIL :</b> The input are invalid.  No VC queue is modified.
+ *
+ * @warning IxAtmSch assumes that the calling software ensures that
+ *          calls to ixAtmSchVcQueueUpdate, ixAtmSchVcQueueClear and
+ *          ixAtmSchTableUpdate are both self and mutually exclusive
+ *          for the same port.
+ *
+ * @sa ixAtmSchVcQueueUpdate(), ixAtmSchVcQueueClear(), ixAtmSchTableUpdate().  */
+PUBLIC IX_STATUS
+ixAtmSchVcQueueClear( IxAtmLogicalPort port,
+                      IxAtmSchedulerVcId vcId);
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @fn ixAtmSchTableUpdate( IxAtmLogicalPort port,
+                     unsigned int maxCells,
+                     IxAtmScheduleTable **rettable)
+ *
+ *  @brief The client calls this function to request an update of the
+ *         schedule table for a particular ATM port.
+ *
+ *  This is called when the client decides it needs a new sequence of
+ *  cells to send (probably because the transmit queue is near to
+ *  empty for this ATM port).  The scheduler will use its stored
+ *  information on the cells submitted for transmit (i.e. data
+ *  supplied via @ref ixAtmSchVcQueueUpdate function) with the traffic
+ *  descriptor information of all established VCs on the ATM port to
+ *  decide the sequence of cells to be sent and fill the schedule
+ *  table for a period of time into the future.
+ *
+ *  IxAtmSch will guarantee a minimum of minCellsToSchedule if there
+ *  is at least one cell ready to send. If there are no cells then
+ *  IX_ATMSCH_RET_QUEUE_EMPTY is returned.
+ *
+ *  This implementation of ixAtmSchTableUpdate uses no operating
+ *  system or external facilities, either directly or indirectly.
+ *  This allows clients to call this function form within an FIQ
+ *  interrupt handler.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Specifies the ATM port for which requested
+ *          schedule table is to be generated.
+ *
+ * @param maxCells unsigned [in] - Specifies the maximum number of cells
+ *          that must be scheduled in the supplied table during any
+ *          call to the interface.
+ *
+ * @param **table @ref IxAtmScheduleTable [out] - A pointer to an area of
+ *              storage is returned which contains the generated
+ *              schedule table.  The client should not modify the
+ *              contents of this table.
+ *
+ * @return
+ *  - <b>IX_SUCCESS :</b> The schedule table has been published.
+ *       Currently there is at least one VC queue that is nonempty.
+ *  - <b>IX_ATMSCH_RET_QUEUE_EMPTY :</b> Currently all VC queues on
+ *       this port are empty.  The schedule table returned is set to
+ *       NULL.  The client is not expected to invoke this function
+ *       again until more cells have been submitted on this port
+ *       through the @ref ixAtmSchVcQueueUpdate function.
+ *  - <b>IX_FAIL :</b> The input are invalid.  No action is taken.
+ *
+ * @warning IxAtmSch assumes that the calling software ensures that
+ *          calls to ixAtmSchVcQueueUpdate, ixAtmSchVcQueueClear and
+ *          ixAtmSchTableUpdate are both self and mutually exclusive
+ *          for the same port.
+ *
+ * @warning Subsequent calls to this function for the same port will
+ *          overwrite the contents of previously supplied schedule
+ *          tables.  The client must be completely finished with the
+ *          previously supplied schedule table before calling this
+ *          function again for the same port.
+ *
+ * @sa ixAtmSchVcQueueUpdate(), ixAtmSchVcQueueClear(), ixAtmSchTableUpdate().  */
+PUBLIC IX_STATUS
+ixAtmSchTableUpdate( IxAtmLogicalPort port,
+                     unsigned int maxCells,
+                     IxAtmScheduleTable **rettable);
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @fn ixAtmSchShow(void)
+ *
+ *  @brief Utility function which will print statistics on the current
+ *         and accumulated state of VCs and traffic in the ATM
+ *         scheduler component.  Output is sent to the default output
+ *         device.
+ *
+ * @param none
+ * @return none
+ */
+PUBLIC void
+ixAtmSchShow(void);
+
+/**  
+ * @ingroup IxAtmSch
+ *
+ * @fn ixAtmSchStatsClear(void)
+ *
+ *  @brief Utility function which will reset all counter statistics in
+ *         the ATM scheduler to zero.
+ *
+ * @param none
+ * @return none
+ */
+PUBLIC void
+ixAtmSchStatsClear(void);
+
+#endif
+/* IXATMSCH_H */
+
+/** @} */

cpu/ixp/npe/include/IxAtmTypes.h

@@ -0,0 +1,409 @@
+/**
+ * @file IxAtmTypes.h
+ *
+ * @date 24-MAR-2002
+ *
+ * @brief This file contains Atm types common to a number of Atm components.
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+/* ------------------------------------------------------
+   Doxygen group definitions
+   ------------------------------------------------------ */
+/**
+ * @defgroup IxAtmTypes IXP400 ATM Types (IxAtmTypes)
+ *
+ * @brief The common set of types used in many Atm components
+ *
+ * @{ */
+
+#ifndef IXATMTYPES_H
+#define IXATMTYPES_H
+
+#include "IxNpeA.h"
+
+/**
+ * @enum IxAtmLogicalPort
+ *
+ * @brief Logical Port Definitions  :
+ *
+ * Only 1 port is available in SPHY configuration
+ * 12 ports are enabled in MPHY configuration
+ *
+ */
+typedef enum
+{
+    IX_UTOPIA_PORT_0 = 0,  /**< Port 0 */
+#ifdef IX_NPE_MPHYMULTIPORT
+    IX_UTOPIA_PORT_1,      /**< Port 1 */
+    IX_UTOPIA_PORT_2,      /**< Port 2 */
+    IX_UTOPIA_PORT_3,      /**< Port 3 */
+    IX_UTOPIA_PORT_4,      /**< Port 4 */
+    IX_UTOPIA_PORT_5,      /**< Port 5 */
+    IX_UTOPIA_PORT_6,      /**< Port 6 */
+    IX_UTOPIA_PORT_7,      /**< Port 7 */
+    IX_UTOPIA_PORT_8,      /**< Port 8 */
+    IX_UTOPIA_PORT_9,      /**< Port 9 */
+    IX_UTOPIA_PORT_10,     /**< Port 10 */
+    IX_UTOPIA_PORT_11,     /**< Port 11 */
+#endif /* IX_NPE_MPHY */
+    IX_UTOPIA_MAX_PORTS    /**< Not a port - just a definition for the
+                           * maximum possible ports
+                           */
+} IxAtmLogicalPort;
+
+/**
+ * @def IX_ATM_CELL_PAYLOAD_SIZE
+ * @brief Size of a ATM cell payload
+ */
+#define IX_ATM_CELL_PAYLOAD_SIZE             (48)
+
+/**
+ * @def IX_ATM_CELL_SIZE
+ * @brief Size of a ATM cell, including header
+ */
+#define IX_ATM_CELL_SIZE                     (53)
+
+/**
+ * @def IX_ATM_CELL_SIZE_NO_HEC
+ * @brief Size of a ATM cell, excluding HEC byte
+ */
+#define IX_ATM_CELL_SIZE_NO_HEC              (IX_ATM_CELL_SIZE - 1)
+
+/**
+ * @def IX_ATM_OAM_CELL_SIZE_NO_HEC
+ * @brief Size of a OAM cell, excluding HEC byte
+ */
+#define IX_ATM_OAM_CELL_SIZE_NO_HEC          IX_ATM_CELL_SIZE_NO_HEC
+
+/**
+ * @def IX_ATM_AAL0_48_CELL_PAYLOAD_SIZE
+ * @brief Size of a AAL0 48 Cell payload
+ */
+#define IX_ATM_AAL0_48_CELL_PAYLOAD_SIZE     IX_ATM_CELL_PAYLOAD_SIZE
+
+/**
+ * @def IX_ATM_AAL5_CELL_PAYLOAD_SIZE
+ * @brief Size of a AAL5 Cell payload
+ */
+#define IX_ATM_AAL5_CELL_PAYLOAD_SIZE        IX_ATM_CELL_PAYLOAD_SIZE
+
+/**
+ * @def IX_ATM_AAL0_52_CELL_SIZE_NO_HEC
+ * @brief Size of a AAL0 52 Cell, excluding HEC byte
+ */
+#define IX_ATM_AAL0_52_CELL_SIZE_NO_HEC      IX_ATM_CELL_SIZE_NO_HEC
+
+
+/**
+ * @def IX_ATM_MAX_VPI
+ * @brief Maximum value of an ATM VPI
+ */
+#define IX_ATM_MAX_VPI 255
+
+/**
+ * @def IX_ATM_MAX_VCI
+ * @brief Maximum value of an ATM VCI
+ */
+#define IX_ATM_MAX_VCI 65535
+
+ /**
+ * @def IX_ATM_MAX_NUM_AAL_VCS
+ * @brief Maximum number of active AAL5/AAL0 VCs in the system
+ */
+#define IX_ATM_MAX_NUM_AAL_VCS 32
+
+/**
+ * @def IX_ATM_MAX_NUM_VC
+ * @brief Maximum number of active AAL5/AAL0 VCs in the system
+ * The use of this macro is depreciated, it is retained for
+ * backward compatiblity. For current software release
+ * and beyond the define IX_ATM_MAX_NUM_AAL_VC should be used.
+ */
+#define IX_ATM_MAX_NUM_VC IX_ATM_MAX_NUM_AAL_VCS
+
+
+
+/**
+ * @def IX_ATM_MAX_NUM_OAM_TX_VCS
+ * @brief Maximum number of active OAM Tx VCs in the system, 
+ *        1 OAM VC per port
+ */
+#define IX_ATM_MAX_NUM_OAM_TX_VCS IX_UTOPIA_MAX_PORTS
+
+/**
+ * @def IX_ATM_MAX_NUM_OAM_RX_VCS
+ * @brief Maximum number of active OAM Rx VCs in the system, 
+ *        1 OAM VC shared accross all ports
+ */
+#define IX_ATM_MAX_NUM_OAM_RX_VCS 1
+
+/**
+ * @def IX_ATM_MAX_NUM_AAL_OAM_TX_VCS
+ * @brief Maximum number of active AAL5/AAL0/OAM Tx VCs in the system
+ */
+#define IX_ATM_MAX_NUM_AAL_OAM_TX_VCS (IX_ATM_MAX_NUM_AAL_VCS + IX_ATM_MAX_NUM_OAM_TX_VCS)
+
+/**
+ * @def IX_ATM_MAX_NUM_AAL_OAM_RX_VCS
+ * @brief Maximum number of active AAL5/AAL0/OAM Rx VCs in the system
+ */
+#define IX_ATM_MAX_NUM_AAL_OAM_RX_VCS (IX_ATM_MAX_NUM_AAL_VCS + IX_ATM_MAX_NUM_OAM_RX_VCS)
+
+/**
+ *  @def IX_ATM_IDLE_CELLS_CONNID
+ *  @brief VC Id used to indicate idle cells in the returned schedule table.
+ */
+#define IX_ATM_IDLE_CELLS_CONNID 0
+
+
+/**
+ *  @def IX_ATM_CELL_HEADER_VCI_GET
+ *  @brief get the VCI field from a cell header
+ */
+#define IX_ATM_CELL_HEADER_VCI_GET(cellHeader) \
+    (((cellHeader) >> 4) & IX_OAM_VCI_BITS_MASK);
+
+/**
+ *  @def IX_ATM_CELL_HEADER_VPI_GET
+ *  @brief get the VPI field from a cell header
+ */
+#define IX_ATM_CELL_HEADER_VPI_GET(cellHeader) \
+    (((cellHeader) >> 20) & IX_OAM_VPI_BITS_MASK);
+
+/**
+ *  @def IX_ATM_CELL_HEADER_PTI_GET
+ *  @brief get the PTI field from a cell header
+ */
+#define IX_ATM_CELL_HEADER_PTI_GET(cellHeader) \
+    ((cellHeader) >> 1) & IX_OAM_PTI_BITS_MASK;
+
+/**
+ * @typedef IxAtmCellHeader
+ *
+ * @brief ATM Cell Header, does not contain 4 byte HEC, added by NPE-A 
+ */
+typedef unsigned int IxAtmCellHeader;
+
+
+/**
+ * @enum IxAtmServiceCategory
+ *
+ * @brief Enumerated type representing available ATM service categories.
+ *   For more informatoin on these categories, see "Traffic Management
+ *   Specification" v4.1, published by the ATM Forum -
+ *   http://www.atmforum.com
+ */
+typedef enum
+{
+    IX_ATM_CBR,    /**< Constant Bit Rate */
+    IX_ATM_RTVBR,  /**< Real Time Variable Bit Rate */
+    IX_ATM_VBR,    /**< Variable Bit Rate */
+    IX_ATM_UBR,    /**< Unspecified Bit Rate */
+    IX_ATM_ABR     /**< Available Bit Rate (not supported) */
+
+} IxAtmServiceCategory;
+
+/**
+ *
+ * @enum IxAtmRxQueueId
+ *
+ * @brief Rx Queue Type for RX traffic
+ *
+ * IxAtmRxQueueId defines the queues involved for receiving data.
+ *
+ * There are two queues to facilitate prioritisation handling
+ * and processing the 2 queues with different algorithms and
+ * constraints
+ *
+ * e.g. : one queue can carry voice (or time-critical traffic), the
+ * other queue can carry non-voice traffic
+ *
+ */
+typedef enum
+{
+    IX_ATM_RX_A = 0,      /**< RX queue A */
+    IX_ATM_RX_B,          /**< RX queue B */
+    IX_ATM_MAX_RX_STREAMS /**< Maximum number of RX streams */
+} IxAtmRxQueueId;
+
+/**
+ * @brief Structure describing an ATM traffic contract for a Virtual
+ *         Connection (VC).
+ *
+ * Structure is used to specify the requested traffic contract for a
+ * VC to the IxAtmSch component using the @ref ixAtmSchVcModelSetup
+ * interface.
+ *
+ * These parameters are defined by the ATM forum working group
+ * (http://www.atmforum.com).
+ *
+ * @note Typical values for a voice channel 64 Kbit/s
+ * - atmService @a IX_ATM_RTVBR
+ * - pcr   400  (include IP overhead, and AAL5 trailer)
+ * - cdvt  5000000 (5 ms)
+ * - scr = pcr
+ *
+ * @note Typical values for a data channel 800 Kbit/s
+ * - atmService @a IX_ATM_UBR
+ * - pcr   1962  (include IP overhead, and AAL5 trailer)
+ * - cdvt  5000000 (5 ms)
+ *
+ */
+typedef struct
+{
+    IxAtmServiceCategory atmService; /**< ATM service category */
+    unsigned pcr;   /**< Peak Cell Rate - cells per second */
+    unsigned cdvt;  /**< Cell Delay Variation Tolerance - in nanoseconds */
+    unsigned scr;   /**< Sustained Cell Rate - cells per second */
+    unsigned mbs;   /**< Max Burst Size - cells */
+    unsigned mcr;   /**< Minimum Cell Rate - cells per second */
+    unsigned mfs;   /**< Max Frame Size - cells */
+} IxAtmTrafficDescriptor;
+
+/**
+ * @typedef IxAtmConnId
+ *
+ * @brief ATM VC data connection identifier.
+ *
+ * This is is generated by IxAtmdAcc when a successful connection is
+ * made on a VC. The is the ID by which IxAtmdAcc knows an active
+ * VC and should be used in IxAtmdAcc API calls to reference a
+ * specific VC.
+ */
+typedef unsigned int IxAtmConnId;
+
+/**
+ * @typedef IxAtmSchedulerVcId
+ *
+ * @brief ATM VC scheduling connection identifier.
+ *
+ * This id is generated and used by ATM Tx controller, generally
+ * the traffic shaper (e.g. IxAtmSch). The IxAtmdAcc component
+ * will request one of these Ids whenever a data connection on
+ * a Tx VC is requested. This ID will be used in callbacks to
+ * the ATM Transmission Ctrl s/w (e.g. IxAtmm) to reference a
+ * particular VC.
+ */
+typedef int IxAtmSchedulerVcId;
+
+/**
+ * @typedef IxAtmNpeRxVcId
+ *
+ * @brief ATM Rx VC identifier used by the ATM Npe.
+ *
+ * This Id is generated by IxAtmdAcc when a successful data connection
+ * is made on a rx VC.
+ */
+typedef unsigned int IxAtmNpeRxVcId;
+
+/**
+ * @brief ATM Schedule Table entry
+ *
+ * This IxAtmScheduleTableEntry is used by an ATM scheduler to inform
+ * IxAtmdAcc about the data to transmit (in term of cells per VC)
+ *
+ * This structure defines
+ * @li the number of cells to be transmitted (numberOfCells)
+ * @li the VC connection to be used for transmission (connId).
+ *
+ * @note - When the connection Id value is IX_ATM_IDLE_CELLS_CONNID, the
+ * corresponding number of idle cells will be transmitted to the hardware.
+ *
+ */
+typedef struct
+{
+    IxAtmConnId connId; /**< connection Id
+                 *
+                 * Identifier of VC from which cells are to be transmitted.
+                 * When this valus is IX_ATM_IDLE_CELLS_CONNID, this indicates
+                 * that the system should transmit the specified number
+                 * of idle cells. Unknown connIds result in the transmission
+                 * idle cells.
+                 */
+    unsigned int numberOfCells; /**< number of cells to transmit
+                 *
+                 * The number of contiguous cells to schedule from this VC
+                 * at this point. The valid range is from 1 to
+                 * @a IX_ATM_SCHEDULETABLE_MAXCELLS_PER_ENTRY. This
+                 * number can swap over mbufs and pdus. OverSchduling results
+                 * in the transmission of idle cells.
+                 */
+} IxAtmScheduleTableEntry;
+
+/**
+ * @brief This structure defines a schedule table which gives details
+ *         on which data (from which VCs) should be transmitted for a
+ *         forthcoming period of time for a particular port and the
+ *         order in which that data should be transmitted.
+ *
+ *  The schedule table consists of a series of entries each of which
+ *  will schedule one or more cells from a particular registered VC.
+ *  The total number of cells scheduled and the total number of
+ *  entries in the table are also indicated.
+ *
+ */
+typedef struct
+{
+    unsigned tableSize;      /**< Number of entries
+                              *
+                              * Indicates the total number of
+                              *   entries in the table.
+                              */
+    unsigned totalCellSlots; /**< Number of cells
+                              *
+                              * Indicates the total number of ATM
+                              *   cells which are scheduled by all the
+                              *   entries in the table.
+                              */
+    IxAtmScheduleTableEntry *table; /**< Pointer to schedule entries
+                                     *
+                                     * Pointer to an array
+                                     *   containing tableSize entries
+                                     */
+} IxAtmScheduleTable;
+
+#endif /* IXATMTYPES_H */
+
+/**
+ * @} defgroup IxAtmTypes
+ */
+
+

