Merge branch 'master' into sh/cachetlb

Conflicts:
	arch/sh/kernel/Makefile_64

Makefile

@@ -1,7 +1,7 @@
 VERSION = 2
 PATCHLEVEL = 6
 SUBLEVEL = 31
-EXTRAVERSION = -rc5
+EXTRAVERSION = -rc6
 NAME = Man-Eating Seals of Antiquity
 
 # *DOCUMENTATION*

arch/sh/boards/mach-se/7724/setup.c

@@ -246,7 +246,7 @@ static struct platform_device ceu1_device = {
 	},
 };
 
-/* KEYSC */
+/* KEYSC in SoC (Needs SW33-2 set to ON) */
 static struct sh_keysc_info keysc_info = {
 	.mode = SH_KEYSC_MODE_1,
 	.scan_timing = 10,
@@ -263,12 +263,13 @@ static struct sh_keysc_info keysc_info = {
 
 static struct resource keysc_resources[] = {
 	[0] = {
-		.start  = 0x1a204000,
-		.end    = 0x1a20400f,
+		.name	= "KEYSC",
+		.start  = 0x044b0000,
+		.end    = 0x044b000f,
 		.flags  = IORESOURCE_MEM,
 	},
 	[1] = {
-		.start  = IRQ0_KEY,
+		.start  = 79,
 		.flags  = IORESOURCE_IRQ,
 	},
 };

arch/sh/drivers/heartbeat.c

@@ -40,14 +40,19 @@ static inline void heartbeat_toggle_bit(struct heartbeat_data *hd,
 	if (inverted)
 		new = ~new;
 
+	new &= hd->mask;
+
 	switch (hd->regsize) {
 	case 32:
+		new |= ioread32(hd->base) & ~hd->mask;
 		iowrite32(new, hd->base);
 		break;
 	case 16:
+		new |= ioread16(hd->base) & ~hd->mask;
 		iowrite16(new, hd->base);
 		break;
 	default:
+		new |= ioread8(hd->base) & ~hd->mask;
 		iowrite8(new, hd->base);
 		break;
 	}
@@ -72,6 +77,7 @@ static int heartbeat_drv_probe(struct platform_device *pdev)
 {
 	struct resource *res;
 	struct heartbeat_data *hd;
+	int i;
 
 	if (unlikely(pdev->num_resources != 1)) {
 		dev_err(&pdev->dev, "invalid number of resources\n");
@@ -107,6 +113,10 @@ static int heartbeat_drv_probe(struct platform_device *pdev)
 		hd->nr_bits = ARRAY_SIZE(default_bit_pos);
 	}
 
+	hd->mask = 0;
+	for (i = 0; i < hd->nr_bits; i++)
+		hd->mask |= (1 << hd->bit_pos[i]);
+
 	if (!hd->regsize)
 		hd->regsize = 8;	/* default access size */
 

arch/sh/include/asm/dwarf.h

@@ -340,6 +340,10 @@ struct dwarf_stack {
 #define DW_CFA_lo_user		0x1c
 #define DW_CFA_hi_user		0x3f
 
+/* GNU extension opcodes  */
+#define DW_CFA_GNU_args_size	0x2e
+#define DW_CFA_GNU_negative_offset_extended 0x2f
+
 /*
  * Some call frame instructions encode their operands in the opcode. We
  * need some helper functions to extract both the opcode and operands

arch/sh/include/asm/heartbeat.h

@@ -11,6 +11,7 @@ struct heartbeat_data {
 	unsigned int nr_bits;
 	struct timer_list timer;
 	unsigned int regsize;
+	unsigned int mask;
 	unsigned long flags;
 };
 

arch/sh/include/mach-kfr2r09/mach/partner-jet-setup.txt

@@ -24,6 +24,7 @@ LIST "setup clocks"
 ED 0xa4150004, 0x00000050
 ED 0xa4150000, 0x91053508
 WAIT 1
+ED 0xa4150050, 0x00000340
 ED 0xa4150024, 0x00005000
 
 LIST "setup pins"

arch/sh/kernel/Makefile

@@ -1,5 +1,40 @@
-ifeq ($(CONFIG_SUPERH32),y)
-include ${srctree}/arch/sh/kernel/Makefile_32
-else
-include ${srctree}/arch/sh/kernel/Makefile_64
+#
+# Makefile for the Linux/SuperH kernel.
+#
+
+extra-y	:= head_$(BITS).o init_task.o vmlinux.lds
+
+ifdef CONFIG_FUNCTION_TRACER
+# Do not profile debug and lowlevel utilities
+CFLAGS_REMOVE_ftrace.o = -pg
 endif
+
+obj-y	:= debugtraps.o dumpstack.o idle.o io.o io_generic.o irq.o	\
+	   machvec.o process_$(BITS).o ptrace_$(BITS).o setup.o 	\
+	   signal_$(BITS).o sys_sh.o sys_sh$(BITS).o syscalls_$(BITS).o	\
+	   time.o topology.o traps.o traps_$(BITS).o unwinder.o
+
+obj-y				+= cpu/
+obj-$(CONFIG_VSYSCALL)		+= vsyscall/
+obj-$(CONFIG_SMP)		+= smp.o
+obj-$(CONFIG_SH_STANDARD_BIOS)	+= sh_bios.o
+obj-$(CONFIG_KGDB)		+= kgdb.o
+obj-$(CONFIG_SH_CPU_FREQ)	+= cpufreq.o
+obj-$(CONFIG_MODULES)		+= sh_ksyms_$(BITS).o module.o
+obj-$(CONFIG_EARLY_PRINTK)	+= early_printk.o
+obj-$(CONFIG_KEXEC)		+= machine_kexec.o relocate_kernel.o
+obj-$(CONFIG_CRASH_DUMP)	+= crash_dump.o
+obj-$(CONFIG_STACKTRACE)	+= stacktrace.o
+obj-$(CONFIG_IO_TRAPPED)	+= io_trapped.o
+obj-$(CONFIG_KPROBES)		+= kprobes.o
+obj-$(CONFIG_GENERIC_GPIO)	+= gpio.o
+obj-$(CONFIG_DYNAMIC_FTRACE)	+= ftrace.o
+obj-$(CONFIG_FTRACE_SYSCALLS)	+= ftrace.o
+obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o
+obj-$(CONFIG_DUMP_CODE)		+= disassemble.o
+obj-$(CONFIG_HIBERNATION)	+= swsusp.o
+obj-$(CONFIG_DWARF_UNWINDER)	+= dwarf.o
+
+obj-$(CONFIG_GENERIC_CLOCKEVENTS_BROADCAST)	+= localtimer.o
+
+EXTRA_CFLAGS += -Werror

arch/sh/kernel/Makefile_32

@@ -1,40 +0,0 @@
-#
-# Makefile for the Linux/SuperH kernel.
-#
-
-extra-y	:= head_32.o init_task.o vmlinux.lds
-
-ifdef CONFIG_FUNCTION_TRACER
-# Do not profile debug and lowlevel utilities
-CFLAGS_REMOVE_ftrace.o = -pg
-endif
-
-obj-y	:= debugtraps.o dumpstack.o idle.o io.o io_generic.o irq.o	\
-	   machvec.o process_32.o ptrace_32.o setup.o signal_32.o	\
-	   sys_sh.o sys_sh32.o syscalls_32.o time.o topology.o		\
-	   traps.o traps_32.o unwinder.o
-
-obj-y				+= cpu/
-obj-$(CONFIG_VSYSCALL)		+= vsyscall/
-obj-$(CONFIG_SMP)		+= smp.o
-obj-$(CONFIG_SH_STANDARD_BIOS)	+= sh_bios.o
-obj-$(CONFIG_KGDB)		+= kgdb.o
-obj-$(CONFIG_SH_CPU_FREQ)	+= cpufreq.o
-obj-$(CONFIG_MODULES)		+= sh_ksyms_32.o module.o
-obj-$(CONFIG_EARLY_PRINTK)	+= early_printk.o
-obj-$(CONFIG_KEXEC)		+= machine_kexec.o relocate_kernel.o
-obj-$(CONFIG_CRASH_DUMP)	+= crash_dump.o
-obj-$(CONFIG_STACKTRACE)	+= stacktrace.o
-obj-$(CONFIG_IO_TRAPPED)	+= io_trapped.o
-obj-$(CONFIG_KPROBES)		+= kprobes.o
-obj-$(CONFIG_GENERIC_GPIO)	+= gpio.o
-obj-$(CONFIG_DYNAMIC_FTRACE)	+= ftrace.o
-obj-$(CONFIG_FTRACE_SYSCALLS)	+= ftrace.o
-obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o
-obj-$(CONFIG_DUMP_CODE)		+= disassemble.o
-obj-$(CONFIG_HIBERNATION)	+= swsusp.o
-obj-$(CONFIG_DWARF_UNWINDER)	+= dwarf.o
-
-obj-$(CONFIG_GENERIC_CLOCKEVENTS_BROADCAST)	+= localtimer.o
-
-EXTRA_CFLAGS += -Werror

arch/sh/kernel/Makefile_64

@@ -1,21 +0,0 @@
-extra-y	:= head_64.o init_task.o vmlinux.lds
-
-obj-y	:= debugtraps.o dumpstack.o idle.o io.o io_generic.o irq.o	\
-	   machvec.o process_64.o ptrace_64.o setup.o signal_64.o	\
-	   sys_sh.o sys_sh64.o syscalls_64.o time.o topology.o 		\
-	   traps.o traps_64.o unwinder.o
-
-obj-y				+= cpu/
-obj-$(CONFIG_SMP)		+= smp.o
-obj-$(CONFIG_SH_CPU_FREQ)	+= cpufreq.o
-obj-$(CONFIG_MODULES)		+= sh_ksyms_64.o module.o
-obj-$(CONFIG_EARLY_PRINTK)	+= early_printk.o
-obj-$(CONFIG_CRASH_DUMP)	+= crash_dump.o
-obj-$(CONFIG_STACKTRACE)	+= stacktrace.o
-obj-$(CONFIG_IO_TRAPPED)	+= io_trapped.o
-obj-$(CONFIG_GENERIC_GPIO)	+= gpio.o
-obj-$(CONFIG_DWARF_UNWINDER)	+= dwarf.o
-
-obj-$(CONFIG_GENERIC_CLOCKEVENTS_BROADCAST)	+= localtimer.o
-
-EXTRA_CFLAGS += -Werror

arch/sh/kernel/cpu/init.c

@@ -3,7 +3,7 @@
  *
  * CPU init code
  *
- * Copyright (C) 2002 - 2007  Paul Mundt
+ * Copyright (C) 2002 - 2009  Paul Mundt
  * Copyright (C) 2003  Richard Curnow
  *
  * This file is subject to the terms and conditions of the GNU General Public
@@ -62,6 +62,37 @@ static void __init speculative_execution_init(void)
 #define speculative_execution_init()	do { } while (0)
 #endif
 
+#ifdef CONFIG_CPU_SH4A
+#define EXPMASK			0xff2f0004
+#define EXPMASK_RTEDS		(1 << 0)
+#define EXPMASK_BRDSSLP		(1 << 1)
+#define EXPMASK_MMCAW		(1 << 4)
+
+static void __init expmask_init(void)
+{
+	unsigned long expmask = __raw_readl(EXPMASK);
+
+	/*
+	 * Future proofing.
+	 *
+	 * Disable support for slottable sleep instruction
+	 * and non-nop instructions in the rte delay slot.
+	 */
+	expmask &= ~(EXPMASK_RTEDS | EXPMASK_BRDSSLP);
+
+	/*
+	 * Enable associative writes to the memory-mapped cache array
+	 * until the cache flush ops have been rewritten.
+	 */
+	expmask |= EXPMASK_MMCAW;
+
+	__raw_writel(expmask, EXPMASK);
+	ctrl_barrier();
+}
+#else
+#define expmask_init()	do { } while (0)
+#endif
+
 /* 2nd-level cache init */
 void __uses_jump_to_uncached __attribute__ ((weak)) l2_cache_init(void)
 {
@@ -319,4 +350,5 @@ asmlinkage void __init sh_cpu_init(void)
 #endif
 
 	speculative_execution_init();
+	expmask_init();
 }

arch/sh/kernel/cpu/sh2/entry.S

@@ -227,8 +227,9 @@ ENTRY(sh_bios_handler)
 	mov.l	@r15+, r14
 	add	#8,r15
 	lds.l	@r15+, pr
+	mov.l	@r15+,r15
 	rte
-	 mov.l	@r15+,r15
+	 nop
 	.align	2
 1:	.long	gdb_vbr_vector
 #endif /* CONFIG_SH_STANDARD_BIOS */

arch/sh/kernel/cpu/sh2a/entry.S

@@ -176,8 +176,9 @@ ENTRY(sh_bios_handler)
 	movml.l	@r15+,r14
 	add	#8,r15
 	lds.l	@r15+, pr
+	mov.l	@r15+,r15
 	rte
-	 mov.l	@r15+,r15
+	 nop
 	.align	2
 1:	.long	gdb_vbr_vector
 #endif /* CONFIG_SH_STANDARD_BIOS */

arch/sh/kernel/cpu/sh3/entry.S

@@ -516,6 +516,14 @@ ENTRY(handle_interrupt)
 	bsr	save_regs	! needs original pr value in k3
 	 mov	#-1, k2		! default vector kept in k2
 
+	stc	sr, r0	! get status register
+	shlr2	r0
+	and	#0x3c, r0
+	cmp/eq	#0x3c, r0
+	bf	9f
+	TRACE_IRQS_OFF
+9:
+
 	! Setup return address and jump to do_IRQ
 	mov.l	4f, r9		! fetch return address
 	lds	r9, pr		! put return address in pr

arch/sh/kernel/cpu/sh4a/clock-sh7724.c

@@ -127,7 +127,7 @@ struct clk *main_clks[] = {
 	&div3_clk,
 };
 
-static int divisors[] = { 2, 0, 4, 6, 8, 12, 16, 0, 24, 32, 36, 48, 0, 72 };
+static int divisors[] = { 2, 3, 4, 6, 8, 12, 16, 0, 24, 32, 36, 48, 0, 72 };
 
 static struct clk_div_mult_table div4_table = {
 	.divisors = divisors,

arch/sh/kernel/cpu/shmobile/cpuidle.c

@@ -21,6 +21,7 @@
 static unsigned long cpuidle_mode[] = {
 	SUSP_SH_SLEEP, /* regular sleep mode */
 	SUSP_SH_SLEEP | SUSP_SH_SF, /* sleep mode + self refresh */
+	SUSP_SH_STANDBY | SUSP_SH_SF, /* software standby mode + self refresh */
 };
 
 static int cpuidle_sleep_enter(struct cpuidle_device *dev,
@@ -96,6 +97,16 @@ void sh_mobile_setup_cpuidle(void)
 	state->flags |= CPUIDLE_FLAG_TIME_VALID;
 	state->enter = cpuidle_sleep_enter;
 
+	state = &dev->states[i++];
+	snprintf(state->name, CPUIDLE_NAME_LEN, "C2");
+	strncpy(state->desc, "SuperH Mobile Standby Mode [SF]", CPUIDLE_DESC_LEN);
+	state->exit_latency = 2300;
+	state->target_residency = 1 * 2;
+	state->power_usage = 1;
+	state->flags = 0;
+	state->flags |= CPUIDLE_FLAG_TIME_VALID;
+	state->enter = cpuidle_sleep_enter;
+
 	dev->state_count = i;
 
 	cpuidle_register_device(dev);

arch/sh/kernel/cpu/shmobile/pm.c

@@ -41,23 +41,11 @@ extern const unsigned int sh_mobile_standby_size;
 
 void sh_mobile_call_standby(unsigned long mode)
 {
-	extern void *vbr_base;
 	void *onchip_mem = (void *)ILRAM_BASE;
-	void (*standby_onchip_mem)(unsigned long) = onchip_mem;
-
-	/* Note: Wake up from sleep may generate exceptions!
-	 * Setup VBR to point to on-chip ram if self-refresh is
-	 * going to be used.
-	 */
-	if (mode & SUSP_SH_SF)
-		asm volatile("ldc %0, vbr" : : "r" (onchip_mem) : "memory");
+	void (*standby_onchip_mem)(unsigned long, unsigned long) = onchip_mem;
 
 	/* Let assembly snippet in on-chip memory handle the rest */
-	standby_onchip_mem(mode);
-
-	/* Put VBR back in System RAM again */
-	if (mode & SUSP_SH_SF)
-		asm volatile("ldc %0, vbr" : : "r" (&vbr_base) : "memory");
+	standby_onchip_mem(mode, ILRAM_BASE);
 }
 
 static int sh_pm_enter(suspend_state_t state)

arch/sh/kernel/cpu/shmobile/sleep.S

@@ -16,18 +16,73 @@
 #include <asm/asm-offsets.h>
 #include <asm/suspend.h>
 
+/*
+ * Kernel mode register usage, see entry.S:
+ *	k0	scratch
+ *	k1	scratch
+ *	k4	scratch
+ */
+#define k0	r0
+#define k1	r1
+#define k4	r4
+
 /* manage self-refresh and enter standby mode.
  * this code will be copied to on-chip memory and executed from there.
  */
 
 	.balign 	4096,0,4096
 ENTRY(sh_mobile_standby)
+
+	/* save original vbr */
+	stc	vbr, r1
+	mova	saved_vbr, r0
+	mov.l	r1, @r0
+
+	/* point vbr to our on-chip memory page */
+	ldc	r5, vbr
+
+	/* save return address */
+	mova	saved_spc, r0
+	sts	pr, r5
+	mov.l	r5, @r0
+
+	/* save sr */
+	mova	saved_sr, r0
+	stc	sr, r5
+	mov.l	r5, @r0
+
+	/* save mode flags */
+	mova	saved_mode, r0
+	mov.l	r4, @r0
+
+	/* put mode flags in r0 */
 	mov	r4, r0
 
 	tst	#SUSP_SH_SF, r0
 	bt	skip_set_sf
+#ifdef CONFIG_CPU_SUBTYPE_SH7724
+	/* DBSC: put memory in self-refresh mode */
+	mov.l	dben_reg, r4
+	mov.l	dben_data0, r1
+	mov.l	r1, @r4
 