cpu/ixp/npe/include/IxAtmdAcc.h

@@ -0,0 +1,1194 @@
+
+/**
+ * @file    IxAtmdAcc.h
+ *
+ * @date    07-Nov-2001
+ *
+ * @brief IxAtmdAcc Public API
+ *
+ * This file contains the public API of IxAtmdAcc, related to the
+ * data functions of the component
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+/* ------------------------------------------------------
+   Doxygen group definitions
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @defgroup IxAtmdAccAPI IXP400 ATM Driver Access (IxAtmdAcc) API
+ *
+ * @brief The public API for the IXP400 Atm Driver Data component
+ *
+ * IxAtmdAcc is the low level interface by which AAL0/AAL5 and
+ * OAM data gets transmitted to,and received from the Utopia bus.
+ *
+ * For AAL0/AAL5 services transmit and receive connections may
+ * be established independantly for unique combinations of
+ * port,VPI,and VCI.
+ *
+ * Two AAL0 services supporting 48 or 52 byte cell data are provided.
+ * Submitted AAL0 PDUs must be a multiple of the cell data size (48/52).
+ * AAL0_52 is a raw cell service the client must format
+ * the PDU with an ATM cell header (excluding HEC) at the start of
+ * each cell, note that AtmdAcc does not validate the cell headers in
+ * a submitted PDU.
+ *
+ * OAM cells cannot be received over the AAL0 service but instead
+ * are received over a dedicated OAM service.
+ *
+ * For the OAM service an "OAM Tx channel" may be enabled for a port
+ * by establishing a single dedicated OAM Tx connection on that port.
+ * A single "OAM Rx channel" for all ports may be  enabled by
+ * establishing a dedicated OAM Rx connection.
+ *
+ * The OAM service allows buffers containing 52 byte OAM F4/F5 cells
+ * to be transmitted and received over the dedicated OAM channels.
+ * HEC is appended/removed, and CRC-10 performed by the NPE. The OAM
+ * service offered by AtmdAcc is a raw cell transport service.
+ * It is assumed that ITU I.610 procedures that make use of this
+ * service are implemented above AtmdAcc.
+ *
+ * Note that the dedicated OAM connections are established on
+ * reserved VPI,VCI, and (in the case of Rx) port values defined below.
+ * These values are used purely to descriminate the dedicated OAM channels
+ * and do not identify a particular OAM F4/F5 flow. F4/F5 flows may be
+ * realised for particluar VPI/VCIs by manipulating the VPI,VCI
+ * fields of the ATM cell headers of cells in the buffers passed
+ * to AtmdAcc. Note that AtmdAcc does not validate the cell headers
+ * in a submitted OAM PDU.
+ *
+ *
+ *
+ * This part is related to the User datapath processing
+ *
+ * @{
+ */
+
+#ifndef IXATMDACC_H
+#define IXATMDACC_H
+
+#include "IxAtmTypes.h"
+
+/* ------------------------------------------------------
+   AtmdAcc Data Types definition
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @def IX_ATMDACC_WARNING
+ *
+ * @brief Warning return code
+ *
+ * This constant is used to tell IxAtmDAcc user about a special case.
+ *
+ */
+#define IX_ATMDACC_WARNING 2
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @def IX_ATMDACC_BUSY
+ *
+ * @brief Busy return code
+ *
+ * This constant is used to tell IxAtmDAcc user that the request
+ * is correct, but cannot be processed because the IxAtmAcc resources
+ * are already used. The user has to retry its request later
+ *
+ */
+#define IX_ATMDACC_BUSY 3
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @def IX_ATMDACC_RESOURCES_STILL_ALLOCATED
+ *
+ * @brief Disconnect return code
+ *
+ * This constant is used to tell IxAtmDAcc user that the disconnect
+ * functions are not complete because the resources used by the driver
+ * are not yet released. The user has to retry the disconnect call
+ * later.
+ *
+ */
+#define IX_ATMDACC_RESOURCES_STILL_ALLOCATED 4
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @def IX_ATMDACC_DEFAULT_REPLENISH_COUNT
+ *
+ * @brief Default resources usage for RxVcFree replenish mechanism
+ *
+ * This constant is used to tell IxAtmDAcc to allocate and use
+ * the minimum of resources for rx free replenish.
+ *
+ * @sa ixAtmdAccRxVcConnect
+ */
+#define IX_ATMDACC_DEFAULT_REPLENISH_COUNT 0
+
+
+/**
+ * @ingroup IxAtmdAccAPI
+ *
+ * @def IX_ATMDACC_OAM_TX_VPI
+ *
+ * @brief The reserved value used for the dedicated OAM
+ * Tx connection. This "well known" value is used by atmdAcc and
+ * its clients to dsicriminate the OAM channel, and should be chosen so
+ * that it does not coencide with the VPI value used in an AAL0/AAL5 connection.
+ * Any attempt to connect a service type other than OAM on this VPI will fail.
+ *
+ *
+ */
+#define IX_ATMDACC_OAM_TX_VPI 0
+
+/**
+ * @ingroup IxAtmdAccAPI
+ *
+ * @def IX_ATMDACC_OAM_TX_VCI
+ *
+ * @brief The reserved value used for the dedicated OAM
+ * Tx connection. This "well known" value is used by atmdAcc and
+ * its clients to dsicriminate the OAM channel, and should be chosen so
+ * that it does not coencide with the VCI value used in an AAL0/AAL5 connection.
+ * Any attempt to connect a service type other than OAM on this VCI will fail.
+ */
+#define IX_ATMDACC_OAM_TX_VCI 0
+
+
+ /**
+ * @ingroup IxAtmdAccAPI
+ *
+ * @def IX_ATMDACC_OAM_RX_PORT
+ *
+ * @brief The reserved dummy PORT used for all dedicated OAM
+ * Rx connections. Note that this is not a real port but must
+ * have a value that lies within the valid range of port values.
+ */
+#define IX_ATMDACC_OAM_RX_PORT IX_UTOPIA_PORT_0
+
+ /**
+ * @ingroup IxAtmdAccAPI
+ *
+ * @def IX_ATMDACC_OAM_RX_VPI
+ *
+ * @brief The reserved value value used for the dedicated OAM
+ * Rx connection. This value should be chosen so that it does not
+ * coencide with the VPI value used in an AAL0/AAL5 connection.
+ * Any attempt to connect a service type other than OAM on this VPI will fail.
+ */
+#define IX_ATMDACC_OAM_RX_VPI 0
+
+/**
+ * @ingroup IxAtmdAccAPI
+ *
+ * @def IX_ATMDACC_OAM_RX_VCI
+ *
+ * @brief The reserved value value used for the dedicated OAM
+ * Rx connection. This value should be chosen so that it does not
+ * coencide with the VCI value used in an AAL0/AAL5 connection.
+ * Any attempt to connect a service type other than OAM on this VCI will fail.
+ */
+#define IX_ATMDACC_OAM_RX_VCI 0
+
+
+/**
+ * @enum IxAtmdAccPduStatus
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @brief IxAtmdAcc Pdu status :
+ *
+ * IxAtmdAccPduStatus is used during a RX operation to indicate
+ * the status of the received PDU
+ *
+ */
+
+typedef enum
+{
+    IX_ATMDACC_AAL0_VALID = 0,    /**< aal0 pdu */
+    IX_ATMDACC_OAM_VALID,         /**< OAM pdu */
+    IX_ATMDACC_AAL2_VALID,        /**< aal2 pdu @b reserved for future use */
+    IX_ATMDACC_AAL5_VALID,        /**< aal5 pdu complete and trailer is valid */
+    IX_ATMDACC_AAL5_PARTIAL,      /**< aal5 pdu not complete, trailer is missing */
+    IX_ATMDACC_AAL5_CRC_ERROR,    /**< aal5 pdu not complete, crc error/length error */
+    IX_ATMDACC_MBUF_RETURN        /**< empty buffer returned to the user */
+} IxAtmdAccPduStatus;
+
+
+/**
+ *
+ * @enum IxAtmdAccAalType
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @brief IxAtmdAcc AAL Service Type :
+ *
+ * IxAtmdAccAalType defines the type of traffic to run on this VC
+ *
+ */
+typedef enum
+{
+    IX_ATMDACC_AAL5,                /**< ITU-T AAL5 */
+    IX_ATMDACC_AAL2,                /**< ITU-T AAL2 @b reserved for future use */
+    IX_ATMDACC_AAL0_48,             /**< AAL0 48 byte payloads (cell header is added by NPE)*/
+    IX_ATMDACC_AAL0_52,             /**< AAL0 52 byte cell data (HEC is added by NPE) */
+    IX_ATMDACC_OAM,                 /**< OAM cell transport service (HEC is added by NPE)*/
+    IX_ATMDACC_MAX_SERVICE_TYPE     /**< not a service, used for parameter validation */
+} IxAtmdAccAalType;
+
+/**
+ *
+ * @enum IxAtmdAccClpStatus
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @brief IxAtmdAcc CLP indication
+ *
+ * IxAtmdAccClpStatus defines the CLP status of the current PDU
+ *
+ */
+typedef enum
+{
+    IX_ATMDACC_CLP_NOT_SET = 0,     /**< CLP indication is not set */
+    IX_ATMDACC_CLP_SET = 1     /**< CLP indication is set */
+} IxAtmdAccClpStatus;
+
+/**
+ * @typedef IxAtmdAccUserId
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @brief User-supplied Id
+ *
+ * IxAtmdAccUserId is passed through callbacks and allows the
+ * IxAtmdAcc user to identify the source of a call back. The range of
+ * this user-owned Id is [0...2^32-1)].
+ *
+ * The user provides this own Ids on a per-channel basis as a parameter
+ * in a call to @a ixAtmdAccRxVcConnect() or @a ixAtmdAccRxVcConnect()
+ *
+ * @sa ixAtmdAccRxVcConnect
+ * @sa ixAtmdAccTxVcConnect
+ *
+ */
+typedef unsigned int IxAtmdAccUserId;
+
+/* ------------------------------------------------------
+   Part of the IxAtmdAcc interface related to RX traffic
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @brief  Rx callback prototype
+ *
+ * IxAtmdAccRxVcRxCallback is the prototype of the Rx callback user
+ * function called once per PDU to pass a receive Pdu to a user on a
+ * partilcular connection. The callback is likely to push the mbufs
+ * to a protocol layer, and recycle the mbufs for a further use.
+ *
+ * @note -This function is called ONLY in the context of
+ *        the @a ixAtmdAccRxDispatch() function
+ *
+ * @sa ixAtmdAccRxDispatch
+ * @sa ixAtmdAccRxVcConnect
+ *
+ * @param port @ref IxAtmLogicalPort [in] - the port on which this PDU was received
+ *        a logical PHY port [@a IX_UTOPIA_PORT_0 .. @a IX_UTOPIA_MAX_PORTS - 1]
+ * @param userId @ref IxAtmdAccUserId [in] - user Id provided in the call
+ *        to @a ixAtmdAccRxVcConnect()
+ * @param status @ref IxAtmdAccPduStatus [in] - an indication about the PDU validity.
+ *        In the case of AAL0 the only possibile value is
+ *        AAL0_VALID, in this case the client may optionally determine
+ *        that an rx timeout occured by checking if the mbuf is
+ *        compleletly or only partially filled, the later case
+ *        indicating a timeout.
+ *        In the case of OAM the only possible value is OAM valid.
+ *        The status is set to @a IX_ATMDACC_MBUF_RETURN when
+ *        the mbuf is released during a disconnect process.
+ * @param clp @ref IxAtmdAccClpStatus [in] - clp indication for this PDU.
+ *        For AAL5/AAL0_48 this information
+ *        is set if the clp bit of any rx cell is set
+ *        For AAL0-52/OAM the client may inspect the CLP in individual
+ *        cell headers in the PDU, and this parameter is set to 0.
+ * @param *mbufPtr @ref IX_OSAL_MBUF [in] - depending on the servive type a pointer to
+ *        an mbuf (AAL5/AAL0/OAM) or mbuf chain (AAL5 only),
+ *        that comprises the complete PDU data.
+ *
+ *        This parameter is guaranteed not to be a null pointer.
+ *
+ */
+typedef void (*IxAtmdAccRxVcRxCallback) (IxAtmLogicalPort port,
+                       IxAtmdAccUserId userId,
+                       IxAtmdAccPduStatus status,
+                       IxAtmdAccClpStatus clp,
+                       IX_OSAL_MBUF * mbufPtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @brief Callback prototype for free buffer level is low.
+ *
+ * IxAtmdAccRxVcFreeLowCallback is the prototype of the user function
+ * which get called on a per-VC basis, when more  mbufs are needed to
+ * continue the ATM data reception. This function is likely to supply
+ * more available mbufs by one or many calls to the replenish function
+ * @a ixAtmdAccRxVcFreeReplenish()
+ *
+ * This function is called when the number of available buffers for
+ * reception is going under the threshold level as defined
+ * in @a ixAtmdAccRxVcFreeLowCallbackRegister()
+ *
+ * This function is called inside an Qmgr dispatch context. No system
+ * resource or interrupt-unsafe feature should be used inside this
+ * callback.
+ *
+ * @sa ixAtmdAccRxVcFreeLowCallbackRegister
+ * @sa IxAtmdAccRxVcFreeLowCallback
+ * @sa ixAtmdAccRxVcFreeReplenish
+ * @sa ixAtmdAccRxVcFreeEntriesQuery
+ * @sa ixAtmdAccRxVcConnect
+ *
+ * @param userId @ref IxAtmdAccUserId [in] - user Id provided in the call
+ * to @a ixAtmdAccRxVcConnect()
+ *
+ * @return None
+ *
+ */
+typedef void (*IxAtmdAccRxVcFreeLowCallback) (IxAtmdAccUserId userId);
+
+/* ------------------------------------------------------
+   Part of the IxAtmdAcc interface related to TX traffic
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @brief  Buffer callback prototype.
+ *
+ * This function is called to relinguish ownership of a transmitted
+ * buffer chain to the user.
+ *
+ * @note -In the case of a chained mbuf the AmtdAcc component can
+ * chain many user buffers together and pass ownership to the user in
+ * one function call.
+ *
+ * @param userId @ref IxAtmdAccUserId [in] - user If provided at registration of this
+ *        callback.
+ * @param mbufPtr @ref IX_OSAL_MBUF [in] - a pointer to mbufs or chain of mbufs and is
+ *        guaranteed not to be a null pointer.
+ *
+ */
+typedef void (*IxAtmdAccTxVcBufferReturnCallback) (IxAtmdAccUserId userId,
+                        IX_OSAL_MBUF * mbufPtr);
+
+/* ------------------------------------------------------
+   Part of the IxAtmdAcc interface related to Initialisation
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @fn ixAtmdAccInit (void)
+ *
+ * @brief Initialise the IxAtmdAcc Component
+ *
+ * This function initialise the IxAtmdAcc component. This function shall
+ * be called before any other function of the API. Its role is to
+ * initialise all internal resources of the IxAtmdAcc component.
+ *
+ * The ixQmgr component needs to be initialized prior the use of
+ * @a ixAtmdAccInit()
+ *
+ * @param none
+ *
+ * Failing to initilialize the IxAtmdAcc API before any use of it will
+ * result in a failed status.
+ * If the specified component is not present, a success status will still be 
+ * returned, however, a warning indicating the NPE to download to is not
+ * present will be issued.
+ *
+ * @return @li IX_SUCCESS initialisation is complete (in case of component not
+ *             being present, a warning is clearly indicated)
+ * @return @li IX_FAIL unable to process this request either
+ *             because this IxAtmdAcc is already initialised
+ *             or some unspecified error has occrred.
+ */
+PUBLIC IX_STATUS ixAtmdAccInit (void);
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @fn ixAtmdAccShow (void)
+ *
+ * @brief Show IxAtmdAcc configuration on a per port basis
+ *
+ * @param none
+ *
+ * @return none
+ *
+ * @note - Display use printf() and are redirected to stdout
+ */
+PUBLIC void
+ixAtmdAccShow (void);
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ * 
+ * @fn ixAtmdAccStatsShow (void)
+ *
+ * @brief Show all IxAtmdAcc stats
+ *
+ * @param none
+ *
+ * @return none
+ *
+ * @note - Stats display use printf() and are redirected to stdout
+ */
+PUBLIC void
+ixAtmdAccStatsShow (void);
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @fn ixAtmdAccStatsReset (void)
+ *
+ * @brief Reset all IxAtmdAcc stats
+ *
+ * @param none
+ *
+ * @return none
+ *
+ */
+PUBLIC void
+ixAtmdAccStatsReset (void);
+
+/* ------------------------------------------------------
+   Part of the IxAtmdAcc interface related to RX traffic
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @fn ixAtmdAccRxVcConnect (IxAtmLogicalPort port,
+                      unsigned int vpi,
+                      unsigned int vci,
+                      IxAtmdAccAalType aalServiceType,
+                      IxAtmRxQueueId rxQueueId,
+                      IxAtmdAccUserId userCallbackId,
+                      IxAtmdAccRxVcRxCallback rxCallback,
+                      unsigned int minimumReplenishCount,
+                      IxAtmConnId * connIdPtr,
+                      IxAtmNpeRxVcId * npeVcIdPtr )
+ *
+ * @brief Connect to a Aal Pdu receive service for a particular
+ *        port/vpi/vci, and service type.
+ *
+ * This function allows a user to connect to an Aal5/Aal0/OAM Pdu receive service
+ * for a particular port/vpi/vci. It registers the callback and allocates
+ * internal resources and a Connection Id to be used in further API calls
+ * related to this VCC.
+ *
+ * The function will setup VC receive service on the specified rx queue.
+ *
+ * This function is blocking and makes use internal locks, and hence
+ * should not be called from an interrupt context.
+ *
+ * On return from @a ixAtmdAccRxVcConnect() with a failure status, the
+ * connection Id parameter is unspecified. Its value cannot be used.
+ * A connId is the reference by which IxAtmdAcc refers to a
+ * connected VC. This identifier is the result of a succesful call
+ * to a connect function. This identifier is invalid after a
+ * sucessful call to a disconnect function.
+ *
+ * Calling this function for the same combination of Vpi, Vci and more
+ * than once without calling @a ixAtmdAccRxVcTryDisconnect() will result in a
+ * failure status.
+ *
+ * If this function returns success the user should supply receive
+ * buffers by calling @a ixAtmdAccRxVcFreeReplenish() and then call
+ * @a ixAtmdAccRxVcEnable() to begin receiving pdus.
+ *
+ * There is a choice of two receive Qs on which the VC pdus could be
+ * receive. The user must associate the VC with one of these. Essentially
+ * having two qs allows more flexible system configuration such as have
+ * high prioriy traffic on one q (e.g. voice) and low priority traffic on
+ * the other (e.g. data). The high priority Q could be serviced in
+ * preference to the low priority Q. One queue may be configured to be
+ * serviced as soon as there is traffic, the other queue may be configured
+ * to be serviced by a polling mechanism running at idle time.
+ *
+ * Two AAL0 services supporting 48 or 52 byte cell data are provided.
+ * Received AAL0 PDUs will be be a multiple of the cell data size (48/52).
+ * AAL0_52 is a raw cell service and includes an ATM cell header
+ * (excluding HEC) at the start of each cell.
+ *
+ * A single "OAM Rx channel" for all ports may be  enabled by
+ * establishing a dedicated OAM Rx connection.
+ *
+ * The OAM service allows buffers containing 52 byte OAM F4/F5 cells
+ * to be transmitted and received over the dedicated OAM channels.
+ * HEC is appended/removed, and CRC-10 performed by the NPE. The OAM
+ * service offered by AtmdAcc is a raw cell transport service.
+ * It is assumed that ITU I.610 procedures that make use of this
+ * service are implemented above AtmdAcc.
+ *
+ * Note that the dedicated OAM connections are established on
+ * reserved VPI,VCI, and (in the case of Rx) port values.
+ * These values are used purely to descriminate the dedicated OAM channels
+ * and do not identify a particular OAM F4/F5 flow. F4/F5 flows may be
+ * realised for particluar VPI/VCIs by manipulating the VPI,VCI
+ * fields of the ATM cell headers of cells in the buffers passed
+ * to AtmdAcc.
+ *
+ * Calling this function prior to enable the port will fail.
+ *
+ * @sa ixAtmdAccRxDispatch
+ * @sa ixAtmdAccRxVcEnable
+ * @sa ixAtmdAccRxVcDisable
+ * @sa ixAtmdAccRxVcTryDisconnect
+ * @sa ixAtmdAccPortEnable
+ *
+ * @param port @ref IxAtmLogicalPort [in] - VC identification : logical PHY port
+ *                  [@a IX_UTOPIA_PORT_0 .. @a IX_UTOPIA_MAX_PORTS - 1]
+ * @param vpi unsigned int [in] - VC identification : ATM Vpi [0..255] or IX_ATMDACC_OAM_VPI
+ * @param vci unsigned int [in] - VC identification : ATM Vci [0..65535] or IX_ATMDACC_OAM_VCI
+ * @param aalServiceType @ref IxAtmdAccAalType [in] - type of service: AAL5, AAL0_48, AAL0_52, or OAM
+ * @param rxQueueId @ref IxAtmRxQueueId [in] - this identifieds which of two Qs the VC
+ *     should use.when icoming traffic is processed
+ * @param userCallbackId @ref IxAtmdAccUserId [in] - user Id used later as a parameter to
+ *     the supplied rxCallback.
+ * @param rxCallback [in] @ref IxAtmdAccRxVxRxCallback - function called when mbufs are received.
+ *     This parameter cannot be a null pointer.
+ * @param bufferFreeCallback [in] - function to be called to return
+ *     ownership of buffers to IxAtmdAcc user.
+ * @param minimumReplenishCount unsigned int [in] -  For AAL5/AAL0 the number of free mbufs
+ *     to be used with this channel. Use a high number when the expected traffic
+ *     rate on this channel is high, or when the user's mbufs are small, or when
+ *     the RxVcFreeLow Notification has to be invoked less often. When this
+ *     value is IX_ATMDACC_DEFAULT_REPLENISH_COUNT, the minimum of
+ *     resources  will be used. Depending on traffic rate, pdu
+ *     size and mbuf size, rxfree queue size, polling/interrupt rate, this value may
+ *     require to be replaced by a different value in the range 1-128
+ *     For OAM the rxFree queue size is fixed by atmdAcc and this parameter is ignored.
+ * @param connIdPtr @ref IxAtmConnId [out] - pointer to a connection Id
+ *     This parameter cannot be a null pointer.
+ * @param npeVcIdPtr @ref IxAtmNpeRxVcId [out] - pointer to an npe Vc Id
+ *     This parameter cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS successful call to IxAtmdAccRxVcConnect
+ * @return @li IX_ATMDACC_BUSY cannot process this request :
+ *         no VC is available
+ * @return @li IX_FAIL
+ *             parameter error,
+ *             VC already in use,
+ *             attempt to connect AAL service on reserved OAM VPI/VCI,
+ *             attempt to connect OAM service on VPI/VCI other than the reserved OAM VPI/VCI,
+ *             port is not initialised,
+ *             or some other error occurs during processing.
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccRxVcConnect (IxAtmLogicalPort port,
+                      unsigned int vpi,
+                      unsigned int vci,
+                      IxAtmdAccAalType aalServiceType,
+                      IxAtmRxQueueId rxQueueId,
+                      IxAtmdAccUserId userCallbackId,
+                      IxAtmdAccRxVcRxCallback rxCallback,
+                      unsigned int minimumReplenishCount,
+                      IxAtmConnId * connIdPtr,
+                      IxAtmNpeRxVcId * npeVcIdPtr );
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ * 
+ * @fn ixAtmdAccRxVcFreeReplenish (IxAtmConnId connId,
+                        IX_OSAL_MBUF * mbufPtr)
+ *
+ * @brief Provide free mbufs for data reception on a connection.
+ *
+ * This function provides mbufs for data reception by the hardware. This
+ * function needs to be called by the user on a regular basis to ensure
+ * no packet loss. Providing free buffers is a connection-based feature;
+ * each connection can have different requirements in terms of buffer size
+ * number of buffers, recycling rate. This function could be invoked from
+ * within the context of a @a IxAtmdAccRxVcFreeLowCallback() callback
+ * for a particular VC
+ *
+ * Mbufs provided through this function call can be chained. They will be
+ * unchained internally. A call to this function with chained mbufs or
+ * multiple calls with unchained mbufs are equivalent, but calls with
+ * unchained mbufs are more efficients.
+ *
+ * Mbufs provided to this interface need to be able to hold at least one
+ * full cell payload (48/52 bytes, depending on service type).
+ * Chained buffers with a size less than the size supported by the hardware
+ * will be returned through the rx callback provided during the connect step.
+ *
+ * Failing to invoke this function prior to enabling the RX traffic
+ * can result in packet loss.
+ *
+ * This function is not reentrant for the same connId.
+ *
+ * This function does not use system resources and can be
+ * invoked from an interrupt context.
+ *
+ * @note - Over replenish is detected, and extra mbufs are returned through
+ *         the rx callback provided during the connect step.
+ *
+ * @note - Mbuf provided to the replenish function should have a length greater or
+ *         equal to 48/52 bytes according to service type.
+ *
+ * @note - The memory cache of mMbuf payload should be invalidated prior to Mbuf
+ *         submission. Flushing the Mbuf headers is handled by IxAtmdAcc.
+ *
+ * @note - When a chained mbuf is provided, this function process the mbufs
+ *         up to the hardware limit and invokes the user-supplied callback
+ *         to release extra buffers.
+ *
+ * @sa ixAtmdAccRxVcFreeLowCallbackRegister
+ * @sa IxAtmdAccRxVcFreeLowCallback
+ * @sa ixAtmdAccRxVcConnect
+ *
+ * @param connId @ref IxAtmConnId [in] - connection Id as returned from a succesfull call to
+ *        @a IxAtmdAccRxVcConnect()
+ * @param mbufPtr @ref IX_OSAL_MBUF [in] - pointer to a mbuf structure to be used for data
+ *        reception. The mbuf pointed to by this parameter can be chained
+ *        to an other mbuf.
+ *
+ * @return @li IX_SUCCESS successful call to @a ixAtmdAccRxVcFreeReplenish()
+ *          and the mbuf is now ready to use for incoming traffic.
+ * @return @li IX_ATMDACC_BUSY cannot process this request because
+ *         the max number of outstanding free buffers has been reached
+ *         or the internal resources have exhausted for this VC.
+ *         The user is responsible for retrying this request later.
+ * @return @li IX_FAIL cannot process this request because of parameter
+ *         errors or some unspecified internal error has occurred.
+ *
+ * @note - It is not always guaranteed the replenish step to be as fast as the
+ *   hardware is consuming Rx Free mbufs. There is nothing in IxAtmdAcc to
+ *   guarantee that replenish reaches the rxFree threshold level. If the
+ *   threshold level is not reached, the next rxFree low notification for
+ *   this channel will not be triggered.
+ *   The preferred ways to replenish can be as follows (depending on
+ *   applications and implementations) :
+ *   @li Replenish in a rxFree low notification until the function
+ *       ixAtmdAccRxVcFreeReplenish() returns IX_ATMDACC_BUSY
+ *   @li Query the queue level using @sa ixAtmdAccRxVcFreeEntriesQuery, then
+ *     , replenish using @a ixAtmdAccRxVcFreeReplenish(), then query the queue
+ *       level again, and replenish if the threshold is still not reached.
+ *   @li Trigger replenish from an other event source and use rxFree starvation
+ *       to throttle the Rx traffic.
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccRxVcFreeReplenish (IxAtmConnId connId,
+                        IX_OSAL_MBUF * mbufPtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ * 
+ * @fn ixAtmdAccRxVcFreeLowCallbackRegister (IxAtmConnId connId,
+                                    unsigned int numberOfMbufs,
+                                    IxAtmdAccRxVcFreeLowCallback callback)
+ *
+ * @brief Configure the RX Free threshold value and register a callback
+ * to handle threshold notifications.
+ *
+ * The function ixAtmdAccRxVcFreeLowCallbackRegister sets the threshold value for
+ * a particular RX VC. When the number of buffers reaches this threshold
+ * the callback is invoked.
+ *
+ * This function should be called once per VC before RX traffic is
+ * enabled.This function will fail if the curent level of the free buffers
+ * is equal or less than the threshold value.
+ *
+ * @sa ixAtmdAccRxVcFreeLowCallbackRegister
+ * @sa IxAtmdAccRxVcFreeLowCallback
+ * @sa ixAtmdAccRxVcFreeReplenish
+ * @sa ixAtmdAccRxVcFreeEntriesQuery
+ * @sa ixAtmdAccRxVcConnect
+ *
+ * @param connId @ref IxAtmConnId [in] - connection Id as resulted from a succesfull call
+ *        to @a IxAtmdAccRxVcConnect()
+ * @param numberOfMbufs unsigned int [in] - threshold number of buffers. This number
+ *        has to be a power of 2, one of the values 0,1,2,4,8,16,32....
+ *        The maximum value cannot be more than half of the rxFree queue
+ *        size (which can be retrieved using @a ixAtmdAccRxVcFreeEntriesQuery()
+ *        before any use of the @a ixAtmdAccRxVcFreeReplenish() function)
+ * @param callback @ref IxAtmdAccRxVcFreeLowCallback [in] - function telling the user that the number of
+ *        free buffers has reduced to the threshold value.
+ *
+ * @return @li IX_SUCCESS Threshold set successfully.
+ * @return @li IX_FAIL parameter error or the current number of free buffers
+ *              is less than or equal to the threshold supplied or some
+ *              unspecified error has occrred.
+ *
+ * @note - the callback will be called when the threshold level will drop from
+ *        exactly (numberOfMbufs + 1) to (numberOfMbufs).
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccRxVcFreeLowCallbackRegister (IxAtmConnId connId,
+    unsigned int numberOfMbufs,
+    IxAtmdAccRxVcFreeLowCallback callback);
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @fn ixAtmdAccRxVcFreeEntriesQuery (IxAtmConnId connId,
+                         unsigned int *numberOfMbufsPtr)
+ *
+ * @brief Get the number of rx mbufs the system can accept to replenish the
+ *       the rx reception mechanism on a particular channel
+ *
+ * The ixAtmdAccRxVcFreeEntriesQuery function is used to retrieve the current
+ * number of available mbuf entries for reception, on a per-VC basis. This
+ * function can be used to know the number of mbufs which can be provided
+ * using @a ixAtmdAccRxVcFreeReplenish().
+ *
+ * This function can be used from a timer context, or can be associated
+ * with a threshold event, or can be used inside an active polling
+ * mechanism which is under user control.
+ *
+ * This function is reentrant and does not use system resources and can
+ * be invoked from an interrupt context.
+ *
+ * @param connId @ref IxAtmConnId [in] - connection Id as resulted from a succesfull call
+ *        to @a IxAtmdAccRxVcConnect()
+ * @param numberOfMbufsPtr unsigned int [out] - Pointer to the number of available entries.
+ *      . This parameter cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS the current number of mbufs not yet used for incoming traffic
+ * @return @li IX_FAIL invalid parameter
+ *
+ * @sa ixAtmdAccRxVcFreeReplenish
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccRxVcFreeEntriesQuery (IxAtmConnId connId,
+                         unsigned int *numberOfMbufsPtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @fn ixAtmdAccRxVcEnable (IxAtmConnId connId)
+ *
+ * @brief Start the RX service on a VC.
+ *
+ * This functions kicks-off the traffic reception for a particular VC.
+ * Once invoked, incoming PDUs will be made available by the hardware
+ * and are eventually directed to the @a IxAtmdAccRxVcRxCallback() callback
+ * registered for the connection.
+ *
+ * If the traffic is already running, this function returns IX_SUCCESS.
+ * This function can be invoked many times.
+ *
+ * IxAtmdAccRxVcFreeLowCallback event will occur only after
+ * @a ixAtmdAccRxVcEnable() function is invoked.
+ *
+ * Before using this function, the @a ixAtmdAccRxVcFreeReplenish() function
+ * has to be used to replenish the RX Free queue. If not, incoming traffic
+ * may be discarded.and in the case of interrupt driven reception the
+ * @a IxAtmdAccRxVcFreeLowCallback() callback may be invoked as a side effect
+ * during a replenish action.
+ *
+ * This function is not reentrant and should not be used inside an
+ * interrupt context.
+ *
+ * For an VC connection this function can be called after a call to
+ * @a ixAtmdAccRxVcDisable() and should not be called after
+ * @a ixAtmdAccRxVcTryDisconnect()
+ *
+ * @sa ixAtmdAccRxVcDisable
+ * @sa ixAtmdAccRxVcConnect
+ * @sa ixAtmdAccRxVcFreeReplenish
+ *
+ * @param connId @ref IxAtmConnId [in] - connection Id as resulted from a succesfull call
+ * to @a IxAtmdAccRxVcConnect()
+ *
+ * @return @li IX_SUCCESS successful call to ixAtmdAccRxVcEnable
+ * @return @li IX_ATMDACC_WARNING the channel is already enabled
+ * @return @li IX_FAIL invalid parameters or some unspecified internal
+ *         error occured.
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccRxVcEnable (IxAtmConnId connId);
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ * 
+ * @fn ixAtmdAccRxVcDisable (IxAtmConnId connId)
+ *
+ * @brief Stop the RX service on a VC.
+ *
+ * This functions stops the traffic reception for a particular VC connection.
+ *
+ * Once invoked, incoming Pdus are discarded by the hardware. Any Pdus
+ * pending will be freed to the user
+ *
+ * Hence once this function returns no more receive callbacks will be
+ * called for that VC. However, buffer free callbacks will be invoked
+ * until such time as all buffers supplied by the user have been freed
+ * back to the user
+ *
+ * Calling this function doe not invalidate the connId.
+ * @a ixAtmdAccRxVcEnable() can be invoked to enable Pdu reception again.
+ *
+ * If the traffic is already stopped, this function returns IX_SUCCESS.
+ *
+ * This function is not reentrant and should not be used inside an
+ * interrupt context.
+ *
+ * @sa ixAtmdAccRxVcConnect
+ * @sa ixAtmdAccRxVcEnable
+ * @sa ixAtmdAccRxVcDisable
+ *
+ * @param connId @ref IxAtmConnId [in] - connection Id as resulted from a succesfull call to @a
+ *     IxAtmdAccRxVcConnect()
+ *
+ * @return @li IX_SUCCESS successful call to @a ixAtmdAccRxVcDisable().
+ * @return @li IX_ATMDACC_WARNING the channel is already disabled
+ * @return @li IX_FAIL invalid parameters or some unspecified internal error occured
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccRxVcDisable (IxAtmConnId connId);
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ * 
+ * @fn ixAtmdAccRxVcTryDisconnect (IxAtmConnId connId)
+ *
+ * @brief Disconnect a VC from the RX service.
+ *
+ * This function deregisters the VC and guarantees that all resources
+ * associated with this VC are free. After its execution, the connection
+ * Id is not available.
+ *
+ * This function will fail until such time as all resources allocated to
+ * the VC connection have been freed. The user is responsible to delay and
+ * call again this function many times until a success status is returned.
+ *
+ * This function needs internal locks and should not be called from an
+ * interrupt context
+ *
+ * @param connId @ref IxAtmConnId [in] - connection Id as resulted from a succesfull call to
+ * @a IxAtmdAccRxVcConnect()
+ *
+ * @return @li IX_SUCCESS successful call to ixAtmdAccRxVcDisable
+ * @return @li IX_ATMDACC_RESOURCES_STILL_ALLOCATED not all resources
+ *         associated with the connection have been freed.
+ * @return @li IX_FAIL cannot process this request because of a parameter
+ *         error
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccRxVcTryDisconnect (IxAtmConnId connId);
+
+/* ------------------------------------------------------
+   Part of the IxAtmdAcc interface related to TX traffic
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ * 
+ * @fn ixAtmdAccTxVcConnect (IxAtmLogicalPort port,
+                      unsigned int vpi,
+                      unsigned int vci,
+                      IxAtmdAccAalType aalServiceType,
+                      IxAtmdAccUserId userId,
+                      IxAtmdAccTxVcBufferReturnCallback bufferFreeCallback,
+                      IxAtmConnId * connIdPtr)
+ *
+ * @brief Connect to a Aal Pdu transmit service for a particular
+ *        port/vpi/vci and service type.
+ *
+ * This function allows a user to connect to an Aal5/Aal0/OAM Pdu transmit service
+ * for a particular port/vpi/vci. It registers the callback and allocates
+ * internal resources and a Connection Id to be used in further API calls
+ * related to this VC.
+ *
+ * The function will setup VC transmit service on the specified on the
+ * specified port. A connId is the reference by which IxAtmdAcc refers to a
+ * connected VC. This identifier is the result of a succesful call
+ * to a connect function. This identifier is invalid after a
+ * sucessful call to a disconnect function.
+ *
+ * This function needs internal locks, and hence  should not be called
+ * from an interrupt context.
+ *
+ * On return from @a ixAtmdAccTxVcConnect() with a failure status, the
+ * connection Id parameter is unspecified. Its value cannot be used.
+ *
+ * Calling this function for the same combination of port, Vpi, Vci and
+ * more than once without calling @a ixAtmdAccTxVcTryDisconnect() will result
+ * in a failure status.
+ *
+ * Two AAL0 services supporting 48 or 52 byte cell data are provided.
+ * Submitted AAL0 PDUs must be a multiple of the cell data size (48/52).
+ * AAL0_52 is a raw cell service the client must format
+ * the PDU with an ATM cell header (excluding HEC) at the start of
+ * each cell, note that AtmdAcc does not validate the cell headers in
+ * a submitted PDU.
+ *
+ * For the OAM service an "OAM Tx channel" may be enabled for a port
+ * by establishing a single dedicated OAM Tx connection on that port.
+ *
+ * The OAM service allows buffers containing 52 byte OAM F4/F5 cells
+ * to be transmitted and received over the dedicated OAM channels.
+ * HEC is appended/removed, and CRC-10 performed by the NPE. The OAM
+ * service offered by AtmdAcc is a raw cell transport service.
+ * It is assumed that ITU I.610 procedures that make use of this
+ * service are implemented above AtmdAcc.
+ *
+ * Note that the dedicated OAM connections are established on
+ * reserved VPI,VCI, and (in the case of Rx) port values.
+ * These values are used purely to descriminate the dedicated OAM channels
+ * and do not identify a particular OAM F4/F5 flow. F4/F5 flows may be
+ * realised for particluar VPI/VCIs by manipulating the VPI,VCI
+ * fields of the ATM cell headers of cells in the buffers passed
+ * to AtmdAcc.
+ *
+ * Calling this function before enabling the port will fail.
+ *
+ * @sa ixAtmdAccTxVcTryDisconnect
+ * @sa ixAtmdAccPortTxScheduledModeEnable
+ * @sa ixAtmdAccPortEnable
+ *
+ * @param port @ref IxAtmLogicalPort [in] - VC identification : logical PHY port
+ *                  [@a IX_UTOPIA_PORT_0 .. @a IX_UTOPIA_MAX_PORTS - 1]
+ * @param vpi unsigned int  [in] - VC identification : ATM Vpi [0..255] or IX_ATMDACC_OAM_VPI
+ * @param vci unsigned int [in] - VC identification : ATM Vci [0..65535] or IX_ATMDACC_OAM_VCI
+ * @param aalServiceType @ref IxAtmdAccAalType [in] - type of service AAL5, AAL0_48, AAL0_52, or OAM
+ * @param userId @ref IxAtmdAccUserId [in] - user id to be used later during callbacks related
+ *        to this channel
+ * @param bufferFreeCallback @ref IxAtmdAccTxVcBufferReturnCallback [in] - function called when mbufs
+ *        transmission is complete. This parameter cannot be a null
+ *        pointer.
+ * @param connIdPtr @ref IxAtmConnId [out] - Pointer to a connection Id.
+ *        This parameter cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS successful call to @a IxAtmdAccRxVcConnect().
+ * @return @li IX_ATMDACC_BUSY cannot process this request
+ *         because no VC is available
+ * @return @li IX_FAIL
+ *             parameter error,
+ *             VC already in use,
+ *             attempt to connect AAL service on reserved OAM VPI/VCI,
+ *             attempt to connect OAM service on VPI/VCI other than the reserved OAM VPI/VCI,
+ *             port is not initialised,
+ *             or some other error occurs during processing.
+ *
+ * @note - Unscheduled mode is not supported in ixp425 1.0. Therefore, the
+ *       function @a ixAtmdAccPortTxScheduledModeEnable() need to be called
+ *       for this port before any establishing a Tx Connection
+ */
+PUBLIC IX_STATUS ixAtmdAccTxVcConnect (IxAtmLogicalPort port,
+                      unsigned int vpi,
+                      unsigned int vci,
+                      IxAtmdAccAalType aalServiceType,
+                      IxAtmdAccUserId userId,
+                      IxAtmdAccTxVcBufferReturnCallback bufferFreeCallback,
+                      IxAtmConnId * connIdPtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @fn ixAtmdAccTxVcPduSubmit (IxAtmConnId connId,
+                    IX_OSAL_MBUF * mbufPtr,
+                    IxAtmdAccClpStatus clp,
+                    unsigned int numberOfCells)
+ *
+ * @brief Submit a Pdu for transmission on connection.
+ *
+ * A data user calls this function to submit an mbufs containing a Pdu
+ * to be transmitted. The buffer supplied can be chained and the Pdu it
+ * contains must be complete.
+ *
+ * The transmission behavior of this call depends on the operational mode
+ * of the port on which the connection is made.
+ *
+ * In unscheduled mode the mbuf will be submitted to the hardware
+ * immediately if sufficent resource is available. Otherwise the function
+ * will return failure.
+ *
+ * In scheduled mode the buffer is queued internally in IxAtmdAcc. The cell
+ * demand is made known to the traffic shaping entity. Cells from the
+ * buffers are MUXed onto the port some time later as dictated by the
+ * traffic shaping entity. The traffic shaping entity does this by sending
+ * transmit schedules to IxAtmdAcc via @a ixAtmdAccPortTxProcess() function call.
+ *
+ * Note that the dedicated OAM channel is scheduled just like any
+ * other channel. This means that any OAM traffic relating to an
+ * active AAL0/AAL5 connection will be scheduled independantly of the
+ * AAL0/AAL5 traffic for that connection.
+ *
+ * When transmission is complete, the TX Done mechanism will give the
+ * owmnership of these buffers back to the customer. The tx done mechanism
+ * must be in operation before transmission is attempted.
+ *
+ * For AAL0/OAM submitted AAL0 PDUs must be a multiple of the cell data
+ * size (48/52). AAL0_52 and OAM are raw cell services, and the client
+ * must format the PDU with an ATM cell header (excluding HEC) at the
+ * start of each cell, note that AtmdAcc does not validate the cell headers in
+ * a submitted PDU.
+ *
+ *
+ * @sa IxAtmdAccTxVcBufferReturnCallback
+ * @sa ixAtmdAccTxDoneDispatch
+ *
+ * @param connId @ref IxAtmConnId [in] - connection Id as resulted from a succesfull call to
+ *        @a ixAtmdAccTxVcConnect()
+ * @param mbufPtr @ref IX_OSAL_MBUF [in] - pointer to a chained structure of mbufs to transmit.
+ *       This parameter cannot be a null pointer.
+ * @param clp @ref IxAtmdAccClpStatus [in] - clp indication for this PDU. All cells of this pdu
+ *       will be sent with the clp bit set
+ * @param numberOfCells unsigned int [in] - number of cells in the PDU.
+ *
+ * @return @li IX_SUCCESS successful call to @a ixAtmdAccTxVcPduSubmit()
+ *             The pdu pointed by the mbufPtr parameter will be
+ *             transmitted
+ * @return @li IX_ATMDACC_BUSY unable to process this request because
+ *             internal resources are all used. The caller is responsible
+ *             for retrying this request later.
+ * @return @li IX_FAIL unable to process this request because of error
+ *             in the parameters (wrong connId supplied,
+ *             or wrong mbuf pointer supplied), the total length of all buffers
+ *             in the chain should be a multiple of the cell size
+ *             ( 48/52 depending on the service type ),
+ *             or unspecified error during processing
+ *
+ * @note - This function in not re-entrant for the same VC (e.g. : two
+ *         thread cannot send PDUs for the same VC). But two threads can
+ *         safely call this function with a different connection Id
+ *
+ * @note - In unscheduled mode, this function is not re-entrant on a per
+ *         port basis. The size of pdus is limited to 8Kb.
+ *
+ * @note - 0-length mbufs should be removed from the chain before submission.
+ *         The total length of the pdu (sdu + padding +trailer) has to be
+ *         updated in the header of the first mbuf of a chain of mbufs.
+ *
+ * @note - Aal5 trailer information (UUI, CPI, SDU length) has to be supplied
+ *         before submission.
+ *
+ * @note - The payload memory cache should be flushed, if needed, prior to
+ *         transmission. Mbuf headers are flushed by IxAtmdAcc
+ *
+ * @note - This function does not use system resources and can be used
+ *         inside an interrupt context
+ */
+PUBLIC IX_STATUS ixAtmdAccTxVcPduSubmit (IxAtmConnId connId,
+                    IX_OSAL_MBUF * mbufPtr,
+                    IxAtmdAccClpStatus clp,
+                    unsigned int numberOfCells);
+
+/**
+ *
+ * @ingroup IxAtmdAccAPI
+ *
+ * @fn ixAtmdAccTxVcTryDisconnect (IxAtmConnId connId)
+ *
+ * @brief Disconnect from a Aal Pdu transmit service for a particular
+ *        port/vpi/vci.
+ *
+ * This function deregisters the VC and guarantees that all resources
+ * associated with this VC are free. After its execution, the connection
+ * Id is not available.
+ *
+ * This function will fail until such time as all resources allocated to
+ * the VC connection have been freed. The user is responsible to delay
+ * and call again this function many times until a success status is
+ * returned.
+ *
+ * After its execution, the connection Id is not available.
+ *
+ * @param connId @ref IxAtmConnId [in] - connection Id as resulted from a succesfull call to
+ *        @a ixAtmdAccTxVcConnect()
+ *
+ * @return @li IX_SUCCESS successful call to @a ixAtmdAccTxVcTryDisconnect()
+ * @return @li IX_ATMDACC_RESOURCES_STILL_ALLOCATED not all resources
+ *     associated with the connection have been freed. This condition will
+ *     disappear after Tx and TxDone is complete for this channel.
+ * @return @li IX_FAIL unable to process this request because of errors
+ *                     in the parameters (wrong connId supplied)
+ *
+ * @note - This function needs internal locks and should not be called
+ *         from an interrupt context
+ *
+ * @note - If the @a IX_ATMDACC_RESOURCES_STILL_ALLOCATED error does not
+ *     clear after a while, this may be linked to a previous problem
+ *     of cell overscheduling. Diabling the port and retry a disconnect
+ *     will free the resources associated with this channel.
+ *
+ * @sa ixAtmdAccPortTxProcess
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccTxVcTryDisconnect (IxAtmConnId connId);
+
+#endif /* IXATMDACC_H */
+
+/**
+ * @} defgroup IxAtmdAccAPI
+ */
+
+

cpu/ixp/npe/include/IxAtmdAccCtrl.h

@@ -0,0 +1,1958 @@
+
+/**
+ * @file    IxAtmdAccCtrl.h
+ *
+ * @date    20-Mar-2002
+ *
+ * @brief IxAtmdAcc Public API
+ *
+ * This file contains the public API of IxAtmdAcc, related to the
+ * control functions of the component.
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+/* ------------------------------------------------------
+   Doxygen group definitions
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @defgroup IxAtmdAccCtrlAPI IXP400 ATM Driver Access (IxAtmdAcc) Control API
+ *
+ * @brief The public API for the IXP400 Atm Driver Control component
+ *
+ * IxAtmdAcc is the low level interface by which AAL PDU get transmitted
+ * to,and received from the Utopia bus
+ *
+ * This part is related to the Control configuration
+ *
+ * @{
+ */
+
+#ifndef IXATMDACCCTRL_H
+#define IXATMDACCCTRL_H
+
+#include "IxAtmdAcc.h"
+
+/* ------------------------------------------------------
+   AtmdAccCtrl Data Types definition
+   ------------------------------------------------------ */
+
+/**
+*
+* @ingroup IxAtmdAccCtrlAPI
+*
+* @def IX_ATMDACC_PORT_DISABLE_IN_PROGRESS
+*
+* @brief Port enable return code
+*
+* This constant is used to tell IxAtmDAcc user that the port disable
+* functions are not complete. The user can call ixAtmdAccPortDisableComplete()
+* to find out when the disable has finished. The port enable can then proceed.
+*
+*/
+#define IX_ATMDACC_PORT_DISABLE_IN_PROGRESS 5
+
+/**
+*
+* @ingroup IxAtmdAccCtrlAPI
+*
+* @def IX_ATMDACC_ALLPDUS
+*
+* @brief All PDUs
+*
+* This constant is used to tell IxAtmDAcc to process all PDUs from
+* the RX queue or the TX Done
+*
+* @sa IxAtmdAccRxDispatcher
+* @sa IxAtmdAccTxDoneDispatcher
+*
+*/
+#define IX_ATMDACC_ALLPDUS 0xffffffff
+
+/* ------------------------------------------------------
+   Part of the IxAtmdAcc interface related to RX traffic
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @brief Callback prototype for notification of available PDUs for
+ * an Rx Q.
+ *
+ * This a protoype for a function which is called when there is at
+ * least one Pdu available for processing on a particular Rx Q.
+ *
+ * This function should call @a ixAtmdAccRxDispatch() with
+ * the aprropriate number of parameters to read and process the Rx Q.
+ *
+ * @sa ixAtmdAccRxDispatch
+ * @sa ixAtmdAccRxVcConnect
+ * @sa ixAtmdAccRxDispatcherRegister
+ *
+ * @param rxQueueId @ref IxAtmRxQueueId [in] indicates which RX queue to has Pdus to process.
+ * @param numberOfPdusToProcess unsigned int [in] indicates the minimum number of
+ *        PDUs available to process all PDUs from the queue.
+ * @param reservedPtr unsigned int* [out] pointer to a int location which can
+ *        be written to, but does not retain written values. This is
+ *        provided to make this prototype compatible
+ *        with @a ixAtmdAccRxDispatch()
+ *
+ * @return @li int - ignored.
+ *
+ */
+typedef IX_STATUS (*IxAtmdAccRxDispatcher) (IxAtmRxQueueId rxQueueId,
+                         unsigned int numberOfPdusToProcess,
+                         unsigned int *reservedPtr);
+
+/* ------------------------------------------------------
+   Part of the IxAtmdAcc interface related to TX traffic
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @brief Callback prototype for transmitted mbuf when threshold level is
+ *        crossed.
+ *
+ * IxAtmdAccTxDoneDispatcher is the prototype of the user function
+ * which get called when pdus are completely transmitted. This function
+ * is likely to call the @a ixAtmdAccTxDoneDispatch() function.
+ *
+ * This function is called when the number of available pdus for
+ * reception is crossing the threshold level as defined
+ * in @a ixAtmdAccTxDoneDispatcherRegister()
+ *
+ * This function is called inside an Qmgr dispatch context. No system
+ * resource or interrupt-unsafe feature should be used inside this
+ * callback.
+ *
+ * Transmitted buffers recycling implementation is a sytem-wide mechanism
+ * and needs to be set before any traffic is started. If this threshold
+ * mechanism is not used, the user is responsible for polling the
+ * transmitted buffers with @a ixAtmdAccTxDoneDispatch()
+ * and @a ixAtmdAccTxDoneLevelQuery() functions.
+ *
+ * @sa ixAtmdAccTxDoneDispatcherRegister
+ * @sa ixAtmdAccTxDoneDispatch
+ * @sa ixAtmdAccTxDoneLevelQuery
+ *
+ * @param numberOfPdusToProcess unsigned int [in] - The current number of pdus currently
+ *        available for recycling
+ * @param *reservedPtr unsigned int [out] - pointer to a int location which can be
+ *        written to but does not retain written values. This is provided
+ *        to make this prototype compatible
+ *        with @a ixAtmdAccTxDoneDispatch()
+ *
+ * @return @li IX_SUCCESS This is provided to make
+ *    this prototype compatible with @a ixAtmdAccTxDoneDispatch()
+ * @return @li IX_FAIL invalid parameters or some unspecified internal
+ *    error occured. This is provided to make
+ *    this prototype compatible with @a ixAtmdAccTxDoneDispatch()
+ *
+ */
+typedef IX_STATUS (*IxAtmdAccTxDoneDispatcher) (unsigned int numberOfPdusToProcess,
+                                                unsigned int *reservedPtr);
+
+/**
+*
+* @ingroup IxAtmdAccCtrlAPI
+*
+* @brief Notification that the threshold number of scheduled cells
+* remains in a port's transmit Q.
+*
+* The is the prototype for of the user notification function which
+* gets called on a per-port basis, when the number of remaining
+* scheduled cells to be transmitted decreases to the threshold level.
+* The number of cells passed as a parameter can be used for scheduling
+* purposes as the maximum number of cells that can be passed in a
+* schedule table to the @a ixAtmdAccPortTxProcess() function.
+*
+* @sa ixAtmdAccPortTxCallbackRegister
+* @sa ixAtmdAccPortTxProcess
+* @sa ixAtmdAccPortTxFreeEntriesQuery
+*
+* @param port @ref IxAtmLogicalPort [in] - logical PHY port [@a IX_UTOPIA_PORT_0 .. @a IX_UTOPIA_MAX_PORTS - 1]
+* @param numberOfAvailableCells unsigned int [in] - number of available
+*        cell entries.for the port
+*
+* @note - This functions shall not use system resources when used
+*         inside an interrupt context.
+*
+*/
+typedef void (*IxAtmdAccPortTxLowCallback) (IxAtmLogicalPort port,
+                           unsigned int numberOfAvailableCells);
+
+/**
+*
+* @ingroup IxAtmdAccCtrlAPI
+*
+* @brief  Prototype to submit cells for transmission
+*
+* IxAtmdAccTxVcDemandUpdateCallback is the prototype of the callback
+* function used by AtmD to notify an ATM Scheduler that the user of
+* a VC has submitted cells for transmission.
+*
+* @sa IxAtmdAccTxVcDemandUpdateCallback
+* @sa IxAtmdAccTxVcDemandClearCallback
+* @sa IxAtmdAccTxSchVcIdGetCallback
+* @sa ixAtmdAccPortTxScheduledModeEnable
+*
+* @param port @ref IxAtmLogicalPort [in] - Specifies the ATM port on which the VC to be updated
+*        is established
+* @param vcId int [in] - Identifies the VC to be updated. This is the value
+*        returned by the @a IxAtmdAccTxSchVcIdGetCallback() call .
+* @param numberOfCells unsigned int [in] - Indicates how many ATM cells should be added
+*        to the queue for this VC.
+*
+* @return @li IX_SUCCESS the function is registering the cell demand for
+*        this VC.
+* @return @li IX_FAIL the function cannot register cell for this VC : the
+*         scheduler maybe overloaded or misconfigured
+*
+*/
+typedef IX_STATUS (*IxAtmdAccTxVcDemandUpdateCallback) (IxAtmLogicalPort port,
+                        int vcId,
+                        unsigned int numberOfCells);
+
+/**
+*
+* @ingroup IxAtmdAccCtrlAPI
+*
+* @brief  prototype to  remove all currently queued cells from a
+* registered VC
+*
+* IxAtmdAccTxVcDemandClearCallback is the prototype of the function
+* to remove all currently queued cells from a registered VC. The
+* pending cell count for the specified VC is reset to zero. After the
+* use of this callback, the scheduler shall not schedule more cells
+* for this VC.
+*
+* This callback function is called during a VC disconnection
+* @a ixAtmdAccTxVcTryDisconnect()
+*
+* @sa IxAtmdAccTxVcDemandUpdateCallback
+* @sa IxAtmdAccTxVcDemandClearCallback
+* @sa IxAtmdAccTxSchVcIdGetCallback
+* @sa ixAtmdAccPortTxScheduledModeEnable
+* @sa ixAtmdAccTxVcTryDisconnect
+*
+* @param port @ref IxAtmLogicalPort [in] - Specifies the ATM port on which the VC to be cleared
+*        is established
+* @param vcId int [in] - Identifies the VC to be cleared. This is the value
+*        returned by the @a IxAtmdAccTxSchVcIdGetCallback() call .
+*
+* @return none
+*
+*/
+typedef void (*IxAtmdAccTxVcDemandClearCallback) (IxAtmLogicalPort port,
+                             int vcId);
+
+/**
+*
+* @ingroup IxAtmdAccCtrlAPI
+*
+* @brief  prototype to get a scheduler vc id
+*
+* IxAtmdAccTxSchVcIdGetCallback is the prototype of the function to get
+* a scheduler vcId
+*
+* @sa IxAtmdAccTxVcDemandUpdateCallback
+* @sa IxAtmdAccTxVcDemandClearCallback
+* @sa IxAtmdAccTxSchVcIdGetCallback
+* @sa ixAtmdAccPortTxScheduledModeEnable
+*
+* @param port @ref IxAtmLogicalPort [in] - Specifies the ATM logical port on which the VC is
+*        established
+* @param vpi unsigned int [in] - For AAL0/AAL5 specifies the ATM vpi on which the 
+*                 VC is established.
+*                 For OAM specifies the dedicated "OAM Tx channel" VPI.
+* @param vci unsigned int [in] - For AAL0/AAL5 specifies the ATM vci on which the 
+*                 VC is established.
+*                 For OAM specifies the dedicated "OAM Tx channel" VCI.
+* @param connId @ref IxAtmConnId [in] - specifies the IxAtmdAcc connection Id already
+*        associated with this VC
+* @param vcId int* [out] - pointer to a vcId
+*
+* @return @li IX_SUCCESS the function is returning a Scheduler vcId for this
+*         VC
+* @return @li IX_FAIL the function cannot process scheduling for this VC.
+*                 the contents of vcId is unspecified
+*
+*/
+typedef IX_STATUS (*IxAtmdAccTxSchVcIdGetCallback) (IxAtmLogicalPort port,
+                               unsigned int vpi,
+                               unsigned int vci,
+                               IxAtmConnId connId,
+                               int *vcId);
+
+/* ------------------------------------------------------
+   Part of the IxAtmdAcc interface related to RX traffic
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @fn ixAtmdAccRxDispatcherRegister (
+                          IxAtmRxQueueId queueId,
+                          IxAtmdAccRxDispatcher callback)
+ *
+ * @brief Register a notification callback to be invoked when there is
+ * at least one entry on a particular Rx queue.
+ *
+ * This function registers a callback to be invoked when there is at
+ * least one entry in a particular queue. The registered callback is
+ * called every time when the hardware adds one or more pdus to the
+ * specified Rx queue.
+ *
+ * This function cannot be used when a Rx Vc using this queue is
+ * already existing.
+ *
+ * @note -The callback function can be the API function
+ *       @a ixAtmdAccRxDispatch() : every time the threhold level
+ *       of the queue is reached, the ixAtmdAccRxDispatch() is
+ *       invoked to remove all entries from the queue.
+ *
+ * @sa ixAtmdAccRxDispatch
+ * @sa IxAtmdAccRxDispatcher
+ *
+ * @param queueId @ref IxAtmRxQueueId [in] RX queue identification
+ * @param callback @ref IxAtmdAccRxDispatcher [in] function triggering the delivery of incoming
+ *        traffic. This parameter cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS Successful call to @a ixAtmdAccRxDispatcherRegister()
+ * @return @li IX_FAIL error in the parameters, or there is an
+ *             already active RX VC for this queue or some unspecified
+ *             internal error occurred.
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccRxDispatcherRegister (
+                          IxAtmRxQueueId queueId,
+                          IxAtmdAccRxDispatcher callback);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @fn ixAtmdAccRxDispatch (IxAtmRxQueueId rxQueueId,
+    unsigned int numberOfPdusToProcess,
+    unsigned int *numberOfPdusProcessedPtr)
+ *
+ *
+ * @brief Control function which executes Rx processing for a particular
+ * Rx stream.
+ *
+ * The @a IxAtmdAccRxDispatch() function is used to process received Pdus
+ * available from one of the two incoming RX streams. When this function
+ * is invoked, the incoming traffic (up to the number of PDUs passed as
+ * a parameter) will be transferred to the IxAtmdAcc users through the
+ * callback @a IxAtmdAccRxVcRxCallback(), as registered during the
+ * @a ixAtmdAccRxVcConnect() call.
+ *
+ * The user receive callbacks will be executed in the context of this
+ * function.
+ *
+ * Failing to use this function on a regular basis when there is traffic
+ * will block incoming traffic and can result in Pdus being dropped by
+ * the hardware.
+ *
+ * This should be used to control when received pdus are handed off from
+ * the hardware to Aal users from a particluar stream. The function can
+ * be used from a timer context, or can be registered as a callback in
+ * response to an rx stream threshold event, or can be used inside an
+ * active polling mechanism which is under user control.
+ *
+ * @note - The signature of this function is directly compatible with the
+ * callback prototype which can be register with @a ixAtmdAccRxDispatcherRegister().
+ *
+ * @sa ixAtmdAccRxDispatcherRegister
+ * @sa IxAtmdAccRxVcRxCallback
+ * @sa ixAtmdAccRxVcFreeEntriesQuery
+ *
+ * @param rxQueueId @ref IxAtmRxQueueId [in] - indicates which RX queue to process.
+ * @param numberOfPdusToProcess unsigned int [in] - indicates the maxiumum number of PDU to
+ *     remove from the RX queue. A value of IX_ATMDACC_ALLPDUS indicates
+ *     to process all PDUs from the queue. This includes at least the PDUs
+ *     in the queue when the fuction is invoked. Because of real-time
+ *     constraints, there is no guarantee thatthe queue will be empty
+ *     when the function exits. If this parameter is greater than the
+ *     number of entries of the queues, the function will succeed
+ *     and the parameter numberOfPdusProcessedPtr will reflect the exact
+ *     number of PDUs processed.
+ * @param *numberOfPdusProcessedPtr unsigned int [out] - indicates the actual number of PDU
+ *     processed during this call. This parameter cannot be a null
+ *     pointer.
+ *
+ * @return @li IX_SUCCESS the number of PDUs as indicated in
+ *     numberOfPdusProcessedPtr are removed from the RX queue and the VC callback
+ *     are called.
+ * @return @li IX_FAIL invalid parameters or some unspecified internal
+ *     error occured.
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccRxDispatch (IxAtmRxQueueId rxQueueId,
+    unsigned int numberOfPdusToProcess,
+    unsigned int *numberOfPdusProcessedPtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ * 
+ * @fn ixAtmdAccRxLevelQuery (IxAtmRxQueueId rxQueueId,
+                     unsigned int *numberOfPdusPtr)
+ *
+ * @brief Query the number of entries in a particular RX queue.
+ *
+ * This function is used to retrieve the number of pdus received by
+ * the hardware and ready for distribution to users.
+ *
+ * @param rxQueueId @ref IxAtmRxQueueId [in] - indicates which of two RX queues to query.
+ * @param numberOfPdusPtr unsigned int* [out] - Pointer to store the number of available
+ *        PDUs in the RX queue. This parameter cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS the value in numberOfPdusPtr specifies the
+ *         number of incoming pdus waiting in this queue
+ * @return @li IX_FAIL an error occurs during processing.
+ *         The value in numberOfPdusPtr is unspecified.
+ *
+ * @note - This function is reentrant, doesn't use system resources
+ *         and can be used from an interrupt context.
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccRxLevelQuery (IxAtmRxQueueId rxQueueId,
+                     unsigned int *numberOfPdusPtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @fn ixAtmdAccRxQueueSizeQuery (IxAtmRxQueueId rxQueueId,
+                     unsigned int *numberOfPdusPtr)
+ *
+ * @brief Query the size of a particular RX queue.
+ *
+ * This function is used to retrieve the number of pdus the system is
+ * able to queue when reception is complete.
+ *
+ * @param rxQueueId @ref IxAtmRxQueueId [in] - indicates which of two RX queues to query.
+ * @param numberOfPdusPtr unsigned int* [out] - Pointer to store the number of pdus
+ *         the system is able to queue in the RX queue. This parameter
+ *         cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS the value in numberOfPdusPtr specifies the
+ *         number of pdus the system is able to queue.
+ * @return @li IX_FAIL an error occurs during processing.
+ *         The value in numberOfPdusPtr is unspecified.
+ *
+ * @note - This function is reentrant, doesn't use system resources
+ *         and can be used from an interrupt context.
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccRxQueueSizeQuery (IxAtmRxQueueId rxQueueId,
+                     unsigned int *numberOfPdusPtr);
+
+/* ------------------------------------------------------
+   Part of the IxAtmdAcc interface related to TX traffic
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @fn ixAtmdAccPortTxFreeEntriesQuery (IxAtmLogicalPort port,
+                         unsigned int *numberOfCellsPtr)
+ *
+ * @brief Get the number of available cells the system can accept for
+ *       transmission.
+ *
+ * The function is used to retrieve the number of cells that can be
+ * queued for transmission to the hardware.
+ *
+ * This number is based on the worst schedule table where one cell
+ * is stored in one schedule table entry, depending on the pdus size
+ * and mbuf size and fragmentation.
+ *
+ * This function doesn't use system resources and can be used from a
+ * timer context, or can be associated with a threshold event, or can
+ * be used inside an active polling mechanism
+ *
+ * @param port @ref IxAtmLogicalPort [in] - logical PHY port [@a IX_UTOPIA_PORT_0 .. @a IX_UTOPIA_MAX_PORTS - 1]
+ * @param numberOfCellsPtr unsigned int* [out] - number of available cells.
+ *                   This parameter cannot be a null pointer.
+ *
+ * @sa ixAtmdAccPortTxProcess
+ *
+ * @return @li IX_SUCCESS numberOfCellsPtr contains the number of cells that can be scheduled
+ *         for this port.
+ * @return @li IX_FAIL error in the parameters, or some processing error
+ *         occured.
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccPortTxFreeEntriesQuery (IxAtmLogicalPort port,
+                         unsigned int *numberOfCellsPtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ * 
+ * @fn ixAtmdAccPortTxCallbackRegister (IxAtmLogicalPort port,
+                       unsigned int numberOfCells,
+                       IxAtmdAccPortTxLowCallback callback)
+ *
+ * @brief Configure the Tx port threshold value and register a callback to handle
+ * threshold notifications.
+ *
+ * This function sets the threshold in cells
+ *
+ * @sa ixAtmdAccPortTxCallbackRegister
+ * @sa ixAtmdAccPortTxProcess
+ * @sa ixAtmdAccPortTxFreeEntriesQuery
+ *
+ * @param port @ref IxAtmLogicalPort [in] - logical PHY port [@a IX_UTOPIA_PORT_0 .. @a IX_UTOPIA_MAX_PORTS - 1]
+ * @param numberOfCells unsigned int [in] - threshold value which triggers the callback
+ *        invocation, This number has to be one of the
+ *        values 0,1,2,4,8,16,32 ....
+ *        The maximum value cannot be more than half of the txVc queue
+ *        size (which can be retrieved using @a ixAtmdAccPortTxFreeEntriesQuery()
+ *        before any Tx traffic is sent for this port)
+ * @param callback @ref IxAtmdAccPortTxLowCallback [in] - callback function to invoke when the threshold
+ *                 level is reached.
+ *                 This parameter cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS Successful call to @a ixAtmdAccPortTxCallbackRegister()
+ * @return @li IX_FAIL error in the parameters, Tx channel already set for this port
+ *             threshold level is not correct or within the range regarding the
+ *             queue size:or unspecified error during processing:
+ *
+ * @note - This callback function get called when the threshold level drops from
+ *         (numberOfCells+1) cells to (numberOfCells) cells
+ *
+ * @note - This function should be called during system initialisation,
+ *         outside an interrupt context
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccPortTxCallbackRegister (IxAtmLogicalPort port,
+                       unsigned int numberOfCells,
+                       IxAtmdAccPortTxLowCallback callback);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @fn ixAtmdAccPortTxScheduledModeEnable (IxAtmLogicalPort port,
+    IxAtmdAccTxVcDemandUpdateCallback vcDemandUpdateCallback,
+    IxAtmdAccTxVcDemandClearCallback vcDemandClearCallback,
+    IxAtmdAccTxSchVcIdGetCallback vcIdGetCallback)
+ *
+ * @brief Put the port into Scheduled Mode
+ *
+ * This function puts the specified port into scheduled mode of
+ * transmission which means an external s/w entity controls the
+ * transmission of cells on this port. This faciltates traffic shaping on
+ * the port.
+ *
+ * Any buffers submitted on a VC for this port will be queued in IxAtmdAcc.
+ * The transmission of these buffers to and by the hardware will be driven
+ * by a transmit schedule submitted regulary in calls to
+ * @a ixAtmdAccPortTxProcess() by traffic shaping entity.
+ *
+ * The transmit schedule is expected to be dynamic in nature based on
+ * the demand in cells for each VC on the port. Hence the callback
+ * parameters provided to this function allow IxAtmdAcc to inform the
+ * shaping entity of demand changes for each VC on the port.
+ *
+ * By default a port is in Unscheduled Mode so if this function is not
+ * called, transmission of data is done without sheduling rules, on a
+ * first-come, first-out basis.
+ *
+ * Once a port is put in scheduled mode it cannot be reverted to
+ * un-scheduled mode. Note that unscheduled mode is not supported
+ * in ixp425 1.0
+ *
+ * @note - This function should be called before any VCs have be
+ * connected on a port. Otherwise this function call will return failure.
+ *
+ * @note - This function uses internal locks and should not be called from
+ * an interrupt context
+ *
+ * @sa IxAtmdAccTxVcDemandUpdateCallback
+ * @sa IxAtmdAccTxVcDemandClearCallback
+ * @sa IxAtmdAccTxSchVcIdGetCallback
+ * @sa ixAtmdAccPortTxProcess
+ *
+ * @param port @ref IxAtmLogicalPort [in] - logical PHY port [@a IX_UTOPIA_PORT_0 .. @a IX_UTOPIA_MAX_PORTS - 1]
+ * @param vcDemandUpdateCallback @ref IxAtmdAccTxVcDemandUpdateCallback [in] - callback function used to update
+ *     the number of outstanding cells for transmission. This parameter
+ *     cannot be a null pointer.
+ * @param vcDemandClearCallback @ref IxAtmdAccTxVcDemandClearCallback [in] - callback function used to remove all
+ *     clear the number of outstanding cells for a VC. This parameter
+ *     cannot be a null pointer.
+ * @param vcIdGetCallback @ref IxAtmdAccTxSchVcIdGetCallback [in] - callback function used to exchange vc
+ *     Identifiers between IxAtmdAcc and the entity supplying the
+ *     transmit schedule. This parameter cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS scheduler registration is complete and the port
+ *         is now in scheduled mode.
+ * @return @li IX_FAIL failed (wrong parameters, or traffic is already
+ *         enabled on this port, possibly without ATM shaping)
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccPortTxScheduledModeEnable (IxAtmLogicalPort port,
+    IxAtmdAccTxVcDemandUpdateCallback vcDemandUpdateCallback,
+    IxAtmdAccTxVcDemandClearCallback vcDemandClearCallback,
+    IxAtmdAccTxSchVcIdGetCallback vcIdGetCallback);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @fn ixAtmdAccPortTxProcess (IxAtmLogicalPort port,
+    IxAtmScheduleTable* scheduleTablePtr)
+ *
+ * @brief Transmit queue cells to the H/W based on the supplied schedule
+ *        table.
+ *
+ * This function @a ixAtmdAccPortTxProcess() process the schedule
+ * table provided as a parameter to the function. As a result cells are
+ * sent to the underlaying hardware for transmission.
+ *
+ * The schedule table is executed in its entirety or not at all. So the
+ * onus is on the caller not to submit a table containing more cells than
+ * can be transmitted at that point. The maximum numbers that can be
+ * transmitted is guaranteed to be the number of cells as returned by the
+ * function @a ixAtmdAccPortTxFreeEntriesQuery().
+ *
+ * When the scheduler is invoked on a threshold level, IxAtmdAcc gives the
+ * minimum number of cells (to ensure the callback will fire again later)
+ * and the maximum number of cells that @a ixAtmdAccPortTxProcess()
+ * will be able to process (assuming the ATM scheduler is able
+ * to produce the worst-case schedule table, i.e. one entry per cell).
+ *
+ * When invoked ouside a threshold level, the overall number of cells of
+ * the schedule table should be less than the number of cells returned
+ * by the @a ixAtmdAccPortTxFreeEntriesQuery() function.
+ *
+ * After invoking the @a ixAtmdAccPortTxProcess() function, it is the
+ * user choice to query again the queue level with the function
+ * @a ixAtmdAccPortTxFreeEntriesQuery() and, depending on a new cell
+ * number, submit an other schedule table.
+ *
+ * IxAtmdAcc will check that the number of cells in the schedule table
+ * is compatible with the current transmit level. If the
+ *
+ * Obsolete or invalid connection Id will be silently discarded.
+ *
+ * This function is not reentrant for the same port.
+ *
+ * This functions doesn't use system resources and can be used inside an
+ * interrupt context.
+ *
+ * This function is used as a response to the hardware requesting more
+ * cells to transmit.
+ *
+ * @sa ixAtmdAccPortTxScheduledModeEnable
+ * @sa ixAtmdAccPortTxFreeEntriesQuery
+ * @sa ixAtmdAccPortTxCallbackRegister
+ * @sa ixAtmdAccPortEnable
+ *
+ * @param port @ref IxAtmLogicalPort [in] - logical PHY port [@a IX_UTOPIA_PORT_0 .. @a IX_UTOPIA_MAX_PORTS - 1]
+ * @param scheduleTablePtr @ref IxAtmScheduleTable* [in] - pointer to a scheduler update table. The
+ *     content of this table is not modified by this function. This
+ *     parameter cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS the schedule table process is complete
+ *             and cells are transmitted to the hardware
+ * @return @li IX_ATMDACC_WARNING : Traffic will be dropped: the schedule table exceed
+ *     the hardware capacity  If this error is ignored, further traffic
+ *     and schedule will work correctly.
+ *     Overscheduling does not occur when the schedule table does
+ *     not contain more entries that the number of free entries returned
+ *     by @a ixAtmdAccPortTxFreeEntriesQuery().
+ *     However, Disconnect attempts just after this error will fail permanently
+ *     with the error code @a IX_ATMDACC_RESOURCES_STILL_ALLOCATED, and it is
+ *     necessary to disable the port to make @a ixAtmdAccTxVcTryDisconnect()
+ *     successful.
+ * @return @li IX_FAIL a wrong parameter is supplied, or the format of
+ *     the schedule table is invalid, or the port is not Enabled, or
+ *     an internal severe error occured. No cells is transmitted to the hardware
+ *
+ * @note - If the failure is linked to an overschedule of data cells
+ *     the result is an inconsistency in the output traffic (one or many
+ *     cells may be missing and the traffic contract is not respected).
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccPortTxProcess (IxAtmLogicalPort port,
+    IxAtmScheduleTable* scheduleTablePtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @fn ixAtmdAccTxDoneDispatch (unsigned int numberOfPdusToProcess,
+                unsigned int *numberOfPdusProcessedPtr)
+ *
+ * @brief Process a number of pending transmit done pdus from the hardware.
+ *
+ * As a by-product of Atm transmit operation buffers which transmission
+ * is complete need to be recycled to users. This function is invoked
+ * to service the oustanding list of transmitted buffers and pass them
+ * to VC users.
+ *
+ * Users are handed back pdus by invoking the free callback registered
+ * during the @a ixAtmdAccTxVcConnect() call.
+ *
+ * There is a single Tx done stream servicing all active Atm Tx ports
+ * which can contain a maximum of 64 entries. If this stream fills port
+ * transmission will stop so this function must be call sufficently
+ * frequently to ensure no disruption to the transmit operation.
+ *
+ * This function can be used from a timer context, or can be associated
+ * with a TxDone level threshold event (see @a ixAtmdAccTxDoneDispatcherRegister() ),
+ * or can be used inside an active polling mechanism under user control.
+ *
+ * For ease of use the signature of this function is compatible with the
+ * TxDone threshold event callback prototype.
+ *
+ * This functions can be used inside an interrupt context.
+ *
+ * @sa ixAtmdAccTxDoneDispatcherRegister
+ * @sa IxAtmdAccTxVcBufferReturnCallback
+ * @sa ixAtmdAccTxDoneLevelQuery
+ *
+ * @param numberOfPdusToProcess unsigned int [in] - maxiumum number of pdus to remove
+ *     from the TX Done queue
+ * @param *numberOfPdusProcessedPtr unsigned int [out] - number of pdus removed from
+ *     the TX Done queue. This parameter cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS the number of pdus as indicated in
+ *     numberOfPdusToProcess are removed from the TX Done hardware
+ *     and passed to the user through the Tx Done callback registered
+ *     during a call to @a ixAtmdAccTxVcConnect()
+ * @return @li IX_FAIL invalid parameters or numberOfPdusProcessedPtr is
+ *     a null pointer or some unspecified internal error occured.
+ *
+ */
+PUBLIC IX_STATUS
+ixAtmdAccTxDoneDispatch (unsigned int numberOfPdusToProcess,
+                unsigned int *numberOfPdusProcessedPtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ * 
+ * @fn ixAtmdAccTxDoneLevelQuery (unsigned int *numberOfPdusPtr)
+ *
+ * @brief Query the current number of transmit pdus ready for
+ *        recycling.
+ *
+ * This function is used to get the number of transmitted pdus which
+ * the hardware is ready to hand back to user.
+ *
+ * This function can be used from a timer context, or can be associated
+ * with a threshold event, on can be used inside an active polling
+ * mechanism
+ *
+ * @sa ixAtmdAccTxDoneDispatch
+ *
+ * @param *numberOfPdusPtr unsigned int [out] - Pointer to the number of pdus transmitted
+ *        at the time of this function call, and ready for recycling
+ *        This parameter cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS numberOfPdusPtr contains the number of pdus
+ *        ready for recycling at the time of this function call
+ *
+ * @return @li IX_FAIL wrong parameter (null pointer as parameter).or
+ *         unspecified rocessing error occurs..The value in numberOfPdusPtr
+ *         is unspecified.
+ *
+ */
+PUBLIC IX_STATUS
+ixAtmdAccTxDoneLevelQuery (unsigned int *numberOfPdusPtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @fn ixAtmdAccTxDoneQueueSizeQuery (unsigned int *numberOfPdusPtr)
+ *
+ * @brief Query the TxDone queue size.
+ *
+ * This function is used to get the number of pdus which
+ * the hardware is able to store after transmission is complete
+ *
+ * The returned value can be used to set a threshold and enable
+ * a callback to be notified when the number of pdus is going over
+ * the threshold.
+ *
+ * @sa ixAtmdAccTxDoneDispatcherRegister
+ *
+ * @param *numberOfPdusPtr unsigned int [out] - Pointer to the number of pdus the system
+ *        is able to queue after transmission
+ *
+ * @return @li IX_SUCCESS numberOfPdusPtr contains the the number of
+ *        pdus the system is able to queue after transmission
+ * @return @li IX_FAIL wrong parameter (null pointer as parameter).or
+ *         unspecified rocessing error occurs..The value in numberOfPdusPtr
+ *         is unspecified.
+ *
+ * @note - This function is reentrant, doesn't use system resources
+ *         and can be used from an interrupt context.
+ */
+PUBLIC IX_STATUS
+ixAtmdAccTxDoneQueueSizeQuery (unsigned int *numberOfPdusPtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ * 
+ * @fn ixAtmdAccTxDoneDispatcherRegister (unsigned int numberOfPdus,
+   IxAtmdAccTxDoneDispatcher notificationCallback)
+ *
+ * @brief Configure the Tx Done stream threshold value and register a
+ * callback to handle threshold notifications.
+ *
+ * This function sets the threshold level in term of number of pdus at
+ * which the supplied notification function should be called.
+ *
+ * The higher the threshold value is, the less events will be necessary
+ * to process transmitted buffers.
+ *
+ * Transmitted buffers recycling implementation is a sytem-wide mechanism
+ * and needs to be set prior any traffic is started. If this threshold
+ * mechanism is not used, the user is responsible for polling the
+ * transmitted buffers thanks to @a ixAtmdAccTxDoneDispatch() and
+ * @a ixAtmdAccTxDoneLevelQuery() functions.
+ *
+ * This function should be called during system initialisation outside
+ * an interrupt context
+ *
+ * @sa ixAtmdAccTxDoneDispatcherRegister
+ * @sa ixAtmdAccTxDoneDispatch
+ * @sa ixAtmdAccTxDoneLevelQuery
+ *
+ * @param numberOfPdus unsigned int [in] - The number of TxDone pdus which triggers the
+ *        callback invocation This number has to be a power of 2, one of the
+ *        values 0,1,2,4,8,16,32 ...
+ *        The maximum value cannot be more than half of the txDone queue
+ *        size (which can be retrieved using @a ixAtmdAccTxDoneQueueSizeQuery())
+ * @param notificationCallback @ref IxAtmdAccTxDoneDispatcher [in] - The function to invoke. (This
+ *        parameter can be @a ixAtmdAccTxDoneDispatch()).This
+ *        parameter ust not be a null pointer.
+ *
+ * @return @li IX_SUCCESS Successful call to ixAtmdAccTxDoneDispatcherRegister
+ * @return @li IX_FAIL error in the parameters:
+ *
+ * @note - The notificationCallback will be called exactly when the threshold level
+ *         will increase from (numberOfPdus) to (numberOfPdus+1)
+ *
+ * @note - If there is no Tx traffic, there is no guarantee that TxDone Pdus will
+ *       be released to the user (when txDone level is permanently under the threshold
+ *       level. One of the preffered way to return resources to the user is to use
+ *       a mix of txDone notifications, used together with a slow
+ *       rate timer and an exclusion mechanism protecting from re-entrancy
+ *
+ * @note - The TxDone threshold will only hand back buffers when the threshold level is
+ *      crossed. Setting this threshold to a great number reduce the interrupt rate
+ *      and the cpu load, but also increase the number of outstanding mbufs and has
+ *      a system wide impact when these mbufs are needed by other components.
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccTxDoneDispatcherRegister (unsigned int numberOfPdus,
+   IxAtmdAccTxDoneDispatcher notificationCallback);
+
+/* ------------------------------------------------------
+   Part of the IxAtmdAcc interface related to Utopia config
+   ------------------------------------------------------ */
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @defgroup IxAtmdAccUtopiaCtrlAPI IXP400 ATM Driver Access (IxAtmdAcc) Utopia Control API
+ *
+ * @brief The public API for the IXP400 Atm Driver Control component
+ *
+ * IxAtmdAcc is the low level interface by which AAL PDU get
+ * transmitted to,and received from the Utopia bus
+ *
+ * This part is related to the UTOPIA configuration.
+ *
+ * @{
+ */
+
+/**
+ *
+ * @brief Utopia configuration
+ *
+ * This structure is used to set the Utopia parameters
+ * @li contains the values of Utopia registers, to be set during initialisation
+ * @li contains debug commands for NPE, to be used during development steps
+ *
+ * @note - the exact description of all parameters is done in the Utopia reference
+ *   documents.
+ *
+ */
+typedef struct
+{
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtTxConfig_
+    * @brief Utopia Tx Config Register
+    */
+    struct UtTxConfig_
+    {
+
+    unsigned int reserved_1:1;  /**< [31] These bits are always 0.*/
+    unsigned int txInterface:1;     /**< [30] Utopia Transmit Interface. The following encoding
+                                   * is used to set the Utopia Transmit interface as ATM master
+                                   * or PHY slave:
+                                   * @li 1 - PHY
+                                   * @li 0 - ATM
+                                   */
+    unsigned int txMode:1;      /**< [29] Utopia Transmit Mode. The following encoding is used
+    *  to set the Utopia Transmit mode to SPHY or MPHY:
+    *  @li 1 - SPHY
+    *  @li 0 - MPHY
+    */
+    unsigned int txOctet:1;  /**< [28] Utopia Transmit cell transfer protocol. Used to set
+    * the Utopia cell transfer protocol to Octet-level handshaking.
+    * Note this is only applicable in SPHY mode.
+    * @li 1 - Octet-handshaking enabled
+    * @li 0 - Cell-handshaking enabled
+    */
+    unsigned int txParity:1;    /**< [27] Utopia Transmit parity enabled when set. TxEvenParity
+    * defines the parity format odd/even.
+    * @li 1 - Enable Parity generation.
+    * @li 0 - ut_op_prty held low.
+    */
+    unsigned int txEvenParity:1; /**< [26] Utopia Transmit Parity Mode
+    * @li 1 - Even Parity Generated.
+    * @li 0 - Odd Parity Generated.
+    */
+    unsigned int txHEC:1; /**< [25] Header Error Check Insertion Mode. Specifies if the transmit
+    * cell header check byte is calculated and inserted when set.
+    * @li 1 - Generate HEC.
+    * @li 0 - Disable HEC generation.
+    */
+    unsigned int txCOSET:1;    /**< [24] If enabled the HEC is Exclusive-ORÆed with the value 0x55 before
+  * being presented on the Utopia bus.
+  * @li 1 - Enable HEC ExOR with value 0x55
+  * @li 0 - Use generated HEC value.