-	/* SDRAM: disable power down and put in self-refresh mode */
+	mov.l	dbrfpdn0_reg, r4
+	mov.l	dbrfpdn0_data0, r1
+	mov.l	r1, @r4
+
+	mov.l	dbcmdcnt_reg, r4
+	mov.l	dbcmdcnt_data0, r1
+	mov.l	r1, @r4
+
+	mov.l	dbcmdcnt_reg, r4
+	mov.l	dbcmdcnt_data1, r1
+	mov.l	r1, @r4
+
+	mov.l	dbrfpdn0_reg, r4
+	mov.l	dbrfpdn0_data1, r1
+	mov.l	r1, @r4
+#else
+	/* SBSC: disable power down and put in self-refresh mode */
 	mov.l	1f, r4
 	mov.l	2f, r1
 	mov.l	@r4, r2
@@ -35,16 +90,9 @@ ENTRY(sh_mobile_standby)
 	mov.l   3f, r3
 	and	r3, r2
 	mov.l	r2, @r4
+#endif
 
 skip_set_sf:
-	tst	#SUSP_SH_SLEEP, r0
-	bt	test_standby
-
-	/* set mode to "sleep mode" */
-	bra	do_sleep
-	 mov	#0x00, r1
-
-test_standby:
 	tst	#SUSP_SH_STANDBY, r0
 	bt	test_rstandby
 
@@ -62,62 +110,135 @@ test_rstandby:
 
 test_ustandby:
 	tst	#SUSP_SH_USTANDBY, r0
-	bt	done_sleep
+	bt	force_sleep
 
 	/* set mode to "u-standby mode" */
-	mov	#0x10, r1
+	bra	do_sleep
+	 mov	#0x10, r1
+
+force_sleep:
 
-	/* fall-through */
+	/* set mode to "sleep mode" */
+	mov	#0x00, r1
 
 do_sleep:
 	/* setup and enter selected standby mode */
 	mov.l	5f, r4
 	mov.l	r1, @r4
+again:
 	sleep
+	bra	again
+	 nop
+
+restore_jump_vbr:
+	/* setup spc with return address to c code */
+	mov.l	saved_spc, k0
+	ldc	k0, spc
+
+	/* restore vbr */
+	mov.l	saved_vbr, k0
+	ldc	k0, vbr
+
+	/* setup ssr with saved sr */
+	mov.l	saved_sr, k0
+	ldc	k0, ssr
+
+	/* get mode flags */
+	mov.l	saved_mode, k0
 
 done_sleep:
 	/* reset standby mode to sleep mode */
-	mov.l	5f, r4
-	mov	#0x00, r1
-	mov.l	r1, @r4
+	mov.l	5f, k4
+	mov	#0x00, k1
+	mov.l	k1, @k4
 
-	tst	#SUSP_SH_SF, r0
+	tst	#SUSP_SH_SF, k0
 	bt	skip_restore_sf
 
-	/* SDRAM: set auto-refresh mode */
-	mov.l	1f, r4
-	mov.l	@r4, r2
-	mov.l   4f, r3
-	and	r3, r2
-	mov.l	r2, @r4
-	mov.l	6f, r4
-	mov.l	7f, r1
-	mov.l	8f, r2
-	mov.l	@r4, r3
-	mov	#-1, r4
-	add	r4, r3
-	or	r2, r3
-	mov.l	r3, @r1
+#ifdef CONFIG_CPU_SUBTYPE_SH7724
+	/* DBSC: put memory in auto-refresh mode */
+	mov.l	dbrfpdn0_reg, k4
+	mov.l	dbrfpdn0_data0, k1
+	mov.l	k1, @k4
+
+	nop /* sleep 140 ns */
+	nop
+	nop
+	nop
+
+	mov.l	dbcmdcnt_reg, k4
+	mov.l	dbcmdcnt_data0, k1
+	mov.l	k1, @k4
+
+	mov.l	dbcmdcnt_reg, k4
+	mov.l	dbcmdcnt_data1, k1
+	mov.l	k1, @k4
+
+	mov.l	dben_reg, k4
+	mov.l	dben_data1, k1
+	mov.l	k1, @k4
+
+	mov.l	dbrfpdn0_reg, k4
+	mov.l	dbrfpdn0_data2, k1
+	mov.l	k1, @k4
+#else
+	/* SBSC: set auto-refresh mode */
+	mov.l	1f, k4
+	mov.l	@k4, k0
+	mov.l   4f, k1
+	and	k1, k0
+	mov.l	k0, @k4
+	mov.l	6f, k4
+	mov.l	8f, k0
+	mov.l	@k4, k1
+	mov	#-1, k4
+	add	k4, k1
+	or	k1, k0
+	mov.l	7f, k1
+	mov.l	k0, @k1
+#endif
 skip_restore_sf:
-	rts
+	/* jump to vbr vector */
+	mov.l	saved_vbr, k0
+	mov.l	offset_vbr, k4
+	add	k4, k0
+	jmp	@k0
 	 nop
 
 	.balign 4
+saved_mode:	.long	0
+saved_spc:	.long	0
+saved_sr:	.long	0
+saved_vbr:	.long	0
+offset_vbr:	.long	0x600
+#ifdef CONFIG_CPU_SUBTYPE_SH7724
+dben_reg:	.long	0xfd000010 /* DBEN */
+dben_data0:	.long	0
+dben_data1:	.long	1
+dbrfpdn0_reg:	.long	0xfd000040 /* DBRFPDN0 */
+dbrfpdn0_data0:	.long	0
+dbrfpdn0_data1:	.long	1
+dbrfpdn0_data2:	.long	0x00010000
+dbcmdcnt_reg:	.long	0xfd000014 /* DBCMDCNT */
+dbcmdcnt_data0:	.long	2
+dbcmdcnt_data1:	.long	4
+#else
 1:	.long	0xfe400008 /* SDCR0 */
 2:	.long	0x00000400
 3:	.long	0xffff7fff
 4:	.long	0xfffffbff
+#endif
 5:	.long	0xa4150020 /* STBCR */
 6:	.long   0xfe40001c /* RTCOR */
 7:	.long   0xfe400018 /* RTCNT */
 8:	.long   0xa55a0000
 
+
 /* interrupt vector @ 0x600 */
 	.balign 	0x400,0,0x400
 	.long	0xdeadbeef
 	.balign 	0x200,0,0x200
-	/* sh7722 will end up here in sleep mode */
-	rte
+	bra	restore_jump_vbr
 	 nop
 sh_mobile_standby_end:
 

arch/sh/kernel/dwarf.c

@@ -330,7 +330,6 @@ struct dwarf_fde *dwarf_lookup_fde(unsigned long pc)
  *	@fde: the FDE for this function
  *	@frame: the instructions calculate the CFA for this frame
  *	@pc: the program counter of the address we're interested in
- *	@define_ra: keep executing insns until the return addr reg is defined?
  *
  *	Execute the Call Frame instruction sequence starting at
  *	@insn_start and ending at @insn_end. The instructions describe
@@ -342,36 +341,17 @@ static int dwarf_cfa_execute_insns(unsigned char *insn_start,
 				   struct dwarf_cie *cie,
 				   struct dwarf_fde *fde,
 				   struct dwarf_frame *frame,
-				   unsigned long pc,
-				   bool define_ra)
+				   unsigned long pc)
 {
 	unsigned char insn;
 	unsigned char *current_insn;
 	unsigned int count, delta, reg, expr_len, offset;
-	bool seen_ra_reg;
 
 	current_insn = insn_start;
 
-	/*
-	 * If we're executing instructions for the dwarf_unwind_stack()
-	 * FDE we need to keep executing instructions until the value of
-	 * DWARF_ARCH_RA_REG is defined. See the comment in
-	 * dwarf_unwind_stack() for more details.
-	 */
-	if (define_ra)
-		seen_ra_reg = false;
-	else
-		seen_ra_reg = true;
-
-	while (current_insn < insn_end && (frame->pc <= pc || !seen_ra_reg) ) {
+	while (current_insn < insn_end && frame->pc <= pc) {
 		insn = __raw_readb(current_insn++);
 
-		if (!seen_ra_reg) {
-			if (frame->num_regs >= DWARF_ARCH_RA_REG &&
-			    frame->regs[DWARF_ARCH_RA_REG].flags)
-				seen_ra_reg = true;
-		}
-
 		/*
 		 * Firstly, handle the opcodes that embed their operands
 		 * in the instructions.
@@ -484,6 +464,19 @@ static int dwarf_cfa_execute_insns(unsigned char *insn_start,
 			frame->regs[reg].flags |= DWARF_REG_OFFSET;
 			frame->regs[reg].addr = offset;
 			break;
+		case DW_CFA_GNU_args_size:
+			count = dwarf_read_uleb128(current_insn, &offset);
+			current_insn += count;
+			break;
+		case DW_CFA_GNU_negative_offset_extended:
+			count = dwarf_read_uleb128(current_insn, &reg);
+			current_insn += count;
+			count = dwarf_read_uleb128(current_insn, &offset);
+			offset *= cie->data_alignment_factor;
+			dwarf_frame_alloc_regs(frame, reg);
+			frame->regs[reg].flags |= DWARF_REG_OFFSET;
+			frame->regs[reg].addr = -offset;
+			break;
 		default:
 			pr_debug("unhandled DWARF instruction 0x%x\n", insn);
 			break;
@@ -510,26 +503,17 @@ struct dwarf_frame *dwarf_unwind_stack(unsigned long pc,
 	struct dwarf_fde *fde;
 	unsigned long addr;
 	int i, offset;
-	bool define_ra = false;
 
 	/*
 	 * If this is the first invocation of this recursive function we
 	 * need get the contents of a physical register to get the CFA
 	 * in order to begin the virtual unwinding of the stack.
 	 *
-	 * Setting "define_ra" to true indictates that we want
-	 * dwarf_cfa_execute_insns() to continue executing instructions
-	 * until we know how to calculate the value of DWARF_ARCH_RA_REG
-	 * (which we need in order to kick off the whole unwinding
-	 * process).
-	 *
 	 * NOTE: the return address is guaranteed to be setup by the
 	 * time this function makes its first function call.
 	 */
-	if (!pc && !prev) {
-		pc = (unsigned long)&dwarf_unwind_stack;
-		define_ra = true;
-	}
+	if (!pc && !prev)
+		pc = (unsigned long)current_text_addr();
 
 	frame = kzalloc(sizeof(*frame), GFP_ATOMIC);
 	if (!frame)
@@ -565,11 +549,11 @@ struct dwarf_frame *dwarf_unwind_stack(unsigned long pc,
 	/* CIE initial instructions */
 	dwarf_cfa_execute_insns(cie->initial_instructions,
 				cie->instructions_end, cie, fde,
-				frame, pc, false);
+				frame, pc);
 
 	/* FDE instructions */
 	dwarf_cfa_execute_insns(fde->instructions, fde->end, cie,
-				fde, frame, pc, define_ra);
+				fde, frame, pc);
 
 	/* Calculate the CFA */
 	switch (frame->flags) {

arch/sh/kernel/entry-common.S

@@ -77,15 +77,6 @@ ENTRY(ret_from_irq)
 	!
 	mov	#OFF_SR, r0
 	mov.l	@(r0,r15), r0	! get status register
-
-	shlr2	r0
-	and	#0x3c, r0
-	cmp/eq	#0x3c, r0
-	bt	9f
-	TRACE_IRQS_ON
-9:
-	mov	#OFF_SR, r0
-	mov.l	@(r0,r15), r0	! get status register
 	shll	r0
 	shll	r0		! kernel space?
 	get_current_thread_info r8, r0
@@ -96,6 +87,7 @@ ENTRY(ret_from_irq)
 	 nop
 ENTRY(resume_kernel)
 	cli
+	TRACE_IRQS_OFF
 	mov.l	@(TI_PRE_COUNT,r8), r0	! current_thread_info->preempt_count
 	tst	r0, r0
 	bf	noresched
@@ -213,12 +205,25 @@ syscall_trace_entry:
 	 mov.l	r0, @(OFF_R0,r15)	! Return value
 
 __restore_all:
-	mov.l	1f, r0
+	mov	#OFF_SR, r0
+	mov.l	@(r0,r15), r0	! get status register
+
+	shlr2	r0
+	and	#0x3c, r0
+	cmp/eq	#0x3c, r0
+	bt	1f
+	TRACE_IRQS_ON
+	bra	2f
+	 nop
+1:
+	TRACE_IRQS_OFF
+2:
+	mov.l	3f, r0
 	jmp	@r0
 	 nop
 
 	.align	2
-1:	.long	restore_all
+3:	.long	restore_all
 
 	.align	2
 syscall_badsys:			! Bad syscall number

arch/x86/Kconfig

@@ -24,6 +24,7 @@ config X86
 	select HAVE_UNSTABLE_SCHED_CLOCK
 	select HAVE_IDE
 	select HAVE_OPROFILE
+	select HAVE_PERF_COUNTERS if (!M386 && !M486)
 	select HAVE_IOREMAP_PROT
 	select HAVE_KPROBES
 	select ARCH_WANT_OPTIONAL_GPIOLIB
@@ -742,7 +743,6 @@ config X86_UP_IOAPIC
 config X86_LOCAL_APIC
 	def_bool y
 	depends on X86_64 || SMP || X86_32_NON_STANDARD || X86_UP_APIC
-	select HAVE_PERF_COUNTERS if (!M386 && !M486)
 
 config X86_IO_APIC
 	def_bool y

arch/x86/kernel/cpu/amd.c

@@ -400,6 +400,13 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c)
 		level = cpuid_eax(1);
 		if((level >= 0x0f48 && level < 0x0f50) || level >= 0x0f58)
 			set_cpu_cap(c, X86_FEATURE_REP_GOOD);
+
+		/*
+		 * Some BIOSes incorrectly force this feature, but only K8
+		 * revision D (model = 0x14) and later actually support it.
+		 */
+		if (c->x86_model < 0x14)
+			clear_cpu_cap(c, X86_FEATURE_LAHF_LM);
 	}
 	if (c->x86 == 0x10 || c->x86 == 0x11)
 		set_cpu_cap(c, X86_FEATURE_REP_GOOD);

arch/x86/kernel/cpu/common.c

@@ -59,7 +59,30 @@ void __init setup_cpu_local_masks(void)
 	alloc_bootmem_cpumask_var(&cpu_sibling_setup_mask);
 }
 
-static const struct cpu_dev *this_cpu __cpuinitdata;
+static void __cpuinit default_init(struct cpuinfo_x86 *c)
+{
+#ifdef CONFIG_X86_64
+	display_cacheinfo(c);
+#else
+	/* Not much we can do here... */
+	/* Check if at least it has cpuid */
+	if (c->cpuid_level == -1) {
+		/* No cpuid. It must be an ancient CPU */
+		if (c->x86 == 4)
+			strcpy(c->x86_model_id, "486");
+		else if (c->x86 == 3)
+			strcpy(c->x86_model_id, "386");
+	}
+#endif
+}
+
+static const struct cpu_dev __cpuinitconst default_cpu = {
+	.c_init		= default_init,
+	.c_vendor	= "Unknown",
+	.c_x86_vendor	= X86_VENDOR_UNKNOWN,
+};
+
+static const struct cpu_dev *this_cpu __cpuinitdata = &default_cpu;
 
 DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = {
 #ifdef CONFIG_X86_64
@@ -332,29 +355,6 @@ void switch_to_new_gdt(int cpu)
 
 static const struct cpu_dev *__cpuinitdata cpu_devs[X86_VENDOR_NUM] = {};
 
-static void __cpuinit default_init(struct cpuinfo_x86 *c)
-{
-#ifdef CONFIG_X86_64
-	display_cacheinfo(c);
-#else
-	/* Not much we can do here... */
-	/* Check if at least it has cpuid */
-	if (c->cpuid_level == -1) {
-		/* No cpuid. It must be an ancient CPU */
-		if (c->x86 == 4)
-			strcpy(c->x86_model_id, "486");
-		else if (c->x86 == 3)
-			strcpy(c->x86_model_id, "386");
-	}
-#endif
-}
-
-static const struct cpu_dev __cpuinitconst default_cpu = {
-	.c_init	= default_init,
-	.c_vendor = "Unknown",
-	.c_x86_vendor = X86_VENDOR_UNKNOWN,
-};
-
 static void __cpuinit get_model_name(struct cpuinfo_x86 *c)
 {
 	unsigned int *v;

arch/x86/kernel/cpu/mcheck/therm_throt.c

@@ -36,6 +36,7 @@
 
 static DEFINE_PER_CPU(__u64, next_check) = INITIAL_JIFFIES;
 static DEFINE_PER_CPU(unsigned long, thermal_throttle_count);
+static DEFINE_PER_CPU(bool, thermal_throttle_active);
 
 static atomic_t therm_throt_en		= ATOMIC_INIT(0);
 
@@ -96,24 +97,27 @@ static int therm_throt_process(int curr)
 {
 	unsigned int cpu = smp_processor_id();
 	__u64 tmp_jiffs = get_jiffies_64();
+	bool was_throttled = __get_cpu_var(thermal_throttle_active);
+	bool is_throttled = __get_cpu_var(thermal_throttle_active) = curr;
 
-	if (curr)
+	if (is_throttled)
 		__get_cpu_var(thermal_throttle_count)++;
 