+  */
+
+    unsigned int reserved_2:1;    /**< [23] These bits are always 0
+    */
+    unsigned int txCellSize:7;    /**< [22:16] Transmit expected cell size. Configures the cell size
+    * for the transmit module: Values between 52-64 are valid.
+    */
+    unsigned int reserved_3:3;  /**< [15:13] These bits are always 0 */
+    unsigned int txAddrRange:5;       /**< [12:8] When configured as an ATM master in MPHY mode this
+    * register specifies the upper limit of the PHY polling logical
+    * range. The number of active PHYs are TxAddrRange + 1.
+    */
+    unsigned int reserved_4:3;      /**< [7:5] These bits are always 0 */
+    unsigned int txPHYAddr:5;     /**< [4:0] When configured as a slave in an MPHY system this register
+    * specifies the physical address of the PHY.
+    */
+    }
+
+    utTxConfig;       /**< Tx config Utopia register */
+
+   /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+   * @struct UtTxStatsConfig_
+   * @brief Utopia Tx stats Register
+    */
+    struct UtTxStatsConfig_
+    {
+
+    unsigned int vpi:12;  /**< [31:20] ATM VPI [11:0] OR GFC [3:0] and VPI [7:0]
+    @li Note: if VCStatsTxGFC is set to 0 the GFC field is ignored in test. */
+
+    unsigned int vci:16;  /**< [19:4] ATM VCI [15:0] or PHY Address[4] */
+
+    unsigned int pti:3;  /**< [3:1] ATM PTI [2:0] or PHY Address[3:1]
+  @li Note: if VCStatsTxPTI is set to 0 the PTI field is ignored in test.
+  @li Note: if VCStatsTxEnb is set to 0 only the transmit PHY port
+  address as defined by this register is used for ATM statistics [4:0]. */
+
+    unsigned int clp:1;  /**< [0] ATM CLP or PHY Address [0]
+  @li Note: if VCStatsTxCLP is set to 0 the CLP field is ignored in test.
+  @li Note: if VCStatsTxEnb is set to 0 only the transmit PHY port
+  address as defined by this register is used for ATM statistics [4:0]. */
+    }
+
+    utTxStatsConfig;       /**< Tx stats config Utopia register */
+
+       /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+       * @struct UtTxDefineIdle_
+       * @brief Utopia Tx idle cells Register
+    */
+    struct UtTxDefineIdle_
+    {
+
+    unsigned int vpi:12;  /**< [31:20] ATM VPI [11:0] OR GFC [3:0] and VPI [7:0]
+    @li Note: if VCIdleTxGFC is set to 0 the GFC field is ignored in test. */
+
+    unsigned int vci:16;  /**< [19:4] ATM VCI [15:0] */
+
+    unsigned int pti:3;  /**< [3:1] ATM PTI PTI [2:0]
+  @li Note: if VCIdleTxPTI is set to 0 the PTI field is ignored in test.*/
+
+    unsigned int clp:1;  /**< [0] ATM CLP [0]
+  @li Note: if VCIdleTxCLP is set to 0 the CLP field is ignored in test.*/
+    }
+
+    utTxDefineIdle;      /**< Tx idle cell config Utopia register */
+
+      /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+      * @struct UtTxEnableFields_
+      * @brief Utopia Tx ienable fields Register
+    */
+    struct UtTxEnableFields_
+    {
+
+    unsigned int defineTxIdleGFC:1;    /**< [31] This register is used to include or exclude the GFC
+    field of the ATM header when testing for Idle cells.
+    @li 1 - GFC field is valid.
+    @li 0 - GFC field ignored.*/
+
+    unsigned int defineTxIdlePTI:1;    /**< [30] This register is used to include or exclude the PTI
+    field of the ATM header when testing for Idle cells.
+    @li 1 - PTI field is valid
+    @li    0 - PTI field ignored.*/
+
+    unsigned int defineTxIdleCLP:1;    /**< [29] This register is used to include or
+    exclude the CLP field of the ATM header when testing for Idle cells.
+    @li 1 - CLP field is valid.
+    @li 0 - CLP field ignored. */
+
+    unsigned int phyStatsTxEnb:1;    /**< [28] This register is used to enable or disable ATM
+  statistics gathering based on the specified PHY address as defined
+  in TxStatsConfig register.
+  @li 1 - Enable statistics for specified transmit PHY address.
+    @li 0 - Disable statistics for specified transmit PHY address. */
+
+    unsigned int vcStatsTxEnb:1;  /**< [27] This register is used to change the ATM
+      statistics-gathering mode from the specified logical PHY address
+      to a specific VPI/VCI address.
+      @li 1 - Enable statistics for specified VPI/VCI address.
+      @li 0 - Disable statistics for specified VPI/VCI address */
+
+    unsigned int vcStatsTxGFC:1;  /**< [26] This register is used to include or exclude the GFC
+      field of the ATM header when ATM VPI/VCI statistics are enabled.
+      GFC is only available at the UNI and uses the first 4-bits of
+      the VPI field.
+      @li 1 - GFC field is valid
+      @li 0 - GFC field ignored.*/
+
+    unsigned int vcStatsTxPTI:1;  /**< [25] This register is used to include or exclude the PTI
+      field of the ATM header when ATM VPI/VCI statistics are enabled.
+      @li 1 - PTI field is valid
+      @li 0 - PTI field ignored.*/
+
+    unsigned int vcStatsTxCLP:1;  /**< [24] This register is used to include or exclude the CLP
+      field of the ATM header when ATM VPI/VCI statistics are enabled.
+      @li 1 - CLP field is valid
+      @li 0 - CLP field ignored. */
+
+    unsigned int reserved_1:3;  /**< [23-21] These bits are always 0 */
+
+    unsigned int txPollStsInt:1;    /**< [20] Enable the assertion of the ucp_tx_poll_sts condition
+  where there is a change in polling status.
+  @li 1 - ucp_tx_poll_sts asserted whenever there is a change in status
+  @li    0 - ucp_tx_poll_sts asserted if ANY transmit PHY is available
+  */
+    unsigned int txCellOvrInt:1;    /**< [19] Enable TxCellCount overflow CBI Transmit Status condition
+      assertion.
+      @li 1 - If TxCellCountOvr is set assert the Transmit Status Condition.
+      @li 0 - No CBI Transmit Status condition assertion */
+
+    unsigned int txIdleCellOvrInt:1;  /**< [18] Enable TxIdleCellCount overflow Transmit Status Condition
+    @li 1 - If TxIdleCellCountOvr is set assert the Transmit Status Condition
+      @li 0 - No CBI Transmit Status condition assertion..*/
+
+    unsigned int enbIdleCellCnt:1;    /**< [17] Enable Transmit Idle Cell Count.
+    @li 1 - Enable count of Idle cells transmitted.
+      @li 0 - No count is maintained. */
+
+    unsigned int enbTxCellCnt:1;    /**< [16] Enable Transmit Valid Cell Count of non-idle/non-error cells
+      @li 1 - Enable count of valid cells transmitted- non-idle/non-error
+      @li 0 - No count is maintained.*/
+
+    unsigned int reserved_2:16;   /**< [15:0] These bits are always 0 */
+    } utTxEnableFields;  /**< Tx enable Utopia register */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtTxTransTable0_
+    * @brief Utopia Tx translation table Register
+    */
+    struct UtTxTransTable0_
+    {
+
+    unsigned int phy0:5;  /**< [31-27] Tx Mapping value of logical phy 0 */
+
+    unsigned int phy1:5;  /**< [26-22] Tx Mapping value of logical phy 1 */
+
+    unsigned int phy2:5;  /**< [21-17] Tx Mapping value of logical phy 2 */
+
+    unsigned int reserved_1:1;  /**< [16] These bits are always 0.*/
+
+    unsigned int phy3:5;  /**< [15-11] Tx Mapping value of logical phy 3 */
+
+    unsigned int phy4:5;  /**< [10-6] Tx Mapping value of logical phy 4 */
+
+    unsigned int phy5:5;  /**< [5-1] Tx Mapping value of logical phy 5 */
+
+    unsigned int reserved_2:1;  /**< [0] These bits are always 0 */
+    } utTxTransTable0;  /**< Tx translation table */
+
+  /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+  * @struct UtTxTransTable1_
+  * @brief Utopia Tx translation table Register
+    */
+    struct UtTxTransTable1_
+    {
+
+    unsigned int phy6:5;  /**< [31-27] Tx Mapping value of logical phy 6 */
+
+    unsigned int phy7:5;  /**< [26-22] Tx Mapping value of logical phy 7 */
+
+    unsigned int phy8:5;  /**< [21-17] Tx Mapping value of logical phy 8 */
+
+    unsigned int reserved_1:1; /**< [16-0] These bits are always 0 */
+
+    unsigned int phy9:5;  /**< [15-11] Tx Mapping value of logical phy 3 */
+
+    unsigned int phy10:5;   /**< [10-6] Tx Mapping value of logical phy 4 */
+
+    unsigned int phy11:5;   /**< [5-1] Tx Mapping value of logical phy 5 */
+
+    unsigned int reserved_2:1;  /**< [0] These bits are always 0 */
+    } utTxTransTable1;  /**< Tx translation table */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtTxTransTable2_
+    * @brief Utopia Tx translation table Register
+    */
+    struct UtTxTransTable2_
+    {
+
+    unsigned int phy12:5;   /**< [31-27] Tx Mapping value of logical phy 6 */
+
+    unsigned int phy13:5;   /**< [26-22] Tx Mapping value of logical phy 7 */
+
+    unsigned int phy14:5;   /**< [21-17] Tx Mapping value of logical phy 8 */
+
+    unsigned int reserved_1:1; /**< [16-0] These bits are always 0 */
+
+    unsigned int phy15:5;   /**< [15-11] Tx Mapping value of logical phy 3 */
+
+    unsigned int phy16:5;   /**< [10-6] Tx Mapping value of logical phy 4 */
+
+    unsigned int phy17:5;   /**< [5-1] Tx Mapping value of logical phy 5 */
+
+    unsigned int reserved_2:1;  /**< [0] These bits are always 0 */
+    } utTxTransTable2;    /**< Tx translation table */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtTxTransTable3_
+    * @brief Utopia Tx translation table Register
+    */
+    struct UtTxTransTable3_
+    {
+
+    unsigned int phy18:5;   /**< [31-27] Tx Mapping value of logical phy 6 */
+
+    unsigned int phy19:5;   /**< [26-22] Tx Mapping value of logical phy 7 */
+
+    unsigned int phy20:5;   /**< [21-17] Tx Mapping value of logical phy 8 */
+
+    unsigned int reserved_1:1; /**< [16-0] These bits are always 0 */
+
+    unsigned int phy21:5;   /**< [15-11] Tx Mapping value of logical phy 3 */
+
+    unsigned int phy22:5;   /**< [10-6] Tx Mapping value of logical phy 4 */
+
+    unsigned int phy23:5;   /**< [5-1] Tx Mapping value of logical phy 5 */
+
+    unsigned int reserved_2:1;  /**< [0] These bits are always 0 */
+    } utTxTransTable3;  /**< Tx translation table */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtTxTransTable4_
+    * @brief Utopia Tx translation table Register
+    */
+    struct UtTxTransTable4_
+    {
+
+    unsigned int phy24:5;   /**< [31-27] Tx Mapping value of logical phy 6 */
+
+    unsigned int phy25:5;   /**< [26-22] Tx Mapping value of logical phy 7 */
+
+    unsigned int phy26:5;   /**< [21-17] Tx Mapping value of logical phy 8 */
+
+    unsigned int reserved_1:1; /**< [16-0] These bits are always 0 */
+
+    unsigned int phy27:5;   /**< [15-11] Tx Mapping value of logical phy 3 */
+
+    unsigned int phy28:5;   /**< [10-6] Tx Mapping value of logical phy 4 */
+
+    unsigned int phy29:5;   /**< [5-1] Tx Mapping value of logical phy 5 */
+
+    unsigned int reserved_2:1;  /**< [0] These bits are always 0 */
+    } utTxTransTable4;  /**< Tx translation table */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtTxTransTable5_
+    * @brief Utopia Tx translation table Register
+    */
+    struct UtTxTransTable5_
+    {
+
+    unsigned int phy30:5;   /**< [31-27] Tx Mapping value of logical phy 6 */
+
+    unsigned int reserved_1:27;     /**< [26-0] These bits are always 0 */
+
+    } utTxTransTable5;  /**< Tx translation table */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtRxConfig_
+    * @brief Utopia Rx config Register
+    */
+    struct UtRxConfig_
+    {
+
+    unsigned int rxInterface:1;    /**< [31] Utopia Receive Interface. The following encoding is used
+ to set the Utopia Receive interface as ATM master or PHY slave:
+ @li 1 - PHY
+    @li 0 - ATM */
+
+    unsigned int rxMode:1;      /**< [30] Utopia Receive Mode. The following encoding is used to set
+      the Utopia Receive mode to SPHY or MPHY:
+      @li 1 - SPHY
+    @li 0 - MPHY */
+
+    unsigned int rxOctet:1;  /**< [29] Utopia Receive cell transfer protocol. Used to set the Utopia
+  cell transfer protocol to Octet-level handshaking. Note this is only
+  applicable in SPHY mode.
+  @li 1 - Octet-handshaking enabled
+    @li 0 - Cell-handshaking enabled */
+
+    unsigned int rxParity:1;    /**< [28] Utopia Receive Parity Checking enable.
+    @li 1 - Parity checking enabled
+    @li 0 - Parity checking disabled */
+
+    unsigned int rxEvenParity:1;/**< [27] Utopia Receive Parity Mode
+    @li 1 - Check for Even Parity
+    @li 0 - Check for Odd Parity.*/
+
+    unsigned int rxHEC:1;    /**< [26] RxHEC    Header Error Check Mode. Enables/disables cell header
+    error checking on the received cell header.
+    @li 1 - HEC checking enabled
+    @li 0 - HEC checking disabled */
+
+    unsigned int rxCOSET:1;  /**< [25] If enabled the HEC is Exclusive-ORÆed with the value 0x55
+  before being tested with the received HEC.
+  @li 1 - Enable HEC ExOR with value 0x55.
+    @li 0 - Use generated HEC value.*/
+
+    unsigned int rxHECpass:1;    /**< [24] Specifies if the incoming cell HEC byte should be transferred
+     after optional processing to the NPE2 Coprocessor Bus Interface or
+     if it should be discarded.
+     @li 1 - HEC maintained 53-byte/UDC cell sent to NPE2.
+    @li 0 - HEC discarded 52-byte/UDC cell sent to NPE2 coprocessor.*/
+
+    unsigned int reserved_1:1;    /**< [23] These bits are always 0 */
+
+    unsigned int rxCellSize:7;    /**< [22:16] Receive cell size. Configures the receive cell size.
+      Values between 52-64 are valid */
+
+    unsigned int rxHashEnbGFC:1;      /**< [15] Specifies if the VPI field [11:8]/GFC field should be
+      included in the Hash data input or if the bits should be padded
+      with 1Æb0.
+      @li 1 - VPI [11:8]/GFC field valid and used in Hash residue calculation.
+      @li 0 - VPI [11:8]/GFC field padded with 1Æb0 */
+
+    unsigned int rxPreHash:1; /**< [14] Enable Pre-hash value generation. Specifies if the
+ incoming cell data should be pre-hashed to allow VPI/VCI header look-up
+ in a hash table.
+ @li 1 - Pre-hashing enabled
+      @li 0 - Pre-hashing disabled */
+
+    unsigned int reserved_2:1;    /**< [13] These bits are always 0 */
+
+    unsigned int rxAddrRange:5;     /**< [12:8] In ATM master, MPHY mode,
+     * this register specifies the upper
+     * limit of the PHY polling logical range. The number of active PHYs are
+     * RxAddrRange + 1.
+     */
+    unsigned int reserved_3:3;    /**< [7-5] These bits are always 0 .*/
+    unsigned int rxPHYAddr:5;     /**< [4:0] When configured as a slave in an MPHY system this register
+      * specifies the physical address of the PHY.
+      */
+    } utRxConfig;  /**< Rx config Utopia register */
+
+      /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+      * @struct UtRxStatsConfig_
+      * @brief Utopia Rx stats config Register
+    */
+    struct UtRxStatsConfig_
+    {
+
+    unsigned int vpi:12;  /**< [31:20] ATM VPI    VPI [11:0] OR GFC [3:0] and VPI [7:0]
+    @li Note: if VCStatsRxGFC is set to 0 the GFC field is ignored in test. */
+
+    unsigned int vci:16;  /**< [19:4] VCI [15:0] or PHY Address [4] */
+
+    unsigned int pti:3;  /**< [3:1] PTI [2:0] or or PHY Address [3:1]
+  @li Note: if VCStatsRxPTI is set to 0 the PTI field is ignored in test.
+  @li Note: if VCStatsRxEnb is set to 0 only the PHY port address is used
+  for statistics gathering.. */
+
+    unsigned int clp:1;  /**< [0] CLP [0] or PHY Address [0]
+  @li Note: if VCStatsRxCLP is set to 0 the CLP field is ignored in test.
+  @li Note: if VCStatsRxEnb is set to 0 only the PHY port address is used
+  for statistics gathering.. */
+    } utRxStatsConfig;  /**< Rx stats config Utopia register */
+
+  /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+  * @struct UtRxDefineIdle_
+  * @brief Utopia Rx idle cells config Register
+    */
+    struct UtRxDefineIdle_
+    {
+
+    unsigned int vpi:12;  /**< [31:20] ATM VPI [11:0] OR GFC [3:0] and VPI [7:0]
+    @li Note: if VCIdleRxGFC is set to 0 the GFC field is ignored in test. */
+
+    unsigned int vci:16;  /**< [19:4] ATM VCI [15:0] */
+
+    unsigned int pti:3;  /**< [3:1] ATM PTI PTI [2:0]
+  @li Note: if VCIdleRxPTI is set to 0 the PTI field is ignored in test.*/
+
+    unsigned int clp:1;  /**< [0] ATM CLP [0]
+  @li Note: if VCIdleRxCLP is set to 0 the CLP field is ignored in test.*/
+    } utRxDefineIdle;      /**< Rx idle cell config Utopia register */
+
+      /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+      * @struct UtRxEnableFields_
+      * @brief Utopia Rx enable Register
+    */
+    struct UtRxEnableFields_
+    {
+
+    unsigned int defineRxIdleGFC:1;/**< [31] This register is used to include or exclude the GFC
+ field of the ATM header when testing for Idle cells.
+ @li 1 - GFC field is valid.
+    @li 0 - GFC field ignored.*/
+
+    unsigned int defineRxIdlePTI:1;/**< [30] This register is used to include or exclude the PTI
+ field of the ATM header when testing for Idle cells.
+ @li 1 - PTI field is valid.
+    @li 0 - PTI field ignored.*/
+
+    unsigned int defineRxIdleCLP:1;/**< [29] This register is used to include or exclude the CLP
+ field of the ATM header when testing for Idle cells.
+ @li 1 - CLP field is valid.
+    @li    0 - CLP field ignored.*/
+
+    unsigned int phyStatsRxEnb:1;/**< [28] This register is used to enable or disable ATM statistics
+     gathering based on the specified PHY address as defined in RxStatsConfig
+     register.
+     @li 1 - Enable statistics for specified receive PHY address.
+    @li 0 - Disable statistics for specified receive PHY address.*/
+
+    unsigned int vcStatsRxEnb:1;/**< [27] This register is used to enable or disable ATM statistics
+    gathering based on a specific VPI/VCI address.
+    @li 1 - Enable statistics for specified VPI/VCI address.
+    @li 0 - Disable statistics for specified VPI/VCI address.*/
+
+    unsigned int vcStatsRxGFC:1;/**< [26] This register is used to include or exclude the GFC field
+    of the ATM header when ATM VPI/VCI statistics are enabled. GFC is only
+    available at the UNI and uses the first 4-bits of the VPI field.
+    @li 1 - GFC field is valid.
+    @li 0 - GFC field ignored. */
+
+    unsigned int vcStatsRxPTI:1;/**< [25] This register is used to include or exclude the PTI field
+    of the ATM header when ATM VPI/VCI statistics are enabled.
+    @li 1 - PTI field is valid.
+    @li 0 - PTI field ignored.*/
+
+    unsigned int vcStatsRxCLP:1;/**< [24] This register is used to include or exclude the CLP field
+    of the ATM header when ATM VPI/VCI statistics are enabled.
+    @li 1 - CLP field is valid.
+    @li 0 - CLP field ignored. */
+
+    unsigned int discardHecErr:1;/**< [23] Discard cells with an invalid HEC.
+     @li 1 - Discard cells with HEC errors
+    @li 0 - Cells with HEC errors are passed */
+
+    unsigned int discardParErr:1;/**< [22] Discard cells containing parity errors.
+     @li 1 - Discard cells with parity errors
+    @li 0 - Cells with parity errors are passed */
+
+    unsigned int discardIdle:1;    /**< [21] Discard Idle Cells based on DefineIdle register values
+ @li    1 - Discard IDLE cells
+    @li 0 - IDLE cells passed */
+
+    unsigned int enbHecErrCnt:1;/**< [20] Enable Receive HEC Error Count.
+    @li 1 - Enable count of received cells containing HEC errors
+    @li 0 - No count is maintained. */
+
+    unsigned int enbParErrCnt:1;/**< [19] Enable Parity Error Count
+    @li    1 - Enable count of received cells containing Parity errors
+    @li 0 - No count is maintained. */
+
+    unsigned int enbIdleCellCnt:1;/**< [18] Enable Receive Idle Cell Count.
+      @li 1 - Enable count of Idle cells received.
+    @li 0 - No count is maintained.*/
+
+    unsigned int enbSizeErrCnt:1;/**< [17] Enable Receive Size Error Count.
+     @li 1 - Enable count of received cells of incorrect size
+    @li    0 - No count is maintained. */
+
+    unsigned int enbRxCellCnt:1;/**< [16] Enable Receive Valid Cell Count of non-idle/non-error cells.
+    @li    1 - Enable count of valid cells received - non-idle/non-error
+    @li 0 - No count is maintained. */
+
+    unsigned int reserved_1:3;    /**< [15:13] These bits are always 0 */
+
+    unsigned int rxCellOvrInt:1;      /**< [12] Enable CBI Utopia Receive Status Condition if the RxCellCount
+      register overflows.
+      @li 1 - CBI Receive Status asserted.
+      @li    0 - No CBI Receive Status asserted.*/
+
+    unsigned int invalidHecOvrInt:1;  /**< [11] Enable CBI Receive Status Condition if the InvalidHecCount
+    register overflows.
+    @li    1 - CBI Receive Condition asserted.
+      @li 0 - No CBI Receive Condition asserted */
+
+    unsigned int invalidParOvrInt:1;  /**< [10] Enable CBI Receive Status Condition if the InvalidParCount
+    register overflows
+    @li    1 - CBI Receive Condition asserted.
+      @li 0 - No CBI Receive Condition asserted */
+
+    unsigned int invalidSizeOvrInt:1;   /**< [9] Enable CBI Receive Status Condition if the InvalidSizeCount
+     register overflows.
+     @li 1 - CBI Receive Status Condition asserted.
+      @li¸0 - No CBI Receive Status asserted */
+
+    unsigned int rxIdleOvrInt:1;      /**< [8] Enable CBI Receive Status Condition if the RxIdleCount overflows.
+      @li    1 - CBI Receive Condition asserted.
+      @li 0 - No CBI Receive Condition asserted */
+
+    unsigned int reserved_2:3;    /**< [7:5] These bits are always 0 */
+
+    unsigned int rxAddrMask:5;    /**< [4:0] This register is used as a mask to allow the user to increase
+    the PHY receive address range. The register should be programmed with
+    the address-range limit, i.e. if set to 0x3 the address range increases
+  to a maximum of 4 addresses. */
+    } utRxEnableFields;      /**< Rx enable Utopia register */
+
+      /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+      * @struct UtRxTransTable0_
+      * @brief Utopia Rx translation table Register
+    */
+    struct UtRxTransTable0_
+    {
+
+    unsigned int phy0:5;  /**< [31-27] Rx Mapping value of logical phy 0 */
+
+    unsigned int phy1:5;  /**< [26-22] Rx Mapping value of logical phy 1 */
+
+    unsigned int phy2:5;  /**< [21-17] Rx Mapping value of logical phy 2 */
+
+    unsigned int reserved_1:1;  /**< [16] These bits are always 0 */
+
+    unsigned int phy3:5;  /**< [15-11] Rx Mapping value of logical phy 3 */
+
+    unsigned int phy4:5;  /**< [10-6] Rx Mapping value of logical phy 4 */
+
+    unsigned int phy5:5;  /**< [5-1] Rx Mapping value of logical phy 5 */
+
+    unsigned int reserved_2:1;  /**< [0] These bits are always 0 */
+    }
+
+    utRxTransTable0;  /**< Rx translation table */
+
+  /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+  * @struct UtRxTransTable1_
+  * @brief Utopia Rx translation table Register
+    */
+    struct UtRxTransTable1_
+    {
+
+    unsigned int phy6:5;  /**< [31-27] Rx Mapping value of logical phy 6 */
+
+    unsigned int phy7:5;  /**< [26-22] Rx Mapping value of logical phy 7 */
+
+    unsigned int phy8:5;  /**< [21-17] Rx Mapping value of logical phy 8 */
+
+    unsigned int reserved_1:1; /**< [16-0] These bits are always 0 */
+
+    unsigned int phy9:5;  /**< [15-11] Rx Mapping value of logical phy 3 */
+
+    unsigned int phy10:5;   /**< [10-6] Rx Mapping value of logical phy 4 */
+
+    unsigned int phy11:5;   /**< [5-1] Rx Mapping value of logical phy 5 */
+
+    unsigned int reserved_2:1;  /**< [0] These bits are always 0 */
+    }
+
+    utRxTransTable1;  /**< Rx translation table */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtRxTransTable2_
+    * @brief Utopia Rx translation table Register
+    */
+    struct UtRxTransTable2_
+    {
+
+    unsigned int phy12:5;   /**< [31-27] Rx Mapping value of logical phy 6 */
+
+    unsigned int phy13:5;   /**< [26-22] Rx Mapping value of logical phy 7 */
+
+    unsigned int phy14:5;   /**< [21-17] Rx Mapping value of logical phy 8 */
+
+    unsigned int reserved_1:1; /**< [16-0] These bits are always 0 */
+
+    unsigned int phy15:5;   /**< [15-11] Rx Mapping value of logical phy 3 */
+
+    unsigned int phy16:5;   /**< [10-6] Rx Mapping value of logical phy 4 */
+
+    unsigned int phy17:5;   /**< [5-1] Rx Mapping value of logical phy 5 */
+
+    unsigned int reserved_2:1;  /**< [0] These bits are always 0 */
+    } utRxTransTable2;    /**< Rx translation table */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtRxTransTable3_
+    * @brief Utopia Rx translation table Register
+    */
+    struct UtRxTransTable3_
+    {
+
+    unsigned int phy18:5;   /**< [31-27] Rx Mapping value of logical phy 6 */
+
+    unsigned int phy19:5;   /**< [26-22] Rx Mapping value of logical phy 7 */
+
+    unsigned int phy20:5;   /**< [21-17] Rx Mapping value of logical phy 8 */
+
+    unsigned int reserved_1:1; /**< [16-0] These bits are always 0 */
+
+    unsigned int phy21:5;   /**< [15-11] Rx Mapping value of logical phy 3 */
+
+    unsigned int phy22:5;   /**< [10-6] Rx Mapping value of logical phy 4 */
+
+    unsigned int phy23:5;   /**< [5-1] Rx Mapping value of logical phy 5 */
+
+    unsigned int reserved_2:1;  /**< [0] These bits are always 0 */
+    } utRxTransTable3;  /**< Rx translation table */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtRxTransTable4_
+    * @brief Utopia Rx translation table Register
+    */
+    struct UtRxTransTable4_
+    {
+
+    unsigned int phy24:5;   /**< [31-27] Rx Mapping value of logical phy 6 */
+
+    unsigned int phy25:5;   /**< [26-22] Rx Mapping value of logical phy 7 */
+
+    unsigned int phy26:5;   /**< [21-17] Rx Mapping value of logical phy 8 */
+
+    unsigned int reserved_1:1; /**< [16-0] These bits are always 0 */
+
+    unsigned int phy27:5;   /**< [15-11] Rx Mapping value of logical phy 3 */
+
+    unsigned int phy28:5;   /**< [10-6] Rx Mapping value of logical phy 4 */
+
+    unsigned int phy29:5;   /**< [5-1] Rx Mapping value of logical phy 5 */
+
+    unsigned int reserved_2:1;  /**< [0] These bits are always 0 */
+    } utRxTransTable4;  /**< Rx translation table */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtRxTransTable5_
+    * @brief Utopia Rx translation table Register
+    */
+    struct UtRxTransTable5_
+    {
+
+    unsigned int phy30:5;   /**< [31-27] Rx Mapping value of logical phy 6 */
+
+    unsigned int reserved_1:27;     /**< [26-0] These bits are always 0 */
+
+    } utRxTransTable5;  /**< Rx translation table */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtSysConfig_
+    * @brief NPE setup Register
+    */
+    struct UtSysConfig_
+    {
+
+    unsigned int reserved_1:2;     /**< [31-30] These bits are always 0 */
+    unsigned int txEnbFSM:1;    /**< [29] Enables the operation ofthe Utopia Transmit FSM
+   * @li 1 - FSM enabled
+   * @li 0 - FSM inactive
+   */
+    unsigned int rxEnbFSM:1;    /**< [28] Enables the operation ofthe Utopia Revieve FSM
+   * @li 1 - FSM enabled
+   * @li 0 - FSM inactive
+   */
+    unsigned int disablePins:1;    /**< [27] Disable Utopia interface I/O pins forcing the signals to an
+   * inactive state.  Note that this bit is set on reset and must be
+   * de-asserted
+   * @li 0 - Normal data transfer
+   * @li    1 - Utopia interface pins are forced inactive
+    */
+    unsigned int tstLoop:1;    /**< [26] Test Loop Back Enable.
+    * @li Note: For loop back to function RxMode and Tx Mode must both be set
+    * to single PHY mode.
+    * @li 0 - Loop back
+    * @li 1 - Normal operating mode
+    */
+
+    unsigned int txReset:1;   /**< [25] Resets the Utopia Coprocessor transmit module to a known state.
+    * @li Note: All transmit configuration and status registers will be reset
+    * to their reset values.
+    * @li 0 - Normal operating mode¸
+    * @li 1 - Reset transmit modules
+    */
+
+    unsigned int rxReset:1;   /**< [24] Resets the Utopia Coprocessor receive module to a known state.
+    * @li Note: All receive configuration and status registers will be reset
+    * to their reset values.
+    * @li    0 - Normal operating mode
+    * @li 1 - Reset receive modules
+    */
+
+    unsigned int reserved_2:24;     /**< [23-0] These bits are always 0 */
+    } utSysConfig;  /**< NPE debug config */
+
+}
+IxAtmdAccUtopiaConfig;
+
+/**
+*
+* @brief Utopia status
+*
+* This structure is used to set/get the Utopia status parameters
+* @li contains debug cell counters, to be accessed during a read operation
+*
+* @note - the exact description of all parameters is done in the Utopia reference
+*   documents.
+*
+*/
+typedef struct
+{
+
+    unsigned int utTxCellCount;  /**< count of cells transmitted */
+
+    unsigned int utTxIdleCellCount;    /**< count of idle cells transmitted */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtTxCellConditionStatus_
+    * @brief Utopia Tx Status Register
+    */
+    struct UtTxCellConditionStatus_
+    {
+
+    unsigned int reserved_1:2;   /**< [31:30] These bits are always 0 */
+    unsigned int txFIFO2Underflow:1; /**< [29] This bit is set if 64-byte
+                                     * Transmit FIFO2 indicates a FIFO underflow
+                                     * error condition.
+                                     */
+    unsigned int txFIFO1Underflow:1; /**< [28] This bit is set if
+                                     * 64-byte Transmit FIFO1 indicates a FIFO
+                                     * underflow error condition.
+                                     */
+    unsigned int txFIFO2Overflow:1;  /**< [27] This bit is set if 64-byte
+                                     * Transmit FIFO2 indicates a FIFO overflow
+                                     * error condition.
+                                     */
+    unsigned int txFIFO1Overflow:1; /**< [26] This bit is set if 64-byte
+                                    * Transmit FIFO1 indicates a FIFO overflow
+                                    * error condition.
+                                    */
+    unsigned int txIdleCellCountOvr:1; /**< [25] This bit is set if the
+                                       * TxIdleCellCount register overflows.
+                                       */
+    unsigned int txCellCountOvr:1; /**< [24] This bit is set if the
+                                   * TxCellCount register overflows
+                                   */
+    unsigned int reserved_2:24;    /**< [23:0] These bits are always 0 */
+    } utTxCellConditionStatus;    /**< Tx cells condition status */
+
+    unsigned int utRxCellCount; /**< count of cell received */
+    unsigned int utRxIdleCellCount;    /**< count of idle cell received */
+    unsigned int utRxInvalidHECount;     /**< count of invalid cell
+                                        * received because of HEC errors
+                                        */
+    unsigned int utRxInvalidParCount;  /**< count of invalid cell received
+                                        * because of parity errors
+                                        */
+    unsigned int utRxInvalidSizeCount;  /**< count of invalid cell
+                                        * received because of cell
+                                        * size errors
+                                        */
+
+    /**
+    * @ingroup IxAtmdAccUtopiaCtrlAPI
+    * @struct UtRxCellConditionStatus_
+    * @brief Utopia Rx Status Register
+    */
+    struct UtRxCellConditionStatus_
+    {
+
+    unsigned int reserved_1:3;  /**< [31:29] These bits are always 0.*/
+    unsigned int rxCellCountOvr:1;      /**< [28] This bit is set if the RxCellCount register overflows. */
+    unsigned int invalidHecCountOvr:1;    /**< [27] This bit is set if the InvalidHecCount register overflows.*/
+    unsigned int invalidParCountOvr:1;    /**< [26] This bit is set if the InvalidParCount register overflows.*/
+    unsigned int invalidSizeCountOvr:1;    /**< [25] This bit is set if the InvalidSizeCount register overflows.*/
+    unsigned int rxIdleCountOvr:1;      /**< [24] This bit is set if the RxIdleCount register overflows.*/
+    unsigned int reserved_2:4;  /**< [23:20] These bits are always 0 */
+    unsigned int rxFIFO2Underflow:1;  /**< [19] This bit is set if 64-byte Receive FIFO2
+                                      * indicates a FIFO underflow error condition.
+                                      */
+    unsigned int rxFIFO1Underflow:1;  /**< [18] This bit is set if 64-byte Receive
+                                      * FIFO1 indicates a FIFO underflow error condition
+                                    . */
+    unsigned int rxFIFO2Overflow:1;    /**< [17] This bit is set if 64-byte Receive FIFO2
+                                       * indicates a FIFO overflow error condition.
+                                        */
+    unsigned int rxFIFO1Overflow:1;    /**< [16] This bit is set if 64-byte Receive FIFO1
+                                    * indicates a FIFO overflow error condition.
+                                    */
+    unsigned int reserved_3:16;      /**< [15:0] These bits are always 0. */
+    } utRxCellConditionStatus;    /**< Rx cells condition status */
+
+} IxAtmdAccUtopiaStatus;
+
+/**
+ * @} defgroup IxAtmdAccUtopiaCtrlAPI
+ */
+
+ /**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @fn ixAtmdAccUtopiaConfigSet (const IxAtmdAccUtopiaConfig *
+                        ixAtmdAccUtopiaConfigPtr)
+ *
+ * @brief Send the configuration structure to the Utopia interface
+ *
+ * This function downloads the @a IxAtmdAccUtopiaConfig structure to
+ * the Utopia and has the following effects
+ *  @li setup the Utopia interface
+ *  @li initialise the NPE
+ *  @li reset the Utopia cell counters and status registers to known values
+ *
+ * This action has to be done once at initialisation. A lock is preventing
+ * the concurrent use of @a ixAtmdAccUtopiaStatusGet() and
+ * @A ixAtmdAccUtopiaConfigSet()
+ *
+ * @param *ixAtmdAccNPEConfigPtr @ref IxAtmdAccUtopiaConfig [in] - pointer to a structure to download to
+ *  Utopia. This parameter cannot be a null pointer.
+ *
+ * @return @li IX_SUCCESS successful download
+ * @return @li IX_FAIL error in the parameters, or configuration is not
+ *         complete or failed
+ *
+ * @sa ixAtmdAccUtopiaStatusGet
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccUtopiaConfigSet (const IxAtmdAccUtopiaConfig *
+                        ixAtmdAccUtopiaConfigPtr);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @fn ixAtmdAccUtopiaStatusGet (IxAtmdAccUtopiaStatus *
+                        ixAtmdAccUtopiaStatus)
+ *
+ * @brief Get the Utopia interface configuration.
+ *
+ * This function reads the Utopia registers and the Cell counts
+ * and fills the @a IxAtmdAccUtopiaStatus structure
+ *
+ * A lock is preventing the concurrent
+ * use of @a ixAtmdAccUtopiaStatusGet() and @A ixAtmdAccUtopiaConfigSet()
+ *
+ * @param ixAtmdAccUtopiaStatus @ref IxAtmdAccUtopiaStatus [out] - pointer to structure to be updated from internal
+ *        hardware counters. This parameter cannot be a NULL pointer.
+ *
+ * @return @li IX_SUCCESS successful read
+ * @return @li IX_FAIL error in the parameters null pointer, or
+ *          configuration read is not complete or failed
+ *
+ * @sa ixAtmdAccUtopiaConfigSet
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccUtopiaStatusGet (IxAtmdAccUtopiaStatus *
+                        ixAtmdAccUtopiaStatus);
+
+/**
+ *
+ * @ingroup IxAtmdAcc
+ *
+ * @fn ixAtmdAccPortEnable (IxAtmLogicalPort port)
+ *
+ * @brief enable a PHY logical port
+ *
+ * This function enables the transmission over one port. It should be
+ * called before accessing any resource from this port and before the
+ * establishment of a VC.
+ *
+ * When a port is enabled, the cell transmission to the Utopia interface
+ * is started. If there is no traffic already running, idle cells are
+ * sent over the interface.
+ *
+ * This function can be called multiple times.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - logical PHY port [@a IX_UTOPIA_PORT_0 .. @a IX_UTOPIA_MAX_PORTS - 1]
+ *
+ * @return @li IX_SUCCESS enable is complete
+ * @return @li IX_ATMDACC_WARNING port already enabled
+ * @return @li IX_FAIL enable failed, wrong parameter, or cannot
+ *         initialise this port (the port is maybe already in use,
+ *         or there is a hardware issue)
+ *
+ * @note - This function needs internal locks and should not be
+ *         called from an interrupt context
+ *
+ * @sa ixAtmdAccPortDisable
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccPortEnable (IxAtmLogicalPort port);
+
+/**
+ *
+ * @ingroup IxAtmdAccCtrlAPI
+ *
+ * @fn ixAtmdAccPortDisable (IxAtmLogicalPort port)
+ *
+ * @brief disable a PHY logical port
+ *
+ * This function disable the transmission over one port.
+ *
+ * When a port is disabled, the cell transmission to the Utopia interface
+ * is stopped.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - logical PHY port [@a IX_UTOPIA_PORT_0 .. @a IX_UTOPIA_MAX_PORTS - 1]
+ *
+ * @return @li IX_SUCCESS disable is complete
+ * @return @li IX_ATMDACC_WARNING port already disabled
+ * @return @li IX_FAIL disable failed, wrong parameter .
+ *
+ * @note - This function needs internal locks and should not be called
+ *         from an interrupt context
+ *
+ * @note - The response from hardware is done through the txDone mechanism
+ *         to ensure the synchrnisation with tx resources. Therefore, the
+ *         txDone mechanism needs to be serviced to make a PortDisable complete.
+ *
+ * @sa ixAtmdAccPortEnable
+ * @sa ixAtmdAccPortDisableComplete
+ * @sa ixAtmdAccTxDoneDispatch
+ *
+ */
+PUBLIC IX_STATUS ixAtmdAccPortDisable (IxAtmLogicalPort port);
+
+/**
+*
+* @ingroup IxAtmdAccCtrlAPI
+*
+* @fn ixAtmdAccPortDisableComplete (IxAtmLogicalPort port)
+*
+* @brief disable a PHY logical port
+*
+* This function indicates if the port disable for a port has completed. This
+* function will return TRUE if the port has never been enabled.
+*
+* @param port @ref IxAtmLogicalPort [in] - logical PHY port [@a IX_UTOPIA_PORT_0 .. @a IX_UTOPIA_MAX_PORTS - 1]
+*
+* @return @li TRUE disable is complete
+* @return @li FALSE disable failed, wrong parameter .
+*
+* @note - This function needs internal locks and should not be called
+*         from an interrupt context
+*
+* @sa ixAtmdAccPortEnable
+* @sa ixAtmdAccPortDisable
+*
+*/
+PUBLIC BOOL ixAtmdAccPortDisableComplete (IxAtmLogicalPort port);
+
+#endif /* IXATMDACCCTRL_H */
+
+/**
+ * @} defgroup IxAtmdAccCtrlAPI
+ */
+
+

cpu/ixp/npe/include/IxAtmm.h

@@ -0,0 +1,795 @@
+/**
+ * @file    IxAtmm.h
+ *
+ * @date    3-DEC-2001
+ *
+ * @brief   Header file for the IXP400 ATM Manager component (IxAtmm)
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+
+/**
+ * @defgroup IxAtmm IXP400 ATM Manager (IxAtmm) API
+ *
+ * @brief IXP400 ATM Manager component Public API
+ *
+ * @{
+ */
+
+#ifndef IXATMM_H
+#define IXATMM_H
+
+/*
+ * Put the user defined include files required
+ */
+#include "IxAtmSch.h"
+#include "IxOsalTypes.h"
+
+/*
+ * #defines and macros used in this file.
+ */
+
+/** 
+ * @def IX_ATMM_RET_ALREADY_INITIALIZED
+ * 
+ * @brief Component has already been initialized 
+ */
+#define IX_ATMM_RET_ALREADY_INITIALIZED 2
+
+/** 
+ * @def IX_ATMM_RET_INVALID_PORT
+ * 
+ * @brief Specified port does not exist or is out of range */
+#define IX_ATMM_RET_INVALID_PORT 3
+
+/** 
+ * @def IX_ATMM_RET_INVALID_VC_DESCRIPTOR
+ * 
+ * @brief The VC description does not adhere to ATM standards */
+#define IX_ATMM_RET_INVALID_VC_DESCRIPTOR 4
+
+/** 
+ * @def IX_ATMM_RET_VC_CONFLICT
+ * 
+ * @brief The VPI/VCI values supplied are either reserved, or they
+ *         conflict with a previously registered VC on this port */
+#define IX_ATMM_RET_VC_CONFLICT 5
+
+/** 
+ * @def IX_ATMM_RET_PORT_CAPACITY_IS_FULL
+ * 
+ * @brief The virtual connection cannot be established on the port
+ *         because the remaining port capacity is not sufficient to
+ *         support it */
+#define IX_ATMM_RET_PORT_CAPACITY_IS_FULL 6
+
+/** 
+ * @def IX_ATMM_RET_NO_SUCH_VC
+ * 
+ * @brief No registered VC, as described by the supplied VCI/VPI or
+ *         VC identifier values, exists on this port */
+#define IX_ATMM_RET_NO_SUCH_VC 7
+
+/** 
+ * @def IX_ATMM_RET_INVALID_VC_ID
+ * 
+ * @brief The specified VC identifier is out of range. */
+#define IX_ATMM_RET_INVALID_VC_ID 8
+
+/** 
+ * @def IX_ATMM_RET_INVALID_PARAM_PTR
+ * 
+ * @brief A pointer parameter was NULL. */
+#define IX_ATMM_RET_INVALID_PARAM_PTR 9
+
+/** 
+ * @def IX_ATMM_UTOPIA_SPHY_ADDR  
+ * 
+ * @brief The phy address when in SPHY mode */
+#define IX_ATMM_UTOPIA_SPHY_ADDR 31
+
+/**
+ * @def IX_ATMM_THREAD_PRI_HIGH
+ *
+ * @brief The value of high priority thread */
+#define IX_ATMM_THREAD_PRI_HIGH 90
+
+/*
+ * Typedefs whose scope is limited to this file.
+ */
+
+/** @brief Definition for use in the @ref IxAtmmVc structure.
+ *         Indicates the direction of a VC */
+typedef enum
+{
+    IX_ATMM_VC_DIRECTION_TX=0, /**< Atmm Vc direction transmit*/
+    IX_ATMM_VC_DIRECTION_RX, /**< Atmm Vc direction receive*/
+    IX_ATMM_VC_DIRECTION_INVALID /**< Atmm Vc direction invalid*/
+} IxAtmmVcDirection;
+
+/** @brief Definition for use with @ref IxAtmmVcChangeCallback
+ *         callback.  Indicates that the event type represented by the
+ *         callback for this VC. */
+typedef enum
+{
+    IX_ATMM_VC_CHANGE_EVENT_REGISTER=0, /**< Atmm Vc event register*/
+    IX_ATMM_VC_CHANGE_EVENT_DEREGISTER, /**< Atmm Vc event de-register*/
+    IX_ATMM_VC_CHANGE_EVENT_INVALID /**< Atmm Vc event invalid*/
+} IxAtmmVcChangeEvent;
+
+/** @brief Definitions for use with @ref ixAtmmUTOPIAInit interface to
+ *         indicate that UTOPIA loopback should be enabled or disabled
+ *         on initialisation. */
+typedef enum
+{
+    IX_ATMM_UTOPIA_LOOPBACK_DISABLED=0, /**< Atmm Utopia loopback mode disabled*/
+    IX_ATMM_UTOPIA_LOOPBACK_ENABLED, /**< Atmm Utopia loopback mode enabled*/
+    IX_ATMM_UTOPIA_LOOPBACK_INVALID /**< Atmm Utopia loopback mode invalid*/
+} IxAtmmUtopiaLoopbackMode;
+
+/** @brief This structure describes the required attributes of a
+ *         virtual connection.
+*/
+typedef struct {
+    unsigned vpi;  /**< VPI value of this virtual connection */
+    unsigned vci;  /**< VCI value of this virtual connection. */
+    IxAtmmVcDirection direction; /**< VC direction */
+
+    /** Traffic descriptor of this virtual connection.  This structure
+     *  is defined by the @ref IxAtmSch component.  */
+    IxAtmTrafficDescriptor trafficDesc;
+} IxAtmmVc;
+
+
+/** @brief Definitions for use with @ref ixAtmmUtopiaInit interface to
+ *         indicate that UTOPIA multi-phy/single-phy mode is used.
+ */
+typedef enum
+{
+    IX_ATMM_MPHY_MODE = 0, /**< Atmm phy mode mphy*/
+    IX_ATMM_SPHY_MODE, /**< Atmm phy mode sphy*/
+    IX_ATMM_PHY_MODE_INVALID /**< Atmm phy mode invalid*/
+} IxAtmmPhyMode;
+
+
+/** @brief Structure contains port-specific information required to
+ *         initialize IxAtmm, and specifically, the IXP400 UTOPIA
+ *         Level-2 device. */
+typedef struct {
+    unsigned reserved_1:11;     /**< [31:21] Should be zero */
+    unsigned UtopiaTxPhyAddr:5; /**< [20:16] Address of the
+     *   transmit (Tx) PHY for this
+     *   port on the 5-bit UTOPIA
+     *   Level-2 address bus */
+    unsigned reserved_2:11;     /**< [15:5] Should be zero */
+    unsigned UtopiaRxPhyAddr:5; /**< [4:0] Address of the receive
+     *   (Rx) PHY for this port on the
+     *   5-bit UTOPIA  Level-2
+     *   address bus */
+} IxAtmmPortCfg;
+
+/** @brief Callback type used with @ref ixAtmmVcChangeCallbackRegister interface
+ *         Defines a callback type  which will be used to notify registered
+ *         users of registration/deregistration events on a particular port
+ *
+ * @param eventType @ref IxAtmmVcChangeEvent [in] - Event indicating
+ *                        whether the VC supplied has been added or
+ *                        removed
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Specifies the port on which the event has
+ *                        occurred
+ *
+ * @param vcChanged @ref IxAtmmVc* [in] - Pointer to a structure which gives
+ *                              details of the VC which has been added
+ *                              or removed on the port
+ */
+typedef void (*IxAtmmVcChangeCallback) (IxAtmmVcChangeEvent eventType,
+					IxAtmLogicalPort port,
+					const IxAtmmVc* vcChanged);
+
+/*
+ * Variable declarations global to this file only. Externs are followed by
+ * static variables.
+ */
+
+/*
+ * Extern function prototypes
+ */
+
+/*
+ * Function declarations
+ */
+
+
+/** 
+ * @ingroup IxAtmm
+ *
+ * @fn ixAtmmInit (void)
+ *
+ * @brief Interface to initialize the IxAtmm software component.  Can
+ *         be called once only.
+ *
+ *  Must be called before any other IxAtmm API is called.
+ *
+ * @param "none"
+ *
+ *  @return @li  IX_SUCCESS : IxAtmm has been successfully initialized.
+ *      Calls to other IxAtmm interfaces may now be performed.
+ *  @return @li  IX_FAIL : IxAtmm has already been initialized.
+ */
+PUBLIC IX_STATUS
+ixAtmmInit (void);
+
+/**  
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmUtopiaInit (unsigned numPorts,
+		  IxAtmmPhyMode phyMode,
+		  IxAtmmPortCfg portCfgs[],
+		  IxAtmmUtopiaLoopbackMode loopbackMode)
+ *
+ * @brief Interface to initialize the UTOPIA Level-2 ATM coprocessor
+ *         for the specified number of physical ports.  The function
+ *         must be called before the ixAtmmPortInitialize interface
+ *         can operate successfully.
+ *
+ * @param numPorts unsigned [in] - Indicates the total number of logical
+ *          ports that are active on the device.  Up to 12 ports are
+ *          supported.
+ *
+ * @param phyMode @ref IxAtmmPhyMode [in] - Put the Utopia coprocessor in SPHY
+ *        or MPHY mode.
+ *
+ * @param portCfgs[] @ref IxAtmmPortCfg [in] - Pointer to an array of elements
+ *          detailing the UTOPIA specific port characteristics.  The
+ *          length of the array must be equal to the number of ports
+ *          activated.  ATM ports are referred to by the relevant
+ *          offset in this array in all subsequent IxAtmm interface
+ *          calls.
+ *
+ * @param loopbackMode @ref IxAtmmUtopiaLoopbackMode [in] - Value must be one of
+ *          @ref IX_ATMM_UTOPIA_LOOPBACK_ENABLED or @ref
+ *          IX_ATMM_UTOPIA_LOOPBACK_DISABLED indicating whether
+ *          loopback should be enabled on the device.  Loopback can
+ *          only be supported on a single PHY, therefore the numPorts
+ *          parameter must be 1 if loopback is enabled.
+ *
+ * @return @li IX_SUCCESS : Indicates that the  UTOPIA device has been
+ *      successfully initialized for the supplied ports.
+ * @return @li IX_ATMM_RET_ALREADY_INITIALIZED : The UTOPIA device has
+ *      already been initialized.
+ * @return @li IX_FAIL : The supplied parameters are invalid or have been
+ *     rejected by the UTOPIA-NPE device.
+ *
+ * @warning
+ * This interface may only be called once.
+ * Port identifiers are assumed to range from 0 to (numPorts - 1) in all 
+ * instances.
+ * In all subsequent calls to interfaces supplied by IxAtmm, the specified
+ * port value is expected to represent the offset in the portCfgs array
+ * specified in this interface.  i.e. The first port in this array will
+ * subsequently be represented as port 0, the second port as port 1,
+ * and so on.*/
+PUBLIC IX_STATUS
+ixAtmmUtopiaInit (unsigned numPorts,
+		  IxAtmmPhyMode phyMode,
+		  IxAtmmPortCfg portCfgs[],
+		  IxAtmmUtopiaLoopbackMode loopbackMode);
+
+
+/**   
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmPortInitialize (IxAtmLogicalPort port,
+		      unsigned txPortRate,
+		      unsigned rxPortRate)
+ *
+ * @brief The interface is called following @ref ixAtmmUtopiaInit ()
+ *         and before calls to any other IxAtmm interface.  It serves
+ *         to activate the registered ATM port with IxAtmm.
+ *
+ *  The transmit and receive port rates are specified in bits per
+ *  second.  This translates to ATM cells per second according to the
+ *  following formula: CellsPerSecond = portRate / (53*8)  The
+ *  IXP400 device supports only 53 byte cells. The client shall make
+ *  sure that the off-chip physical layer device has already been
+ *  initialized.
+ *
+ *  IxAtmm will configure IxAtmdAcc and IxAtmSch to enable scheduling
+ *  on the port.
+ *
+ *  This interface must be called once for each active port in the
+ *  system.  The first time the interface is invoked, it will configure
+ *  the mechanism by which the handling of transmit, transmit-done and
+ *  receive are driven with the IxAtmdAcc component.
+ *
+ *  This function is reentrant.
+ *
+ *  @note The minimum tx rate that will be accepted is 424 bit/s which equates
+ *        to 1 cell (53 bytes) per second.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Identifies the port which is to be
+ *          initialized.
+ *
+ * @param txPortRate unsigned [in] - Value specifies the
+ *          transmit port rate for this port in
+ *          bits/second.  This value is used by the ATM Scheduler
+ *          component is evaluating VC access requests for the port.
+ *
+ * @param rxPortRate unsigned [in] - Value specifies the
+ *          receive port rate for this port in bits/second.
+ *
+ * @return @li IX_SUCCESS : The specificed ATM port has been successfully
+ *       initialized. IxAtmm is ready to accept VC registrations on
+ *       this port.
+ *
+ * @return @li IX_ATMM_RET_ALREADY_INITIALIZED : ixAtmmPortInitialize has
+ *       already been called successfully on this port.  The current
+ *       call is rejected.
+ *
+ * @return @li IX_ATMM_RET_INVALID_PORT : The port value indicated in the
+ *       input is not valid.  The request is rejected.
+ *
+ * @return @li IX_FAIL : IxAtmm could not initialize the port because the
+ * inputs are not understood.
+ *
+ * @sa ixAtmmPortEnable, ixAtmmPortDisable
+ *
+ */
+PUBLIC IX_STATUS
+ixAtmmPortInitialize (IxAtmLogicalPort port,
+		      unsigned txPortRate,
+		      unsigned rxPortRate);
+
+/**    
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmPortModify (IxAtmLogicalPort port,
+		  unsigned txPortRate,
+		  unsigned rxPortRate)
+ *
+ * @brief A client may call this interface to change the existing
+ *         port rate (expressed in bits/second) on an established ATM
+ *         port.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Identifies the port which is to be
+ *          initialized.
+ *
+ * @param txPortRate unsigned [in] -  Value specifies the``
+ *          transmit port rate for this port in
+ *          bits/second.  This value is used by the ATM Scheduler
+ *          component is evaluating VC access requests for the port.
+ *
+ * @param rxPortRate unsigned [in] - Value specifies the
+ *          receive port rate for this port in
+ *          bits/second.
+ *
+ * @return @li IX_SUCCESS : The indicated ATM port rates have been
+ *      successfully modified.
+ *
+ * @return @li IX_ATMM_RET_INVALID_PORT : The port value indicated in the
+ *       input is not valid.  The request is rejected.
+ *
+ * @return @li IX_FAIL : IxAtmm could not update the port because the
+ *       inputs are not understood, or the interface was called before
+ * the port was initialized.  */
+PUBLIC IX_STATUS
+ixAtmmPortModify (IxAtmLogicalPort port,
+		  unsigned txPortRate,
+		  unsigned rxPortRate);
+
+/**    
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmPortQuery (IxAtmLogicalPort port,
+		 unsigned *txPortRate,
+		 unsigned *rxPortRate);
+
+ *
+ * @brief The client may call this interface to request details on
+ *          currently registered transmit and receive rates for an ATM
+ *          port.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Value identifies the port from which the
+ *          rate details are requested.
+ *
+ * @param *txPortRate unsigned [out] - Pointer to a value
+ *          which will be filled with the value of the transmit port
+ *          rate specified in bits/second.
+ *
+ * @param *rxPortRate unsigned [out] - Pointer to a value
+ *          which will be filled with the value of the receive port
+ *          rate specified in bits/second.
+ *
+ * @return @li IX_SUCCESS : The information requested on the specified
+ *       port has been successfully supplied in the output.
+ *
+ * @return @li IX_ATMM_RET_INVALID_PORT : The port value indicated in the
+ *       input is not valid.  The request is rejected.
+ *
+ * @return @li IX_ATMM_RET_INVALID_PARAM_PTR : A pointer parameter was
+ *       NULL.
+ *
+ * @return @li IX_FAIL : IxAtmm could not update the port because the
+ *       inputs are not understood, or the interface was called before
+ *       the port was initialized.  */
+PUBLIC IX_STATUS
+ixAtmmPortQuery (IxAtmLogicalPort port,
+		 unsigned *txPortRate,
+		 unsigned *rxPortRate);
+
+/**    
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmPortEnable(IxAtmLogicalPort port)
+ *
+ * @brief The client call this interface to enable transmit for an ATM
+ *          port. At initialisation, all the ports are disabled.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Value identifies the port
+ *
+ * @return @li IX_SUCCESS : Transmission over this port is started.
+ *
+ * @return @li IX_FAIL : The port parameter is not valid, or the
+ *       port is already enabled
+ *
+ * @note - When a port is disabled, Rx and Tx VC Connect requests will fail
+ *
+ * @note - This function uses system resources and should not be used
+ *        inside an interrupt context.
+ *
+ * @sa ixAtmmPortDisable  */
+PUBLIC IX_STATUS
+ixAtmmPortEnable(IxAtmLogicalPort port);
+
+/**    
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmPortDisable(IxAtmLogicalPort port)
+ *
+ * @brief The client call this interface to disable transmit for an ATM
+ *          port. At initialisation, all the ports are disabled.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Value identifies the port
+ *
+ * @return @li IX_SUCCESS : Transmission over this port is stopped.
+ *
+ * @return @li IX_FAIL : The port parameter is not valid, or the
+ *       port is already disabled
+ *
+ * @note - When a port is disabled, Rx and Tx VC Connect requests will fail
+ *
+ * @note - This function call does not stop RX traffic. It is supposed
+ *        that this function is invoked when a serious problem
+ *        is detected (e.g. physical layer broken). Then, the RX traffic
+ *        is not passing.
+ *
+ * @note - This function is blocking until the hw acknowledge that the
+ *        transmission is stopped.
+ *
+ * @note - This function uses system resources and should not be used
+ *        inside an interrupt context.
+ *
+ * @sa ixAtmmPortEnable  */
+PUBLIC IX_STATUS
+ixAtmmPortDisable(IxAtmLogicalPort port);
+
+/**    
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmVcRegister (IxAtmLogicalPort port,
+		  IxAtmmVc *vcToAdd,
+		  IxAtmSchedulerVcId *vcId)
+ *
+ * @brief This interface is used to register an ATM Virtual
+ *         Connection on the specified ATM port.
+ *
+ *  Each call to this interface registers a unidirectional virtual
+ *  connection with the parameters specified.  If a bi-directional VC
+ *  is needed, the function should be called twice (once for each
+ *  direction, Tx & Rx) where the VPI and VCI and port parameters in
+ *  each call are identical.
+ *
+ *  With the addition of each new VC to a port, a series of
+ *  callback functions are invoked by the IxAtmm component to notify
+ *  possible external components of the change.  The callback functions
+ *  are registered using the @ref ixAtmmVcChangeCallbackRegister interface.
+ *
+ *  The IxAtmSch component is notified of the registration of transmit
+ *  VCs.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Identifies port on which the specified VC is
+ *          to be registered.
+ *
+ * @param *vcToAdd @ref IxAtmmVc [in] -  Pointer to an @ref IxAtmmVc structure
+ *          containing a description of the VC to be registered. The
+ *          client shall fill the vpi, vci and direction and relevant
+ *          trafficDesc members of this structure before calling this
+ *          function.
+ *
+ * @param *vcId @ref IxAtmSchedulerVcId [out] - Pointer to an integer value which is filled
+ *              with the per-port unique identifier value for this VC.
+ *              This identifier will be required when a request is
+ *              made to deregister or change this VC.  VC identifiers
+ *              for transmit VCs will have a value between 0-43,
+ *              i.e. 32 data Tx VCs + 12 OAM Tx Port VCs.
+ *              Receive VCs will have a value between 44-66,
+ *              i.e. 32 data Rx VCs + 1 OAM Rx VC.
+ *
+ * @return @li IX_SUCCESS : The VC has been successfully registered on
+ *       this port. The VC is ready for a client to configure IxAtmdAcc
+ *       for receive and transmit operations on the VC.
+ * @return @li IX_ATMM_RET_INVALID_PORT : The port value indicated in the
+ *       input is not valid or has not been initialized.  The request
+ *       is rejected.
+ * @return @li IX_ATMM_RET_INVALID_VC_DESCRIPTOR : The descriptor
+ *       pointed to by vcToAdd is invalid.  The registration request
+ *       is rejected.
+ * @return @li IX_ATMM_RET_VC_CONFLICT : The VC requested conflicts with
+ *      reserved VPI and/or VCI values or with another VC already activated
+ *      on this port.
+ * @return @li IX_ATMM_RET_PORT_CAPACITY_IS_FULL : The VC cannot be
+ *       registered in the port becuase the port capacity is
+ *       insufficient to support the requested ATM traffic contract.
+ *       The registration request is rejected.
+ * @return @li IX_ATMM_RET_INVALID_PARAM_PTR : A pointer parameter was
+ *       NULL.
+ *
+ * @warning IxAtmm has no capability of signaling or negotiating a virtual
+ *          connection. Negotiation of the admission of the VC to the network
+ *          is beyond the scope of this function.  This is assumed to be
+ *          performed by the calling client, if appropriate,
+ *          before or after this function is called.
+ */
+PUBLIC IX_STATUS
+ixAtmmVcRegister (IxAtmLogicalPort port,
+		  IxAtmmVc *vcToAdd,
+		  IxAtmSchedulerVcId *vcId);
+
+/**    
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmVcDeregister (IxAtmLogicalPort port, IxAtmSchedulerVcId vcId)
+ *
+ * @brief Function called by a client to deregister a VC from the
+ *         system.
+ *
+ *  With the removal of each new VC from a port, a series of
+ *  registered callback functions are invoked by the IxAtmm component
+ *  to notify possible external components of the change.  The callback
+ *  functions are registered using the @ref ixAtmmVcChangeCallbackRegister.
+ *
+ *  The IxAtmSch component is notified of the removal of transmit VCs.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Identifies port on which the VC to be
+ *          removed is currently registered.
+ *
+ * @param vcId @ref IxAtmSchedulerVcId [in] - VC identifier value of the VC to
+ *          be deregistered.  This value was supplied to the client when
+            the VC was originally registered.  This value can also be
+	    queried from the IxAtmm component through the @ref ixAtmmVcQuery
+ *          interface.
+ *
+ * @return @li IX_SUCCESS : The specified VC has been successfully
+ *       removed from this port.
+ * @return @li IX_ATMM_RET_INVALID_PORT : The port value indicated in the
+ *       input is not valid or has not been initialized.  The request
+ *       is rejected.
+ * @return @li IX_FAIL : There is no registered VC associated with the
+ *       supplied identifier registered on this port. */
+PUBLIC IX_STATUS
+ixAtmmVcDeregister (IxAtmLogicalPort port, IxAtmSchedulerVcId vcId);
+
+/**    
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmVcQuery (IxAtmLogicalPort port,
+	       unsigned vpi,
+	       unsigned vci,
+	       IxAtmmVcDirection direction,
+	       IxAtmSchedulerVcId *vcId,
+	       IxAtmmVc *vcDesc)
+ *
+ * @brief This interface supplies information about an active VC on a
+ *         particular port when supplied with the VPI, VCI and
+ *         direction of that VC.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Identifies port on which the VC to be
+ *          queried is currently registered.
+ *
+ * @param vpi unsigned [in] - ATM VPI value of the requested VC.
+ *
+ * @param vci unsigned [in] - ATM VCI value of the requested VC.
+ *
+ * @param direction @ref IxAtmmVcDirection [in] - One of @ref
+ *          IX_ATMM_VC_DIRECTION_TX or @ref IX_ATMM_VC_DIRECTION_RX
+ *          indicating the direction (Tx or Rx) of the requested VC.
+ *
+ * @param *vcId @ref IxAtmSchedulerVcId [out] - Pointer to an integer value which will be
+ *              filled with the VC identifier value for the requested
+ *              VC (as returned by @ref ixAtmmVcRegister), if it
+ *              exists on this port.
+ *
+ * @param *vcDesc @ref IxAtmmVc [out] - Pointer to an @ref IxAtmmVc structure
+ *              which will be filled with the specific details of the
+ *              requested VC, if it exists on this port.
+ *
+ * @return @li IX_SUCCESS : The specified VC has been found on this port
+ *       and the requested details have been returned.
+ * @return @li IX_ATMM_RET_INVALID_PORT : The port value indicated in the
+ *       input is not valid or has not been initialized.  The request
+ *       is rejected.
+ * @return @li IX_ATMM_RET_NO_SUCH_VC : No VC exists on the specified
+ *       port which matches the search criteria (VPI, VCI, direction)
+ *       given.  No data is returned.
+ * @return @li IX_ATMM_RET_INVALID_PARAM_PTR : A pointer parameter was
+ *       NULL.
+ *
+ */
+PUBLIC IX_STATUS
+ixAtmmVcQuery (IxAtmLogicalPort port,
+	       unsigned vpi,
+	       unsigned vci,
+	       IxAtmmVcDirection direction,
+	       IxAtmSchedulerVcId *vcId,
+	       IxAtmmVc *vcDesc);
+
+
+/**    
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmVcIdQuery (IxAtmLogicalPort port, IxAtmSchedulerVcId vcId, IxAtmmVc *vcDesc)
+ *
+ * @brief This interface supplies information about an active VC on a
+ *         particular port when supplied with a vcId for that VC.
+ *
+ * @param port @ref IxAtmLogicalPort [in] - Identifies port on which the VC to be
+ *          queried is currently registered.
+ *
+ * @param vcId @ref IxAtmSchedulerVcId [in] - Value returned by @ref ixAtmmVcRegister which
+ *          uniquely identifies the requested VC on this port.
+ *
+ * @param *vcDesc @ref IxAtmmVc [out] - Pointer to an @ref IxAtmmVc structure
+ *              which will be filled with the specific details of the
+ *              requested VC, if it exists on this port.
+ *
+ * @return @li IX_SUCCESS : The specified VC has been found on this port
+ *       and the requested details have been returned.
+ * @return @li IX_ATMM_RET_INVALID_PORT : The port value indicated in the
+ *       input is not valid or has not been initialized.  The request
+ *       is rejected.
+ * @return @li IX_ATMM_RET_NO_SUCH_VC : No VC exists on the specified
+ *       port which matches the supplied identifier.  No data is
+ *       returned.
+ * @return @li IX_ATMM_RET_INVALID_PARAM_PTR : A pointer parameter was
+ *       NULL.
+ */
+PUBLIC IX_STATUS
+ixAtmmVcIdQuery (IxAtmLogicalPort port, IxAtmSchedulerVcId vcId, IxAtmmVc *vcDesc);
+
+/**    
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmVcChangeCallbackRegister (IxAtmmVcChangeCallback callback)
+ *
+ * @brief This interface is invoked to supply a function to IxAtmm
+ *         which will be called to notify the client if a new VC is
+ *         registered with IxAtmm or an existing VC is removed.
+ *
+ * The callback, when invoked, will run within the context of the call
+ * to @ref ixAtmmVcRegister or @ref ixAtmmVcDeregister which caused
+ * the change of state.
+ *
+ * A maximum of 32 calbacks may be registered in with IxAtmm.
+ *
+ * @param callback @ref IxAtmmVcChangeCallback [in] - Callback which complies
+ *          with the @ref IxAtmmVcChangeCallback definition.  This
+ *          function will be invoked by IxAtmm with the appropiate
+ *          parameters for the relevant VC when any VC has been
+ *          registered or deregistered with IxAtmm.
+ *
+ * @return @li IX_SUCCESS : The specified callback has been registered
+ *      successfully with IxAtmm and will be invoked when appropriate.
+ * @return @li IX_FAIL : Either the supplied callback is invalid, or
+ *      IxAtmm has already registered 32 and connot accommodate
+ *      any further registrations of this type.  The request is
+ *      rejected.
+ *
+ * @warning The client must not call either the @ref
+ *          ixAtmmVcRegister or @ref ixAtmmVcDeregister interfaces
+ *          from within the supplied callback function.  */
+PUBLIC IX_STATUS ixAtmmVcChangeCallbackRegister (IxAtmmVcChangeCallback callback);
+
+
+/**    
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmVcChangeCallbackDeregister (IxAtmmVcChangeCallback callback)
+ *
+ * @brief This interface is invoked to deregister a previously supplied
+ *         callback function.
+ *
+ * @param callback @ref IxAtmmVcChangeCallback [in] - Callback which complies
+ *          with the @ref IxAtmmVcChangeCallback definition.  This
+ *          function will removed from the table of callbacks.
+ *
+ * @return @li IX_SUCCESS : The specified callback has been deregistered
+ *      successfully from IxAtmm.
+ * @return @li IX_FAIL : Either the supplied callback is invalid, or
+ *      is not currently registered with IxAtmm.
+ */
+PUBLIC IX_STATUS
+ixAtmmVcChangeCallbackDeregister (IxAtmmVcChangeCallback callback);
+
+/**    
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmUtopiaStatusShow (void)
+ * 
+ *  @brief Display utopia status counters
+ *
+ * @param "none"
+ *
+ * @return @li IX_SUCCESS : Show function was successful
+ * @return @li IX_FAIL : Internal failure
+ */
+PUBLIC IX_STATUS
+ixAtmmUtopiaStatusShow (void);
+
+/**     
+ * @ingroup IxAtmm
+ * 
+ * @fn ixAtmmUtopiaCfgShow (void)
+ *
+ * @brief Display utopia information(config registers and status registers)
+ *
+ * @param "none"
+ *
+ * @return @li IX_SUCCESS : Show function was successful
+ * @return @li IX_FAIL : Internal failure
+ */
+PUBLIC IX_STATUS
+ixAtmmUtopiaCfgShow (void);
+
+#endif
+/* IXATMM_H */
+
+/** @} */

cpu/ixp/npe/include/IxDmaAcc.h

@@ -0,0 +1,260 @@
+/**
+ * @file IxDmaAcc.h
+ *
+ * @date	15 October 2002 
+ *
+ * @brief   API of the IXP400 DMA Access Driver Component (IxDma)
+ *
+ *
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+/*---------------------------------------------------------------------
+   Doxygen group definitions
+  ---------------------------------------------------------------------*/
+
+#ifndef IXDMAACC_H
+#define IXDMAACC_H
+
+#include "IxOsal.h"
+#include "IxNpeDl.h"
+/**
+ * @defgroup IxDmaTypes IXP400 DMA Types (IxDmaTypes)
+ * @brief The common set of types used in the DMA component
+ * @{
+ */
+
+/** 
+ * @ingroup IxDmaTypes
+ * @enum IxDmaReturnStatus
+ * @brief Dma return status definitions
+ */
+typedef enum
+{
+    IX_DMA_SUCCESS = IX_SUCCESS,  /**< DMA Transfer Success */
+    IX_DMA_FAIL = IX_FAIL,        /**< DMA Transfer Fail */
+    IX_DMA_INVALID_TRANSFER_WIDTH, /**< Invalid transfer width */
+    IX_DMA_INVALID_TRANSFER_LENGTH, /**< Invalid transfer length */
+    IX_DMA_INVALID_TRANSFER_MODE, /**< Invalid transfer mode */
+    IX_DMA_INVALID_ADDRESS_MODE, /**< Invalid address mode */
+    IX_DMA_REQUEST_FIFO_FULL  /**< DMA request queue is full */
+} IxDmaReturnStatus;
+
+/** 
+ * @ingroup IxDmaTypes
+ * @enum IxDmaTransferMode
+ * @brief Dma transfer mode definitions
+ * @note Copy and byte swap, and copy and reverse modes only support multiples of word data length.
+ */
+typedef enum
+{
+    IX_DMA_COPY_CLEAR = 0,      /**< copy and clear source*/
+    IX_DMA_COPY,                /**< copy */
+    IX_DMA_COPY_BYTE_SWAP,      /**< copy and byte swap (endian) */
+    IX_DMA_COPY_REVERSE,        /**< copy and reverse */
+    IX_DMA_TRANSFER_MODE_INVALID /**< Invalid transfer mode */
+} IxDmaTransferMode;
+
+/** 
+ * @ingroup IxDmaTypes
+ * @enum IxDmaAddressingMode
+ * @brief Dma addressing mode definitions
+ * @note Fixed source address to fixed destination address addressing mode is not supported.
+ */
+typedef enum
+{
+    IX_DMA_INC_SRC_INC_DST = 0, /**< Incremental source address to incremental destination address */
+    IX_DMA_INC_SRC_FIX_DST,     /**< Incremental source address to incremental destination address */
+    IX_DMA_FIX_SRC_INC_DST,     /**< Incremental source address to incremental destination address */
+    IX_DMA_FIX_SRC_FIX_DST,     /**< Incremental source address to incremental destination address */
+    IX_DMA_ADDRESSING_MODE_INVALID /**< Invalid Addressing Mode */
+} IxDmaAddressingMode;
+
+/** 
+ * @ingroup IxDmaTypes
+ * @enum IxDmaTransferWidth
+ * @brief Dma transfer width definitions
+ * @Note Fixed addresses (either source or destination) do not support burst transfer width.
+ */
+typedef enum
+{
+    IX_DMA_32_SRC_32_DST = 0,  /**< 32-bit src to 32-bit dst */
+    IX_DMA_32_SRC_16_DST,      /**< 32-bit src to 16-bit dst */
+    IX_DMA_32_SRC_8_DST,       /**< 32-bit src to 8-bit dst */
+    IX_DMA_16_SRC_32_DST,      /**< 16-bit src to 32-bit dst */
+    IX_DMA_16_SRC_16_DST,      /**< 16-bit src to 16-bit dst */
+    IX_DMA_16_SRC_8_DST,       /**< 16-bit src to 8-bit dst */
+    IX_DMA_8_SRC_32_DST,       /**< 8-bit src to 32-bit dst */
+    IX_DMA_8_SRC_16_DST,       /**< 8-bit src to 16-bit dst */
+    IX_DMA_8_SRC_8_DST,        /**< 8-bit src to 8-bit dst */
+    IX_DMA_8_SRC_BURST_DST,    /**< 8-bit src to burst dst - Not supported for fixed destination address */
+    IX_DMA_16_SRC_BURST_DST,   /**< 16-bit src to burst dst - Not supported for fixed destination address */
+    IX_DMA_32_SRC_BURST_DST,   /**< 32-bit src to burst dst - Not supported for fixed destination address */
+    IX_DMA_BURST_SRC_8_DST,    /**< burst src to 8-bit dst  - Not supported for fixed source address */
+    IX_DMA_BURST_SRC_16_DST,   /**< burst src to 16-bit dst - Not supported for fixed source address */
+    IX_DMA_BURST_SRC_32_DST,   /**< burst src to 32-bit dst - Not supported for fixed source address*/
+    IX_DMA_BURST_SRC_BURST_DST, /**< burst src to burst dst  - Not supported for fixed source and destination address
+*/
+    IX_DMA_TRANSFER_WIDTH_INVALID /**< Invalid transfer width */
+} IxDmaTransferWidth;
+
+/** 
+ * @ingroup IxDmaTypes
+ * @enum IxDmaNpeId
+ * @brief NpeId numbers to identify NPE A, B or C
+ */
+typedef enum
+{
+    IX_DMA_NPEID_NPEA = 0, /**< Identifies NPE A */
+    IX_DMA_NPEID_NPEB,     /**< Identifies NPE B */
+    IX_DMA_NPEID_NPEC,     /**< Identifies NPE C */
+    IX_DMA_NPEID_MAX       /**< Total Number of NPEs */
+} IxDmaNpeId;
+/* @} */
+/**
+ * @defgroup IxDmaAcc IXP400 DMA Access Driver (IxDmaAcc) API
+ *
+ * @brief The public API for the IXP400 IxDmaAcc component
+ *
+ * @{
+ */
+
+/**
+ * @ingroup IxDmaAcc
+ * @brief DMA Request Id type
+ */
+typedef UINT32 IxDmaAccRequestId;
+
+/**
+ * @ingroup IxDmaAcc
+ * @def IX_DMA_REQUEST_FULL
+ * @brief DMA request queue is full
+ * This constant is a return value used to tell the user that the IxDmaAcc
+ * queue is full.
+ *
+ */
+#define IX_DMA_REQUEST_FULL 16
+
+/**
+ * @ingroup 	IxDmaAcc
+ * @brief       DMA completion notification
+ * This function is called to notify a client that the DMA has been completed
+ * @param status @ref IxDmaReturnStatus [out] - reporting to client
+ *
+ */
+typedef void (*IxDmaAccDmaCompleteCallback) (IxDmaReturnStatus status);
+
+/**
+ * @ingroup 	IxDmaAcc
+ * 
+ * @fn ixDmaAccInit(IxNpeDlNpeId npeId)
+ * 
+ * @brief 	Initialise the DMA Access component
+ * This function will initialise the DMA Access component internals
+ * @param npeId @ref IxNpeDlNpeId [in] - NPE to use for Dma Transfer
+ * @return @li IX_SUCCESS succesfully initialised the component
+ * @return @li IX_FAIL Initialisation failed for some unspecified
+ * internal reason.
+ */
+PUBLIC IX_STATUS
+ixDmaAccInit(IxNpeDlNpeId npeId);
+
+/**
+ * @ingroup 	IxDmaAcc
+ * 
+ * @fn ixDmaAccDmaTransfer(
+    IxDmaAccDmaCompleteCallback callback,
+    UINT32 SourceAddr,
+    UINT32 DestinationAddr,
+    UINT16 TransferLength,
+    IxDmaTransferMode TransferMode,
+    IxDmaAddressingMode AddressingMode,
+    IxDmaTransferWidth TransferWidth)
+ *
+ * @brief       Perform DMA transfer
+ * This function will perform DMA transfer between devices within the
+ * IXP400 memory map.
+ * @note The following are restrictions for IxDmaAccDmaTransfer:
+ *      @li The function is non re-entrant.
+ *      @li The function assumes host devices are operating in big-endian mode.
+ *      @li Fixed address does not suport burst transfer width
+ *      @li Fixed source address to fixed destinatiom address mode is not suported
+ *      @li The incrementing source address for expansion bus will not support a burst transfer width and copy and clear mode
+ *
+ * @param callback @ref IxDmaAccDmaCompleteCallback [in] - function pointer to be stored and called when the DMA transfer is completed. This cannot be NULL.
+ * @param SourceAddr UINT32 [in] -	Starting address of DMA source. Must be a valid IXP400 memory map address.
+ * @param DestinationAddr UINT32 [in] - Starting address of DMA destination. Must be a valid IXP400 memory map address.
+ * @param TransferLength UINT16 [in] - The size of DMA data transfer. The range must be from 1-64Kbyte
+ * @param TransferMode @ref IxDmaTransferMode [in] - The DMA transfer mode
+ * @param AddressingMode @ref IxDmaAddressingMode [in] - The DMA addressing mode
+ * @param TransferWidth	@ref IxDmaTransferWidth [in] - The DMA transfer width
+ *
+ * @return @li IX_DMA_SUCCESS 	Notification that the DMA request is succesful
+ * @return @li IX_DMA_FAIL 	IxDmaAcc not yet initialised or some internal error has occured
+ * @return @li IX_DMA_INVALID_TRANSFER_WIDTH Transfer width is nit valid
+ * @return @li IX_DMA_INVALID_TRANSFER_LENGTH Transfer length outside of valid range
+ * @return @li IX_DMA_INVALID_TRANSFER_MODE Transfer Mode not valid
+ * @return @li IX_DMA_REQUEST_FIFO_FULL IxDmaAcc request queue is full
+ */
+PUBLIC IxDmaReturnStatus
+ixDmaAccDmaTransfer(
+    IxDmaAccDmaCompleteCallback callback,
+    UINT32 SourceAddr,
+    UINT32 DestinationAddr,
+    UINT16 TransferLength,
+    IxDmaTransferMode TransferMode,
+    IxDmaAddressingMode AddressingMode,
+    IxDmaTransferWidth TransferWidth);
+/**
+ * @ingroup IxDmaAcc
+ *
+ * @fn ixDmaAccShow(void)
+ *
+ * @brief Display some component information for debug purposes
+ * Show some internal operation information relating to the DMA service.
+ * At a minimum the following will show.
+ * - the number of the DMA pend (in queue)
+ * @param None
+ * @return @li None
+ */
+PUBLIC IX_STATUS
+ixDmaAccShow(void);
+
+#endif /* IXDMAACC_H */
+