-	if (time_before64(tmp_jiffs, __get_cpu_var(next_check)))
+	if (!(was_throttled ^ is_throttled) &&
+	    time_before64(tmp_jiffs, __get_cpu_var(next_check)))
 		return 0;
 
 	__get_cpu_var(next_check) = tmp_jiffs + CHECK_INTERVAL;
 
 	/* if we just entered the thermal event */
-	if (curr) {
+	if (is_throttled) {
 		printk(KERN_CRIT "CPU%d: Temperature above threshold, "
-		       "cpu clock throttled (total events = %lu)\n", cpu,
-		       __get_cpu_var(thermal_throttle_count));
+		       "cpu clock throttled (total events = %lu)\n",
+		       cpu, __get_cpu_var(thermal_throttle_count));
 
 		add_taint(TAINT_MACHINE_CHECK);
-	} else {
-		printk(KERN_CRIT "CPU%d: Temperature/speed normal\n", cpu);
+	} else if (was_throttled) {
+		printk(KERN_INFO "CPU%d: Temperature/speed normal\n", cpu);
 	}
 
 	return 1;

arch/x86/kernel/cpu/perf_counter.c

@@ -55,6 +55,7 @@ struct x86_pmu {
 	int		num_counters_fixed;
 	int		counter_bits;
 	u64		counter_mask;
+	int		apic;
 	u64		max_period;
 	u64		intel_ctrl;
 };
@@ -72,8 +73,8 @@ static const u64 p6_perfmon_event_map[] =
 {
   [PERF_COUNT_HW_CPU_CYCLES]		= 0x0079,
   [PERF_COUNT_HW_INSTRUCTIONS]		= 0x00c0,
-  [PERF_COUNT_HW_CACHE_REFERENCES]	= 0x0000,
-  [PERF_COUNT_HW_CACHE_MISSES]		= 0x0000,
+  [PERF_COUNT_HW_CACHE_REFERENCES]	= 0x0f2e,
+  [PERF_COUNT_HW_CACHE_MISSES]		= 0x012e,
   [PERF_COUNT_HW_BRANCH_INSTRUCTIONS]	= 0x00c4,
   [PERF_COUNT_HW_BRANCH_MISSES]		= 0x00c5,
   [PERF_COUNT_HW_BUS_CYCLES]		= 0x0062,
@@ -613,6 +614,7 @@ static DEFINE_MUTEX(pmc_reserve_mutex);
 
 static bool reserve_pmc_hardware(void)
 {
+#ifdef CONFIG_X86_LOCAL_APIC
 	int i;
 
 	if (nmi_watchdog == NMI_LOCAL_APIC)
@@ -627,9 +629,11 @@ static bool reserve_pmc_hardware(void)
 		if (!reserve_evntsel_nmi(x86_pmu.eventsel + i))
 			goto eventsel_fail;
 	}
+#endif
 
 	return true;
 
+#ifdef CONFIG_X86_LOCAL_APIC
 eventsel_fail:
 	for (i--; i >= 0; i--)
 		release_evntsel_nmi(x86_pmu.eventsel + i);
@@ -644,10 +648,12 @@ perfctr_fail:
 		enable_lapic_nmi_watchdog();
 
 	return false;
+#endif
 }
 
 static void release_pmc_hardware(void)
 {
+#ifdef CONFIG_X86_LOCAL_APIC
 	int i;
 
 	for (i = 0; i < x86_pmu.num_counters; i++) {
@@ -657,6 +663,7 @@ static void release_pmc_hardware(void)
 
 	if (nmi_watchdog == NMI_LOCAL_APIC)
 		enable_lapic_nmi_watchdog();
+#endif
 }
 
 static void hw_perf_counter_destroy(struct perf_counter *counter)
@@ -748,6 +755,15 @@ static int __hw_perf_counter_init(struct perf_counter *counter)
 		hwc->sample_period = x86_pmu.max_period;
 		hwc->last_period = hwc->sample_period;
 		atomic64_set(&hwc->period_left, hwc->sample_period);
+	} else {
+		/*
+		 * If we have a PMU initialized but no APIC
+		 * interrupts, we cannot sample hardware
+		 * counters (user-space has to fall back and
+		 * sample via a hrtimer based software counter):
+		 */
+		if (!x86_pmu.apic)
+			return -EOPNOTSUPP;
 	}
 
 	counter->destroy = hw_perf_counter_destroy;
@@ -1449,18 +1465,22 @@ void smp_perf_pending_interrupt(struct pt_regs *regs)
 
 void set_perf_counter_pending(void)
 {
+#ifdef CONFIG_X86_LOCAL_APIC
 	apic->send_IPI_self(LOCAL_PENDING_VECTOR);
+#endif
 }
 
 void perf_counters_lapic_init(void)
 {
-	if (!x86_pmu_initialized())
+#ifdef CONFIG_X86_LOCAL_APIC
+	if (!x86_pmu.apic || !x86_pmu_initialized())
 		return;
 
 	/*
 	 * Always use NMI for PMU
 	 */
 	apic_write(APIC_LVTPC, APIC_DM_NMI);
+#endif
 }
 
 static int __kprobes
@@ -1484,7 +1504,9 @@ perf_counter_nmi_handler(struct notifier_block *self,
 
 	regs = args->regs;
 
+#ifdef CONFIG_X86_LOCAL_APIC
 	apic_write(APIC_LVTPC, APIC_DM_NMI);
+#endif
 	/*
 	 * Can't rely on the handled return value to say it was our NMI, two
 	 * counters could trigger 'simultaneously' raising two back-to-back NMIs.
@@ -1515,6 +1537,7 @@ static struct x86_pmu p6_pmu = {
 	.event_map		= p6_pmu_event_map,
 	.raw_event		= p6_pmu_raw_event,
 	.max_events		= ARRAY_SIZE(p6_perfmon_event_map),
+	.apic			= 1,
 	.max_period		= (1ULL << 31) - 1,
 	.version		= 0,
 	.num_counters		= 2,
@@ -1541,6 +1564,7 @@ static struct x86_pmu intel_pmu = {
 	.event_map		= intel_pmu_event_map,
 	.raw_event		= intel_pmu_raw_event,
 	.max_events		= ARRAY_SIZE(intel_perfmon_event_map),
+	.apic			= 1,
 	/*
 	 * Intel PMCs cannot be accessed sanely above 32 bit width,
 	 * so we install an artificial 1<<31 period regardless of
@@ -1564,6 +1588,7 @@ static struct x86_pmu amd_pmu = {
 	.num_counters		= 4,
 	.counter_bits		= 48,
 	.counter_mask		= (1ULL << 48) - 1,
+	.apic			= 1,
 	/* use highest bit to detect overflow */
 	.max_period		= (1ULL << 47) - 1,
 };
@@ -1589,13 +1614,14 @@ static int p6_pmu_init(void)
 		return -ENODEV;
 	}
 
+	x86_pmu = p6_pmu;
+
 	if (!cpu_has_apic) {
-		pr_info("no Local APIC, try rebooting with lapic");
-		return -ENODEV;
+		pr_info("no APIC, boot with the \"lapic\" boot parameter to force-enable it.\n");
+		pr_info("no hardware sampling interrupt available.\n");
+		x86_pmu.apic = 0;
 	}
 
-	x86_pmu				= p6_pmu;
-
 	return 0;
 }
 

arch/x86/kernel/reboot.c

@@ -418,20 +418,20 @@ static int __init set_pci_reboot(const struct dmi_system_id *d)
 }
 
 static struct dmi_system_id __initdata pci_reboot_dmi_table[] = {
-	{	/* Handle problems with rebooting on Apple MacBook5,2 */
+	{	/* Handle problems with rebooting on Apple MacBook5 */
 		.callback = set_pci_reboot,
-		.ident = "Apple MacBook",
+		.ident = "Apple MacBook5",
 		.matches = {
 			DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "MacBook5,2"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "MacBook5"),
 		},
 	},
-	{	/* Handle problems with rebooting on Apple MacBookPro5,1 */
+	{	/* Handle problems with rebooting on Apple MacBookPro5 */
 		.callback = set_pci_reboot,
-		.ident = "Apple MacBookPro5,1",
+		.ident = "Apple MacBookPro5",
 		.matches = {
 			DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "MacBookPro5,1"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "MacBookPro5"),
 		},
 	},
 	{ }

drivers/clocksource/sh_cmt.c

@@ -40,6 +40,7 @@ struct sh_cmt_priv {
 	struct platform_device *pdev;
 
 	unsigned long flags;
+	unsigned long flags_suspend;
 	unsigned long match_value;
 	unsigned long next_match_value;
 	unsigned long max_match_value;
@@ -667,11 +668,38 @@ static int __devexit sh_cmt_remove(struct platform_device *pdev)
 	return -EBUSY; /* cannot unregister clockevent and clocksource */
 }
 
+static int sh_cmt_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct sh_cmt_priv *p = platform_get_drvdata(pdev);
+
+	/* save flag state and stop CMT channel */
+	p->flags_suspend = p->flags;
+	sh_cmt_stop(p, p->flags);
+	return 0;
+}
+
+static int sh_cmt_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct sh_cmt_priv *p = platform_get_drvdata(pdev);
+
+	/* start CMT channel from saved state */
+	sh_cmt_start(p, p->flags_suspend);
+	return 0;
+}
+
+static struct dev_pm_ops sh_cmt_dev_pm_ops = {
+	.suspend = sh_cmt_suspend,
+	.resume = sh_cmt_resume,
+};
+
 static struct platform_driver sh_cmt_device_driver = {
 	.probe		= sh_cmt_probe,
 	.remove		= __devexit_p(sh_cmt_remove),
 	.driver		= {
 		.name	= "sh_cmt",
+		.pm	= &sh_cmt_dev_pm_ops,
 	}
 };
 

drivers/md/md.c

@@ -359,6 +359,7 @@ static mddev_t * mddev_find(dev_t unit)
 	else
 		new->md_minor = MINOR(unit) >> MdpMinorShift;
 
+	mutex_init(&new->open_mutex);
 	mutex_init(&new->reconfig_mutex);
 	INIT_LIST_HEAD(&new->disks);
 	INIT_LIST_HEAD(&new->all_mddevs);
@@ -1974,17 +1975,14 @@ repeat:
 		/* otherwise we have to go forward and ... */
 		mddev->events ++;
 		if (!mddev->in_sync || mddev->recovery_cp != MaxSector) { /* not clean */
-			/* .. if the array isn't clean, insist on an odd 'events' */
-			if ((mddev->events&1)==0) {
-				mddev->events++;
+			/* .. if the array isn't clean, an 'even' event must also go
+			 * to spares. */
+			if ((mddev->events&1)==0)
 				nospares = 0;
-			}
 		} else {
-			/* otherwise insist on an even 'events' (for clean states) */
-			if ((mddev->events&1)) {
-				mddev->events++;
+			/* otherwise an 'odd' event must go to spares */
+			if ((mddev->events&1))
 				nospares = 0;
-			}
 		}
 	}
 
@@ -3601,6 +3599,7 @@ max_sync_store(mddev_t *mddev, const char *buf, size_t len)
 		if (max < mddev->resync_min)
 			return -EINVAL;
 		if (max < mddev->resync_max &&
+		    mddev->ro == 0 &&
 		    test_bit(MD_RECOVERY_RUNNING, &mddev->recovery))
 			return -EBUSY;
 
@@ -4304,12 +4303,11 @@ static int do_md_stop(mddev_t * mddev, int mode, int is_open)
 	struct gendisk *disk = mddev->gendisk;
 	mdk_rdev_t *rdev;
 
+	mutex_lock(&mddev->open_mutex);
 	if (atomic_read(&mddev->openers) > is_open) {
 		printk("md: %s still in use.\n",mdname(mddev));
-		return -EBUSY;
-	}
-
-	if (mddev->pers) {
+		err = -EBUSY;
+	} else if (mddev->pers) {
 
 		if (mddev->sync_thread) {
 			set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
@@ -4367,7 +4365,10 @@ static int do_md_stop(mddev_t * mddev, int mode, int is_open)
 			set_disk_ro(disk, 1);
 		clear_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
 	}
-
+out:
+	mutex_unlock(&mddev->open_mutex);
+	if (err)
+		return err;
 	/*
 	 * Free resources if final stop
 	 */
@@ -4433,7 +4434,6 @@ static int do_md_stop(mddev_t * mddev, int mode, int is_open)
 	blk_integrity_unregister(disk);
 	md_new_event(mddev);
 	sysfs_notify_dirent(mddev->sysfs_state);
-out:
 	return err;
 }
 
@@ -5518,12 +5518,12 @@ static int md_open(struct block_device *bdev, fmode_t mode)
 	}
 	BUG_ON(mddev != bdev->bd_disk->private_data);
 
-	if ((err = mutex_lock_interruptible_nested(&mddev->reconfig_mutex, 1)))
+	if ((err = mutex_lock_interruptible(&mddev->open_mutex)))
 		goto out;
 
 	err = 0;
 	atomic_inc(&mddev->openers);
-	mddev_unlock(mddev);
+	mutex_unlock(&mddev->open_mutex);
 
 	check_disk_change(bdev);
  out:

drivers/md/md.h

@@ -223,6 +223,16 @@ struct mddev_s
 							    * so we don't loop trying */
 
 	int				in_sync;	/* know to not need resync */
+	/* 'open_mutex' avoids races between 'md_open' and 'do_md_stop', so
+	 * that we are never stopping an array while it is open.
+	 * 'reconfig_mutex' protects all other reconfiguration.
+	 * These locks are separate due to conflicting interactions
+	 * with bdev->bd_mutex.
+	 * Lock ordering is:
+	 *  reconfig_mutex -> bd_mutex : e.g. do_md_run -> revalidate_disk
+	 *  bd_mutex -> open_mutex:  e.g. __blkdev_get -> md_open
+	 */
+	struct mutex			open_mutex;
 	struct mutex			reconfig_mutex;
 	atomic_t			active;		/* general refcount */
 	atomic_t			openers;	/* number of active opens */

drivers/md/raid5.c

@@ -3785,7 +3785,7 @@ static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped
 		    conf->reshape_progress < raid5_size(mddev, 0, 0)) {
 			sector_nr = raid5_size(mddev, 0, 0)
 				- conf->reshape_progress;
-		} else if (mddev->delta_disks > 0 &&
+		} else if (mddev->delta_disks >= 0 &&
 			   conf->reshape_progress > 0)
 			sector_nr = conf->reshape_progress;
 		sector_div(sector_nr, new_data_disks);
@@ -4509,7 +4509,26 @@ static int run(mddev_t *mddev)
 			   (old_disks-max_degraded));
 		/* here_old is the first stripe that we might need to read
 		 * from */
-		if (here_new >= here_old) {
+		if (mddev->delta_disks == 0) {
+			/* We cannot be sure it is safe to start an in-place
+			 * reshape.  It is only safe if user-space if monitoring
+			 * and taking constant backups.
+			 * mdadm always starts a situation like this in
+			 * readonly mode so it can take control before
+			 * allowing any writes.  So just check for that.
+			 */
+			if ((here_new * mddev->new_chunk_sectors != 
+			     here_old * mddev->chunk_sectors) ||
+			    mddev->ro == 0) {
+				printk(KERN_ERR "raid5: in-place reshape must be started"
+				       " in read-only mode - aborting\n");
+				return -EINVAL;
+			}
+		} else if (mddev->delta_disks < 0
+		    ? (here_new * mddev->new_chunk_sectors <=
+		       here_old * mddev->chunk_sectors)
+		    : (here_new * mddev->new_chunk_sectors >=
+		       here_old * mddev->chunk_sectors)) {
 			/* Reading from the same stripe as writing to - bad */
 			printk(KERN_ERR "raid5: reshape_position too early for "
 			       "auto-recovery - aborting.\n");
@@ -5078,8 +5097,15 @@ static void raid5_finish_reshape(mddev_t *mddev)
 					mddev->degraded--;
 			for (d = conf->raid_disks ;
 			     d < conf->raid_disks - mddev->delta_disks;
-			     d++)
-				raid5_remove_disk(mddev, d);
+			     d++) {
+				mdk_rdev_t *rdev = conf->disks[d].rdev;
+				if (rdev && raid5_remove_disk(mddev, d) == 0) {
+					char nm[20];
+					sprintf(nm, "rd%d", rdev->raid_disk);
+					sysfs_remove_link(&mddev->kobj, nm);
+					rdev->raid_disk = -1;
+				}
+			}
 		}
 		mddev->layout = conf->algorithm;
 		mddev->chunk_sectors = conf->chunk_sectors;

drivers/video/sh_mobile_lcdcfb.c

@@ -477,6 +477,9 @@ static int sh_mobile_lcdc_start(struct sh_mobile_lcdc_priv *priv)
 	/* tell the board code to enable the panel */
 	for (k = 0; k < ARRAY_SIZE(priv->ch); k++) {
 		ch = &priv->ch[k];
+		if (!ch->enabled)
+			continue;
+
 		board_cfg = &ch->cfg.board_cfg;
 		if (board_cfg->display_on)
 			board_cfg->display_on(board_cfg->board_data);
@@ -494,6 +497,8 @@ static void sh_mobile_lcdc_stop(struct sh_mobile_lcdc_priv *priv)
 	/* clean up deferred io and ask board code to disable panel */
 	for (k = 0; k < ARRAY_SIZE(priv->ch); k++) {
 		ch = &priv->ch[k];
+		if (!ch->enabled)
+			continue;
 