cpu/ixp/npe/include/IxEthAcc.h

@@ -0,0 +1,2512 @@
+/** @file    IxEthAcc.h
+ *
+ * @brief this file contains the public API of @ref IxEthAcc component
+ *
+ * Design notes:
+ * The IX_OSAL_MBUF address is to be specified on bits [31-5] and must 
+ * be cache aligned (bits[4-0] cleared)
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ *
+ */
+
+#ifndef IxEthAcc_H
+#define IxEthAcc_H
+
+#include <IxOsBuffMgt.h>
+#include <IxTypes.h>
+
+/**
+ * @defgroup IxEthAcc IXP400 Ethernet Access (IxEthAcc) API
+ *
+ * @brief ethAcc is a library that does provides access to the internal IXP400 10/100Bt Ethernet MACs.
+ *
+ *@{
+ */
+
+/**
+ * @ingroup IxEthAcc
+ * @brief Definition of the Ethernet Access status
+ */
+typedef enum /* IxEthAccStatus */
+{
+    IX_ETH_ACC_SUCCESS = IX_SUCCESS, /**< return success*/
+    IX_ETH_ACC_FAIL = IX_FAIL, /**< return fail*/
+    IX_ETH_ACC_INVALID_PORT, /**< return invalid port*/
+    IX_ETH_ACC_PORT_UNINITIALIZED, /**< return uninitialized*/
+    IX_ETH_ACC_MAC_UNINITIALIZED, /**< return MAC uninitialized*/
+    IX_ETH_ACC_INVALID_ARG, /**< return invalid arg*/
+    IX_ETH_TX_Q_FULL, /**< return tx queue is full*/
+    IX_ETH_ACC_NO_SUCH_ADDR /**< return no such address*/
+} IxEthAccStatus;
+
+/**
+ * @ingroup IxEthAcc
+ * @enum IxEthAccPortId
+ * @brief Definition of the IXP400 Mac Ethernet device.
+ */
+typedef enum  
+{
+	IX_ETH_PORT_1 = 0, /**< Ethernet Port 1 */
+	IX_ETH_PORT_2 = 1  /**< Ethernet port 2 */
+	,IX_ETH_PORT_3 = 2 /**< Ethernet port 3 */
+} IxEthAccPortId;
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETH_ACC_NUMBER_OF_PORTS
+ *
+ * @brief  Definition of the number of ports
+ *
+ */
+#ifdef __ixp46X
+#define IX_ETH_ACC_NUMBER_OF_PORTS (3)
+#else
+#define IX_ETH_ACC_NUMBER_OF_PORTS (2)
+#endif
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_IEEE803_MAC_ADDRESS_SIZE
+ *
+ * @brief  Definition of the size of the MAC address
+ *
+ */
+#define IX_IEEE803_MAC_ADDRESS_SIZE (6)
+
+
+/**
+ *
+ * @brief Definition of the IEEE 802.3 Ethernet MAC address structure.
+ *
+ * The data should be packed with bytes xx:xx:xx:xx:xx:xx 
+ * @note
+ * The data must be packed in network byte order.
+ */
+typedef struct  
+{
+    UINT8 macAddress[IX_IEEE803_MAC_ADDRESS_SIZE]; /**< MAC address */
+} IxEthAccMacAddr;
+
+/**
+ * @ingroup IxEthAcc
+ * @def IX_ETH_ACC_NUM_TX_PRIORITIES
+ * @brief Definition of the number of transmit priorities
+ * 
+ */
+#define IX_ETH_ACC_NUM_TX_PRIORITIES (8)
+
+/**
+ * @ingroup IxEthAcc
+ * @enum IxEthAccTxPriority
+ * @brief Definition of the relative priority used to transmit a frame
+ * 
+ */
+typedef enum  
+{
+	IX_ETH_ACC_TX_PRIORITY_0 = 0, /**<Lowest Priority submission */
+	IX_ETH_ACC_TX_PRIORITY_1 = 1, /**<submission prority of 1 (0 is lowest)*/
+	IX_ETH_ACC_TX_PRIORITY_2 = 2, /**<submission prority of 2 (0 is lowest)*/
+	IX_ETH_ACC_TX_PRIORITY_3 = 3, /**<submission prority of 3 (0 is lowest)*/
+	IX_ETH_ACC_TX_PRIORITY_4 = 4, /**<submission prority of 4 (0 is lowest)*/
+	IX_ETH_ACC_TX_PRIORITY_5 = 5, /**<submission prority of 5 (0 is lowest)*/
+	IX_ETH_ACC_TX_PRIORITY_6 = 6, /**<submission prority of 6 (0 is lowest)*/
+	IX_ETH_ACC_TX_PRIORITY_7 = 7, /**<Highest priority submission */
+
+	IX_ETH_ACC_TX_DEFAULT_PRIORITY = IX_ETH_ACC_TX_PRIORITY_0 /**< By default send all 
+								 packets with lowest priority */
+} IxEthAccTxPriority;
+
+/**
+ * @ingroup IxEthAcc
+ * @enum IxEthAccRxFrameType
+ * @brief Identify the type of a frame.
+ * 
+ * @sa IX_ETHACC_NE_FLAGS
+ * @sa IX_ETHACC_NE_LINKMASK
+ */
+typedef enum  
+{
+	IX_ETHACC_RX_LLCTYPE = 0x00, /**< 802.3 - 8802, with LLC/SNAP */
+	IX_ETHACC_RX_ETHTYPE = 0x10, /**< 802.3 (Ethernet) without LLC/SNAP */
+	IX_ETHACC_RX_STATYPE = 0x20, /**< 802.11, AP <=> STA */
+	IX_ETHACC_RX_APTYPE  = 0x30  /**< 802.11, AP <=> AP */
+} IxEthAccRxFrameType;
+
+/**
+ * @ingroup IxEthAcc
+ * @enum IxEthAccDuplexMode
+ * @brief Definition to provision the duplex mode of the MAC. 
+ * 
+ */
+typedef enum
+{
+    IX_ETH_ACC_FULL_DUPLEX, /**< Full duplex operation of the MAC */
+    IX_ETH_ACC_HALF_DUPLEX  /**< Half duplex operation of the MAC */
+} IxEthAccDuplexMode;
+
+
+/**
+ * @ingroup IxEthAcc
+ * @struct IxEthAccNe
+ * @brief Definition of service-specific informations.
+ * 
+ * This structure defines the Ethernet service-specific informations
+ * and enable QoS and VLAN features.
+ */
+typedef struct
+{
+    UINT32 ixReserved_next;    /**< reserved for chaining */
+    UINT32 ixReserved_lengths; /**< reserved for buffer lengths */
+    UINT32 ixReserved_data;    /**< reserved for buffer pointer */
+    UINT8  ixDestinationPortId; /**< Destination portId for this packet, if known by NPE */
+    UINT8  ixSourcePortId; /**< Source portId for this packet */
+    UINT16 ixFlags;        /**< BitField of option for this frame */
+    UINT8  ixQoS;          /**< QoS class of the frame */
+    UINT8  ixReserved;     /**< reserved */
+    UINT16 ixVlanTCI;      /**< Vlan TCI */
+    UINT8  ixDestMac[IX_IEEE803_MAC_ADDRESS_SIZE]; /**< Destination MAC address */
+    UINT8  ixSourceMac[IX_IEEE803_MAC_ADDRESS_SIZE]; /**< Source MAC address */
+} IxEthAccNe;
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_PORT_UNKNOWN
+ *
+ * @brief  Contents of the field @a IX_ETHACC_NE_DESTPORTID when no
+ * destination port can be found by the NPE for this frame.
+ *
+ */
+#define IX_ETHACC_NE_PORT_UNKNOWN   (0xff)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_DESTMAC
+ *
+ * @brief The location of the destination MAC address in the Mbuf header.
+ *
+ */
+#define IX_ETHACC_NE_DESTMAC(mBufPtr) ((IxEthAccNe *)&((mBufPtr)->ix_ne))->ixDestMac
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_SOURCEMAC
+ *
+ * @brief The location of the source MAC address in the Mbuf header.
+ *
+ */
+#define IX_ETHACC_NE_SOURCEMAC(mBufPtr) ((IxEthAccNe *)&((mBufPtr)->ix_ne))->ixSourceMac
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_VLANTCI
+ *
+ * @brief The VLAN Tag Control Information associated with this frame
+ * 
+ * The VLAN Tag Control Information associated with this frame. On Rx
+ * path, this field is extracted from the packet header.
+ * On Tx path, the value of this field is inserted in the frame when
+ * the port is configured to insert or replace vlan tags in the 
+ * egress frames.
+ *
+ * @sa IX_ETHACC_NE_FLAGS
+ */
+#define IX_ETHACC_NE_VLANTCI(mBufPtr) ((IxEthAccNe *)&((mBufPtr)->ix_ne))->ixVlanTCI
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_SOURCEPORTID
+ *
+ * @brief The port where this frame came from.
+ *
+ * The port where this frame came from. This field is set on receive
+ * with the port information. This field is ignored on Transmit path.
+ */
+#define IX_ETHACC_NE_SOURCEPORTID(mBufPtr) ((IxEthAccNe *)&((mBufPtr)->ix_ne))->ixSourcePortId
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_DESTPORTID
+ *
+ * @brief The destination port where this frame should be sent.
+ *
+ * The destination port where this frame should be sent.
+ *
+ * @li In the transmit direction, this field contains the destination port
+ * and is ignored unless @a IX_ETHACC_NE_FLAG_DST is set.
+ * 
+ * @li In the receive direction, this field contains the port where the
+ * destination MAC addresses has been learned. If the destination
+ * MAC address is unknown, then this value is set to the reserved value
+ * @a IX_ETHACC_NE_PORT_UNKNOWN
+ *
+ */
+#define IX_ETHACC_NE_DESTPORTID(mBufPtr) ((IxEthAccNe *)&((mBufPtr)->ix_ne))->ixDestinationPortId
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_QOS
+ *
+ * @brief QualityOfService class (QoS) for this received frame.
+ *
+ */
+#define IX_ETHACC_NE_QOS(mBufPtr) ((IxEthAccNe *)&((mBufPtr)->ix_ne))->ixQoS
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_FLAGS
+ *
+ * @brief Bit Mask of the different flags associated with a frame
+ * 
+ * The flags are the bit-oring combination 
+ * of the following different fields :
+ *
+ *      @li IP flag (Rx @a IX_ETHACC_NE_IPMASK)
+ *      @li Spanning Tree flag (Rx @a IX_ETHACC_NE_STMASK)
+ *      @li Link layer type (Rx and Tx @a IX_ETHACC_NE_LINKMASK)
+ *      @li VLAN Tagged Frame (Rx @a IX_ETHACC_NE_VLANMASK)
+ *      @li New source MAC address (Rx @a IX_ETHACC_NE_NEWSRCMASK)
+ *      @li Multicast flag (Rx @a IX_ETHACC_NE_MCASTMASK)
+ *      @li Broadcast flag (Rx @a IX_ETHACC_NE_BCASTMASK)
+ *      @li Destination port flag (Tx @a IX_ETHACC_NE_PORTMASK)
+ *      @li Tag/Untag Tx frame (Tx @a IX_ETHACC_NE_TAGMODEMASK)
+ *      @li Overwrite destination port (Tx @a IX_ETHACC_NE_PORTOVERMASK)
+ *      @li Filtered frame (Rx @a IX_ETHACC_NE_STMASK)
+ *      @li VLAN Enabled (Rx and Tx @a IX_ETHACC_NE_VLANENABLEMASK)
+ */
+#define IX_ETHACC_NE_FLAGS(mBufPtr) ((IxEthAccNe *)&((mBufPtr)->ix_ne))->ixFlags
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_BCASTMASK
+ *
+ * @brief This mask defines if a received frame is a broadcast frame.
+ *
+ * This mask defines if a received frame is a broadcast frame.
+ * The BCAST flag is set when the destination MAC address of 
+ * a frame is broadcast.
+ *
+ * @sa IX_ETHACC_NE_FLAGS 
+ *
+ */
+#define IX_ETHACC_NE_BCASTMASK      (0x1)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_MCASTMASK
+ *
+ * @brief This mask defines if a received frame is a multicast frame.
+ *
+ * This mask defines if a received frame is a multicast frame.
+ * The MCAST flag is set when the destination MAC address of 
+ * a frame is multicast.
+ *
+ * @sa IX_ETHACC_NE_FLAGS 
+ *
+ */
+#define IX_ETHACC_NE_MCASTMASK      (0x1 << 1)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_IPMASK
+ *
+ * @brief This mask defines if a received frame is a IP frame.
+ *
+ * This mask applies to @a IX_ETHACC_NE_FLAGS and defines if a received 
+ * frame is a IP frame. The IP flag is set on Rx direction, depending on 
+ * the frame contents. The flag is set when the length/type field of a 
+ * received frame is 0x8000.
+ *
+ * @sa IX_ETHACC_NE_FLAGS
+ *
+ */
+#define IX_ETHACC_NE_IPMASK         (0x1 << 2)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_VLANMASK
+ *
+ * @brief This mask defines if a received frame is VLAN tagged.
+ *
+ * This mask defines if a received frame is VLAN tagged.
+ * When set, the Rx frame is VLAN-tagged and the tag value 
+ * is available thru @a IX_ETHACC_NE_VLANID.
+ * Note that when sending frames which are already tagged
+ * this flag should be set, to avoid inserting another VLAN tag.
+ *
+ * @sa IX_ETHACC_NE_FLAGS 
+ * @sa IX_ETHACC_NE_VLANID
+ *
+ */
+#define IX_ETHACC_NE_VLANMASK       (0x1 << 3)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_LINKMASK
+ *
+ * @brief This mask is the link layer protocol indicator
+ *
+ * This mask applies to @a IX_ETHACC_NE_FLAGS.
+ * It reflects the state of a frame as it exits an NPE on the Rx path
+ * or enters an NPE on the Tx path. Its values are as follows:
+ *      @li 0x00 - IEEE802.3 - 8802 (Rx) / IEEE802.3 - 8802 (Tx)
+ *      @li 0x01 - IEEE802.3 - Ethernet (Rx) / IEEE802.3 - Ethernet (Tx)
+ *      @li 0x02 - IEEE802.11 AP -> STA (Rx) / IEEE802.11 STA -> AP (Tx)
+ *      @li 0x03 - IEEE802.11 AP -> AP (Rx) / IEEE802.11 AP->AP (Tx)
+ *
+ * @sa IX_ETHACC_NE_FLAGS
+ *
+ */
+#define IX_ETHACC_NE_LINKMASK       (0x3 << 4)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_STMASK
+ *
+ * @brief This mask defines if a received frame is a Spanning Tree frame.
+ *
+ * This mask applies to @a IX_ETHACC_NE_FLAGS.
+ * On rx direction, it defines if a received if frame is a Spanning Tree frame.
+ * Setting this fkag on transmit direction overrides the port settings 
+ * regarding the VLAN options and 
+ *
+ * @sa IX_ETHACC_NE_FLAGS
+ *
+ */
+#define IX_ETHACC_NE_STMASK         (0x1 << 6)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_FILTERMASK
+ *
+ * @brief This bit indicates whether a frame has been filtered by the Rx service.
+ *
+ * This mask applies to @a IX_ETHACC_NE_FLAGS.
+ * Certain frames, which should normally be fully filtered by the NPE to due
+ * the destination MAC address being on the same segment as the Rx port are
+ * still forwarded to the XScale (although the payload is invalid) in order
+ * to learn the MAC address of the transmitting station, if this is unknown.
+ * Normally EthAcc will filter and recycle these framess internally and no
+ * frames with the FILTER bit set will be received by the client.
+ *
+ * @sa IX_ETHACC_NE_FLAGS
+ *
+ */
+#define IX_ETHACC_NE_FILTERMASK     (0x1 << 7)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_PORTMASK
+ *
+ * @brief This mask defines the rule to transmit a frame
+ *
+ * This mask defines the rule to transmit a frame. When set, a frame
+ * is transmitted to the destination port as set by the macro
+ * @a IX_ETHACC_NE_DESTPORTID. If not set, the destination port 
+ * is searched using the destination MAC address.
+ *
+ * @note This flag is meaningful only for multiport Network Engines.
+ * 
+ * @sa IX_ETHACC_NE_FLAGS 
+ * @sa IX_ETHACC_NE_DESTPORTID
+ *
+ */
+#define IX_ETHACC_NE_PORTOVERMASK   (0x1 << 8)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_TAGMODEMASK
+ *
+ * @brief This mask defines the tagging rules to apply to a transmit frame.
+ *
+ * This mask defines the tagging rules to apply to a transmit frame
+ * regardless of the default setting for a port. When used together 
+ * with @a IX_ETHACC_NE_TAGOVERMASK and when set, the 
+ * frame will be tagged prior to transmission. When not set,
+ * the frame will be untagged prior to transmission. This is accomplished
+ * irrespective of the Egress tagging rules, constituting a per-frame override.
+ *
+ * @sa IX_ETHACC_NE_FLAGS
+ * @sa IX_ETHACC_NE_TAGOVERMASK 
+ *
+ */
+#define IX_ETHACC_NE_TAGMODEMASK    (0x1 << 9)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_TAGOVERMASK
+ *
+ * @brief This mask defines the rule to transmit a frame
+ *
+ * This mask defines the rule to transmit a frame. When set, the
+ * default transmit rules of a port are overriden.
+ * When not set, the default rules as set by @ref IxEthDB should apply.
+ *
+ * @sa IX_ETHACC_NE_FLAGS
+ * @sa IX_ETHACC_NE_TAGMODEMASK
+ *
+ */
+#define IX_ETHACC_NE_TAGOVERMASK    (0x1 << 10)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_VLANENABLEMASK
+ *
+ * @brief This mask defines if a frame is a VLAN frame or not
+ *
+ * When set, frames undergo normal VLAN processing on the Tx path
+ * (membership filtering, tagging, tag removal etc). If this flag is
+ * not set, the frame is considered to be a regular non-VLAN frame
+ * and no VLAN processing will be performed.
+ *
+ * Note that VLAN-enabled NPE images will always set this flag in all
+ * Rx frames, and images which are not VLAN enabled will clear this
+ * flag for all received frames.
+ *
+ * @sa IX_ETHACC_NE_FLAGS
+ *
+ */
+#define IX_ETHACC_NE_VLANENABLEMASK (0x1 << 14)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IX_ETHACC_NE_NEWSRCMASK
+ *
+ * @brief This mask defines if a received frame has been learned.
+ *
+ * This mask defines if the source MAC address of a frame is 
+ * already known. If the bit is set, the source MAC address was
+ * unknown to the NPE at the time the frame was received.
+ *
+ * @sa IX_ETHACC_NE_FLAGS 
+ *
+ */
+#define IX_ETHACC_NE_NEWSRCMASK     (0x1 << 15)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @brief This defines the recommanded minimum size of MBUF's submitted
+ * to the frame receive service.
+ *
+ */
+#define IX_ETHACC_RX_MBUF_MIN_SIZE (2048)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @brief This defines the highest MII address of any attached PHYs 
+ * 
+ * The maximum number for PHY address is 31, add on for range checking.
+ *
+ */
+#define IXP425_ETH_ACC_MII_MAX_ADDR   32
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccInit(void)
+ * 
+ * @brief Initializes the IXP400 Ethernet Access Service.
+ * 
+ * @li Reentrant    - no
+ * @li ISR Callable - no
+ * 
+ * This should be called once per module initialization.
+ * @pre
+ *   The NPE must first be downloaded with the required microcode which supports all
+ *   required features.
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_FAIL	:  Service has failed to initialize.
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus ixEthAccInit(void);
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccUnload(void)
+ * 
+ * @brief Unload the Ethernet Access Service.
+ * 
+ * @li Reentrant    - no
+ * @li ISR Callable - no
+ *
+ * @return void 
+ *
+ * <hr>
+ */
+PUBLIC void ixEthAccUnload(void);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortInit( IxEthAccPortId portId)
+ *
+ * @brief Initializes an NPE/Ethernet MAC Port.
+ *
+ * The NPE/Ethernet port initialisation includes the following steps
+ * @li Initialize the NPE/Ethernet MAC hardware.
+ * @li Verify NPE downloaded and operational.
+ * @li The NPE shall be available for usage once this API returns.
+ * @li Verify that the Ethernet port is present before initializing
+ *
+ * @li Reentrant    - no
+ * @li ISR Callable - no
+ *
+ * This should be called once per mac device.
+ * The NPE/MAC shall be in disabled state after init.
+ *
+ * @pre
+ *   The component must be initialized via @a ixEthAccInit
+ *   The NPE must first be downloaded with the required microcode which supports all
+ *   required features.
+ *
+ * Dependant on Services: (Must be initialized before using this service may be initialized)
+ * 	ixNPEmh - NPE Message handling service.
+ * 	ixQmgr	- Queue Manager component.
+ *
+ * @param portId  @ref IxEthAccPortId [in]
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS: if the ethernet port is not present, a warning is issued.
+ * @li @a IX_ETH_ACC_FAIL : The NPE processor has failed to initialize.
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus ixEthAccPortInit(IxEthAccPortId portId);
+
+
+/*************************************************************************
+
+ #####     ##     #####    ##            #####     ##     #####  #    #
+ #    #   #  #      #     #  #           #    #   #  #      #    #    #
+ #    #  #    #     #    #    #          #    #  #    #     #    ######
+ #    #  ######     #    ######          #####   ######     #    #    #
+ #    #  #    #     #    #    #          #       #    #     #    #    #
+ #####   #    #     #    #    #          #       #    #     #    #    #
+
+*************************************************************************/
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortTxFrameSubmit( 
+    IxEthAccPortId portId,
+    IX_OSAL_MBUF *buffer, 
+    IxEthAccTxPriority priority)
+ * 
+ * @brief This function shall be used to submit MBUFs buffers for transmission on a particular MAC device. 
+ *
+ * When the frame is transmitted, the buffer shall be returned thru the 
+ * callback @a IxEthAccPortTxDoneCallback.
+ *
+ * In case of over-submitting, the order of the frames on the 
+ * network may be modified.
+ *
+ * Buffers shall be not queued for transmission if the port is disabled.
+ * The port can be enabled using @a ixEthAccPortEnable
+ *
+ * 
+ * @li Reentrant    - yes
+ * @li ISR Callable - yes
+ *
+ *
+ * @pre 
+ *  @a ixEthAccPortTxDoneCallbackRegister must be called to register a function to allow this service to
+ *   return the buffer to the calling service. 
+ * 
+ * @note 
+ *  If the buffer submit fails for any reason the user has retained ownership of the buffer.
+ *
+ * @param portId @ref IxEthAccPortId [in] - MAC port ID to transmit Ethernet frame on.
+ * @param buffer @ref IX_OSAL_MBUF [in] - pointer to an MBUF formatted buffer. Chained buffers are supported for transmission.
+ *             Chained packets are not supported and the field IX_OSAL_MBUF_NEXT_PKT_IN_CHAIN_PTR is ignored. 
+ * @param priority @ref IxEthAccTxPriority [in]
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_FAIL  : Failed to queue frame for transmission. 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+
+PUBLIC IxEthAccStatus ixEthAccPortTxFrameSubmit( 
+    IxEthAccPortId portId,
+    IX_OSAL_MBUF *buffer, 
+    IxEthAccTxPriority priority);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @brief Function prototype for Ethernet Tx Buffer Done callback. Registered 
+ *  via @a ixEthAccTxBufferDoneCallbackRegister 
+ * 
+ * This function is called once the previously submitted buffer is no longer required by this service.
+ * It may be returned upon successful transmission of the frame or during the shutdown of 
+ * the port prior to the transmission of a queued frame.
+ * The calling of this registered function is not a guarantee of successful transmission of the buffer.
+ *
+ *  
+ * @li Reentrant    - yes , The user provided function should be reentrant.
+ * @li ISR Callable - yes , The user provided function must be callable from an ISR.
+ *
+ *
+ * <b>Calling Context </b>: 
+ * @par
+ *   This callback is called in the context of the queue manager dispatch loop @a ixQmgrgrDispatcherLoopRun
+ *   within the @ref IxQMgrAPI component. The calling context may be from interrupt or high priority thread. 
+ *   The decision is system specific.
+ *
+ * @param callbackTag UINT32 [in] - This tag is that provided when the callback was registered for a particular MAC 
+ * via @a ixEthAccPortTxDoneCallbackRegister. It allows the same callback to be used for multiple MACs.
+ * @param mbuf @ref IX_OSAL_MBUF [in] - Pointer to the Tx mbuf descriptor. 
+ * 
+ * @return void
+ *
+ * @note
+ * The field IX_OSAL_MBUF_NEXT_PKT_IN_CHAIN_PTR is modified by the access layer and reset to NULL.
+ *
+ * <hr>
+ */
+typedef void (*IxEthAccPortTxDoneCallback) ( UINT32 callbackTag, IX_OSAL_MBUF *buffer );
+
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortTxDoneCallbackRegister( IxEthAccPortId portId, 
+					   IxEthAccPortTxDoneCallback txCallbackFn, 
+ 					   UINT32 callbackTag)
+ *
+ * @brief Register a callback function to allow 
+ * the transmitted buffers to return to the user.
+ * 
+ * This function registers the transmit buffer done function callback for a particular port.
+ *
+ * The registered callback function is called once the previously submitted buffer is no longer required by this service.
+ * It may be returned upon successful transmission of the frame or  shutdown of port prior to submission.
+ * The calling of this registered function is not a guarantee of successful transmission of the buffer.
+ *
+ * If called several times the latest callback shall be registered for a particular port.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - yes
+ *
+ * @pre
+ *	The port must be initialized via @a ixEthAccPortInit
+ *
+ *
+ * @param portId @ref IxEthAccPortId [in] - Register callback for a particular MAC device.
+ * @param txCallbackFn @ref IxEthAccPortTxDoneCallback [in] - Function to be called to return transmit buffers to the user.
+ * @param callbackTag UINT32 [in] -  This tag shall be provided to the callback function.
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ * @li @a IX_ETH_ACC_INVALID_ARG : An argument other than portId is invalid.
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortTxDoneCallbackRegister(IxEthAccPortId portId,
+								   IxEthAccPortTxDoneCallback txCallbackFn,
+								   UINT32 callbackTag);
+
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @brief Function prototype for Ethernet Frame Rx callback. Registered via @a ixEthAccPortRxCallbackRegister 
+ * 
+ * It is the responsibility of the user function to free any MBUF's which it receives.
+ *  
+ * @li Reentrant    - yes , The user provided function should be reentrant.
+ * @li ISR Callable - yes , The user provided function must be callable from an ISR.
+ * @par
+ *
+ * This function dispatches frames to the user level
+ * via the provided function. The invocation shall be made for each
+ * frame dequeued from the Ethernet QM queue. The user is required to free any MBUF's 
+ * supplied via this callback. In addition the registered callback must free up MBUF's
+ * from the receive free queue when the port is disabled 
+ * 
+ * If called several times the latest callback shall be registered for a particular port.
+ *
+ * <b>Calling Context </b>: 
+ * @par
+ *   This callback is called in the context of the queue manager dispatch loop @a ixQmgrgrDispatcherLoopRun
+ *   within the @ref IxQMgrAPI component. The calling context may be from interrupt or high priority thread. 
+ *   The decision is system specific.
+ *
+ *
+ * @param callbackTag UINT32 [in] - This tag is that provided when the callback was registered for a particular MAC 
+ * via @a ixEthAccPortRxCallbackRegister. It allows the same callback to be used for multiple MACs.
+ * @param mbuf @ref IX_OSAL_MBUF [in] - Pointer to the Rx mbuf header. Mbufs may be chained if 
+ *               the frame length is greater than the supplied mbuf length.
+ * @param reserved [in] - deprecated parameter The information is passed 
+ *      thru the IxEthAccNe header destination port ID field 
+ *      (@sa IX_ETHACC_NE_DESTPORTID). For backward 
+ *      compatibility,the value is equal to IX_ETH_DB_UNKNOWN_PORT (0xff). 
+ * 
+ * @return void
+ *
+ * @note
+ * Buffers may not be filled up to the length supplied in 
+ * @a ixEthAccPortRxFreeReplenish(). The firmware fills
+ * them to the previous 64 bytes boundary. The user has to be aware 
+ * that the length of the received mbufs may be smaller than the length
+ * of the supplied mbufs. 
+ * The mbuf header contains the following modified field
+ * @li @a IX_OSAL_MBUF_PKT_LEN is set in the header of the first mbuf and indicates
+ *  the total frame size
+ * @li @a IX_OSAL_MBUF_MLEN is set each mbuf header and indicates the payload length
+ * @li @a IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR contains a pointer to the next 
+ *     mbuf, or NULL at the end of a chain.
+ * @li @a IX_OSAL_MBUF_NEXT_PKT_IN_CHAIN_PTR is modified. Its value is reset to NULL
+ * @li @a IX_OSAL_MBUF_FLAGS contains the bit 4 set for a broadcast packet and the bit 5
+ *     set for a multicast packet. Other bits are unmodified.
+ *
+ * <hr>
+ */
+typedef void (*IxEthAccPortRxCallback) (UINT32 callbackTag, IX_OSAL_MBUF *buffer, UINT32 reserved);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @brief Function prototype for Ethernet Frame Rx callback. Registered via @a ixEthAccPortMultiBufferRxCallbackRegister 
+ * 
+ * It is the responsibility of the user function to free any MBUF's which it receives.
+ *  
+ * @li Reentrant    - yes , The user provided function should be reentrant.
+ * @li ISR Callable - yes , The user provided function must be callable from an ISR.
+ * @par
+ *
+ * This function dispatches many frames to the user level
+ * via the provided function. The invocation shall be made for multiple frames
+ * dequeued from the Ethernet QM queue. The user is required to free any MBUF's 
+ * supplied via this callback. In addition the registered callback must free up MBUF's
+ * from the receive free queue when the port is disabled 
+ * 
+ * If called several times the latest callback shall be registered for a particular port.
+ *
+ * <b>Calling Context </b>: 
+ * @par
+ *   This callback is called in the context of the queue manager dispatch loop @a ixQmgrDispatcherLoopRun
+ *   within the @ref IxQMgrAPI component. The calling context may be from interrupt or high priority thread. 
+ *   The decision is system specific.
+ *
+ *
+ * @param callbackTag - This tag is that provided when the callback was registered for a particular MAC 
+ * via @a ixEthAccPortMultiBufferRxCallbackRegister. It allows the same callback to be used for multiple MACs.
+ * @param mbuf - Pointer to an array of Rx mbuf headers. Mbufs 
+ *               may be chained if 
+ *               the frame length is greater than the supplied mbuf length.
+ *               The end of the array contains a zeroed entry (NULL pointer).
+ *
+ * @return void
+ *
+ * @note The mbufs passed to this callback have the same structure than the
+ *  buffers passed to @a IxEthAccPortRxCallback interfac. 
+ *
+ * @note The usage of this callback is exclusive with the usage of
+ *  @a ixEthAccPortRxCallbackRegister and @a IxEthAccPortRxCallback 
+ *
+ * @sa ixEthAccPortMultiBufferRxCallbackRegister
+ * @sa IxEthAccPortMultiBufferRxCallback
+ * @sa ixEthAccPortRxCallbackRegister
+ * @sa IxEthAccPortRxCallback
+ * <hr>
+ */
+
+typedef void (*IxEthAccPortMultiBufferRxCallback) (UINT32 callbackTag, IX_OSAL_MBUF **buffer);
+
+
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortRxCallbackRegister( IxEthAccPortId portId, IxEthAccPortRxCallback rxCallbackFn, UINT32 callbackTag)
+ *
+ * @brief Register a callback function to allow 
+ * the reception of frames.
+ *
+ * The registered callback function is called once a frame is received  by this service.
+ *
+ * If called several times the latest callback shall be registered for a particular port.
+ *
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - yes
+ *
+ *
+ * @param portId @ref IxEthAccPortId [in] - Register callback for a particular MAC device.
+ * @param rxCallbackFn @ref IxEthAccPortRxCallback [in] - Function to be called when Ethernet frames are availble.
+ * @param callbackTag UINT32 [in] -  This tag shall be provided to the callback function.
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ * @li @a IX_ETH_ACC_INVALID_ARG : An argument other than portId is invalid.
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortRxCallbackRegister(IxEthAccPortId portId,
+							   IxEthAccPortRxCallback rxCallbackFn,
+							   UINT32 callbackTag);
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortMultiBufferRxCallbackRegister( IxEthAccPortId portId, IxEthAccPortMultiBufferRxCallback rxCallbackFn, UINT32 callbackTag)
+ *
+ * @brief Register a callback function to allow 
+ * the reception of frames.
+ * 
+ * The registered callback function is called once a frame is 
+ * received  by this service. If many frames are already received, 
+ * the function is called once.
+ *
+ * If called several times the latest callback shall be registered for a particular port.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - yes
+ *
+ *
+ * @param portId - Register callback for a particular MAC device.
+ * @param rxCallbackFn - @a IxEthAccMultiBufferRxCallbackFn - Function to be called when Ethernet frames are availble.
+ * @param callbackTag -  This tag shall be provided to the callback function.
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ * @li @a IX_ETH_ACC_INVALID_ARG : An argument other than portId is invalid.
+ *
+ * @sa ixEthAccPortMultiBufferRxCallbackRegister
+ * @sa IxEthAccPortMultiBufferRxCallback
+ * @sa ixEthAccPortRxCallbackRegister
+ * @sa IxEthAccPortRxCallback
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortMultiBufferRxCallbackRegister(IxEthAccPortId portId,
+										  IxEthAccPortMultiBufferRxCallback rxCallbackFn,
+										  UINT32 callbackTag);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortRxFreeReplenish( IxEthAccPortId portId, IX_OSAL_MBUF *buffer)
+ *
+ * @brief This function provides buffers for the Ethernet receive path. 
+ *
+ * This component does not have a buffer management mechanisms built in. All Rx buffers must be supplied to it
+ * via this interface. 
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - yes
+ *
+ * @param portId @ref IxEthAccPortId [in] - Provide buffers only to specific Rx MAC. 
+ * @param buffer @ref IX_OSAL_MBUF [in] - Provide an MBUF to the Ethernet receive mechanism. 
+ *                 Buffers size smaller than IX_ETHACC_RX_MBUF_MIN_SIZE may result in poor
+ *                 performances and excessive buffer chaining. Buffers
+ *                 larger than this size may be suitable for jumbo frames.
+ *                 Chained packets are not supported and the field IX_OSAL_MBUF_NEXT_PKT_IN_CHAIN_PTR must be NULL. 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_FAIL : Buffer has was not able to queue the 
+ *                     buffer in the receive service.
+ * @li @a IX_ETH_ACC_FAIL : Buffer size is less than IX_ETHACC_RX_MBUF_MIN_SIZE
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * @note
+ * If the buffer replenish operation fails it is the responsibility 
+ * of the user to free the buffer.
+ *
+ * @note
+ * Sufficient buffers must be supplied to the component to maintain
+ * receive throughput and avoid rx buffer underflow conditions.
+ * To meet this goal, It is expected that the user preload the 
+ * component with a sufficent number of buffers prior to enabling the
+ * NPE Ethernet receive path. The recommended minimum number of 
+ * buffers is 8.
+ *
+ * @note
+ * For maximum performances, the mbuf size should be greater 
+ * than the maximum frame size (Ethernet header, payload and FCS) + 64. 
+ * Supplying smaller mbufs to the service results in mbuf
+ * chaining and degraded performances. The recommended size
+ * is @a IX_ETHACC_RX_MBUF_MIN_SIZE, which is
+ * enough to take care of 802.3 frames and "baby jumbo" frames without
+ * chaining, and "jumbo" frame within chaining.
+ *
+ * @note
+ * Buffers may not be filled up to their length. The firware fills
+ * them up to the previous 64 bytes boundary. The user has to be aware 
+ * that the length of the received mbufs may be smaller than the length
+ * of the supplied mbufs.
+ *
+ * @warning This function checks the parameters if the NDEBUG 
+ * flag is not defined. Turning on the argument checking (disabled by 
+ * default) results in a lower EthAcc performance as this function
+ * is part of the data path.
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortRxFreeReplenish( IxEthAccPortId portId, IX_OSAL_MBUF *buffer);
+
+
+
+/***************************************************************
+
+  ####    ####   #    #   #####  #####    ####   #
+ #    #  #    #  ##   #     #    #    #  #    #  #
+ #       #    #  # #  #     #    #    #  #    #  #
+ #       #    #  #  # #     #    #####   #    #  #
+ #    #  #    #  #   ##     #    #   #   #    #  #
+  ####    ####   #    #     #    #    #   ####   ######
+
+
+         #####   #         ##    #    #  ######
+         #    #  #        #  #   ##   #  #
+         #    #  #       #    #  # #  #  #####
+         #####   #       ######  #  # #  #
+         #       #       #    #  #   ##  #
+         #       ######  #    #  #    #  ######
+
+***************************************************************/
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortEnable(IxEthAccPortId portId)
+ *
+ * @brief This enables an Ethernet port for both Tx and Rx. 
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @pre The port must first be initialized via @a ixEthAccPortInit and the MAC address 
+ * must be set using @a ixEthAccUnicastMacAddressSet before enabling it
+ * The rx and Tx Done callbacks registration via @a
+ * ixEthAccPortTxDoneCallbackRegister amd @a  ixEthAccPortRxCallbackRegister
+ * has to be done before enabling the traffic.
+ * 
+ * @param  portId @ref IxEthAccPortId [in] - Port id to act upon.
+ * 
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is not initialized
+ * @li @a IX_ETH_ACC_MAC_UNINITIALIZED : port MAC address is not initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus ixEthAccPortEnable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortDisable(IxEthAccPortId portId)
+ *
+ * @brief This disables an Ethernet port for both Tx and Rx. 
+ *
+ * Free MBufs are returned to the user via the registered callback when the port is disabled 
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @pre The port must be enabled with @a ixEthAccPortEnable, otherwise this
+ * function has no effect
+ *
+ * @param  portId @ref IxEthAccPortId [in] - Port id to act upon.
+ * 
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is not initialized
+ * @li @a IX_ETH_ACC_MAC_UNINITIALIZED : port MAC address is not initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus ixEthAccPortDisable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortEnabledQuery(IxEthAccPortId portId, BOOL *enabled)
+ *
+ * @brief Get the enabled state of a port.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - yes
+ *
+ * @pre The port must first be initialized via @a ixEthAccPortInit
+ *
+ * @param  portId @ref IxEthAccPortId [in] - Port id to act upon.
+ * @param  enabled BOOL [out] - location to store the state of the port
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortEnabledQuery(IxEthAccPortId portId, BOOL *enabled);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortPromiscuousModeClear(IxEthAccPortId portId)
+ *
+ * @brief Put the Ethernet MAC device in non-promiscuous mode.
+ * 
+ * In non-promiscuous mode the MAC filters all frames other than 
+ * destination MAC address which matches the following criteria:
+ * @li Unicast address provisioned via @a ixEthAccUnicastMacAddressSet
+ * @li All broadcast frames.
+ * @li Multicast addresses provisioned via @a ixEthAccMulticastAddressJoin
+ *
+ * Other functions modify the MAC filtering
+ *
+ * @li @a ixEthAccPortMulticastAddressJoinAll() - all multicast
+ *     frames are forwarded to the application
+ * @li @a ixEthAccPortMulticastAddressLeaveAll() - rollback the
+ *     effects of @a ixEthAccPortMulticastAddressJoinAll()
+ * @li @a ixEthAccPortMulticastAddressLeave() - unprovision a new 
+ *     filtering address
+ * @li @a ixEthAccPortMulticastAddressJoin() - provision a new 
+ *     filtering address
+ * @li @a ixEthAccPortPromiscuousModeSet() - all frames are 
+ *     forwarded to the application regardless of the multicast 
+ *     address provisioned
+ * @li @a ixEthAccPortPromiscuousModeClear() - frames are forwarded 
+ *     to the application following the multicast address provisioned
+ *
+ * In all cases, unicast and broadcast addresses are forwarded to 
+ * the application.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ * 
+ * @sa ixEthAccPortPromiscuousModeSet
+ * 
+ * @param portId @ref IxEthAccPortId [in] - Ethernet port id.
+ * 
+ * @return IxEthAccStatus 
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus ixEthAccPortPromiscuousModeClear(IxEthAccPortId portId);
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn  ixEthAccPortPromiscuousModeSet(IxEthAccPortId portId)
+ *
+ * @brief Put the MAC device in promiscuous mode.
+ * 
+ * If the device is in promiscuous mode then all all received frames shall be forwared
+ * to the NPE for processing.
+ *
+ * Other functions modify the MAC filtering
+ *
+ * @li @a ixEthAccPortMulticastAddressJoinAll() - all multicast
+ *     frames are forwarded to the application
+ * @li @a ixEthAccPortMulticastAddressLeaveAll() - rollback the
+ *     effects of @a ixEthAccPortMulticastAddressJoinAll()
+ * @li @a ixEthAccPortMulticastAddressLeave() - unprovision a new 
+ *     filtering address
+ * @li @a ixEthAccPortMulticastAddressJoin() - provision a new 
+ *     filtering address
+ * @li @a ixEthAccPortPromiscuousModeSet() - all frames are 
+ *     forwarded to the application regardless of the multicast 
+ *     address provisioned
+ * @li @a ixEthAccPortPromiscuousModeClear() - frames are forwarded 
+ *     to the application following the multicast address provisioned
+ *
+ * In all cases, unicast and broadcast addresses are forwarded to 
+ * the application.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ * 
+ * @sa ixEthAccPortPromiscuousModeClear
+ *
+ * @param portId @ref IxEthAccPortId [in] - Ethernet port id.
+ * 
+ * @return IxEthAccStatus 
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus ixEthAccPortPromiscuousModeSet(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortUnicastMacAddressSet(      IxEthAccPortId portId,
+                                                  IxEthAccMacAddr *macAddr)
+ *
+ * @brief Configure unicast MAC address for a particular port
+ *
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in] - Ethernet port id.
+ * @param *macAddr @ref IxEthAccMacAddr [in] - Ethernet Mac address.
+ *
+ * @return IxEthAccStatus 
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus ixEthAccPortUnicastMacAddressSet(IxEthAccPortId portId,
+													   IxEthAccMacAddr *macAddr);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortUnicastMacAddressGet(	IxEthAccPortId portId, 
+					IxEthAccMacAddr *macAddr)
+ *
+ * @brief Get unicast MAC address for a particular MAC port 
+ *
+ * @pre
+ * The MAC address must first be set via @a ixEthAccMacPromiscuousModeSet
+ * If the MAC address has not been set, the function returns a 
+ * IX_ETH_ACC_MAC_UNINITIALIZED status
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in] - Ethernet port id.
+ * @param *macAddr @ref IxEthAccMacAddr [out] - Ethernet MAC address.
+ *
+ * @return  IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_MAC_UNINITIALIZED : port MAC address is not initialized.
+ * @li @a IX_ETH_ACC_FAIL : macAddr is invalid.
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortUnicastMacAddressGet(IxEthAccPortId portId,
+								 IxEthAccMacAddr *macAddr);
+
+
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortMulticastAddressJoin(      IxEthAccPortId portId,
+                                             IxEthAccMacAddr *macAddr)
+ *
+ * @brief Add a multicast address to the MAC address table.
+ *
+ *  @note
+ *  Due to the operation of the Ethernet MAC multicast filtering mechanism, frames which do not 
+ *  have a multicast destination address which were provisioned via this API may be forwarded 
+ *  to the NPE's. This is a result of the hardware comparison  algorithm used in the destination mac address logic
+ *  within the Ethernet MAC. 
+ *
+ *  See Also: IXP425 hardware development manual.
+ * 
+ * Other functions modify the MAC filtering
+ *
+ * @li @a ixEthAccPortMulticastAddressJoinAll() - all multicast
+ *     frames are forwarded to the application
+ * @li @a ixEthAccPortMulticastAddressLeaveAll() - rollback the
+ *     effects of @a ixEthAccPortMulticastAddressJoinAll()
+ * @li @a ixEthAccPortMulticastAddressLeave() - unprovision a new 
+ *     filtering address
+ * @li @a ixEthAccPortMulticastAddressJoin() - provision a new 
+ *     filtering address
+ * @li @a ixEthAccPortPromiscuousModeSet() - all frames are 
+ *     forwarded to the application regardless of the multicast 
+ *     address provisioned
+ * @li @a ixEthAccPortPromiscuousModeClear() - frames are forwarded 
+ *     to the application following the multicast address provisioned
+ *
+ * In all cases, unicast and broadcast addresses are forwarded to 
+ * the application.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in] - Ethernet port id.
+ * @param *macAddr @ref IxEthAccMacAddr [in] - Ethernet Mac address.
+ *
+ * @return IxEthAccStatus 
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_FAIL : Error writing to the MAC registers
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortMulticastAddressJoin(IxEthAccPortId portId,
+								 IxEthAccMacAddr *macAddr);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortMulticastAddressJoinAll(  IxEthAccPortId portId)
+ *
+ * @brief Filter all frames with multicast dest.
+ *
+ * This function clears the MAC address table, and then sets
+ * the MAC to forward ALL multicast frames to the NPE.
+ * Specifically, it forwards all frames whose destination address
+ * has the LSB of the highest byte set  (01:00:00:00:00:00).  This
+ * bit is commonly referred to as the "multicast bit".
+ * Broadcast frames will still be forwarded. 
+ *
+ * Other functions modify the MAC filtering
+ *
+ * @li @a ixEthAccPortMulticastAddressJoinAll() - all multicast
+ *     frames are forwarded to the application
+ * @li @a ixEthAccPortMulticastAddressLeaveAll() - rollback the
+ *     effects of @a ixEthAccPortMulticastAddressJoinAll()
+ * @li @a ixEthAccPortMulticastAddressLeave() - unprovision a new 
+ *     filtering address
+ * @li @a ixEthAccPortMulticastAddressJoin() - provision a new 
+ *     filtering address
+ * @li @a ixEthAccPortPromiscuousModeSet() - all frames are 
+ *     forwarded to the application regardless of the multicast 
+ *     address provisioned
+ * @li @a ixEthAccPortPromiscuousModeClear() - frames are forwarded 
+ *     to the application following the multicast address provisioned
+ *
+ * In all cases, unicast and broadcast addresses are forwarded to 
+ * the application.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in] - Ethernet port id.
+ *
+ * @return IxEthAccStatus 
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortMulticastAddressJoinAll(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortMulticastAddressLeave( IxEthAccPortId portId,
+                                         IxEthAccMacAddr *macAddr)
+ *
+ * @brief Remove a multicast address from the MAC address table.
+ *
+ * Other functions modify the MAC filtering
+ *
+ * @li @a ixEthAccPortMulticastAddressJoinAll() - all multicast
+ *     frames are forwarded to the application
+ * @li @a ixEthAccPortMulticastAddressLeaveAll() - rollback the
+ *     effects of @a ixEthAccPortMulticastAddressJoinAll()
+ * @li @a ixEthAccPortMulticastAddressLeave() - unprovision a new 
+ *     filtering address
+ * @li @a ixEthAccPortMulticastAddressJoin() - provision a new 
+ *     filtering address
+ * @li @a ixEthAccPortPromiscuousModeSet() - all frames are 
+ *     forwarded to the application regardless of the multicast 
+ *     address provisioned
+ * @li @a ixEthAccPortPromiscuousModeClear() - frames are forwarded 
+ *     to the application following the multicast address provisioned
+ *
+ * In all cases, unicast and broadcast addresses are forwarded to 
+ * the application.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in] - Ethernet port id.
+ * @param *macAddr @ref IxEthAccMacAddr [in] - Ethernet Mac address.
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_NO_SUCH_ADDR :  Failed if MAC address was not in the table.
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortMulticastAddressLeave(IxEthAccPortId portId,
+								  IxEthAccMacAddr *macAddr);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortMulticastAddressLeaveAll( IxEthAccPortId portId)
+ *
+ * @brief This function unconfigures the multicast filtering settings
+ *
+ * This function first clears the MAC address table, and then sets
+ * the MAC as configured by the promiscuous mode current settings.
+ *
+ * Other functions modify the MAC filtering
+ *
+ * @li @a ixEthAccPortMulticastAddressJoinAll() - all multicast
+ *     frames are forwarded to the application
+ * @li @a ixEthAccPortMulticastAddressLeaveAll() - rollback the
+ *     effects of @a ixEthAccPortMulticastAddressJoinAll()
+ * @li @a ixEthAccPortMulticastAddressLeave() - unprovision a new 
+ *     filtering address
+ * @li @a ixEthAccPortMulticastAddressJoin() - provision a new 
+ *     filtering address
+ * @li @a ixEthAccPortPromiscuousModeSet() - all frames are 
+ *     forwarded to the application regardless of the multicast 
+ *     address provisioned
+ * @li @a ixEthAccPortPromiscuousModeClear() - frames are forwarded 
+ *     to the application following the multicast address provisioned
+ *
+ * In all cases, unicast and broadcast addresses are forwarded to 
+ * the application.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in] - Ethernet port id.
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortMulticastAddressLeaveAll(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortUnicastAddressShow(IxEthAccPortId portId)
+ *
+ * @brief Displays unicast MAC address
+ *
+ * Displays unicast address which is configured using 
+ * @a ixEthAccUnicastMacAddressSet. This function also displays the MAC filter used
+ * to filter multicast frames.
+ *
+ * Other functions modify the MAC filtering
+ *
+ * @li @a ixEthAccPortMulticastAddressJoinAll() - all multicast
+ *     frames are forwarded to the application
+ * @li @a ixEthAccPortMulticastAddressLeaveAll() - rollback the
+ *     effects of @a ixEthAccPortMulticastAddressJoinAll()
+ * @li @a ixEthAccPortMulticastAddressLeave() - unprovision a new 
+ *     filtering address
+ * @li @a ixEthAccPortMulticastAddressJoin() - provision a new 
+ *     filtering address
+ * @li @a ixEthAccPortPromiscuousModeSet() - all frames are 
+ *     forwarded to the application regardless of the multicast 
+ *     address provisioned
+ * @li @a ixEthAccPortPromiscuousModeClear() - frames are forwarded 
+ *     to the application following the multicast address provisioned
+ *
+ * In all cases, unicast and broadcast addresses are forwarded to 
+ * the application.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in] - Ethernet port id.
+ *
+ * @return void
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus ixEthAccPortUnicastAddressShow(IxEthAccPortId portId);
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortMulticastAddressShow( IxEthAccPortId portId)
+ *
+ * @brief Displays multicast MAC address
+ *
+ * Displays multicast address which have been configured using @a ixEthAccMulticastAddressJoin
+ * 
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ * 
+ * @param portId @ref IxEthAccPortId [in] - Ethernet port id.
+ *
+ * @return void
+ *
+ * <hr>
+ */
+PUBLIC void ixEthAccPortMulticastAddressShow( IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortDuplexModeSet( IxEthAccPortId portId, IxEthAccDuplexMode mode )
+ *
+ * @brief  Set the duplex mode for the MAC.
+ *
+ * Configure the IXP400 MAC to either full or half duplex. 
+ *
+ * @note 
+ * The configuration should match that provisioned on the PHY.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in]
+ * @param mode @ref IxEthAccDuplexMode [in]
+ *
+ * @return IxEthAccStatus 
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus 
+ixEthAccPortDuplexModeSet(IxEthAccPortId portId,IxEthAccDuplexMode mode);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortDuplexModeGet( IxEthAccPortId portId, IxEthAccDuplexMode *mode )
+ *
+ * @brief  Get the duplex mode for the MAC.
+ *
+ * return the duplex configuration of the IXP400 MAC.
+ *
+ * @note
+ * The configuration should match that provisioned on the PHY.
+ * See @a ixEthAccDuplexModeSet
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in]
+ * @param *mode @ref IxEthAccDuplexMode [out]
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ *
+ */
+PUBLIC IxEthAccStatus 
+ixEthAccPortDuplexModeGet(IxEthAccPortId portId,IxEthAccDuplexMode *mode );
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortTxFrameAppendPaddingEnable( IxEthAccPortId portId)
+ *
+ * @brief  Enable padding bytes to be appended to runt frames submitted to
+ * this port
+ * 
+ * Enable up to 60 null-bytes padding bytes to be appended to runt frames 
+ * submitted to this port. This is the default behavior of the access 
+ * component.
+ *
+ * @warning Do not change this behaviour while the port is enabled.
+ *
+ * @note When Tx padding is enabled, Tx FCS generation is turned on
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @sa ixEthAccPortTxFrameAppendFCSDusable
+ *
+ * @param portId @ref IxEthAccPortId [in]
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortTxFrameAppendPaddingEnable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortTxFrameAppendPaddingDisable( IxEthAccPortId portId)
+ *
+ * @brief  Disable padding bytes to be appended to runt frames submitted to
+ * this port
+ * 
+ * Disable padding bytes to be appended to runt frames 
+ * submitted to this port. This is not the default behavior of the access 
+ * component.
+ *
+ * @warning Do not change this behaviour while the port is enabled.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in] 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortTxFrameAppendPaddingDisable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortTxFrameAppendFCSEnable( IxEthAccPortId portId)
+ *
+ * @brief  Enable the appending of Ethernet FCS to all frames submitted to this port
+ * 
+ * When enabled, the FCS is added to the submitted frames. This is the default 
+ * behavior of the access component.
+ * Do not change this behaviour while the port is enabled.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in] 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortTxFrameAppendFCSEnable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortTxFrameAppendFCSDisable( IxEthAccPortId portId)
+ *
+ * @brief  Disable the appending of Ethernet FCS to all frames submitted to this port.
+ * 
+ * When disabled, the Ethernet FCS is not added to the submitted frames. 
+ * This is not the default
+ * behavior of the access component.
+ *
+ * @note Since the FCS is not appended to the frame it is expected that the frame submitted to the 
+ * component includes a valid FCS at the end of the data, although this will not be validated.
+ *
+ * The component shall forward the frame to the Ethernet MAC WITHOUT modification.
+ *
+ * Do not change this behaviour while the port is enabled.
+ *
+ * @note Tx FCS append is not disabled while Tx padding is enabled.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @sa ixEthAccPortTxFrameAppendPaddingEnable
+ *
+ * @param portId @ref IxEthAccPortId [in] 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortTxFrameAppendFCSDisable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortRxFrameAppendFCSEnable( IxEthAccPortId portId)
+ *
+ * @brief Forward frames with FCS included in the receive buffer.
+ *
+ * The FCS is not striped from the receive buffer. 
+ * The received frame length includes the FCS size (4 bytes). ie. 
+ * A minimum sized ethernet frame shall have a length of 64bytes.
+ *
+ * Frame FCS validity checks are still carried out on all received frames.
+ *
+ * This is not the default
+ * behavior of the access component.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in]
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortRxFrameAppendFCSEnable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccPortRxFrameAppendFCSDisable( IxEthAccPortId portId)
+ *
+ * @brief  Do not forward the FCS portion of the received Ethernet frame to the user. 
+ * The FCS is striped from the receive buffer. 
+ * The received frame length does not include the FCS size (4 bytes).
+ * Frame FCS validity checks are still carried out on all received frames.
+ *
+ * This is the default behavior of the component.
+ * Do not change this behaviour while the port is enabled.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in]
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS 
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccPortRxFrameAppendFCSDisable(IxEthAccPortId portId);
+
+
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @enum IxEthAccSchedulerDiscipline
+ *
+ * @brief  Definition for the port scheduling discipline
+ *
+ * Select the port scheduling discipline on receive and transmit path
+ * @li FIFO : No Priority : In this configuration all frames are processed
+ *                       in the access component in the strict order in which 
+ *                        the component received them.
+ * @li FIFO : Priority : This shall be a very simple priority mechanism. 
+ *                     Higher prior-ity frames shall be forwarded 
+ *                     before lower priority frames. There shall be no 
+ *                     fairness mechanisms applied across different 
+ *                     priorities. Higher priority frames could starve 
+ *                     lower priority frames indefinitely.
+ */
+typedef  enum 
+{
+    FIFO_NO_PRIORITY, /**<frames submitted with no priority*/
+    FIFO_PRIORITY /**<higher prority frames submitted before lower priority*/
+}IxEthAccSchedulerDiscipline;
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def IxEthAccTxSchedulerDiscipline
+ *
+ * @brief  Deprecated definition for the port transmit scheduling discipline
+ */ 
+#define IxEthAccTxSchedulerDiscipline IxEthAccSchedulerDiscipline
+
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccTxSchedulingDisciplineSet( IxEthAccPortId portId, IxEthAccSchedulerDiscipline sched)
+ *
+ * @brief Set the port scheduling to one of @a IxEthAccSchedulerDiscipline
+ *
+ * The default behavior of the component is @a FIFO_NO_PRIORITY.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @pre
+ *
+ *
+ * @param portId @ref IxEthAccPortId [in] 
+ * @param sched @ref IxEthAccSchedulerDiscipline [in] 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS : Set appropriate discipline.
+ * @li @a IX_ETH_ACC_FAIL :  Invalid/unsupported discipline.
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccTxSchedulingDisciplineSet(IxEthAccPortId portId, 
+								  IxEthAccSchedulerDiscipline sched);
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccRxSchedulingDisciplineSet(IxEthAccSchedulerDiscipline sched)
+ *
+ * @brief Set the Rx scheduling to one of @a IxEthAccSchedulerDiscipline
+ *
+ * The default behavior of the component is @a FIFO_NO_PRIORITY.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @pre
+ *
+ * @param sched : @a IxEthAccSchedulerDiscipline 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS : Set appropriate discipline.
+ * @li @a IX_ETH_ACC_FAIL :  Invalid/unsupported discipline.
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccRxSchedulingDisciplineSet(IxEthAccSchedulerDiscipline sched);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn IxEthAccStatus ixEthAccNpeLoopbackEnable(IxEthAccPortId portId)
+ *
+ * @brief Enable NPE loopback
+ *
+ * When this loopback mode is enabled all the transmitted frames are
+ * received on the same port, without payload.
+ *
+ * This function is recommended for power-up diagnostic checks and
+ * should never be used under normal Ethernet traffic operations.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @pre
+ *
+ * @param portId : ID of the port 
+ *
+ * @note Calling ixEthAccPortDisable followed by ixEthAccPortEnable is
+ * guaranteed to restore correct Ethernet Tx/Rx operation.
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS : NPE loopback mode enabled
+ * @li @a IX_ETH_ACC_FAIL : Invalid port or Ethernet service not initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus 
+ixEthAccPortNpeLoopbackEnable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn IxEthAccStatus ixEthAccPortNpeLoopbackDisable(IxEthAccPortId portId)
+ *
+ * @brief Disable NPE loopback
+ *
+ * This function is used to disable the NPE loopback if previously
+ * enabled using ixEthAccNpeLoopbackEnable.
+ *
+ * This function is recommended for power-up diagnostic checks and
+ * should never be used under normal Ethernet traffic operations.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @pre
+ *
+ * @note Calling ixEthAccPortDisable followed by ixEthAccPortEnable is
+ * guaranteed to restore correct Ethernet Tx/Rx operation.
+ *
+ * @param portId : ID of the port 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS : NPE loopback successfully disabled
+ * @li @a IX_ETH_ACC_FAIL : Invalid port or Ethernet service not initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus 
+ixEthAccPortNpeLoopbackDisable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn IxEthAccStatus ixEthAccPortTxEnable(IxEthAccPortId portId)
+ *
+ * @brief Enable Tx on the port
+ *
+ * This function is the complement of ixEthAccPortTxDisable and should
+ * be used only after Tx was disabled. A MAC core reset is required before
+ * this function is called (see @a ixEthAccPortMacReset).
+ *
+ * This function is the recommended usage scenario for emergency security
+ * shutdown and hardware failure recovery and should never be used for throttling 
+ * traffic.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @pre
+ *
+ * @note Calling ixEthAccPortDisable followed by ixEthAccPortEnable is
+ * guaranteed to restore correct Ethernet Tx/Rx operation.
+ *
+ * @param portId : ID of the port 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS : Tx successfully enabled
+ * @li @a IX_ETH_ACC_FAIL : Invalid port or Ethernet service not initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus 
+ixEthAccPortTxEnable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn IxEthAccStatus ixEthAccPortTxDisable(IxEthAccPortId portId)
+ *
+ * @brief Disable Tx on the port
+ *
+ * This function can be used to disable Tx in the MAC core.
+ * Tx can be re-enabled, although this is not guaranteed, by performing
+ * a MAC core reset (@a ixEthAccPortMacReset) and calling ixEthAccPortTxEnable.
+ * Note that using this function is not recommended, except for shutting
+ * down Tx for emergency reasons. For proper port shutdown and re-enabling
+ * see ixEthAccPortEnable and ixEthAccPortDisable.
+ *
+ * This function is the recommended usage scenario for emergency security
+ * shutdown and hardware failure recovery and should never be used for throttling 
+ * traffic.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @note Calling ixEthAccPortDisable followed by ixEthAccPortEnable is
+ * guaranteed to restore correct Ethernet Tx/Rx operation.
+ *
+ * @pre
+ *
+ * @param portId : ID of the port 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS : Tx successfully disabled
+ * @li @a IX_ETH_ACC_FAIL : Invalid port or Ethernet service not initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus 
+ixEthAccPortTxDisable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn IxEthAccStatus ixEthAccPortRxEnable(IxEthAccPortId portId)
+ *
+ * @brief Enable Rx on the port
+ *
+ * This function is the complement of ixEthAccPortRxDisable and should
+ * be used only after Rx was disabled.
+ *
+ * This function is the recommended usage scenario for emergency security
+ * shutdown and hardware failure recovery and should never be used for throttling 
+ * traffic.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @note Calling ixEthAccPortDisable followed by ixEthAccPortEnable is
+ * guaranteed to restore correct Ethernet Tx/Rx operation.
+ *
+ * @pre
+ *
+ * @param portId : ID of the port 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS : Rx successfully enabled
+ * @li @a IX_ETH_ACC_FAIL : Invalid port or Ethernet service not initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus 
+ixEthAccPortRxEnable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn IxEthAccStatus ixEthAccPortRxDisable(IxEthAccPortId portId)
+ *
+ * @brief Disable Rx on the port
+ *
+ * This function can be used to disable Rx in the MAC core.
+ * Rx can be re-enabled, although this is not guaranteed, by performing
+ * a MAC core reset (@a ixEthAccPortMacReset) and calling ixEthAccPortRxEnable.
+ * Note that using this function is not recommended, except for shutting
+ * down Rx for emergency reasons. For proper port shutdown and re-enabling
+ * see ixEthAccPortEnable and ixEthAccPortDisable.
+ *
+ * This function is the recommended usage scenario for emergency security
+ * shutdown and hardware failure recovery and should never be used for throttling 
+ * traffic.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @pre
+ *
+ * @note Calling ixEthAccPortDisable followed by ixEthAccPortEnable is
+ * guaranteed to restore correct Ethernet Tx/Rx operation.
+ *
+ * @param portId : ID of the port 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS : Rx successfully disabled
+ * @li @a IX_ETH_ACC_FAIL : Invalid port or Ethernet service not initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus 
+ixEthAccPortRxDisable(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn IxEthAccStatus ixEthAccPortMacReset(IxEthAccPortId portId)
+ *
+ * @brief Reset MAC core on the port
+ *
+ * This function will perform a MAC core reset (NPE Ethernet coprocessor).
+ * This function is inherently unsafe and the NPE recovery is not guaranteed
+ * after this function is called. The proper manner of performing port disable
+ * and enable (which will reset the MAC as well) is ixEthAccPortEnable/ixEthAccPortDisable.
+ *
+ * This function is the recommended usage scenario for hardware failure recovery
+ * and should never be used for throttling traffic.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @pre
+ *
+ * @note Calling ixEthAccPortDisable followed by ixEthAccPortEnable is
+ * guaranteed to restore correct Ethernet Tx/Rx operation.
+ *
+ * @param portId : ID of the port 
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS : MAC core reset
+ * @li @a IX_ETH_ACC_FAIL : Invalid port or Ethernet service not initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus 
+ixEthAccPortMacReset(IxEthAccPortId portId);
+
+/*********************************************************************************
+  ####    #####    ##     #####     #     ####    #####     #     ####    ####
+ #          #     #  #      #       #    #          #       #    #    #  #
+  ####      #    #    #     #       #     ####      #       #    #        ####
+      #     #    ######     #       #         #     #       #    #            #
+ #    #     #    #    #     #       #    #    #     #       #    #    #  #    #
+  ####      #    #    #     #       #     ####      #       #     ####    ####
+**********************************************************************************/
+
+
+/**
+ *
+ * @brief This struct defines the statistics returned by this component.
+ *
+ * The component returns MIB2 EthObj variables which are obtained from the 
+ * hardware or maintained by this component. 
+ *
+ *
+ */
+typedef struct   
+{
+    UINT32 dot3StatsAlignmentErrors;            /**< link error count (rx) */
+    UINT32 dot3StatsFCSErrors;                  /**< link error count (rx) */
+    UINT32 dot3StatsInternalMacReceiveErrors;   /**< link error count (rx) */
+    UINT32 RxOverrunDiscards;                   /**< NPE: discarded frames count (rx) */
+    UINT32 RxLearnedEntryDiscards;                /**< NPE: discarded frames count(rx)  */
+    UINT32 RxLargeFramesDiscards;                 /**< NPE: discarded frames count(rx)  */
+    UINT32 RxSTPBlockedDiscards;                  /**< NPE: discarded frames count(rx)  */
+    UINT32 RxVLANTypeFilterDiscards;              /**< NPE: discarded frames count (rx) */
+    UINT32 RxVLANIdFilterDiscards;                /**< NPE: discarded frames count (rx) */
+    UINT32 RxInvalidSourceDiscards;               /**< NPE: discarded frames count (rx) */
+    UINT32 RxBlackListDiscards;                   /**< NPE: discarded frames count (rx) */
+    UINT32 RxWhiteListDiscards;                   /**< NPE: discarded frames count (rx) */
+    UINT32 RxUnderflowEntryDiscards;              /**< NPE: discarded frames count (rx) */
+    UINT32 dot3StatsSingleCollisionFrames;      /**< link error count (tx) */
+    UINT32 dot3StatsMultipleCollisionFrames;    /**< link error count (tx) */
+    UINT32 dot3StatsDeferredTransmissions;      /**< link error count (tx) */
+    UINT32 dot3StatsLateCollisions;             /**< link error count (tx) */
+    UINT32 dot3StatsExcessiveCollsions;         /**< link error count (tx) */
+    UINT32 dot3StatsInternalMacTransmitErrors;  /**< link error count (tx) */
+    UINT32 dot3StatsCarrierSenseErrors;         /**< link error count (tx) */
+    UINT32 TxLargeFrameDiscards;                /**< NPE: discarded frames count (tx) */
+    UINT32 TxVLANIdFilterDiscards;              /**< NPE: discarded frames count (tx) */
+
+}IxEthEthObjStats;
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccMibIIStatsGet(IxEthAccPortId portId ,IxEthEthObjStats *retStats )
+ *
+ * @brief  Returns the statistics maintained for a port.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @pre
+ *
+ *
+ * @param portId @ref IxEthAccPortId [in] 
+ * @param retStats @ref IxEthEthObjStats [out]
+ * @note Please note the user is responsible for cache coheriency of the retStat
+ * buffer. The data is actually populated via the NPE's. As such cache safe
+ * memory should be used in the retStats argument.
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_FAIL : Invalid arguments.
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccMibIIStatsGet(IxEthAccPortId portId, IxEthEthObjStats *retStats );
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccMibIIStatsGetClear(IxEthAccPortId portId, IxEthEthObjStats *retStats)
+ * 
+ * @brief  Returns and clears the statistics maintained for a port. 
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - yes
+ *
+ * @pre
+ *
+ * @param portId @ref IxEthAccPortId [in] 
+ * @param retStats @ref IxEthEthObjStats [out]
+ * @note Please note the user is responsible for cache coheriency of the retStats
+ * buffer. The data is actually populated via the NPE's. As such cache safe
+ * memory should be used in the retStats argument.
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_FAIL : invalid arguments.
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccMibIIStatsGetClear(IxEthAccPortId portId, IxEthEthObjStats *retStats);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccMibIIStatsClear(IxEthAccPortId portId)
+ *
+ * @brief   Clears the statistics maintained for a port.
+ *
+ * @li Reentrant    - yes
+ * @li ISR Callable - no
+ *
+ * @pre
+ *
+ * @param portId @ref IxEthAccPortId [in]
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_FAIL : Invalid arguments.
+ * @li @a IX_ETH_ACC_INVALID_PORT : portId is invalid.
+ * @li @a IX_ETH_ACC_PORT_UNINITIALIZED : portId is un-initialized
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus ixEthAccMibIIStatsClear(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccMacInit(IxEthAccPortId portId)
+ * 
+ * @brief Initializes the ethernet MAC settings 
+ * 
+ * @li Reentrant    - no
+ * @li ISR Callable - no
+ *
+ * @param portId @ref IxEthAccPortId [in]
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_INVALID_PORT	:  portId is invalid.
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus ixEthAccMacInit(IxEthAccPortId portId);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccStatsShow(IxEthAccPortId portId)
+ *
+ *
+ * @brief Displays a ports statistics on the standard io console using printf.
+ *
+ * @li Reentrant    - no
+ * @li ISR Callable - no
+ *
+ * @pre
+ *
+ * @param portId @ref IxEthAccPortId [in]
+ *
+ * @return void
+ *
+ * <hr>
+ */
+PUBLIC void ixEthAccStatsShow(IxEthAccPortId portId);
+
+/*************************************************************************
+
+ #    #     #       #            #    #  #####      #     ####
+ ##  ##     #       #            ##  ##  #    #     #    #    #
+ # ## #     #       #            # ## #  #    #     #    #    #
+ #    #     #       #            #    #  #    #     #    #    #
+ #    #     #       #            #    #  #    #     #    #    #
+ #    #     #       #            #    #  #####      #     ####
+
+*************************************************************************/
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccMiiReadRtn (UINT8 phyAddr, 
+                           UINT8 phyReg, 
+	                   UINT16 *value)
+ *
+ *
+ * @brief Reads a 16 bit value from a PHY
+ *
+ * Reads a 16-bit word from a register of a MII-compliant PHY. Reading
+ * is performed through the MII management interface.  This function returns
+ * when the read operation has successfully completed, or when a timeout has elapsed.
+ *
+ * @li Reentrant    - no
+ * @li ISR Callable - no
+ *
+ * @pre The MAC on Ethernet Port 2 (NPE C) must be initialised, and generating the MDIO clock.
+ *   
+ * @param phyAddr UINT8 [in] - the address of the Ethernet PHY (0-31)
+ * @param phyReg UINT8 [in] -  the number of the MII register to read (0-31)
+ * @param value UINT16 [in] -  the value read from the register
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_FAIL : failed to read the register.
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccMiiReadRtn (UINT8 phyAddr, UINT8 phyReg, UINT16 *value);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccMiiWriteRtn (UINT8 phyAddr, 
+                            UINT8 phyReg, 
+	                    UINT16 value)
+ *
+ *
+ * @brief Writes a 16 bit value to a PHY
+ *
+ * Writes a 16-bit word from a register of a MII-compliant PHY. Writing
+ * is performed through the MII management interface.  This function returns
+ * when the write operation has successfully completed, or when a timeout has elapsed.
+ *
+ * @li Reentrant    - no
+ * @li ISR Callable - no
+ *
+ * @pre The MAC on Ethernet Port 2 (NPE C) must be initialised, and generating the MDIO clock.
+ *   
+ * @param phyAddr UINT8 [in] - the address of the Ethernet PHY (0-31)
+ * @param phyReg UINT8 [in] -  the number of the MII register to write (0-31)
+ * @param value UINT16 [out] -  the value to write to the register
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_FAIL : failed to write register.
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccMiiWriteRtn (UINT8 phyAddr, UINT8 phyReg, UINT16 value);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccMiiAccessTimeoutSet(UINT32 timeout)
+ *
+ * @brief Overrides the default timeout value and retry count when reading or 
+ * writing MII registers using ixEthAccMiiWriteRtn or ixEthAccMiiReadRtn
+ *
+ * The default behavior of the component is to use a IX_ETH_ACC_MII_10TH_SEC_IN_MILLIS ms
+ * timeout (declared as 100 in IxEthAccMii_p.h) and a retry count of IX_ETH_ACC_MII_TIMEOUT_10TH_SECS
+ * (declared as 5 in IxEthAccMii_p.h).
+ *
+ * The MII read and write functions will attempt to read the status of the register up
+ * to the retry count times, delaying between each attempt with the timeout value.
+ *
+ * @li Reentrant    - no
+ * @li ISR Callable - no
+ *
+ * @pre
+ *
+ * @param timeout UINT32 [in] - new timeout value, in milliseconds
+ * @param timeout UINT32 [in] - new retry count (a minimum value of 1 must be used)
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_FAIL : invalid parameter(s)
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus
+ixEthAccMiiAccessTimeoutSet(UINT32 timeout, UINT32 retryCount);
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @fn ixEthAccMiiStatsShow (UINT32 phyAddr)
+ *
+ *
+ * @brief Displays detailed information on a specified PHY
+ *
+ * Displays the current values of the first eigth MII registers for a PHY, 
+ *
+ * @li Reentrant    - no
+ * @li ISR Callable - no
+ *
+ * @pre The MAC on Ethernet Port 2 (NPE C) must be initialised, and 
+ *      generating the MDIO clock.
+ *   
+ * @param phyAddr UINT32 [in] - the address of the Ethernet PHY (0-31)
+ *
+ * @return IxEthAccStatus
+ * @li @a IX_ETH_ACC_SUCCESS
+ * @li @a IX_ETH_ACC_FAIL : invalid arguments.
+ *
+ * <hr>
+ */
+PUBLIC IxEthAccStatus ixEthAccMiiStatsShow (UINT32 phyAddr);
+
+
+
+/******* BOARD SPECIFIC DEPRECATED API *********/
+
+/* The following functions are high level functions which rely
+ * on the properties and interface of some Ethernet PHYs. The
+ * implementation is hardware specific and has been moved to 
+ * the hardware-specific component IxEthMii.
+ */
+
+ #include "IxEthMii.h"
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def  ixEthAccMiiPhyScan
+ *
+ * @brief : deprecated API entry point. This definition 
+ * ensures backward compatibility
+ *
+ * See @ref ixEthMiiPhyScan
+ *
+ * @note this feature is board specific
+ *
+ */
+#define ixEthAccMiiPhyScan(phyPresent) ixEthMiiPhyScan(phyPresent,IXP425_ETH_ACC_MII_MAX_ADDR)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def ixEthAccMiiPhyConfig
+ *
+ * @brief : deprecated API entry point. This definition 
+ * ensures backward compatibility
+ *
+ * See @ref ixEthMiiPhyConfig
+ *
+ * @note this feature is board specific
+ */
+#define ixEthAccMiiPhyConfig(phyAddr,speed100,fullDuplex,autonegotiate) \
+           ixEthMiiPhyConfig(phyAddr,speed100,fullDuplex,autonegotiate)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def ixEthAccMiiPhyReset
+ *
+ * @brief : deprecated API entry point. This definition 
+ * ensures backward compatibility
+ *
+ * See @ref ixEthMiiPhyReset
+ *
+ * @note this feature is board specific
+ */
+#define ixEthAccMiiPhyReset(phyAddr) \
+           ixEthMiiPhyReset(phyAddr)
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def ixEthAccMiiLinkStatus
+ *
+ * @brief : deprecated API entry point. This definition 
+ * ensures backward compatibility
+ *
+ * See @ref ixEthMiiLinkStatus
+ *
+ * @note this feature is board specific
+ */
+#define ixEthAccMiiLinkStatus(phyAddr,linkUp,speed100,fullDuplex,autoneg) \
+           ixEthMiiLinkStatus(phyAddr,linkUp,speed100,fullDuplex,autoneg)
+
+
+
+/**
+ * @ingroup IxEthAcc
+ *
+ * @def ixEthAccMiiShow  
+ *
+ * @brief : deprecated API entry point. This definition 
+ * ensures backward compatibility
+ *
+ * See @ref ixEthMiiPhyShow
+ *
+ * @note this feature is board specific
+ */
+#define ixEthAccMiiShow(phyAddr) \
+        ixEthMiiPhyShow(phyAddr)
+
+#endif /* ndef IxEthAcc_H */
+/**
+ *@}
+ */