 		/* deferred io mode:
 		 * flush frame, and wait for frame end interrupt

fs/ocfs2/alloc.c

@@ -1914,7 +1914,8 @@ static void ocfs2_adjust_adjacent_records(struct ocfs2_extent_rec *left_rec,
 	 * immediately to their right.
 	 */
 	left_clusters = le32_to_cpu(right_child_el->l_recs[0].e_cpos);
-	if (ocfs2_is_empty_extent(&right_child_el->l_recs[0])) {
+	if (!ocfs2_rec_clusters(right_child_el, &right_child_el->l_recs[0])) {
+		BUG_ON(right_child_el->l_tree_depth);
 		BUG_ON(le16_to_cpu(right_child_el->l_next_free_rec) <= 1);
 		left_clusters = le32_to_cpu(right_child_el->l_recs[1].e_cpos);
 	}
@@ -2476,15 +2477,37 @@ out_ret_path:
 	return ret;
 }
 
-static void ocfs2_update_edge_lengths(struct inode *inode, handle_t *handle,
-				      struct ocfs2_path *path)
+static int ocfs2_update_edge_lengths(struct inode *inode, handle_t *handle,
+				     int subtree_index, struct ocfs2_path *path)
 {
-	int i, idx;
+	int i, idx, ret;
 	struct ocfs2_extent_rec *rec;
 	struct ocfs2_extent_list *el;
 	struct ocfs2_extent_block *eb;
 	u32 range;
 
+	/*
+	 * In normal tree rotation process, we will never touch the
+	 * tree branch above subtree_index and ocfs2_extend_rotate_transaction
+	 * doesn't reserve the credits for them either.
+	 *
+	 * But we do have a special case here which will update the rightmost
+	 * records for all the bh in the path.
+	 * So we have to allocate extra credits and access them.
+	 */
+	ret = ocfs2_extend_trans(handle,
+				 handle->h_buffer_credits + subtree_index);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
+	}
+
+	ret = ocfs2_journal_access_path(inode, handle, path);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
+	}
+
 	/* Path should always be rightmost. */
 	eb = (struct ocfs2_extent_block *)path_leaf_bh(path)->b_data;
 	BUG_ON(eb->h_next_leaf_blk != 0ULL);
@@ -2505,6 +2528,8 @@ static void ocfs2_update_edge_lengths(struct inode *inode, handle_t *handle,
 
 		ocfs2_journal_dirty(handle, path->p_node[i].bh);
 	}
+out:
+	return ret;
 }
 
 static void ocfs2_unlink_path(struct inode *inode, handle_t *handle,
@@ -2717,7 +2742,12 @@ static int ocfs2_rotate_subtree_left(struct inode *inode, handle_t *handle,
 	if (del_right_subtree) {
 		ocfs2_unlink_subtree(inode, handle, left_path, right_path,
 				     subtree_index, dealloc);
-		ocfs2_update_edge_lengths(inode, handle, left_path);
+		ret = ocfs2_update_edge_lengths(inode, handle, subtree_index,
+						left_path);
+		if (ret) {
+			mlog_errno(ret);
+			goto out;
+		}
 
 		eb = (struct ocfs2_extent_block *)path_leaf_bh(left_path)->b_data;
 		ocfs2_et_set_last_eb_blk(et, le64_to_cpu(eb->h_blkno));
@@ -3034,7 +3064,12 @@ static int ocfs2_remove_rightmost_path(struct inode *inode, handle_t *handle,
 
 		ocfs2_unlink_subtree(inode, handle, left_path, path,
 				     subtree_index, dealloc);
-		ocfs2_update_edge_lengths(inode, handle, left_path);
+		ret = ocfs2_update_edge_lengths(inode, handle, subtree_index,
+						left_path);
+		if (ret) {
+			mlog_errno(ret);
+			goto out;
+		}
 
 		eb = (struct ocfs2_extent_block *)path_leaf_bh(left_path)->b_data;
 		ocfs2_et_set_last_eb_blk(et, le64_to_cpu(eb->h_blkno));

fs/ocfs2/aops.c

@@ -193,6 +193,7 @@ static int ocfs2_get_block(struct inode *inode, sector_t iblock,
 			     (unsigned long long)OCFS2_I(inode)->ip_blkno);
 			mlog(ML_ERROR, "Size %llu, clusters %u\n", (unsigned long long)i_size_read(inode), OCFS2_I(inode)->ip_clusters);
 			dump_stack();
+			goto bail;
 		}
 
 		past_eof = ocfs2_blocks_for_bytes(inode->i_sb, i_size_read(inode));
@@ -894,18 +895,17 @@ struct ocfs2_write_cluster_desc {
 	 */
 	unsigned	c_new;
 	unsigned	c_unwritten;
+	unsigned	c_needs_zero;
 };
 
-static inline int ocfs2_should_zero_cluster(struct ocfs2_write_cluster_desc *d)
-{
-	return d->c_new || d->c_unwritten;
-}
-
 struct ocfs2_write_ctxt {
 	/* Logical cluster position / len of write */
 	u32				w_cpos;
 	u32				w_clen;
 
+	/* First cluster allocated in a nonsparse extend */
+	u32				w_first_new_cpos;
+
 	struct ocfs2_write_cluster_desc	w_desc[OCFS2_MAX_CLUSTERS_PER_PAGE];
 
 	/*
@@ -983,6 +983,7 @@ static int ocfs2_alloc_write_ctxt(struct ocfs2_write_ctxt **wcp,
 		return -ENOMEM;
 
 	wc->w_cpos = pos >> osb->s_clustersize_bits;
+	wc->w_first_new_cpos = UINT_MAX;
 	cend = (pos + len - 1) >> osb->s_clustersize_bits;
 	wc->w_clen = cend - wc->w_cpos + 1;
 	get_bh(di_bh);
@@ -1217,20 +1218,18 @@ out:
  */
 static int ocfs2_write_cluster(struct address_space *mapping,
 			       u32 phys, unsigned int unwritten,
+			       unsigned int should_zero,
 			       struct ocfs2_alloc_context *data_ac,
 			       struct ocfs2_alloc_context *meta_ac,
 			       struct ocfs2_write_ctxt *wc, u32 cpos,
 			       loff_t user_pos, unsigned user_len)
 {
-	int ret, i, new, should_zero = 0;
+	int ret, i, new;
 	u64 v_blkno, p_blkno;
 	struct inode *inode = mapping->host;
 	struct ocfs2_extent_tree et;
 
 	new = phys == 0 ? 1 : 0;
-	if (new || unwritten)
-		should_zero = 1;
-
 	if (new) {
 		u32 tmp_pos;
 
@@ -1301,7 +1300,7 @@ static int ocfs2_write_cluster(struct address_space *mapping,
 		if (tmpret) {
 			mlog_errno(tmpret);
 			if (ret == 0)
-				tmpret = ret;
+				ret = tmpret;
 		}
 	}
 
@@ -1341,7 +1340,9 @@ static int ocfs2_write_cluster_by_desc(struct address_space *mapping,
 			local_len = osb->s_clustersize - cluster_off;
 
 		ret = ocfs2_write_cluster(mapping, desc->c_phys,
-					  desc->c_unwritten, data_ac, meta_ac,
+					  desc->c_unwritten,
+					  desc->c_needs_zero,
+					  data_ac, meta_ac,
 					  wc, desc->c_cpos, pos, local_len);
 		if (ret) {
 			mlog_errno(ret);
@@ -1391,14 +1392,14 @@ static void ocfs2_set_target_boundaries(struct ocfs2_super *osb,
 		 * newly allocated cluster.
 		 */
 		desc = &wc->w_desc[0];
-		if (ocfs2_should_zero_cluster(desc))
+		if (desc->c_needs_zero)
 			ocfs2_figure_cluster_boundaries(osb,
 							desc->c_cpos,
 							&wc->w_target_from,
 							NULL);
 
 		desc = &wc->w_desc[wc->w_clen - 1];
-		if (ocfs2_should_zero_cluster(desc))
+		if (desc->c_needs_zero)
 			ocfs2_figure_cluster_boundaries(osb,
 							desc->c_cpos,
 							NULL,
@@ -1466,13 +1467,28 @@ static int ocfs2_populate_write_desc(struct inode *inode,
 			phys++;
 		}
 
+		/*
+		 * If w_first_new_cpos is < UINT_MAX, we have a non-sparse
+		 * file that got extended.  w_first_new_cpos tells us
+		 * where the newly allocated clusters are so we can
+		 * zero them.
+		 */
+		if (desc->c_cpos >= wc->w_first_new_cpos) {
+			BUG_ON(phys == 0);
+			desc->c_needs_zero = 1;
+		}
+
 		desc->c_phys = phys;
 		if (phys == 0) {
 			desc->c_new = 1;
+			desc->c_needs_zero = 1;
 			*clusters_to_alloc = *clusters_to_alloc + 1;
 		}
-		if (ext_flags & OCFS2_EXT_UNWRITTEN)
+
+		if (ext_flags & OCFS2_EXT_UNWRITTEN) {
 			desc->c_unwritten = 1;
+			desc->c_needs_zero = 1;
+		}
 
 		num_clusters--;
 	}
@@ -1632,10 +1648,13 @@ static int ocfs2_expand_nonsparse_inode(struct inode *inode, loff_t pos,
 	if (newsize <= i_size_read(inode))
 		return 0;
 
-	ret = ocfs2_extend_no_holes(inode, newsize, newsize - len);
+	ret = ocfs2_extend_no_holes(inode, newsize, pos);
 	if (ret)
 		mlog_errno(ret);
 
+	wc->w_first_new_cpos =
+		ocfs2_clusters_for_bytes(inode->i_sb, i_size_read(inode));
+
 	return ret;
 }
 
@@ -1644,7 +1663,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
 			     struct page **pagep, void **fsdata,
 			     struct buffer_head *di_bh, struct page *mmap_page)
 {
-	int ret, credits = OCFS2_INODE_UPDATE_CREDITS;
+	int ret, cluster_of_pages, credits = OCFS2_INODE_UPDATE_CREDITS;
 	unsigned int clusters_to_alloc, extents_to_split;
 	struct ocfs2_write_ctxt *wc;
 	struct inode *inode = mapping->host;
@@ -1722,8 +1741,19 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
 
 	}
 
-	ocfs2_set_target_boundaries(osb, wc, pos, len,
-				    clusters_to_alloc + extents_to_split);
+	/*
+	 * We have to zero sparse allocated clusters, unwritten extent clusters,
+	 * and non-sparse clusters we just extended.  For non-sparse writes,
+	 * we know zeros will only be needed in the first and/or last cluster.
+	 */
+	if (clusters_to_alloc || extents_to_split ||
+	    wc->w_desc[0].c_needs_zero ||
+	    wc->w_desc[wc->w_clen - 1].c_needs_zero)
+		cluster_of_pages = 1;
+	else
+		cluster_of_pages = 0;
+
+	ocfs2_set_target_boundaries(osb, wc, pos, len, cluster_of_pages);
 
 	handle = ocfs2_start_trans(osb, credits);
 	if (IS_ERR(handle)) {
@@ -1756,8 +1786,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
 	 * extent.
 	 */
 	ret = ocfs2_grab_pages_for_write(mapping, wc, wc->w_cpos, pos,
-					 clusters_to_alloc + extents_to_split,
-					 mmap_page);
+					 cluster_of_pages, mmap_page);
 	if (ret) {
 		mlog_errno(ret);
 		goto out_quota;

fs/ocfs2/dcache.c

@@ -310,22 +310,19 @@ out_attach:
 	return ret;
 }
 
-static DEFINE_SPINLOCK(dentry_list_lock);
+DEFINE_SPINLOCK(dentry_list_lock);
 
 /* We limit the number of dentry locks to drop in one go. We have
  * this limit so that we don't starve other users of ocfs2_wq. */
 #define DL_INODE_DROP_COUNT 64
 
 /* Drop inode references from dentry locks */
-void ocfs2_drop_dl_inodes(struct work_struct *work)
+static void __ocfs2_drop_dl_inodes(struct ocfs2_super *osb, int drop_count)
 {
-	struct ocfs2_super *osb = container_of(work, struct ocfs2_super,
-					       dentry_lock_work);
 	struct ocfs2_dentry_lock *dl;
-	int drop_count = DL_INODE_DROP_COUNT;
 
 	spin_lock(&dentry_list_lock);
-	while (osb->dentry_lock_list && drop_count--) {
+	while (osb->dentry_lock_list && (drop_count < 0 || drop_count--)) {
 		dl = osb->dentry_lock_list;
 		osb->dentry_lock_list = dl->dl_next;
 		spin_unlock(&dentry_list_lock);
@@ -333,11 +330,32 @@ void ocfs2_drop_dl_inodes(struct work_struct *work)
 		kfree(dl);
 		spin_lock(&dentry_list_lock);
 	}
-	if (osb->dentry_lock_list)
+	spin_unlock(&dentry_list_lock);
+}
+
+void ocfs2_drop_dl_inodes(struct work_struct *work)
+{
+	struct ocfs2_super *osb = container_of(work, struct ocfs2_super,
+					       dentry_lock_work);
+
+	__ocfs2_drop_dl_inodes(osb, DL_INODE_DROP_COUNT);
+	/*
+	 * Don't queue dropping if umount is in progress. We flush the
+	 * list in ocfs2_dismount_volume
+	 */
+	spin_lock(&dentry_list_lock);
+	if (osb->dentry_lock_list &&
+	    !ocfs2_test_osb_flag(osb, OCFS2_OSB_DROP_DENTRY_LOCK_IMMED))
 		queue_work(ocfs2_wq, &osb->dentry_lock_work);
 	spin_unlock(&dentry_list_lock);
 }
 
+/* Flush the whole work queue */
+void ocfs2_drop_all_dl_inodes(struct ocfs2_super *osb)
+{
+	__ocfs2_drop_dl_inodes(osb, -1);
+}
+
 /*
  * ocfs2_dentry_iput() and friends.
  *
@@ -368,7 +386,8 @@ static void ocfs2_drop_dentry_lock(struct ocfs2_super *osb,
 	/* We leave dropping of inode reference to ocfs2_wq as that can
 	 * possibly lead to inode deletion which gets tricky */
 	spin_lock(&dentry_list_lock);
-	if (!osb->dentry_lock_list)
+	if (!osb->dentry_lock_list &&
+	    !ocfs2_test_osb_flag(osb, OCFS2_OSB_DROP_DENTRY_LOCK_IMMED))
 		queue_work(ocfs2_wq, &osb->dentry_lock_work);
 	dl->dl_next = osb->dentry_lock_list;
 	osb->dentry_lock_list = dl;

fs/ocfs2/dcache.h

@@ -49,10 +49,13 @@ struct ocfs2_dentry_lock {
 int ocfs2_dentry_attach_lock(struct dentry *dentry, struct inode *inode,
 			     u64 parent_blkno);
 
+extern spinlock_t dentry_list_lock;
+
 void ocfs2_dentry_lock_put(struct ocfs2_super *osb,
 			   struct ocfs2_dentry_lock *dl);
 
 void ocfs2_drop_dl_inodes(struct work_struct *work);
+void ocfs2_drop_all_dl_inodes(struct ocfs2_super *osb);
 
 struct dentry *ocfs2_find_local_alias(struct inode *inode, u64 parent_blkno,
 				      int skip_unhashed);

fs/ocfs2/dlm/dlmast.c

@@ -103,7 +103,6 @@ static void __dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 		     lock->ast_pending, lock->ml.type);
 		BUG();
 	}
-	BUG_ON(!list_empty(&lock->ast_list));
 	if (lock->ast_pending)
 		mlog(0, "lock has an ast getting flushed right now\n");
 

fs/ocfs2/dlm/dlmrecovery.c

@@ -1118,7 +1118,7 @@ static int dlm_send_mig_lockres_msg(struct dlm_ctxt *dlm,
 
 	mlog(0, "%s:%.*s: sending mig lockres (%s) to %u\n",
 	     dlm->name, res->lockname.len, res->lockname.name,
-	     orig_flags & DLM_MRES_MIGRATION ? "migrate" : "recovery",
+	     orig_flags & DLM_MRES_MIGRATION ? "migration" : "recovery",
 	     send_to);
 
 	/* send it */

fs/ocfs2/file.c

@@ -1851,6 +1851,7 @@ relock:
 		if (ret)
 			goto out_dio;
 
+		count = ocount;
 		ret = generic_write_checks(file, ppos, &count,
 					   S_ISBLK(inode->i_mode));
 		if (ret)
@@ -1918,8 +1919,10 @@ out_sems:
 
 	mutex_unlock(&inode->i_mutex);
 
+	if (written)
+		ret = written;
 	mlog_exit(ret);
-	return written ? written : ret;
+	return ret;
 }
 
 static int ocfs2_splice_to_file(struct pipe_inode_info *pipe,

fs/ocfs2/journal.c

@@ -1954,10 +1954,16 @@ void ocfs2_orphan_scan_init(struct ocfs2_super *osb)
 	os->os_osb = osb;
 	os->os_count = 0;
 	os->os_seqno = 0;
-	os->os_scantime = CURRENT_TIME;
 	mutex_init(&os->os_lock);
 	INIT_DELAYED_WORK(&os->os_orphan_scan_work, ocfs2_orphan_scan_work);
+}
 
+void ocfs2_orphan_scan_start(struct ocfs2_super *osb)
+{
+	struct ocfs2_orphan_scan *os;
+
+	os = &osb->osb_orphan_scan;
+	os->os_scantime = CURRENT_TIME;
 	if (ocfs2_is_hard_readonly(osb) || ocfs2_mount_local(osb))
 		atomic_set(&os->os_state, ORPHAN_SCAN_INACTIVE);
 	else {

fs/ocfs2/journal.h

@@ -145,6 +145,7 @@ static inline void ocfs2_inode_set_new(struct ocfs2_super *osb,
 
 /* Exported only for the journal struct init code in super.c. Do not call. */
 void ocfs2_orphan_scan_init(struct ocfs2_super *osb);
+void ocfs2_orphan_scan_start(struct ocfs2_super *osb);
 void ocfs2_orphan_scan_stop(struct ocfs2_super *osb);
 void ocfs2_orphan_scan_exit(struct ocfs2_super *osb);
 