cpu/ixp/npe/include/IxEthAccDataPlane_p.h

@@ -0,0 +1,245 @@
+/**
+ * @file IxEthAccDataPlane_p.h
+ *
+ * @author Intel Corporation
+ * @date 12-Feb-2002
+ *
+ * @brief  Internal Header file for IXP425 Ethernet Access component.
+ *
+ * Design Notes:
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+
+
+#ifndef IxEthAccDataPlane_p_H
+#define IxEthAccDataPlane_p_H
+
+#include <IxOsal.h>
+#include <IxQMgr.h>
+
+/**
+ * @addtogroup IxEthAccPri
+ *@{
+ */
+
+/* typedefs global to this file*/
+
+typedef struct
+{
+    IX_OSAL_MBUF *pHead;
+    IX_OSAL_MBUF *pTail;
+}IxEthAccDataPlaneQList;
+
+
+/**
+ * @struct  IxEthAccDataPlaneStats
+ * @brief   Statistics data structure associated with the data plane
+ *
+ */
+typedef struct
+{
+    UINT32 addToSwQ;
+    UINT32 removeFromSwQ;
+    UINT32 unchainedTxMBufs;
+    UINT32 chainedTxMBufs;
+    UINT32 unchainedTxDoneMBufs;
+    UINT32 chainedTxDoneMBufs;
+    UINT32 unchainedRxMBufs;
+    UINT32 chainedRxMBufs;
+    UINT32 unchainedRxFreeMBufs;
+    UINT32 chainedRxFreeMBufs;
+    UINT32 rxCallbackCounter;
+    UINT32 rxCallbackBurstRead;
+    UINT32 txDoneCallbackCounter;
+    UINT32 unexpectedError;
+} IxEthAccDataPlaneStats;
+
+/**
+ * @fn ixEthAccMbufFromSwQ
+ * @brief  used during disable steps to convert mbufs from 
+ *  swq format, ready to be pushed into hw queues for NPE, 
+ *  back into XScale format 
+ */
+IX_OSAL_MBUF *ixEthAccMbufFromSwQ(IX_OSAL_MBUF *mbuf);
+
+/**
+ * @fn ixEthAccDataPlaneShow
+ * @brief  Show function (for data plane statistics
+ */
+void ixEthAccDataPlaneShow(void);
+
+/*
+ * lock dataplane when atomic operation is required
+ */
+#define IX_ETH_ACC_DATA_PLANE_LOCK(arg) arg = ixOsalIrqLock();
+#define IX_ETH_ACC_DATA_PLANE_UNLOCK(arg) ixOsalIrqUnlock(arg);
+
+/*
+ * Use MBUF fields
+ */
+#define IX_ETHACC_NE_SHARED(mBufPtr) \
+ ((IxEthAccNe *)&((mBufPtr)->ix_ne))
+
+#if 1
+
+#define IX_ETHACC_NE_NEXT(mBufPtr) (mBufPtr)->ix_ne.reserved[0]
+
+/* tm - wrong!! len and pkt_len are in the second word - #define IX_ETHACC_NE_LEN(mBufPtr) (mBufPtr)->ix_ne.reserved[3] */
+#define IX_ETHACC_NE_LEN(mBufPtr) (mBufPtr)->ix_ne.reserved[1]
+
+#define IX_ETHACC_NE_DATA(mBufPtr)(mBufPtr)->ix_ne.reserved[2]
+
+#else
+
+#define IX_ETHACC_NE_NEXT(mBufPtr) \
+  IX_ETHACC_NE_SHARED(mBufPtr)->ixReserved_next
+
+#define IX_ETHACC_NE_LEN(mBufPtr) \
+  IX_ETHACC_NE_SHARED(mBufPtr)->ixReserved_lengths
+
+#define IX_ETHACC_NE_DATA(mBufPtr) \
+  IX_ETHACC_NE_SHARED(mBufPtr)->ixReserved_data
+#endif
+
+/*
+ * Use MBUF  next pointer field to chain data.
+ */
+#define IX_ETH_ACC_MBUF_NEXT_PKT_CHAIN_MEMBER(mbuf) (mbuf)->ix_ctrl.ix_chain
+
+
+
+#define IX_ETH_ACC_DATAPLANE_IS_Q_EMPTY(mbuf_list) ((mbuf_list.pHead) == NULL)
+    
+
+#define IX_ETH_ACC_DATAPLANE_ADD_MBUF_TO_Q_HEAD(mbuf_list,mbuf_to_add) 		\
+  do {										\
+    int lockVal;								\
+    IX_ETH_ACC_DATA_PLANE_LOCK(lockVal);                                    	\
+    IX_ETH_ACC_STATS_INC(ixEthAccDataStats.addToSwQ);                           \
+    if ( (mbuf_list.pHead) != NULL ) 						\
+    {										\
+      (IX_ETH_ACC_MBUF_NEXT_PKT_CHAIN_MEMBER((mbuf_to_add))) = (mbuf_list.pHead);\
+      (mbuf_list.pHead) = (mbuf_to_add);					\
+    } 										\
+    else {									\
+      (mbuf_list.pTail) = (mbuf_list.pHead) = (mbuf_to_add);			\
+      IX_ETH_ACC_MBUF_NEXT_PKT_CHAIN_MEMBER((mbuf_to_add)) = NULL;		\
+    } 										\
+    IX_ETH_ACC_DATA_PLANE_UNLOCK(lockVal);					\
+  } while(0)
+
+
+#define IX_ETH_ACC_DATAPLANE_ADD_MBUF_TO_Q_TAIL(mbuf_list,mbuf_to_add)   	\
+  do {										\
+    int lockVal;								\
+    IX_ETH_ACC_DATA_PLANE_LOCK(lockVal);                                    	\
+    IX_ETH_ACC_STATS_INC(ixEthAccDataStats.addToSwQ);                           \
+    if ( (mbuf_list.pHead) == NULL ) 						\
+    {										\
+      (mbuf_list.pHead) = mbuf_to_add;						\
+      IX_ETH_ACC_MBUF_NEXT_PKT_CHAIN_MEMBER((mbuf_to_add)) = NULL;		\
+    } 										\
+    else {									\
+      IX_ETH_ACC_MBUF_NEXT_PKT_CHAIN_MEMBER((mbuf_list.pTail)) = (mbuf_to_add);	\
+      IX_ETH_ACC_MBUF_NEXT_PKT_CHAIN_MEMBER((mbuf_to_add)) = NULL;		\
+    } 										\
+    (mbuf_list.pTail) = mbuf_to_add;						\
+    IX_ETH_ACC_DATA_PLANE_UNLOCK(lockVal);					\
+  } while (0)
+
+
+#define IX_ETH_ACC_DATAPLANE_REMOVE_MBUF_FROM_Q_HEAD(mbuf_list,mbuf_to_rem)   	\
+  do {										\
+    int lockVal;								\
+    IX_ETH_ACC_DATA_PLANE_LOCK(lockVal);                                    	\
+    if ( (mbuf_list.pHead) != NULL ) 						\
+    {										\
+      IX_ETH_ACC_STATS_INC(ixEthAccDataStats.removeFromSwQ);                    \
+      (mbuf_to_rem) = (mbuf_list.pHead) ;					\
+      (mbuf_list.pHead) = (IX_ETH_ACC_MBUF_NEXT_PKT_CHAIN_MEMBER((mbuf_to_rem)));\
+    } 										\
+    else {									\
+      (mbuf_to_rem) = NULL;							\
+    } 										\
+    IX_ETH_ACC_DATA_PLANE_UNLOCK(lockVal);					\
+  } while (0)
+
+
+/**
+ * @brief message handler QManager entries for NPE id => port ID conversion (NPE_B => 0, NPE_C => 1)
+ */
+#define IX_ETH_ACC_PORT_TO_NPE_ID(port) \
+   ixEthAccPortData[(port)].npeId
+
+#define IX_ETH_ACC_NPE_TO_PORT_ID(npe) ((npe == 0 ? 2 : (npe == 1 ? 0 : ( npe == 2 ? 1 : -1 ))))
+
+#define IX_ETH_ACC_PORT_TO_TX_Q_ID(port)  \
+   ixEthAccPortData[(port)].ixEthAccTxData.txQueue
+
+#define IX_ETH_ACC_PORT_TO_RX_FREE_Q_ID(port) \
+   ixEthAccPortData[(port)].ixEthAccRxData.rxFreeQueue
+
+#define IX_ETH_ACC_PORT_TO_TX_Q_SOURCE(port)    (port == IX_ETH_PORT_1 ? IX_ETH_ACC_TX_FRAME_ENET0_Q_SOURCE : (port == IX_ETH_PORT_2 ? IX_ETH_ACC_TX_FRAME_ENET1_Q_SOURCE : IX_ETH_ACC_TX_FRAME_ENET2_Q_SOURCE))
+
+#define IX_ETH_ACC_PORT_TO_RX_FREE_Q_SOURCE(port) (port == IX_ETH_PORT_1 ? IX_ETH_ACC_RX_FREE_BUFF_ENET0_Q_SOURCE : (port == IX_ETH_PORT_2 ? IX_ETH_ACC_RX_FREE_BUFF_ENET1_Q_SOURCE : IX_ETH_ACC_RX_FREE_BUFF_ENET2_Q_SOURCE ))
+
+/* Flush the mbufs chain and all data pointed to by the mbuf */
+
+#ifndef NDEBUG
+#define IX_ETH_ACC_STATS_INC(x) (x++)
+#else
+#define IX_ETH_ACC_STATS_INC(x)
+#endif
+
+#define IX_ETH_ACC_MAX_TX_FRAMES_TO_SUBMIT 128
+
+void ixEthRxFrameQMCallback(IxQMgrQId qId, IxQMgrCallbackId callbackId);
+void ixEthRxMultiBufferQMCallback(IxQMgrQId qId, IxQMgrCallbackId callbackId);
+void ixEthTxFrameDoneQMCallback(IxQMgrQId qId, IxQMgrCallbackId callbackId);
+
+#endif /* IxEthAccDataPlane_p_H */
+
+
+/**
+ *@}
+ */
+