@@ -329,20 +330,27 @@ int                  ocfs2_journal_dirty(handle_t *handle,
 /* extended attribute block update */
 #define OCFS2_XATTR_BLOCK_UPDATE_CREDITS 1
 
+/* Update of a single quota block */
+#define OCFS2_QUOTA_BLOCK_UPDATE_CREDITS 1
+
 /* global quotafile inode update, data block */
-#define OCFS2_QINFO_WRITE_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 1)
+#define OCFS2_QINFO_WRITE_CREDITS (OCFS2_INODE_UPDATE_CREDITS + \
+				   OCFS2_QUOTA_BLOCK_UPDATE_CREDITS)
 
+#define OCFS2_LOCAL_QINFO_WRITE_CREDITS OCFS2_QUOTA_BLOCK_UPDATE_CREDITS
 /*
  * The two writes below can accidentally see global info dirty due
  * to set_info() quotactl so make them prepared for the writes.
  */
 /* quota data block, global info */
 /* Write to local quota file */
-#define OCFS2_QWRITE_CREDITS (OCFS2_QINFO_WRITE_CREDITS + 1)
+#define OCFS2_QWRITE_CREDITS (OCFS2_QINFO_WRITE_CREDITS + \
+			      OCFS2_QUOTA_BLOCK_UPDATE_CREDITS)
 
 /* global quota data block, local quota data block, global quota inode,
  * global quota info */
-#define OCFS2_QSYNC_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 3)
+#define OCFS2_QSYNC_CREDITS (OCFS2_QINFO_WRITE_CREDITS + \
+			     2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS)
 
 static inline int ocfs2_quota_trans_credits(struct super_block *sb)
 {
@@ -355,11 +363,6 @@ static inline int ocfs2_quota_trans_credits(struct super_block *sb)
 	return credits;
 }
 
-/* Number of credits needed for removing quota structure from file */
-int ocfs2_calc_qdel_credits(struct super_block *sb, int type);
-/* Number of credits needed for initialization of new quota structure */
-int ocfs2_calc_qinit_credits(struct super_block *sb, int type);
-
 /* group extend. inode update and last group update. */
 #define OCFS2_GROUP_EXTEND_CREDITS	(OCFS2_INODE_UPDATE_CREDITS + 1)
 

fs/ocfs2/ocfs2.h

@@ -224,10 +224,12 @@ enum ocfs2_mount_options
 	OCFS2_MOUNT_GRPQUOTA = 1 << 10, /* We support group quotas */
 };
 
-#define OCFS2_OSB_SOFT_RO	0x0001
-#define OCFS2_OSB_HARD_RO	0x0002
-#define OCFS2_OSB_ERROR_FS	0x0004
-#define OCFS2_DEFAULT_ATIME_QUANTUM	60
+#define OCFS2_OSB_SOFT_RO			0x0001
+#define OCFS2_OSB_HARD_RO			0x0002
+#define OCFS2_OSB_ERROR_FS			0x0004
+#define OCFS2_OSB_DROP_DENTRY_LOCK_IMMED	0x0008
+
+#define OCFS2_DEFAULT_ATIME_QUANTUM		60
 
 struct ocfs2_journal;
 struct ocfs2_slot_info;
@@ -490,6 +492,18 @@ static inline void ocfs2_set_osb_flag(struct ocfs2_super *osb,
 	spin_unlock(&osb->osb_lock);
 }
 
+
+static inline unsigned long  ocfs2_test_osb_flag(struct ocfs2_super *osb,
+						 unsigned long flag)
+{
+	unsigned long ret;
+
+	spin_lock(&osb->osb_lock);
+	ret = osb->osb_flags & flag;
+	spin_unlock(&osb->osb_lock);
+	return ret;
+}
+
 static inline void ocfs2_set_ro_flag(struct ocfs2_super *osb,
 				     int hard)
 {

fs/ocfs2/quota.h

@@ -50,7 +50,6 @@ struct ocfs2_mem_dqinfo {
 	unsigned int dqi_chunks;	/* Number of chunks in local quota file */
 	unsigned int dqi_blocks;	/* Number of blocks allocated for local quota file */
 	unsigned int dqi_syncms;	/* How often should we sync with other nodes */
-	unsigned int dqi_syncjiff;	/* Precomputed dqi_syncms in jiffies */
 	struct list_head dqi_chunk;	/* List of chunks */
 	struct inode *dqi_gqinode;	/* Global quota file inode */
 	struct ocfs2_lock_res dqi_gqlock;	/* Lock protecting quota information structure */

fs/ocfs2/quota_global.c

@@ -69,6 +69,7 @@ static void ocfs2_global_mem2diskdqb(void *dp, struct dquot *dquot)
 	d->dqb_curspace = cpu_to_le64(m->dqb_curspace);
 	d->dqb_btime = cpu_to_le64(m->dqb_btime);
 	d->dqb_itime = cpu_to_le64(m->dqb_itime);
+	d->dqb_pad1 = d->dqb_pad2 = 0;
 }
 
 static int ocfs2_global_is_id(void *dp, struct dquot *dquot)
@@ -211,14 +212,13 @@ ssize_t ocfs2_quota_write(struct super_block *sb, int type,
 
 	mutex_lock_nested(&gqinode->i_mutex, I_MUTEX_QUOTA);
 	if (gqinode->i_size < off + len) {
-		down_write(&OCFS2_I(gqinode)->ip_alloc_sem);
-		err = ocfs2_extend_no_holes(gqinode, off + len, off);
-		up_write(&OCFS2_I(gqinode)->ip_alloc_sem);
-		if (err < 0)
-			goto out;
+		loff_t rounded_end =
+				ocfs2_align_bytes_to_blocks(sb, off + len);
+
+		/* Space is already allocated in ocfs2_global_read_dquot() */
 		err = ocfs2_simple_size_update(gqinode,
 					       oinfo->dqi_gqi_bh,
-					       off + len);
+					       rounded_end);
 		if (err < 0)
 			goto out;
 		new = 1;
@@ -234,7 +234,7 @@ ssize_t ocfs2_quota_write(struct super_block *sb, int type,
 	}
 	if (err) {
 		mlog_errno(err);
-		return err;
+		goto out;
 	}
 	lock_buffer(bh);
 	if (new)
@@ -342,7 +342,6 @@ int ocfs2_global_read_info(struct super_block *sb, int type)
 	info->dqi_bgrace = le32_to_cpu(dinfo.dqi_bgrace);
 	info->dqi_igrace = le32_to_cpu(dinfo.dqi_igrace);
 	oinfo->dqi_syncms = le32_to_cpu(dinfo.dqi_syncms);
-	oinfo->dqi_syncjiff = msecs_to_jiffies(oinfo->dqi_syncms);
 	oinfo->dqi_gi.dqi_blocks = le32_to_cpu(dinfo.dqi_blocks);
 	oinfo->dqi_gi.dqi_free_blk = le32_to_cpu(dinfo.dqi_free_blk);
 	oinfo->dqi_gi.dqi_free_entry = le32_to_cpu(dinfo.dqi_free_entry);
@@ -352,7 +351,7 @@ int ocfs2_global_read_info(struct super_block *sb, int type)
 	oinfo->dqi_gi.dqi_qtree_depth = qtree_depth(&oinfo->dqi_gi);
 	INIT_DELAYED_WORK(&oinfo->dqi_sync_work, qsync_work_fn);
 	queue_delayed_work(ocfs2_quota_wq, &oinfo->dqi_sync_work,
-			   oinfo->dqi_syncjiff);
+			   msecs_to_jiffies(oinfo->dqi_syncms));
 
 out_err:
 	mlog_exit(status);
@@ -402,13 +401,36 @@ int ocfs2_global_write_info(struct super_block *sb, int type)
 	return err;
 }
 
+static int ocfs2_global_qinit_alloc(struct super_block *sb, int type)
+{
+	struct ocfs2_mem_dqinfo *oinfo = sb_dqinfo(sb, type)->dqi_priv;
+
+	/*
+	 * We may need to allocate tree blocks and a leaf block but not the
+	 * root block
+	 */
+	return oinfo->dqi_gi.dqi_qtree_depth;
+}
+
+static int ocfs2_calc_global_qinit_credits(struct super_block *sb, int type)
+{
+	/* We modify all the allocated blocks, tree root, and info block */
+	return (ocfs2_global_qinit_alloc(sb, type) + 2) *
+			OCFS2_QUOTA_BLOCK_UPDATE_CREDITS;
+}
+
 /* Read in information from global quota file and acquire a reference to it.
  * dquot_acquire() has already started the transaction and locked quota file */
 int ocfs2_global_read_dquot(struct dquot *dquot)
 {
 	int err, err2, ex = 0;
-	struct ocfs2_mem_dqinfo *info =
-			sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv;
+	struct super_block *sb = dquot->dq_sb;
+	int type = dquot->dq_type;
+	struct ocfs2_mem_dqinfo *info = sb_dqinfo(sb, type)->dqi_priv;
+	struct ocfs2_super *osb = OCFS2_SB(sb);
+	struct inode *gqinode = info->dqi_gqinode;
+	int need_alloc = ocfs2_global_qinit_alloc(sb, type);
+	handle_t *handle = NULL;
 
 	err = ocfs2_qinfo_lock(info, 0);
 	if (err < 0)
@@ -419,14 +441,33 @@ int ocfs2_global_read_dquot(struct dquot *dquot)
 	OCFS2_DQUOT(dquot)->dq_use_count++;
 	OCFS2_DQUOT(dquot)->dq_origspace = dquot->dq_dqb.dqb_curspace;
 	OCFS2_DQUOT(dquot)->dq_originodes = dquot->dq_dqb.dqb_curinodes;
+	ocfs2_qinfo_unlock(info, 0);
+
 	if (!dquot->dq_off) {	/* No real quota entry? */
-		/* Upgrade to exclusive lock for allocation */
-		ocfs2_qinfo_unlock(info, 0);
-		err = ocfs2_qinfo_lock(info, 1);
-		if (err < 0)
-			goto out_qlock;
 		ex = 1;
+		/*
+		 * Add blocks to quota file before we start a transaction since
+		 * locking allocators ranks above a transaction start
+		 */
+		WARN_ON(journal_current_handle());
+		down_write(&OCFS2_I(gqinode)->ip_alloc_sem);
+		err = ocfs2_extend_no_holes(gqinode,
+			gqinode->i_size + (need_alloc << sb->s_blocksize_bits),
+			gqinode->i_size);
+		up_write(&OCFS2_I(gqinode)->ip_alloc_sem);
+		if (err < 0)
+			goto out;
 	}
+
+	handle = ocfs2_start_trans(osb,
+				   ocfs2_calc_global_qinit_credits(sb, type));
+	if (IS_ERR(handle)) {
+		err = PTR_ERR(handle);
+		goto out;
+	}
+	err = ocfs2_qinfo_lock(info, ex);
+	if (err < 0)
+		goto out_trans;
 	err = qtree_write_dquot(&info->dqi_gi, dquot);
 	if (ex && info_dirty(sb_dqinfo(dquot->dq_sb, dquot->dq_type))) {
 		err2 = __ocfs2_global_write_info(dquot->dq_sb, dquot->dq_type);
@@ -438,6 +479,9 @@ out_qlock:
 		ocfs2_qinfo_unlock(info, 1);
 	else
 		ocfs2_qinfo_unlock(info, 0);
+out_trans:
+	if (handle)
+		ocfs2_commit_trans(osb, handle);
 out:
 	if (err < 0)
 		mlog_errno(err);
@@ -607,7 +651,7 @@ static void qsync_work_fn(struct work_struct *work)
 
 	dquot_scan_active(sb, ocfs2_sync_dquot_helper, oinfo->dqi_type);
 	queue_delayed_work(ocfs2_quota_wq, &oinfo->dqi_sync_work,
-			   oinfo->dqi_syncjiff);
+			   msecs_to_jiffies(oinfo->dqi_syncms));
 }
 
 /*
@@ -635,20 +679,18 @@ out:
 	return status;
 }
 
-int ocfs2_calc_qdel_credits(struct super_block *sb, int type)
+static int ocfs2_calc_qdel_credits(struct super_block *sb, int type)
 {
-	struct ocfs2_mem_dqinfo *oinfo;
-	int features[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA,
-				    OCFS2_FEATURE_RO_COMPAT_GRPQUOTA };
-
-	if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, features[type]))
-		return 0;
-
-	oinfo = sb_dqinfo(sb, type)->dqi_priv;
-	/* We modify tree, leaf block, global info, local chunk header,
-	 * global and local inode */
-	return oinfo->dqi_gi.dqi_qtree_depth + 2 + 1 +
-	       2 * OCFS2_INODE_UPDATE_CREDITS;
+	struct ocfs2_mem_dqinfo *oinfo = sb_dqinfo(sb, type)->dqi_priv;
+	/*
+	 * We modify tree, leaf block, global info, local chunk header,
+	 * global and local inode; OCFS2_QINFO_WRITE_CREDITS already
+	 * accounts for inode update
+	 */
+	return (oinfo->dqi_gi.dqi_qtree_depth + 2) *
+	       OCFS2_QUOTA_BLOCK_UPDATE_CREDITS +
+	       OCFS2_QINFO_WRITE_CREDITS +
+	       OCFS2_INODE_UPDATE_CREDITS;
 }
 
 static int ocfs2_release_dquot(struct dquot *dquot)
@@ -680,33 +722,10 @@ out:
 	return status;
 }
 
-int ocfs2_calc_qinit_credits(struct super_block *sb, int type)
-{
-	struct ocfs2_mem_dqinfo *oinfo;
-	int features[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA,
-				    OCFS2_FEATURE_RO_COMPAT_GRPQUOTA };
-	struct ocfs2_dinode *lfe, *gfe;
-
-	if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, features[type]))
-		return 0;
-
-	oinfo = sb_dqinfo(sb, type)->dqi_priv;
-	gfe = (struct ocfs2_dinode *)oinfo->dqi_gqi_bh->b_data;
-	lfe = (struct ocfs2_dinode *)oinfo->dqi_lqi_bh->b_data;
-	/* We can extend local file + global file. In local file we
-	 * can modify info, chunk header block and dquot block. In
-	 * global file we can modify info, tree and leaf block */
-	return ocfs2_calc_extend_credits(sb, &lfe->id2.i_list, 0) +
-	       ocfs2_calc_extend_credits(sb, &gfe->id2.i_list, 0) +
-	       3 + oinfo->dqi_gi.dqi_qtree_depth + 2;
-}
-
 static int ocfs2_acquire_dquot(struct dquot *dquot)
 {
-	handle_t *handle;
 	struct ocfs2_mem_dqinfo *oinfo =
 			sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv;
-	struct ocfs2_super *osb = OCFS2_SB(dquot->dq_sb);
 	int status = 0;
 
 	mlog_entry("id=%u, type=%d", dquot->dq_id, dquot->dq_type);
@@ -715,16 +734,7 @@ static int ocfs2_acquire_dquot(struct dquot *dquot)
 	status = ocfs2_lock_global_qf(oinfo, 1);
 	if (status < 0)
 		goto out;
-	handle = ocfs2_start_trans(osb,
-		ocfs2_calc_qinit_credits(dquot->dq_sb, dquot->dq_type));
-	if (IS_ERR(handle)) {
-		status = PTR_ERR(handle);
-		mlog_errno(status);
-		goto out_ilock;
-	}
 	status = dquot_acquire(dquot);
-	ocfs2_commit_trans(osb, handle);
-out_ilock:
 	ocfs2_unlock_global_qf(oinfo, 1);
 out:
 	mlog_exit(status);

fs/ocfs2/quota_local.c

@@ -20,6 +20,7 @@
 #include "sysfile.h"
 #include "dlmglue.h"
 #include "quota.h"
+#include "uptodate.h"
 
 /* Number of local quota structures per block */
 static inline unsigned int ol_quota_entries_per_block(struct super_block *sb)
@@ -100,7 +101,8 @@ static int ocfs2_modify_bh(struct inode *inode, struct buffer_head *bh,
 	handle_t *handle;
 	int status;
 
-	handle = ocfs2_start_trans(OCFS2_SB(sb), 1);
+	handle = ocfs2_start_trans(OCFS2_SB(sb),
+				   OCFS2_QUOTA_BLOCK_UPDATE_CREDITS);
 	if (IS_ERR(handle)) {
 		status = PTR_ERR(handle);
 		mlog_errno(status);
@@ -610,7 +612,8 @@ int ocfs2_finish_quota_recovery(struct ocfs2_super *osb,
 			goto out_bh;
 		/* Mark quota file as clean if we are recovering quota file of
 		 * some other node. */
-		handle = ocfs2_start_trans(osb, 1);
+		handle = ocfs2_start_trans(osb,
+					   OCFS2_LOCAL_QINFO_WRITE_CREDITS);
 		if (IS_ERR(handle)) {
 			status = PTR_ERR(handle);
 			mlog_errno(status);
@@ -940,7 +943,7 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk(
 	struct ocfs2_local_disk_chunk *dchunk;
 	int status;
 	handle_t *handle;
-	struct buffer_head *bh = NULL;
+	struct buffer_head *bh = NULL, *dbh = NULL;
 	u64 p_blkno;
 
 	/* We are protected by dqio_sem so no locking needed */
@@ -964,32 +967,35 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk(
 		mlog_errno(status);
 		goto out;
 	}
+	/* Local quota info and two new blocks we initialize */
+	handle = ocfs2_start_trans(OCFS2_SB(sb),
+			OCFS2_LOCAL_QINFO_WRITE_CREDITS +
+			2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS);
+	if (IS_ERR(handle)) {
+		status = PTR_ERR(handle);
+		mlog_errno(status);
+		goto out;
+	}
 
+	/* Initialize chunk header */
 	down_read(&OCFS2_I(lqinode)->ip_alloc_sem);
 	status = ocfs2_extent_map_get_blocks(lqinode, oinfo->dqi_blocks,
 					     &p_blkno, NULL, NULL);
 	up_read(&OCFS2_I(lqinode)->ip_alloc_sem);
 	if (status < 0) {
 		mlog_errno(status);
-		goto out;
+		goto out_trans;
 	}
 	bh = sb_getblk(sb, p_blkno);
 	if (!bh) {
 		status = -ENOMEM;
 		mlog_errno(status);
-		goto out;
+		goto out_trans;
 	}
 	dchunk = (struct ocfs2_local_disk_chunk *)bh->b_data;
-
-	handle = ocfs2_start_trans(OCFS2_SB(sb), 2);
-	if (IS_ERR(handle)) {
-		status = PTR_ERR(handle);
-		mlog_errno(status);
-		goto out;
-	}
-
+	ocfs2_set_new_buffer_uptodate(lqinode, bh);
 	status = ocfs2_journal_access_dq(handle, lqinode, bh,
-					 OCFS2_JOURNAL_ACCESS_WRITE);
+					 OCFS2_JOURNAL_ACCESS_CREATE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto out_trans;
@@ -999,7 +1005,6 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk(
 	memset(dchunk->dqc_bitmap, 0,
 	       sb->s_blocksize - sizeof(struct ocfs2_local_disk_chunk) -
 	       OCFS2_QBLK_RESERVED_SPACE);
-	set_buffer_uptodate(bh);
 	unlock_buffer(bh);
 	status = ocfs2_journal_dirty(handle, bh);
 	if (status < 0) {
@@ -1007,6 +1012,38 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk(
 		goto out_trans;
 	}
 
+	/* Initialize new block with structures */
+	down_read(&OCFS2_I(lqinode)->ip_alloc_sem);
+	status = ocfs2_extent_map_get_blocks(lqinode, oinfo->dqi_blocks + 1,
+					     &p_blkno, NULL, NULL);
+	up_read(&OCFS2_I(lqinode)->ip_alloc_sem);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+	dbh = sb_getblk(sb, p_blkno);
+	if (!dbh) {
+		status = -ENOMEM;
+		mlog_errno(status);
+		goto out_trans;
+	}
+	ocfs2_set_new_buffer_uptodate(lqinode, dbh);
+	status = ocfs2_journal_access_dq(handle, lqinode, dbh,
+					 OCFS2_JOURNAL_ACCESS_CREATE);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+	lock_buffer(dbh);
+	memset(dbh->b_data, 0, sb->s_blocksize - OCFS2_QBLK_RESERVED_SPACE);
+	unlock_buffer(dbh);
+	status = ocfs2_journal_dirty(handle, dbh);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+
+	/* Update local quotafile info */
 	oinfo->dqi_blocks += 2;
 	oinfo->dqi_chunks++;
 	status = ocfs2_local_write_info(sb, type);
@@ -1031,6 +1068,7 @@ out_trans:
 	ocfs2_commit_trans(OCFS2_SB(sb), handle);
 out:
 	brelse(bh);
+	brelse(dbh);
 	kmem_cache_free(ocfs2_qf_chunk_cachep, chunk);
 	return ERR_PTR(status);
 }
@@ -1048,6 +1086,8 @@ static struct ocfs2_quota_chunk *ocfs2_extend_local_quota_file(
 	struct ocfs2_local_disk_chunk *dchunk;
 	int epb = ol_quota_entries_per_block(sb);
 	unsigned int chunk_blocks;
+	struct buffer_head *bh;
+	u64 p_blkno;
 	int status;
 	handle_t *handle;
 
@@ -1075,12 +1115,49 @@ static struct ocfs2_quota_chunk *ocfs2_extend_local_quota_file(
 		mlog_errno(status);
 		goto out;
 	}
-	handle = ocfs2_start_trans(OCFS2_SB(sb), 2);
+
+	/* Get buffer from the just added block */
+	down_read(&OCFS2_I(lqinode)->ip_alloc_sem);
+	status = ocfs2_extent_map_get_blocks(lqinode, oinfo->dqi_blocks,
+					     &p_blkno, NULL, NULL);
+	up_read(&OCFS2_I(lqinode)->ip_alloc_sem);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+	bh = sb_getblk(sb, p_blkno);
+	if (!bh) {
+		status = -ENOMEM;
+		mlog_errno(status);
+		goto out;
+	}
+	ocfs2_set_new_buffer_uptodate(lqinode, bh);
+
+	/* Local quota info, chunk header and the new block we initialize */
+	handle = ocfs2_start_trans(OCFS2_SB(sb),
+			OCFS2_LOCAL_QINFO_WRITE_CREDITS +
+			2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS);
 	if (IS_ERR(handle)) {
 		status = PTR_ERR(handle);
 		mlog_errno(status);
 		goto out;
 	}
+	/* Zero created block */
+	status = ocfs2_journal_access_dq(handle, lqinode, bh,
+				 OCFS2_JOURNAL_ACCESS_CREATE);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+	lock_buffer(bh);
+	memset(bh->b_data, 0, sb->s_blocksize);
+	unlock_buffer(bh);
+	status = ocfs2_journal_dirty(handle, bh);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+	/* Update chunk header */
 	status = ocfs2_journal_access_dq(handle, lqinode, chunk->qc_headerbh,
 				 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
@@ -1097,6 +1174,7 @@ static struct ocfs2_quota_chunk *ocfs2_extend_local_quota_file(
 		mlog_errno(status);
 		goto out_trans;
 	}
+	/* Update file header */
 	oinfo->dqi_blocks++;
 	status = ocfs2_local_write_info(sb, type);
 	if (status < 0) {

fs/ocfs2/stack_o2cb.c

@@ -17,6 +17,7 @@
  * General Public License for more details.
  */
 
+#include <linux/kernel.h>
 #include <linux/crc32.h>
 #include <linux/module.h>
 
@@ -153,7 +154,7 @@ static int status_map[] = {
 
 static int dlm_status_to_errno(enum dlm_status status)
 {
-	BUG_ON(status > (sizeof(status_map) / sizeof(status_map[0])));
+	BUG_ON(status < 0 || status >= ARRAY_SIZE(status_map));
 
 	return status_map[status];
 }

fs/ocfs2/super.c

@@ -777,6 +777,7 @@ static int ocfs2_sb_probe(struct super_block *sb,
 		}
 		di = (struct ocfs2_dinode *) (*bh)->b_data;
 		memset(stats, 0, sizeof(struct ocfs2_blockcheck_stats));
+		spin_lock_init(&stats->b_lock);
 		status = ocfs2_verify_volume(di, *bh, blksize, stats);
 		if (status >= 0)
 			goto bail;
@@ -1182,7 +1183,7 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent)
 	wake_up(&osb->osb_mount_event);
 
 	/* Start this when the mount is almost sure of being successful */
-	ocfs2_orphan_scan_init(osb);
+	ocfs2_orphan_scan_start(osb);
 
 	mlog_exit(status);
 	return status;
@@ -1213,14 +1214,27 @@ static int ocfs2_get_sb(struct file_system_type *fs_type,
 			   mnt);
 }
 
+static void ocfs2_kill_sb(struct super_block *sb)
+{
+	struct ocfs2_super *osb = OCFS2_SB(sb);
+
+	/* Prevent further queueing of inode drop events */
+	spin_lock(&dentry_list_lock);
+	ocfs2_set_osb_flag(osb, OCFS2_OSB_DROP_DENTRY_LOCK_IMMED);
+	spin_unlock(&dentry_list_lock);
+	/* Wait for work to finish and/or remove it */
+	cancel_work_sync(&osb->dentry_lock_work);
+
+	kill_block_super(sb);
+}
+
 static struct file_system_type ocfs2_fs_type = {
 	.owner          = THIS_MODULE,
 	.name           = "ocfs2",
 	.get_sb         = ocfs2_get_sb, /* is this called when we mount
 					* the fs? */
-	.kill_sb        = kill_block_super, /* set to the generic one
-					     * right now, but do we
-					     * need to change that? */
+	.kill_sb        = ocfs2_kill_sb,
+
 	.fs_flags       = FS_REQUIRES_DEV|FS_RENAME_DOES_D_MOVE,
 	.next           = NULL
 };
@@ -1819,6 +1833,12 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err)
 
 	debugfs_remove(osb->osb_ctxt);
 
+	/*
+	 * Flush inode dropping work queue so that deletes are
+	 * performed while the filesystem is still working
+	 */
+	ocfs2_drop_all_dl_inodes(osb);
+
 	/* Orphan scan should be stopped as early as possible */
 	ocfs2_orphan_scan_stop(osb);
 
@@ -1981,6 +2001,8 @@ static int ocfs2_initialize_super(struct super_block *sb,
 	snprintf(osb->dev_str, sizeof(osb->dev_str), "%u,%u",
 		 MAJOR(osb->sb->s_dev), MINOR(osb->sb->s_dev));
 
+	ocfs2_orphan_scan_init(osb);
+
 	status = ocfs2_recovery_init(osb);
 	if (status) {
 		mlog(ML_ERROR, "Unable to initialize recovery state\n");

fs/ocfs2/xattr.c

@@ -1052,7 +1052,8 @@ static int ocfs2_xattr_block_get(struct inode *inode,
 	struct ocfs2_xattr_block *xb;
 	struct ocfs2_xattr_value_root *xv;
 	size_t size;
-	int ret = -ENODATA, name_offset, name_len, block_off, i;
+	int ret = -ENODATA, name_offset, name_len, i;
+	int uninitialized_var(block_off);
 
 	xs->bucket = ocfs2_xattr_bucket_new(inode);
 	if (!xs->bucket) {

include/linux/perf_counter.h

@@ -115,7 +115,7 @@ enum perf_counter_sample_format {
 	PERF_SAMPLE_TID				= 1U << 1,
 	PERF_SAMPLE_TIME			= 1U << 2,
 	PERF_SAMPLE_ADDR			= 1U << 3,
-	PERF_SAMPLE_GROUP			= 1U << 4,
+	PERF_SAMPLE_READ			= 1U << 4,
 	PERF_SAMPLE_CALLCHAIN			= 1U << 5,
 	PERF_SAMPLE_ID				= 1U << 6,
 	PERF_SAMPLE_CPU				= 1U << 7,
@@ -127,16 +127,32 @@ enum perf_counter_sample_format {
 };
 
 /*
- * Bits that can be set in attr.read_format to request that
- * reads on the counter should return the indicated quantities,
- * in increasing order of bit value, after the counter value.
+ * The format of the data returned by read() on a perf counter fd,
+ * as specified by attr.read_format:
+ *
+ * struct read_format {
+ * 	{ u64		value;
+ * 	  { u64		time_enabled; } && PERF_FORMAT_ENABLED
+ * 	  { u64		time_running; } && PERF_FORMAT_RUNNING
+ * 	  { u64		id;           } && PERF_FORMAT_ID
+ * 	} && !PERF_FORMAT_GROUP
+ *
+ * 	{ u64		nr;
+ * 	  { u64		time_enabled; } && PERF_FORMAT_ENABLED
+ * 	  { u64		time_running; } && PERF_FORMAT_RUNNING
+ * 	  { u64		value;
+ * 	    { u64	id;           } && PERF_FORMAT_ID
+ * 	  }		cntr[nr];
+ * 	} && PERF_FORMAT_GROUP
+ * };
  */
 enum perf_counter_read_format {
 	PERF_FORMAT_TOTAL_TIME_ENABLED		= 1U << 0,
 	PERF_FORMAT_TOTAL_TIME_RUNNING		= 1U << 1,
 	PERF_FORMAT_ID				= 1U << 2,
+	PERF_FORMAT_GROUP			= 1U << 3,
 
-	PERF_FORMAT_MAX = 1U << 3, 		/* non-ABI */
+	PERF_FORMAT_MAX = 1U << 4, 		/* non-ABI */
 };
 
 #define PERF_ATTR_SIZE_VER0	64	/* sizeof first published struct */
@@ -343,10 +359,8 @@ enum perf_event_type {
 	 * struct {
 	 * 	struct perf_event_header	header;
 	 * 	u32				pid, tid;
-	 * 	u64				value;
-	 * 	{ u64		time_enabled; 	} && PERF_FORMAT_ENABLED
-	 * 	{ u64		time_running; 	} && PERF_FORMAT_RUNNING
-	 * 	{ u64		parent_id;	} && PERF_FORMAT_ID
+	 *
+	 * 	struct read_format		values;
 	 * };
 	 */
 	PERF_EVENT_READ			= 8,
@@ -364,11 +378,22 @@ enum perf_event_type {
 	 *	{ u32			cpu, res; } && PERF_SAMPLE_CPU
 	 * 	{ u64			period;   } && PERF_SAMPLE_PERIOD
 	 *
-	 *	{ u64			nr;
-	 *	  { u64 id, val; }	cnt[nr];  } && PERF_SAMPLE_GROUP
+	 *	{ struct read_format	values;	  } && PERF_SAMPLE_READ
 	 *
 	 *	{ u64			nr,
 	 *	  u64			ips[nr];  } && PERF_SAMPLE_CALLCHAIN
+	 *
+	 * 	#
+	 * 	# The RAW record below is opaque data wrt the ABI
+	 * 	#
+	 * 	# That is, the ABI doesn't make any promises wrt to
+	 * 	# the stability of its content, it may vary depending
+	 * 	# on event, hardware, kernel version and phase of
+	 * 	# the moon.
+	 * 	#
+	 * 	# In other words, PERF_SAMPLE_RAW contents are not an ABI.
+	 * 	#
+	 *
 	 *	{ u32			size;
 	 *	  char                  data[size];}&& PERF_SAMPLE_RAW
 	 * };
@@ -694,6 +719,8 @@ struct perf_sample_data {
 
 extern int perf_counter_overflow(struct perf_counter *counter, int nmi,
 				 struct perf_sample_data *data);
+extern void perf_counter_output(struct perf_counter *counter, int nmi,
+				struct perf_sample_data *data);
 
 /*
  * Return 1 for a software counter, 0 for a hardware counter

include/linux/wait.h

@@ -77,7 +77,14 @@ struct task_struct;
 #define __WAIT_BIT_KEY_INITIALIZER(word, bit)				\
 	{ .flags = word, .bit_nr = bit, }
 
-extern void init_waitqueue_head(wait_queue_head_t *q);
+extern void __init_waitqueue_head(wait_queue_head_t *q, struct lock_class_key *);
+
+#define init_waitqueue_head(q)				\
+	do {						\
+		static struct lock_class_key __key;	\
+							\
+		__init_waitqueue_head((q), &__key);	\
+	} while (0)
 
 #ifdef CONFIG_LOCKDEP
 # define __WAIT_QUEUE_HEAD_INIT_ONSTACK(name) \

kernel/futex.c

@@ -1010,15 +1010,19 @@ void requeue_futex(struct futex_q *q, struct futex_hash_bucket *hb1,
  * requeue_pi_wake_futex() - Wake a task that acquired the lock during requeue
  * q:	the futex_q
  * key:	the key of the requeue target futex
+ * hb:  the hash_bucket of the requeue target futex
  *
  * During futex_requeue, with requeue_pi=1, it is possible to acquire the
  * target futex if it is uncontended or via a lock steal.  Set the futex_q key
  * to the requeue target futex so the waiter can detect the wakeup on the right
  * futex, but remove it from the hb and NULL the rt_waiter so it can detect
- * atomic lock acquisition.  Must be called with the q->lock_ptr held.
+ * atomic lock acquisition.  Set the q->lock_ptr to the requeue target hb->lock
+ * to protect access to the pi_state to fixup the owner later.  Must be called
+ * with both q->lock_ptr and hb->lock held.
  */
 static inline
-void requeue_pi_wake_futex(struct futex_q *q, union futex_key *key)
+void requeue_pi_wake_futex(struct futex_q *q, union futex_key *key,
+			   struct futex_hash_bucket *hb)
 {
 	drop_futex_key_refs(&q->key);
 	get_futex_key_refs(key);
@@ -1030,6 +1034,11 @@ void requeue_pi_wake_futex(struct futex_q *q, union futex_key *key)
 	WARN_ON(!q->rt_waiter);
 	q->rt_waiter = NULL;
 
+	q->lock_ptr = &hb->lock;
+#ifdef CONFIG_DEBUG_PI_LIST
+	q->list.plist.lock = &hb->lock;
+#endif
+
 	wake_up_state(q->task, TASK_NORMAL);
 }
 
@@ -1088,7 +1097,7 @@ static int futex_proxy_trylock_atomic(u32 __user *pifutex,
 	ret = futex_lock_pi_atomic(pifutex, hb2, key2, ps, top_waiter->task,
 				   set_waiters);
 	if (ret == 1)
-		requeue_pi_wake_futex(top_waiter, key2);
+		requeue_pi_wake_futex(top_waiter, key2, hb2);
 
 	return ret;
 }
@@ -1247,8 +1256,15 @@ retry_private:
 		if (!match_futex(&this->key, &key1))
 			continue;
 
-		WARN_ON(!requeue_pi && this->rt_waiter);
-		WARN_ON(requeue_pi && !this->rt_waiter);
+		/*
+		 * FUTEX_WAIT_REQEUE_PI and FUTEX_CMP_REQUEUE_PI should always
+		 * be paired with each other and no other futex ops.
+		 */
+		if ((requeue_pi && !this->rt_waiter) ||
+		    (!requeue_pi && this->rt_waiter)) {
+			ret = -EINVAL;
+			break;
+		}
 