cpu/ixp/npe/include/IxEthAccMac_p.h

@@ -0,0 +1,248 @@
+/*
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+*/
+
+
+#ifndef IxEthAccMac_p_H
+#define IxEthAccMac_p_H
+
+#include "IxOsal.h"
+
+#define IX_ETH_ACC_MAX_MULTICAST_ADDRESSES 256
+#define IX_ETH_ACC_NUM_PORTS 3
+#define IX_ETH_ACC_MAX_FRAME_SIZE_DEFAULT 1536
+#define IX_ETH_ACC_MAX_FRAME_SIZE_UPPER_RANGE (65536-64)
+#define IX_ETH_ACC_MAX_FRAME_SIZE_LOWER_RANGE 64
+
+/*
+ * 
+ * MAC register definitions
+ *
+ */
+#define IX_ETH_ACC_MAC_0_BASE IX_OSAL_IXP400_ETHA_PHYS_BASE
+#define IX_ETH_ACC_MAC_1_BASE IX_OSAL_IXP400_ETHB_PHYS_BASE
+#define IX_ETH_ACC_MAC_2_BASE IX_OSAL_IXP400_ETH_NPEA_PHYS_BASE
+
+#define IX_ETH_ACC_MAC_TX_CNTRL1       0x000
+#define IX_ETH_ACC_MAC_TX_CNTRL2       0x004
+#define IX_ETH_ACC_MAC_RX_CNTRL1       0x010
+#define IX_ETH_ACC_MAC_RX_CNTRL2       0x014
+#define IX_ETH_ACC_MAC_RANDOM_SEED     0x020
+#define IX_ETH_ACC_MAC_THRESH_P_EMPTY  0x030
+#define IX_ETH_ACC_MAC_THRESH_P_FULL   0x038
+#define IX_ETH_ACC_MAC_BUF_SIZE_TX     0x040
+#define IX_ETH_ACC_MAC_TX_DEFER        0x050
+#define IX_ETH_ACC_MAC_RX_DEFER        0x054
+#define IX_ETH_ACC_MAC_TX_TWO_DEFER_1  0x060
+#define IX_ETH_ACC_MAC_TX_TWO_DEFER_2  0x064
+#define IX_ETH_ACC_MAC_SLOT_TIME       0x070
+#define IX_ETH_ACC_MAC_MDIO_CMD_1      0x080
+#define IX_ETH_ACC_MAC_MDIO_CMD_2      0x084
+#define IX_ETH_ACC_MAC_MDIO_CMD_3      0x088
+#define IX_ETH_ACC_MAC_MDIO_CMD_4      0x08c
+#define IX_ETH_ACC_MAC_MDIO_STS_1      0x090
+#define IX_ETH_ACC_MAC_MDIO_STS_2      0x094
+#define IX_ETH_ACC_MAC_MDIO_STS_3      0x098
+#define IX_ETH_ACC_MAC_MDIO_STS_4      0x09c
+#define IX_ETH_ACC_MAC_ADDR_MASK_1     0x0A0
+#define IX_ETH_ACC_MAC_ADDR_MASK_2     0x0A4
+#define IX_ETH_ACC_MAC_ADDR_MASK_3     0x0A8
+#define IX_ETH_ACC_MAC_ADDR_MASK_4     0x0AC
+#define IX_ETH_ACC_MAC_ADDR_MASK_5     0x0B0
+#define IX_ETH_ACC_MAC_ADDR_MASK_6     0x0B4
+#define IX_ETH_ACC_MAC_ADDR_1          0x0C0
+#define IX_ETH_ACC_MAC_ADDR_2          0x0C4
+#define IX_ETH_ACC_MAC_ADDR_3          0x0C8
+#define IX_ETH_ACC_MAC_ADDR_4          0x0CC
+#define IX_ETH_ACC_MAC_ADDR_5          0x0D0
+#define IX_ETH_ACC_MAC_ADDR_6          0x0D4
+#define IX_ETH_ACC_MAC_INT_CLK_THRESH  0x0E0
+#define IX_ETH_ACC_MAC_UNI_ADDR_1      0x0F0
+#define IX_ETH_ACC_MAC_UNI_ADDR_2      0x0F4
+#define IX_ETH_ACC_MAC_UNI_ADDR_3      0x0F8
+#define IX_ETH_ACC_MAC_UNI_ADDR_4      0x0FC
+#define IX_ETH_ACC_MAC_UNI_ADDR_5      0x100
+#define IX_ETH_ACC_MAC_UNI_ADDR_6      0x104
+#define IX_ETH_ACC_MAC_CORE_CNTRL      0x1FC
+
+
+/*
+ *
+ *Bit definitions
+ *
+ */
+
+/* TX Control Register 1*/
+
+#define IX_ETH_ACC_TX_CNTRL1_TX_EN         BIT(0)
+#define IX_ETH_ACC_TX_CNTRL1_DUPLEX        BIT(1)
+#define IX_ETH_ACC_TX_CNTRL1_RETRY         BIT(2)
+#define IX_ETH_ACC_TX_CNTRL1_PAD_EN        BIT(3)
+#define IX_ETH_ACC_TX_CNTRL1_FCS_EN        BIT(4)
+#define IX_ETH_ACC_TX_CNTRL1_2DEFER        BIT(5)
+#define IX_ETH_ACC_TX_CNTRL1_RMII          BIT(6)
+
+/* TX Control Register 2 */
+#define IX_ETH_ACC_TX_CNTRL2_RETRIES_MASK  0xf
+
+/* RX Control Register 1 */
+#define IX_ETH_ACC_RX_CNTRL1_RX_EN         BIT(0)
+#define IX_ETH_ACC_RX_CNTRL1_PADSTRIP_EN   BIT(1)
+#define IX_ETH_ACC_RX_CNTRL1_CRC_EN        BIT(2)
+#define IX_ETH_ACC_RX_CNTRL1_PAUSE_EN      BIT(3)
+#define IX_ETH_ACC_RX_CNTRL1_LOOP_EN       BIT(4)
+#define IX_ETH_ACC_RX_CNTRL1_ADDR_FLTR_EN  BIT(5)
+#define IX_ETH_ACC_RX_CNTRL1_RX_RUNT_EN    BIT(6)
+#define IX_ETH_ACC_RX_CNTRL1_BCAST_DIS     BIT(7)
+
+/* RX Control Register 2 */
+#define IX_ETH_ACC_RX_CNTRL2_DEFER_EN      BIT(0)
+
+
+
+/* Core Control Register */
+#define IX_ETH_ACC_CORE_RESET              BIT(0)
+#define IX_ETH_ACC_CORE_RX_FIFO_FLUSH      BIT(1)
+#define IX_ETH_ACC_CORE_TX_FIFO_FLUSH      BIT(2)
+#define IX_ETH_ACC_CORE_SEND_JAM           BIT(3)
+#define IX_ETH_ACC_CORE_MDC_EN             BIT(4)
+
+/* 1st bit of 1st MAC octet */
+#define IX_ETH_ACC_ETH_MAC_BCAST_MCAST_BIT ( 1) 
+
+
+/*
+ *
+ * Default values
+ *
+ */
+
+
+#define IX_ETH_ACC_TX_CNTRL1_DEFAULT  (IX_ETH_ACC_TX_CNTRL1_TX_EN | \
+ IX_ETH_ACC_TX_CNTRL1_RETRY  | \
+ IX_ETH_ACC_TX_CNTRL1_FCS_EN | \
+ IX_ETH_ACC_TX_CNTRL1_2DEFER | \
+ IX_ETH_ACC_TX_CNTRL1_PAD_EN)
+
+#define IX_ETH_ACC_TX_MAX_RETRIES_DEFAULT      0x0f 
+
+#define IX_ETH_ACC_RX_CNTRL1_DEFAULT    (IX_ETH_ACC_RX_CNTRL1_CRC_EN    \
+					 | IX_ETH_ACC_RX_CNTRL1_RX_EN)
+
+#define IX_ETH_ACC_RX_CNTRL2_DEFAULT           0x0
+
+/* Thresholds determined by NPE firmware FS */
+#define IX_ETH_ACC_MAC_THRESH_P_EMPTY_DEFAULT  0x12
+#define IX_ETH_ACC_MAC_THRESH_P_FULL_DEFAULT   0x30
+
+/* Number of bytes that must be in the tx fifo before
+   transmission commences*/
+#define IX_ETH_ACC_MAC_BUF_SIZE_TX_DEFAULT     0x8
+
+/* One-part deferral values */
+#define IX_ETH_ACC_MAC_TX_DEFER_DEFAULT        0x15
+#define IX_ETH_ACC_MAC_RX_DEFER_DEFAULT        0x16
+
+/* Two-part deferral values... */
+#define IX_ETH_ACC_MAC_TX_TWO_DEFER_1_DEFAULT  0x08
+#define IX_ETH_ACC_MAC_TX_TWO_DEFER_2_DEFAULT  0x07
+
+/* This value applies to MII */
+#define IX_ETH_ACC_MAC_SLOT_TIME_DEFAULT       0x80
+
+/* This value applies to RMII */
+#define IX_ETH_ACC_MAC_SLOT_TIME_RMII_DEFAULT  0xFF
+
+#define IX_ETH_ACC_MAC_ADDR_MASK_DEFAULT       0xFF
+
+#define IX_ETH_ACC_MAC_INT_CLK_THRESH_DEFAULT  0x1
+/*The following is a value chosen at random*/
+#define IX_ETH_ACC_RANDOM_SEED_DEFAULT         0x8
+
+/*By default we must configure the MAC to generate the 
+  MDC clock*/
+#define IX_ETH_ACC_CORE_DEFAULT                (IX_ETH_ACC_CORE_MDC_EN)
+
+#define IXP425_ETH_ACC_MAX_PHY 2
+#define IXP425_ETH_ACC_MAX_AN_ENTRIES 20
+#define IX_ETH_ACC_MAC_RESET_DELAY    1
+
+#define IX_ETH_ACC_MAC_ALL_BITS_SET   0xFF
+
+#define IX_ETH_ACC_MAC_MSGID_SHL      24
+
+#define IX_ETH_ACC_PORT_DISABLE_DELAY_MSECS 20
+#define IX_ETH_ACC_PORT_DISABLE_DELAY_COUNT 200  /* 4 seconds timeout */
+#define IX_ETH_ACC_PORT_DISABLE_RETRY_COUNT 3
+#define IX_ETH_ACC_MIB_STATS_DELAY_MSECS 2000 /* 2 seconds delay for ethernet stats */ 
+
+/*Register access macros*/
+#if  (CPU == SIMSPARCSOLARIS)
+extern void registerWriteStub (UINT32 base, UINT32 offset, UINT32 val);
+extern UINT32 registerReadStub (UINT32 base, UINT32 offset);
+
+#define REG_WRITE(b,o,v) registerWriteStub(b, o, v)
+#define REG_READ(b,o,v)  do { v = registerReadStub(b, o); } while (0)
+#else
+#define REG_WRITE(b,o,v) IX_OSAL_WRITE_LONG((volatile UINT32 *)(b + o), v)
+#define REG_READ(b,o,v)  (v = IX_OSAL_READ_LONG((volatile UINT32 *)(b + o)))
+
+#endif
+
+void ixEthAccMacUnload(void);
+IxEthAccStatus ixEthAccMacMemInit(void);
+
+/* MAC core loopback */
+IxEthAccStatus ixEthAccPortLoopbackEnable(IxEthAccPortId portId);
+IxEthAccStatus ixEthAccPortLoopbackDisable(IxEthAccPortId portId);
+
+/* MAC core traffic control */
+IxEthAccStatus ixEthAccPortTxEnablePriv(IxEthAccPortId portId);
+IxEthAccStatus ixEthAccPortTxDisablePriv(IxEthAccPortId portId);
+IxEthAccStatus ixEthAccPortRxEnablePriv(IxEthAccPortId portId);
+IxEthAccStatus ixEthAccPortRxDisablePriv(IxEthAccPortId portId);
+IxEthAccStatus ixEthAccPortMacResetPriv(IxEthAccPortId portId);
+
+/* NPE software loopback */
+IxEthAccStatus ixEthAccNpeLoopbackDisablePriv(IxEthAccPortId portId);
+IxEthAccStatus ixEthAccNpeLoopbackEnablePriv(IxEthAccPortId portId);
+
+#endif /*IxEthAccMac_p_H*/
+

cpu/ixp/npe/include/IxEthAccMii_p.h

@@ -0,0 +1,97 @@
+/**
+ * @file IxEthAccMii_p.h
+ *
+ * @author Intel Corporation
+ * @date 
+ *
+ * @brief  MII Header file
+ *
+ * Design Notes:
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+#ifndef IxEthAccMii_p_H
+#define IxEthAccMii_p_H
+
+/* MII definitions - these have been verified against the LXT971 and LXT972 PHYs*/
+
+#define IXP425_ETH_ACC_MII_MAX_REG      32      /* max register per phy */
+
+#define IX_ETH_ACC_MII_REG_SHL    16
+#define IX_ETH_ACC_MII_ADDR_SHL   21
+
+/* Definitions for MII access routines*/
+ 
+#define IX_ETH_ACC_MII_GO                  BIT(31)
+#define IX_ETH_ACC_MII_WRITE               BIT(26)
+#define IX_ETH_ACC_MII_TIMEOUT_10TH_SECS        5    
+#define IX_ETH_ACC_MII_10TH_SEC_IN_MILLIS     100
+#define IX_ETH_ACC_MII_READ_FAIL           BIT(31)
+ 
+#define IX_ETH_ACC_MII_PHY_DEF_DELAY   300  /* max delay before link up, etc. */
+#define IX_ETH_ACC_MII_PHY_NO_DELAY    0x0  /* do not delay */
+#define IX_ETH_ACC_MII_PHY_NULL        0xff /* PHY is not present */
+#define IX_ETH_ACC_MII_PHY_DEF_ADDR    0x0  /* default PHY's logical address */
+
+#ifndef IX_ETH_ACC_MII_MONITOR_DELAY
+#   define IX_ETH_ACC_MII_MONITOR_DELAY   0x5    /* in seconds */
+#endif
+
+/* Register definition */  
+
+#define IX_ETH_ACC_MII_CTRL_REG	    0x0	/* Control Register */
+#define IX_ETH_ACC_MII_STAT_REG	    0x1	/* Status Register */
+#define IX_ETH_ACC_MII_PHY_ID1_REG  0x2	/* PHY identifier 1 Register */
+#define IX_ETH_ACC_MII_PHY_ID2_REG  0x3	/* PHY identifier 2 Register */
+#define IX_ETH_ACC_MII_AN_ADS_REG   0x4	/* Auto-Negotiation 	  */
+					/* Advertisement Register */
+#define IX_ETH_ACC_MII_AN_PRTN_REG  0x5	/* Auto-Negotiation 	    */
+					/* partner ability Register */
+#define IX_ETH_ACC_MII_AN_EXP_REG   0x6	/* Auto-Negotiation   */
+					/* Expansion Register */
+#define IX_ETH_ACC_MII_AN_NEXT_REG  0x7	/* Auto-Negotiation 	       */
+					/* next-page transmit Register */
+
+IxEthAccStatus ixEthAccMdioShow (void);
+IxEthAccStatus ixEthAccMiiInit(void);
+void ixEthAccMiiUnload(void);
+
+#endif  /*IxEthAccMii_p_H*/

cpu/ixp/npe/include/IxEthAccQueueAssign_p.h

@@ -0,0 +1,137 @@
+/**
+ * @file IxEthAccQueueAssign_p.h
+ *
+ * @author Intel Corporation
+ * @date 06-Mar-2002
+ *
+ * @brief   Mapping from QMgr Q's to internal assignment
+ *
+ * Design Notes:
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+/**
+ * @addtogroup IxEthAccPri
+ *@{
+ */
+
+/*
+ * Os/System dependancies.
+ */
+#include "IxOsal.h"
+
+/*
+ * Intermodule dependancies
+ */
+#include "IxQMgr.h"
+#include "IxQueueAssignments.h"
+
+/* Check range of Q's assigned to this component. */
+#if IX_ETH_ACC_RX_FRAME_ETH_Q >= (IX_QMGR_MIN_QUEUPP_QID ) |    	\
+ IX_ETH_ACC_RX_FREE_BUFF_ENET0_Q  >=  (IX_QMGR_MIN_QUEUPP_QID) | 	\
+ IX_ETH_ACC_RX_FREE_BUFF_ENET1_Q  >=  (IX_QMGR_MIN_QUEUPP_QID) | 	\
+ IX_ETH_ACC_TX_FRAME_ENET0_Q >=   (IX_QMGR_MIN_QUEUPP_QID) |  		\
+ IX_ETH_ACC_TX_FRAME_ENET1_Q >=   (IX_QMGR_MIN_QUEUPP_QID) | 		\
+ IX_ETH_ACC_TX_FRAME_DONE_ETH_Q  >=  (IX_QMGR_MIN_QUEUPP_QID)  
+#error "Not all Ethernet Access Queues are betweem 1-31, requires full functionalty Q's unless otherwise validated "
+#endif
+
+/**
+*
+* @typedef  IxEthAccQregInfo
+*
+* @brief 
+*
+*/
+typedef struct 
+{
+   IxQMgrQId qId;
+   char *qName;
+   IxQMgrCallback qCallback;
+   IxQMgrCallbackId callbackTag;
+   IxQMgrQSizeInWords qSize;
+   IxQMgrQEntrySizeInWords qWords; 
+   BOOL           qNotificationEnableAtStartup;
+   IxQMgrSourceId qConditionSource; 
+   IxQMgrWMLevel  AlmostEmptyThreshold;
+   IxQMgrWMLevel  AlmostFullThreshold;
+
+} IxEthAccQregInfo;
+
+/*
+ * Prototypes for all QM callbacks.
+ */
+
+/* 
+ * Rx Callbacks 
+ */
+IX_ETH_ACC_PUBLIC
+void  ixEthRxFrameQMCallback(IxQMgrQId, IxQMgrCallbackId);
+
+IX_ETH_ACC_PUBLIC
+void  ixEthRxMultiBufferQMCallback(IxQMgrQId, IxQMgrCallbackId);
+
+IX_ETH_ACC_PUBLIC
+void  ixEthRxFreeQMCallback(IxQMgrQId, IxQMgrCallbackId);
+
+/* 
+ * Tx Callback.
+ */
+IX_ETH_ACC_PUBLIC
+void  ixEthTxFrameQMCallback(IxQMgrQId, IxQMgrCallbackId);
+
+IX_ETH_ACC_PUBLIC
+void  ixEthTxFrameDoneQMCallback(IxQMgrQId, IxQMgrCallbackId );
+
+
+#define IX_ETH_ACC_QM_QUEUE_DISPATCH_PRIORITY (IX_QMGR_Q_PRIORITY_0) /* Highest priority */
+
+/*
+ * Queue watermarks
+ */
+#define IX_ETH_ACC_RX_FRAME_ETH_Q_SOURCE 		(IX_QMGR_Q_SOURCE_ID_NOT_E   )
+#define IX_ETH_ACC_RX_FREE_BUFF_ENET0_Q_SOURCE 		(IX_QMGR_Q_SOURCE_ID_E   )   
+#define IX_ETH_ACC_RX_FREE_BUFF_ENET1_Q_SOURCE 		(IX_QMGR_Q_SOURCE_ID_E   )
+#define IX_ETH_ACC_RX_FREE_BUFF_ENET2_Q_SOURCE 		(IX_QMGR_Q_SOURCE_ID_E   )
+#define IX_ETH_ACC_TX_FRAME_ENET0_Q_SOURCE 		(IX_QMGR_Q_SOURCE_ID_E   ) 
+#define IX_ETH_ACC_TX_FRAME_ENET1_Q_SOURCE 		(IX_QMGR_Q_SOURCE_ID_E   )
+#define IX_ETH_ACC_TX_FRAME_ENET2_Q_SOURCE 		(IX_QMGR_Q_SOURCE_ID_E   )
+#define IX_ETH_ACC_TX_FRAME_DONE_ETH_Q_SOURCE 		(IX_QMGR_Q_SOURCE_ID_NOT_E   )

cpu/ixp/npe/include/IxEthAcc_p.h

@@ -0,0 +1,325 @@
+/**
+ * @file IxEthAcc_p.h
+ *
+ * @author Intel Corporation
+ * @date 12-Feb-2002
+ *
+ * @brief  Internal Header file for IXP425 Ethernet Access component.
+ *
+ * Design Notes:
+ *
+ * 
+ * @par
+ * IXP400 SW Release version 2.0
+ * 
+ * -- Copyright Notice --
+ * 
+ * @par
+ * Copyright 2001-2005, Intel Corporation.
+ * All rights reserved.
+ * 
+ * @par
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of the Intel Corporation nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * 
+ * @par
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS''
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ * 
+ * @par
+ * -- End of Copyright Notice --
+ */
+
+/**
+ * @addtogroup IxEthAccPri
+ *@{
+ */
+
+#ifndef IxEthAcc_p_H
+#define IxEthAcc_p_H
+
+/*
+ * Os/System dependancies.
+ */
+#include "IxOsal.h"
+
+/*
+ * Intermodule dependancies
+ */
+#include "IxNpeDl.h"
+#include "IxQMgr.h"
+
+#include "IxEthNpe.h"
+
+/* 
+ * Intra module dependancies
+ */
+
+#include "IxEthAccDataPlane_p.h"
+#include "IxEthAccMac_p.h"
+
+
+#define INLINE __inline__
+
+#ifdef NDEBUG
+
+#define IX_ETH_ACC_PRIVATE static
+
+#else
+
+#define IX_ETH_ACC_PRIVATE
+
+#endif /* ndef NDEBUG */
+
+#define IX_ETH_ACC_PUBLIC
+
+
+#define IX_ETH_ACC_IS_PORT_VALID(port) ((port) <  IX_ETH_ACC_NUMBER_OF_PORTS  ? TRUE : FALSE )
+
+
+
+#ifndef NDEBUG
+#define IX_ETH_ACC_FATAL_LOG(a,b,c,d,e,f,g)   { ixOsalLog ( IX_OSAL_LOG_LVL_FATAL,IX_OSAL_LOG_DEV_STDOUT,a,b,c,d,e,f,g);}
+#define IX_ETH_ACC_WARNING_LOG(a,b,c,d,e,f,g) { ixOsalLog ( IX_OSAL_LOG_LVL_WARNING,IX_OSAL_LOG_DEV_STDOUT,a,b,c,d,e,f,g);}
+#define IX_ETH_ACC_DEBUG_LOG(a,b,c,d,e,f,g)   { ixOsalLog ( IX_OSAL_LOG_LVL_FATAL,IX_OSAL_LOG_DEV_STDOUT,a,b,c,d,e,f,g);}
+#else
+#define IX_ETH_ACC_FATAL_LOG(a,b,c,d,e,f,g)   { ixOsalLog ( IX_OSAL_LOG_LVL_FATAL,IX_OSAL_LOG_DEV_STDOUT,a,b,c,d,e,f,g);}
+#define IX_ETH_ACC_WARNING_LOG(a,b,c,d,e,f,g) { ixOsalLog ( IX_OSAL_LOG_LVL_WARNING,IX_OSAL_LOG_DEV_STDOUT,a,b,c,d,e,f,g);}
+#define IX_ETH_ACC_DEBUG_LOG(a,b,c,d,e,f,g)   {}
+#endif
+
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccInitDataPlane(void);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccQMgrQueuesConfig(void);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccQMgrRxCallbacksRegister(IxQMgrCallback ixQMgrCallback);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccSingleEthNpeCheck(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC void ixEthAccQMgrRxQEntryGet(UINT32 *numRxQueueEntries);
+
+/* prototypes for the private control plane functions (used by the control interface wrapper) */
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortEnablePriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortDisablePriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortEnabledQueryPriv(IxEthAccPortId portId, BOOL *enabled);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortPromiscuousModeClearPriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortPromiscuousModeSetPriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortUnicastMacAddressSetPriv(IxEthAccPortId portId, IxEthAccMacAddr *macAddr);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortUnicastMacAddressGetPriv(IxEthAccPortId portId, IxEthAccMacAddr *macAddr);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortMulticastAddressJoinPriv(IxEthAccPortId portId, IxEthAccMacAddr *macAddr);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortMulticastAddressJoinAllPriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortMulticastAddressLeavePriv(IxEthAccPortId portId, IxEthAccMacAddr *macAddr);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortMulticastAddressLeaveAllPriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortUnicastAddressShowPriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC void ixEthAccPortMulticastAddressShowPriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortDuplexModeSetPriv(IxEthAccPortId portId, IxEthAccDuplexMode mode);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortDuplexModeGetPriv(IxEthAccPortId portId, IxEthAccDuplexMode *mode);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortTxFrameAppendPaddingEnablePriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortTxFrameAppendPaddingDisablePriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortTxFrameAppendFCSEnablePriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortTxFrameAppendFCSDisablePriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortRxFrameAppendFCSEnablePriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccPortRxFrameAppendFCSDisablePriv(IxEthAccPortId portId);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccTxSchedulingDisciplineSetPriv(IxEthAccPortId portId, IxEthAccSchedulerDiscipline sched);
+IX_ETH_ACC_PUBLIC IxEthAccStatus ixEthAccRxSchedulingDisciplineSetPriv(IxEthAccSchedulerDiscipline sched);
+
+/**
+ * @struct  ixEthAccRxDataStats
+ * @brief   Stats data structures for data path. - Not obtained from h/w
+ *
+ */
+typedef struct
+{ 
+    UINT32 rxFrameClientCallback;
+    UINT32 rxFreeRepOK;
+    UINT32 rxFreeRepDelayed;
+    UINT32 rxFreeRepFromSwQOK;
+    UINT32 rxFreeRepFromSwQDelayed;
+    UINT32 rxFreeLateNotificationEnabled;
+    UINT32 rxFreeLowCallback;
+    UINT32 rxFreeOverflow;
+    UINT32 rxFreeLock;
+    UINT32 rxDuringDisable;
+    UINT32 rxSwQDuringDisable;
+    UINT32 rxUnlearnedMacAddress;
+    UINT32 rxPriority[IX_ETH_ACC_TX_PRIORITY_7 + 1];
+    UINT32 rxUnexpectedError;
+    UINT32 rxFiltered;
+} IxEthAccRxDataStats;
+
+/**
+ * @struct  IxEthAccTxDataStats
+ * @brief   Stats data structures for data path. - Not obtained from h/w
+ *
+ */
+typedef struct
+{   
+    UINT32 txQOK;
+    UINT32 txQDelayed;
+    UINT32 txFromSwQOK;
+    UINT32 txFromSwQDelayed;
+    UINT32 txLowThreshCallback;
+    UINT32 txDoneClientCallback;
+    UINT32 txDoneClientCallbackDisable;
+    UINT32 txOverflow;
+    UINT32 txLock;
+    UINT32 txPriority[IX_ETH_ACC_TX_PRIORITY_7 + 1];
+    UINT32 txLateNotificationEnabled;
+    UINT32 txDoneDuringDisable;
+    UINT32 txDoneSwQDuringDisable;
+    UINT32 txUnexpectedError;
+} IxEthAccTxDataStats;
+
+/* port Disable state machine : list of states */
+typedef enum
+{
+    /* general port states */
+    DISABLED = 0,
+    ACTIVE,
+
+    /* particular Tx/Rx states */
+    REPLENISH,
+    RECEIVE,
+    TRANSMIT,
+    TRANSMIT_DONE
+} IxEthAccPortDisableState;
+
+typedef struct
+{
+    BOOL fullDuplex;
+    BOOL rxFCSAppend;
+    BOOL txFCSAppend;
+    BOOL txPADAppend;
+    BOOL enabled;
+    BOOL promiscuous;
+    BOOL joinAll;
+    IxOsalMutex ackMIBStatsLock;
+    IxOsalMutex ackMIBStatsResetLock;
+    IxOsalMutex MIBStatsGetAccessLock;
+    IxOsalMutex MIBStatsGetResetAccessLock;
+    IxOsalMutex npeLoopbackMessageLock;
+    IxEthAccMacAddr mcastAddrsTable[IX_ETH_ACC_MAX_MULTICAST_ADDRESSES];
+    UINT32 mcastAddrIndex;
+    IX_OSAL_MBUF *portDisableTxMbufPtr;
+    IX_OSAL_MBUF *portDisableRxMbufPtr;
+
+    volatile IxEthAccPortDisableState portDisableState;
+    volatile IxEthAccPortDisableState rxState;
+    volatile IxEthAccPortDisableState txState;
+
+    BOOL initDone;
+    BOOL macInitialised;
+} IxEthAccMacState;
+
+/**
+ * @struct  IxEthAccRxInfo
+ * @brief   System-wide data structures associated with the data plane.
+ *
+ */
+typedef struct
+{
+  IxQMgrQId higherPriorityQueue[IX_QMGR_MAX_NUM_QUEUES]; /**< higher priority queue list */
+  IxEthAccSchedulerDiscipline schDiscipline; /**< Receive Xscale QoS type */
+} IxEthAccInfo; 
+
+/**
+ * @struct  IxEthAccRxDataInfo
+ * @brief   Per Port data structures associated with the receive data plane.
+ *
+ */
+typedef struct
+{
+  IxQMgrQId rxFreeQueue; /**< rxFree Queue for this port */
+  IxEthAccPortRxCallback rxCallbackFn;
+  UINT32  rxCallbackTag;
+  IxEthAccDataPlaneQList freeBufferList;
+  IxEthAccPortMultiBufferRxCallback rxMultiBufferCallbackFn;
+  UINT32  rxMultiBufferCallbackTag;
+  BOOL rxMultiBufferCallbackInUse;
+  IxEthAccRxDataStats stats; /**< Receive s/w stats */
+} IxEthAccRxDataInfo; 
+
+/**
+ * @struct  IxEthAccTxDataInfo
+ * @brief   Per Port data structures associated with the transmit data plane.
+ *
+ */
+typedef struct
+{
+  IxEthAccPortTxDoneCallback  txBufferDoneCallbackFn;
+  UINT32  txCallbackTag;
+  IxEthAccDataPlaneQList 	txQ[IX_ETH_ACC_NUM_TX_PRIORITIES]; /**< Transmit Q */
+  IxEthAccSchedulerDiscipline schDiscipline; /**< Transmit Xscale QoS */
+  IxQMgrQId txQueue; /**< txQueue for this port */
+  IxEthAccTxDataStats stats; /**< Transmit s/w stats */
+} IxEthAccTxDataInfo; 
+
+
+/**
+ * @struct  IxEthAccPortDataInfo
+ * @brief   Per Port data structures associated with the port data plane.
+ *
+ */
+typedef struct
+{
+    BOOL               portInitialized;
+    UINT32 npeId; /**< NpeId for this port */
+    IxEthAccTxDataInfo ixEthAccTxData; /**< Transmit data control structures */
+    IxEthAccRxDataInfo ixEthAccRxData; /**< Recieve data control structures */
+} IxEthAccPortDataInfo; 
+
+extern IxEthAccPortDataInfo  ixEthAccPortData[];
+#define IX_ETH_IS_PORT_INITIALIZED(port) (ixEthAccPortData[port].portInitialized)
+
+extern BOOL ixEthAccServiceInit;
+#define IX_ETH_ACC_IS_SERVICE_INITIALIZED() (ixEthAccServiceInit == TRUE )
+
+/* 
+ * Maximum number of frames to consume from the Rx Frame Q.
+ */
+
+#define IX_ETH_ACC_MAX_RX_FRAME_CONSUME_PER_CALLBACK (128)
+
+/*
+ * Max number of times to load the Rx Free Q from callback.  
+ */
+#define IX_ETH_ACC_MAX_RX_FREE_BUFFERS_LOAD (256)  /* Set greater than depth of h/w Q + drain time at line rate */
+
+/*
+ *  Max number of times to read from the Tx Done Q in one sitting.
+ */
+
+#define IX_ETH_ACC_MAX_TX_FRAME_DONE_CONSUME_PER_CALLBACK (256)
+
+/*
+ *  Max number of times to take buffers from S/w queues and write them to the H/w Tx
+ *  queues on receipt of a Tx low threshold callback 
+ */
+
+#define IX_ETH_ACC_MAX_TX_FRAME_TX_CONSUME_PER_CALLBACK (16)
+
+
+#define IX_ETH_ACC_FLUSH_CACHE(addr,size)  IX_OSAL_CACHE_FLUSH((addr),(size))
+#define IX_ETH_ACC_INVALIDATE_CACHE(addr,size)  IX_OSAL_CACHE_INVALIDATE((addr),(size))
+
+
+#define IX_ETH_ACC_MEMSET(start,value,size) memset(start,value,size)
+
+#endif /* ndef IxEthAcc_p_H */
+
+
+