 		/*
 		 * Wake nr_wake waiters.  For requeue_pi, if we acquired the
@@ -1273,7 +1289,7 @@ retry_private:
 							this->task, 1);
 			if (ret == 1) {
 				/* We got the lock. */
-				requeue_pi_wake_futex(this, &key2);
+				requeue_pi_wake_futex(this, &key2, hb2);
 				continue;
 			} else if (ret) {
 				/* -EDEADLK */

kernel/futex_compat.c

@@ -180,7 +180,8 @@ asmlinkage long compat_sys_futex(u32 __user *uaddr, int op, u32 val,
 	int cmd = op & FUTEX_CMD_MASK;
 
 	if (utime && (cmd == FUTEX_WAIT || cmd == FUTEX_LOCK_PI ||
-		      cmd == FUTEX_WAIT_BITSET)) {
+		      cmd == FUTEX_WAIT_BITSET ||
+		      cmd == FUTEX_WAIT_REQUEUE_PI)) {
 		if (get_compat_timespec(&ts, utime))
 			return -EFAULT;
 		if (!timespec_valid(&ts))
@@ -191,7 +192,8 @@ asmlinkage long compat_sys_futex(u32 __user *uaddr, int op, u32 val,
 			t = ktime_add_safe(ktime_get(), t);
 		tp = &t;
 	}
-	if (cmd == FUTEX_REQUEUE || cmd == FUTEX_CMP_REQUEUE)
+	if (cmd == FUTEX_REQUEUE || cmd == FUTEX_CMP_REQUEUE ||
+	    cmd == FUTEX_CMP_REQUEUE_PI || cmd == FUTEX_WAKE_OP)
 		val2 = (int) (unsigned long) utime;
 
 	return do_futex(uaddr, op, val, tp, uaddr2, val2, val3);

kernel/irq/manage.c

@@ -761,7 +761,6 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id)
 {
 	struct irq_desc *desc = irq_to_desc(irq);
 	struct irqaction *action, **action_ptr;
-	struct task_struct *irqthread;
 	unsigned long flags;
 
 	WARN(in_interrupt(), "Trying to free IRQ %d from IRQ context!\n", irq);
@@ -809,9 +808,6 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id)
 			desc->chip->disable(irq);
 	}
 
-	irqthread = action->thread;
-	action->thread = NULL;
-
 	spin_unlock_irqrestore(&desc->lock, flags);
 
 	unregister_handler_proc(irq, action);
@@ -819,12 +815,6 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id)
 	/* Make sure it's not being used on another CPU: */
 	synchronize_irq(irq);
 
-	if (irqthread) {
-		if (!test_bit(IRQTF_DIED, &action->thread_flags))
-			kthread_stop(irqthread);
-		put_task_struct(irqthread);
-	}
-
 #ifdef CONFIG_DEBUG_SHIRQ
 	/*
 	 * It's a shared IRQ -- the driver ought to be prepared for an IRQ
@@ -840,6 +830,13 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id)
 		local_irq_restore(flags);
 	}
 #endif
+
+	if (action->thread) {
+		if (!test_bit(IRQTF_DIED, &action->thread_flags))
+			kthread_stop(action->thread);
+		put_task_struct(action->thread);
+	}
+
 	return action;
 }
 

kernel/perf_counter.c

@@ -88,6 +88,7 @@ void __weak hw_perf_disable(void)		{ barrier(); }
 void __weak hw_perf_enable(void)		{ barrier(); }
 
 void __weak hw_perf_counter_setup(int cpu)	{ barrier(); }
+void __weak hw_perf_counter_setup_online(int cpu)	{ barrier(); }
 
 int __weak
 hw_perf_group_sched_in(struct perf_counter *group_leader,
@@ -306,6 +307,10 @@ counter_sched_out(struct perf_counter *counter,
 		return;
 
 	counter->state = PERF_COUNTER_STATE_INACTIVE;
+	if (counter->pending_disable) {
+		counter->pending_disable = 0;
+		counter->state = PERF_COUNTER_STATE_OFF;
+	}
 	counter->tstamp_stopped = ctx->time;
 	counter->pmu->disable(counter);
 	counter->oncpu = -1;
@@ -1691,7 +1696,32 @@ static int perf_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-static u64 perf_counter_read_tree(struct perf_counter *counter)
+static int perf_counter_read_size(struct perf_counter *counter)
+{
+	int entry = sizeof(u64); /* value */
+	int size = 0;
+	int nr = 1;
+
+	if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED)
+		size += sizeof(u64);
+
+	if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING)
+		size += sizeof(u64);
+
+	if (counter->attr.read_format & PERF_FORMAT_ID)
+		entry += sizeof(u64);
+
+	if (counter->attr.read_format & PERF_FORMAT_GROUP) {
+		nr += counter->group_leader->nr_siblings;
+		size += sizeof(u64);
+	}
+
+	size += entry * nr;
+
+	return size;
+}
+
+static u64 perf_counter_read_value(struct perf_counter *counter)
 {
 	struct perf_counter *child;
 	u64 total = 0;
@@ -1703,14 +1733,96 @@ static u64 perf_counter_read_tree(struct perf_counter *counter)
 	return total;
 }
 
+static int perf_counter_read_entry(struct perf_counter *counter,
+				   u64 read_format, char __user *buf)
+{
+	int n = 0, count = 0;
+	u64 values[2];
+
+	values[n++] = perf_counter_read_value(counter);
+	if (read_format & PERF_FORMAT_ID)
+		values[n++] = primary_counter_id(counter);
+
+	count = n * sizeof(u64);
+
+	if (copy_to_user(buf, values, count))
+		return -EFAULT;
+
+	return count;
+}
+
+static int perf_counter_read_group(struct perf_counter *counter,
+				   u64 read_format, char __user *buf)
+{
+	struct perf_counter *leader = counter->group_leader, *sub;
+	int n = 0, size = 0, err = -EFAULT;
+	u64 values[3];
+
+	values[n++] = 1 + leader->nr_siblings;
+	if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) {
+		values[n++] = leader->total_time_enabled +
+			atomic64_read(&leader->child_total_time_enabled);
+	}
+	if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) {
+		values[n++] = leader->total_time_running +
+			atomic64_read(&leader->child_total_time_running);
+	}
+
+	size = n * sizeof(u64);
+
+	if (copy_to_user(buf, values, size))
+		return -EFAULT;
+
+	err = perf_counter_read_entry(leader, read_format, buf + size);
+	if (err < 0)
+		return err;
+
+	size += err;
+
+	list_for_each_entry(sub, &leader->sibling_list, list_entry) {
+		err = perf_counter_read_entry(counter, read_format,
+				buf + size);
+		if (err < 0)
+			return err;
+
+		size += err;
+	}
+
+	return size;
+}
+
+static int perf_counter_read_one(struct perf_counter *counter,
+				 u64 read_format, char __user *buf)
+{
+	u64 values[4];
+	int n = 0;
+
+	values[n++] = perf_counter_read_value(counter);
+	if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) {
+		values[n++] = counter->total_time_enabled +
+			atomic64_read(&counter->child_total_time_enabled);
+	}
+	if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) {
+		values[n++] = counter->total_time_running +
+			atomic64_read(&counter->child_total_time_running);
+	}
+	if (read_format & PERF_FORMAT_ID)
+		values[n++] = primary_counter_id(counter);
+
+	if (copy_to_user(buf, values, n * sizeof(u64)))
+		return -EFAULT;
+
+	return n * sizeof(u64);
+}
+
 /*
  * Read the performance counter - simple non blocking version for now
  */
 static ssize_t
 perf_read_hw(struct perf_counter *counter, char __user *buf, size_t count)
 {
-	u64 values[4];
-	int n;
+	u64 read_format = counter->attr.read_format;
+	int ret;
 
 	/*
 	 * Return end-of-file for a read on a counter that is in
@@ -1720,28 +1832,18 @@ perf_read_hw(struct perf_counter *counter, char __user *buf, size_t count)
 	if (counter->state == PERF_COUNTER_STATE_ERROR)
 		return 0;
 
+	if (count < perf_counter_read_size(counter))
+		return -ENOSPC;
+
 	WARN_ON_ONCE(counter->ctx->parent_ctx);
 	mutex_lock(&counter->child_mutex);
-	values[0] = perf_counter_read_tree(counter);
-	n = 1;
-	if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED)
-		values[n++] = counter->total_time_enabled +
-			atomic64_read(&counter->child_total_time_enabled);
-	if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING)
-		values[n++] = counter->total_time_running +
-			atomic64_read(&counter->child_total_time_running);
-	if (counter->attr.read_format & PERF_FORMAT_ID)
-		values[n++] = primary_counter_id(counter);
+	if (read_format & PERF_FORMAT_GROUP)
+		ret = perf_counter_read_group(counter, read_format, buf);
+	else
+		ret = perf_counter_read_one(counter, read_format, buf);
 	mutex_unlock(&counter->child_mutex);
 
-	if (count < n * sizeof(u64))
-		return -EINVAL;
-	count = n * sizeof(u64);
-
-	if (copy_to_user(buf, values, count))
-		return -EFAULT;
-
-	return count;
+	return ret;
 }
 
 static ssize_t
@@ -2245,7 +2347,7 @@ static void perf_pending_counter(struct perf_pending_entry *entry)
 
 	if (counter->pending_disable) {
 		counter->pending_disable = 0;
-		perf_counter_disable(counter);
+		__perf_counter_disable(counter);
 	}
 
 	if (counter->pending_wakeup) {
@@ -2630,7 +2732,80 @@ static u32 perf_counter_tid(struct perf_counter *counter, struct task_struct *p)
 	return task_pid_nr_ns(p, counter->ns);
 }
 
-static void perf_counter_output(struct perf_counter *counter, int nmi,
+static void perf_output_read_one(struct perf_output_handle *handle,
+				 struct perf_counter *counter)
+{
+	u64 read_format = counter->attr.read_format;
+	u64 values[4];
+	int n = 0;
+
+	values[n++] = atomic64_read(&counter->count);
+	if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) {
+		values[n++] = counter->total_time_enabled +
+			atomic64_read(&counter->child_total_time_enabled);
+	}
+	if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) {
+		values[n++] = counter->total_time_running +
+			atomic64_read(&counter->child_total_time_running);
+	}
+	if (read_format & PERF_FORMAT_ID)
+		values[n++] = primary_counter_id(counter);
+
+	perf_output_copy(handle, values, n * sizeof(u64));
+}
+
+/*
+ * XXX PERF_FORMAT_GROUP vs inherited counters seems difficult.
+ */
+static void perf_output_read_group(struct perf_output_handle *handle,
+			    struct perf_counter *counter)
+{
+	struct perf_counter *leader = counter->group_leader, *sub;
+	u64 read_format = counter->attr.read_format;
+	u64 values[5];
+	int n = 0;
+
+	values[n++] = 1 + leader->nr_siblings;
+
+	if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED)
+		values[n++] = leader->total_time_enabled;
+
+	if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING)
+		values[n++] = leader->total_time_running;
+
+	if (leader != counter)
+		leader->pmu->read(leader);
+
+	values[n++] = atomic64_read(&leader->count);
+	if (read_format & PERF_FORMAT_ID)
+		values[n++] = primary_counter_id(leader);
+
+	perf_output_copy(handle, values, n * sizeof(u64));
+
+	list_for_each_entry(sub, &leader->sibling_list, list_entry) {
+		n = 0;
+
+		if (sub != counter)
+			sub->pmu->read(sub);
+
+		values[n++] = atomic64_read(&sub->count);
+		if (read_format & PERF_FORMAT_ID)
+			values[n++] = primary_counter_id(sub);
+
+		perf_output_copy(handle, values, n * sizeof(u64));
+	}
+}
+
+static void perf_output_read(struct perf_output_handle *handle,
+			     struct perf_counter *counter)
+{
+	if (counter->attr.read_format & PERF_FORMAT_GROUP)
+		perf_output_read_group(handle, counter);
+	else
+		perf_output_read_one(handle, counter);
+}
+
+void perf_counter_output(struct perf_counter *counter, int nmi,
 				struct perf_sample_data *data)
 {
 	int ret;
@@ -2641,10 +2816,6 @@ static void perf_counter_output(struct perf_counter *counter, int nmi,
 	struct {
 		u32 pid, tid;
 	} tid_entry;
-	struct {
-		u64 id;
-		u64 counter;
-	} group_entry;
 	struct perf_callchain_entry *callchain = NULL;
 	int callchain_size = 0;
 	u64 time;
@@ -2699,10 +2870,8 @@ static void perf_counter_output(struct perf_counter *counter, int nmi,
 	if (sample_type & PERF_SAMPLE_PERIOD)
 		header.size += sizeof(u64);
 
-	if (sample_type & PERF_SAMPLE_GROUP) {
-		header.size += sizeof(u64) +
-			counter->nr_siblings * sizeof(group_entry);
-	}
+	if (sample_type & PERF_SAMPLE_READ)
+		header.size += perf_counter_read_size(counter);
 
 	if (sample_type & PERF_SAMPLE_CALLCHAIN) {
 		callchain = perf_callchain(data->regs);
@@ -2759,26 +2928,8 @@ static void perf_counter_output(struct perf_counter *counter, int nmi,
 	if (sample_type & PERF_SAMPLE_PERIOD)
 		perf_output_put(&handle, data->period);
 
-	/*
-	 * XXX PERF_SAMPLE_GROUP vs inherited counters seems difficult.
-	 */
-	if (sample_type & PERF_SAMPLE_GROUP) {
-		struct perf_counter *leader, *sub;
-		u64 nr = counter->nr_siblings;
-
-		perf_output_put(&handle, nr);
-
-		leader = counter->group_leader;
-		list_for_each_entry(sub, &leader->sibling_list, list_entry) {
-			if (sub != counter)
-				sub->pmu->read(sub);
-
-			group_entry.id = primary_counter_id(sub);
-			group_entry.counter = atomic64_read(&sub->count);
-
-			perf_output_put(&handle, group_entry);
-		}
-	}
+	if (sample_type & PERF_SAMPLE_READ)
+		perf_output_read(&handle, counter);
 
 	if (sample_type & PERF_SAMPLE_CALLCHAIN) {
 		if (callchain)
@@ -2817,8 +2968,6 @@ struct perf_read_event {
 
 	u32				pid;
 	u32				tid;
-	u64				value;
-	u64				format[3];
 };
 
 static void
@@ -2830,34 +2979,20 @@ perf_counter_read_event(struct perf_counter *counter,
 		.header = {
 			.type = PERF_EVENT_READ,
 			.misc = 0,
-			.size = sizeof(event) - sizeof(event.format),
+			.size = sizeof(event) + perf_counter_read_size(counter),
 		},
 		.pid = perf_counter_pid(counter, task),
 		.tid = perf_counter_tid(counter, task),
-		.value = atomic64_read(&counter->count),
 	};
-	int ret, i = 0;
-
-	if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) {
-		event.header.size += sizeof(u64);
-		event.format[i++] = counter->total_time_enabled;
-	}
-
-	if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) {
-		event.header.size += sizeof(u64);
-		event.format[i++] = counter->total_time_running;
-	}
-
-	if (counter->attr.read_format & PERF_FORMAT_ID) {
-		event.header.size += sizeof(u64);
-		event.format[i++] = primary_counter_id(counter);
-	}
+	int ret;
 
 	ret = perf_output_begin(&handle, counter, event.header.size, 0, 0);
 	if (ret)
 		return;
 
-	perf_output_copy(&handle, &event, event.header.size);
+	perf_output_put(&handle, event);
+	perf_output_read(&handle, counter);
+
 	perf_output_end(&handle);
 }
 
@@ -2893,10 +3028,10 @@ static void perf_counter_task_output(struct perf_counter *counter,
 		return;
 
 	task_event->event.pid = perf_counter_pid(counter, task);
-	task_event->event.ppid = perf_counter_pid(counter, task->real_parent);
+	task_event->event.ppid = perf_counter_pid(counter, current);
 
 	task_event->event.tid = perf_counter_tid(counter, task);
-	task_event->event.ptid = perf_counter_tid(counter, task->real_parent);
+	task_event->event.ptid = perf_counter_tid(counter, current);
 
 	perf_output_put(&handle, task_event->event);
 	perf_output_end(&handle);
@@ -3443,40 +3578,32 @@ static void perf_swcounter_add(struct perf_counter *counter, u64 nr,
 
 static int perf_swcounter_is_counting(struct perf_counter *counter)
 {
-	struct perf_counter_context *ctx;
-	unsigned long flags;
-	int count;
-
+	/*
+	 * The counter is active, we're good!
+	 */
 	if (counter->state == PERF_COUNTER_STATE_ACTIVE)
 		return 1;
 
+	/*
+	 * The counter is off/error, not counting.
+	 */
 	if (counter->state != PERF_COUNTER_STATE_INACTIVE)
 		return 0;
 
 	/*
-	 * If the counter is inactive, it could be just because
-	 * its task is scheduled out, or because it's in a group
-	 * which could not go on the PMU.  We want to count in
-	 * the first case but not the second.  If the context is
-	 * currently active then an inactive software counter must
-	 * be the second case.  If it's not currently active then
-	 * we need to know whether the counter was active when the
-	 * context was last active, which we can determine by
-	 * comparing counter->tstamp_stopped with ctx->time.
-	 *
-	 * We are within an RCU read-side critical section,
-	 * which protects the existence of *ctx.
+	 * The counter is inactive, if the context is active
+	 * we're part of a group that didn't make it on the 'pmu',
+	 * not counting.
 	 */
-	ctx = counter->ctx;
-	spin_lock_irqsave(&ctx->lock, flags);
-	count = 1;
-	/* Re-check state now we have the lock */
-	if (counter->state < PERF_COUNTER_STATE_INACTIVE ||
-	    counter->ctx->is_active ||
-	    counter->tstamp_stopped < ctx->time)
-		count = 0;
-	spin_unlock_irqrestore(&ctx->lock, flags);
-	return count;
+	if (counter->ctx->is_active)
+		return 0;
+
+	/*
+	 * We're inactive and the context is too, this means the
+	 * task is scheduled out, we're counting events that happen
+	 * to us, like migration events.
+	 */
+	return 1;
 }
 
 static int perf_swcounter_match(struct perf_counter *counter,
@@ -3928,9 +4055,9 @@ perf_counter_alloc(struct perf_counter_attr *attr,
 	atomic64_set(&hwc->period_left, hwc->sample_period);
 
 	/*
-	 * we currently do not support PERF_SAMPLE_GROUP on inherited counters
+	 * we currently do not support PERF_FORMAT_GROUP on inherited counters
 	 */
-	if (attr->inherit && (attr->sample_type & PERF_SAMPLE_GROUP))
+	if (attr->inherit && (attr->read_format & PERF_FORMAT_GROUP))
 		goto done;
 
 	switch (attr->type) {
@@ -4592,6 +4719,11 @@ perf_cpu_notify(struct notifier_block *self, unsigned long action, void *hcpu)
 		perf_counter_init_cpu(cpu);
 		break;
 
+	case CPU_ONLINE:
+	case CPU_ONLINE_FROZEN:
+		hw_perf_counter_setup_online(cpu);
+		break;
+
 	case CPU_DOWN_PREPARE:
 	case CPU_DOWN_PREPARE_FROZEN:
 		perf_counter_exit_cpu(cpu);
@@ -4616,6 +4748,8 @@ void __init perf_counter_init(void)
 {
 	perf_cpu_notify(&perf_cpu_nb, (unsigned long)CPU_UP_PREPARE,
 			(void *)(long)smp_processor_id());
+	perf_cpu_notify(&perf_cpu_nb, (unsigned long)CPU_ONLINE,
+			(void *)(long)smp_processor_id());
 	register_cpu_notifier(&perf_cpu_nb);
 }
 

kernel/wait.c

@@ -10,13 +10,14 @@
 #include <linux/wait.h>
 #include <linux/hash.h>
 
-void init_waitqueue_head(wait_queue_head_t *q)
+void __init_waitqueue_head(wait_queue_head_t *q, struct lock_class_key *key)
 {
 	spin_lock_init(&q->lock);
+	lockdep_set_class(&q->lock, key);
 	INIT_LIST_HEAD(&q->task_list);
 }
 
-EXPORT_SYMBOL(init_waitqueue_head);
+EXPORT_SYMBOL(__init_waitqueue_head);
 
 void add_wait_queue(wait_queue_head_t *q, wait_queue_t *wait)
 {

net/socket.c

@@ -736,7 +736,7 @@ static ssize_t sock_sendpage(struct file *file, struct page *page,
 	if (more)
 		flags |= MSG_MORE;
 
-	return sock->ops->sendpage(sock, page, offset, size, flags);
+	return kernel_sendpage(sock, page, offset, size, flags);
 }
 
 static ssize_t sock_splice_read(struct file *file, loff_t *ppos,

tools/perf/Makefile

@@ -382,22 +382,29 @@ endif
 ifdef NO_DEMANGLE
 	BASIC_CFLAGS += -DNO_DEMANGLE
 else
-
 	has_bfd := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) -lbfd > /dev/null 2>&1 && echo y")
 
-	has_bfd_iberty := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) -lbfd -liberty > /dev/null 2>&1 && echo y")
-
-	has_bfd_iberty_z := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) -lbfd -liberty -lz > /dev/null 2>&1 && echo y")
-
 	ifeq ($(has_bfd),y)
 		EXTLIBS += -lbfd
-	else ifeq ($(has_bfd_iberty),y)
-		EXTLIBS += -lbfd -liberty
-	else ifeq ($(has_bfd_iberty_z),y)
-		EXTLIBS += -lbfd -liberty -lz
 	else
-		msg := $(warning No bfd.h/libbfd found, install binutils-dev[el] to gain symbol demangling)
-		BASIC_CFLAGS += -DNO_DEMANGLE
+		has_bfd_iberty := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) -lbfd -liberty > /dev/null 2>&1 && echo y")
+		ifeq ($(has_bfd_iberty),y)
+			EXTLIBS += -lbfd -liberty
+		else
+			has_bfd_iberty_z := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) -lbfd -liberty -lz > /dev/null 2>&1 && echo y")
+			ifeq ($(has_bfd_iberty_z),y)
+				EXTLIBS += -lbfd -liberty -lz
+			else
+				has_cplus_demangle := $(shell sh -c "(echo 'extern char *cplus_demangle(const char *, int);'; echo 'int main(void) { cplus_demangle(0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) -liberty > /dev/null 2>&1 && echo y")
+				ifeq ($(has_cplus_demangle),y)
+					EXTLIBS += -liberty
+					BASIC_CFLAGS += -DHAVE_CPLUS_DEMANGLE
+				else
+					msg := $(warning No bfd.h/libbfd found, install binutils-dev[el] to gain symbol demangling)
+					BASIC_CFLAGS += -DNO_DEMANGLE
+				endif
+			endif
+		endif
 	endif
 endif
 

tools/perf/builtin-list.c

@@ -10,11 +10,12 @@
 
 #include "perf.h"
 
-#include "util/parse-options.h"
 #include "util/parse-events.h"
+#include "util/cache.h"
 
 int cmd_list(int argc __used, const char **argv __used, const char *prefix __used)
 {
+	setup_pager();
 	print_events();
 	return 0;
 }

tools/perf/builtin-record.c

@@ -34,7 +34,9 @@ static int			output;
 static const char		*output_name			= "perf.data";
 static int			group				= 0;
 static unsigned int		realtime_prio			= 0;
+static int			raw_samples			= 0;
 static int			system_wide			= 0;
+static int			profile_cpu			= -1;
 static pid_t			target_pid			= -1;
 static int			inherit				= 1;
 static int			force				= 0;
@@ -203,46 +205,48 @@ static void sig_atexit(void)
 	kill(getpid(), signr);
 }
 
-static void pid_synthesize_comm_event(pid_t pid, int full)
+static pid_t pid_synthesize_comm_event(pid_t pid, int full)
 {
 	struct comm_event comm_ev;
 	char filename[PATH_MAX];
 	char bf[BUFSIZ];
-	int fd;
-	size_t size;
-	char *field, *sep;
+	FILE *fp;
+	size_t size = 0;
 	DIR *tasks;
 	struct dirent dirent, *next;
+	pid_t tgid = 0;
 
-	snprintf(filename, sizeof(filename), "/proc/%d/stat", pid);
+	snprintf(filename, sizeof(filename), "/proc/%d/status", pid);
 
-	fd = open(filename, O_RDONLY);
-	if (fd < 0) {
+	fp = fopen(filename, "r");
+	if (fd == NULL) {
 		/*
 		 * We raced with a task exiting - just return:
 		 */
 		if (verbose)
 			fprintf(stderr, "couldn't open %s\n", filename);
-		return;
-	}
-	if (read(fd, bf, sizeof(bf)) < 0) {
-		fprintf(stderr, "couldn't read %s\n", filename);
-		exit(EXIT_FAILURE);
+		return 0;
 	}
-	close(fd);
 
-	/* 9027 (cat) R 6747 9027 6747 34816 9027 ... */
 	memset(&comm_ev, 0, sizeof(comm_ev));
-	field = strchr(bf, '(');
-	if (field == NULL)
-		goto out_failure;
-	sep = strchr(++field, ')');
-	if (sep == NULL)
-		goto out_failure;
-	size = sep - field;
-	memcpy(comm_ev.comm, field, size++);
-
-	comm_ev.pid = pid;
+	while (!comm_ev.comm[0] || !comm_ev.pid) {
+		if (fgets(bf, sizeof(bf), fp) == NULL)
+			goto out_failure;
+
+		if (memcmp(bf, "Name:", 5) == 0) {
+			char *name = bf + 5;
+			while (*name && isspace(*name))
+				++name;
+			size = strlen(name) - 1;
+			memcpy(comm_ev.comm, name, size++);
+		} else if (memcmp(bf, "Tgid:", 5) == 0) {
+			char *tgids = bf + 5;
+			while (*tgids && isspace(*tgids))
+				++tgids;
+			tgid = comm_ev.pid = atoi(tgids);
+		}
+	}
+
 	comm_ev.header.type = PERF_EVENT_COMM;
 	size = ALIGN(size, sizeof(u64));
 	comm_ev.header.size = sizeof(comm_ev) - (sizeof(comm_ev.comm) - size);
@@ -251,7 +255,7 @@ static void pid_synthesize_comm_event(pid_t pid, int full)
 		comm_ev.tid = pid;
 
 		write_output(&comm_ev, comm_ev.header.size);
-		return;
+		goto out_fclose;
 	}
 
 	snprintf(filename, sizeof(filename), "/proc/%d/task", pid);
@@ -268,7 +272,10 @@ static void pid_synthesize_comm_event(pid_t pid, int full)
 		write_output(&comm_ev, comm_ev.header.size);
 	}
 	closedir(tasks);
-	return;
+
+out_fclose:
+	fclose(fp);
+	return tgid;
 
 out_failure:
 	fprintf(stderr, "couldn't get COMM and pgid, malformed %s\n",
@@ -276,7 +283,7 @@ out_failure:
 	exit(EXIT_FAILURE);
 }
 
-static void pid_synthesize_mmap_samples(pid_t pid)
+static void pid_synthesize_mmap_samples(pid_t pid, pid_t tgid)
 {
 	char filename[PATH_MAX];
 	FILE *fp;
@@ -328,7 +335,7 @@ static void pid_synthesize_mmap_samples(pid_t pid)
 			mmap_ev.len -= mmap_ev.start;
 			mmap_ev.header.size = (sizeof(mmap_ev) -
 					       (sizeof(mmap_ev.filename) - size));
-			mmap_ev.pid = pid;
+			mmap_ev.pid = tgid;
 			mmap_ev.tid = pid;
 
 			write_output(&mmap_ev, mmap_ev.header.size);
@@ -347,14 +354,14 @@ static void synthesize_all(void)
 
 	while (!readdir_r(proc, &dirent, &next) && next) {
 		char *end;
-		pid_t pid;
+		pid_t pid, tgid;
 
 		pid = strtol(dirent.d_name, &end, 10);
 		if (*end) /* only interested in proper numerical dirents */
 			continue;
 
-		pid_synthesize_comm_event(pid, 1);
-		pid_synthesize_mmap_samples(pid);
+		tgid = pid_synthesize_comm_event(pid, 1);
+		pid_synthesize_mmap_samples(pid, tgid);
 	}
 
 	closedir(proc);
@@ -392,7 +399,7 @@ static void create_counter(int counter, int cpu, pid_t pid)
 				  PERF_FORMAT_TOTAL_TIME_RUNNING |
 				  PERF_FORMAT_ID;
 
-	attr->sample_type	= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
+	attr->sample_type	|= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
 
 	if (freq) {
 		attr->sample_type	|= PERF_SAMPLE_PERIOD;
@@ -412,6 +419,8 @@ static void create_counter(int counter, int cpu, pid_t pid)
 	if (call_graph)
 		attr->sample_type	|= PERF_SAMPLE_CALLCHAIN;
 
+	if (raw_samples)
+		attr->sample_type	|= PERF_SAMPLE_RAW;
 
 	attr->mmap		= track;
 	attr->comm		= track;
@@ -426,6 +435,8 @@ try_again:
 
 		if (err == EPERM)
 			die("Permission error - are you root?\n");
+		else if (err ==  ENODEV && profile_cpu != -1)
+			die("No such device - did you specify an out-of-range profile CPU?\n");
 
 		/*
 		 * If it's cycles then fall back to hrtimer
@@ -559,16 +570,22 @@ static int __cmd_record(int argc, const char **argv)
 		if (pid == -1)
 			pid = getpid();
 
-		open_counters(-1, pid);
-	} else for (i = 0; i < nr_cpus; i++)
-		open_counters(i, target_pid);
+		open_counters(profile_cpu, pid);
+	} else {
+		if (profile_cpu != -1) {
+			open_counters(profile_cpu, target_pid);
+		} else {
+			for (i = 0; i < nr_cpus; i++)
+				open_counters(i, target_pid);
+		}
+	}
 
 	if (file_new)
 		perf_header__write(header, output);
 
 	if (!system_wide) {
-		pid_synthesize_comm_event(pid, 0);
-		pid_synthesize_mmap_samples(pid);
+		pid_t tgid = pid_synthesize_comm_event(pid, 0);
+		pid_synthesize_mmap_samples(pid, tgid);
 	} else
 		synthesize_all();
 
@@ -636,10 +653,14 @@ static const struct option options[] = {
 		    "record events on existing pid"),
 	OPT_INTEGER('r', "realtime", &realtime_prio,
 		    "collect data with this RT SCHED_FIFO priority"),
+	OPT_BOOLEAN('R', "raw-samples", &raw_samples,
+		    "collect raw sample records from all opened counters"),
 	OPT_BOOLEAN('a', "all-cpus", &system_wide,
 			    "system-wide collection from all CPUs"),
 	OPT_BOOLEAN('A', "append", &append_file,
 			    "append to the output file to do incremental profiling"),
+	OPT_INTEGER('C', "profile_cpu", &profile_cpu,
+			    "CPU to profile on"),
 	OPT_BOOLEAN('f', "force", &force,
 			"overwrite existing data file"),
 	OPT_LONG('c', "count", &default_interval,

tools/perf/builtin-report.c

@@ -1526,11 +1526,11 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 		more_data += sizeof(u64);
 	}
 
-	dprintf("%p [%p]: PERF_EVENT_SAMPLE (IP, %d): %d: %p period: %Ld\n",
+	dprintf("%p [%p]: PERF_EVENT_SAMPLE (IP, %d): %d/%d: %p period: %Ld\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
 		event->header.misc,
-		event->ip.pid,
+		event->ip.pid, event->ip.tid,
 		(void *)(long)ip,
 		(long long)period);
 
@@ -1590,10 +1590,11 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 	if (show & show_mask) {
 		struct symbol *sym = resolve_symbol(thread, &map, &dso, &ip);
 
-		if (dso_list && dso && dso->name && !strlist__has_entry(dso_list, dso->name))
+		if (dso_list && (!dso || !dso->name ||
+				 !strlist__has_entry(dso_list, dso->name)))
 			return 0;
 
-		if (sym_list && sym && !strlist__has_entry(sym_list, sym->name))
+		if (sym_list && (!sym || !strlist__has_entry(sym_list, sym->name)))
 			return 0;
 
 		if (hist_entry__add(thread, map, dso, sym, ip, chain, level, period)) {
@@ -1612,10 +1613,11 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 	struct thread *thread = threads__findnew(event->mmap.pid);
 	struct map *map = map__new(&event->mmap);
 
-	dprintf("%p [%p]: PERF_EVENT_MMAP %d: [%p(%p) @ %p]: %s\n",
+	dprintf("%p [%p]: PERF_EVENT_MMAP %d/%d: [%p(%p) @ %p]: %s\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
 		event->mmap.pid,
+		event->mmap.tid,
 		(void *)(long)event->mmap.start,
 		(void *)(long)event->mmap.len,
 		(void *)(long)event->mmap.pgoff,

tools/perf/util/parse-events.c

@@ -379,6 +379,7 @@ static int parse_tracepoint_event(const char **strp,
 				    struct perf_counter_attr *attr)
 {
 	const char *evt_name;
+	char *flags;
 	char sys_name[MAX_EVENT_LENGTH];
 	char id_buf[4];
 	int fd;
@@ -400,6 +401,15 @@ static int parse_tracepoint_event(const char **strp,
 	strncpy(sys_name, *strp, sys_length);
 	sys_name[sys_length] = '\0';
 	evt_name = evt_name + 1;
+
+	flags = strchr(evt_name, ':');
+	if (flags) {
+		*flags = '\0';
+		flags++;
+		if (!strncmp(flags, "record", strlen(flags)))
+			attr->sample_type |= PERF_SAMPLE_RAW;
+	}
+
 	evt_length = strlen(evt_name);
 	if (evt_length >= MAX_EVENT_LENGTH)
 		return 0;

tools/perf/util/symbol.c

@@ -7,23 +7,8 @@
 #include <gelf.h>
 #include <elf.h>
 
-#ifndef NO_DEMANGLE
-#include <bfd.h>
-#else
-static inline
-char *bfd_demangle(void __used *v, const char __used *c, int __used i)
-{
-	return NULL;
-}
-#endif
-
 const char *sym_hist_filter;
 
-#ifndef DMGL_PARAMS
-#define DMGL_PARAMS      (1 << 0)       /* Include function args */
-#define DMGL_ANSI        (1 << 1)       /* Include const, volatile, etc */
-#endif
-
 enum dso_origin {
 	DSO__ORIG_KERNEL = 0,
 	DSO__ORIG_JAVA_JIT,
@@ -816,6 +801,8 @@ more:
 	}
 out:
 	free(name);
+	if (ret < 0 && strstr(self->name, " (deleted)") != NULL)
+		return 0;
 	return ret;
 }
 

tools/perf/util/symbol.h

@@ -7,6 +7,30 @@
 #include <linux/rbtree.h>
 #include "module.h"
 
+#ifdef HAVE_CPLUS_DEMANGLE
+extern char *cplus_demangle(const char *, int);
+
+static inline char *bfd_demangle(void __used *v, const char *c, int i)
+{
+	return cplus_demangle(c, i);
+}
+#else
+#ifdef NO_DEMANGLE
+static inline char *bfd_demangle(void __used *v, const char __used *c,
+				 int __used i)
+{
+	return NULL;
+}
+#else
+#include <bfd.h>
+#endif
+#endif
+
+#ifndef DMGL_PARAMS
+#define DMGL_PARAMS      (1 << 0)       /* Include function args */
+#define DMGL_ANSI        (1 << 1)       /* Include const, volatile, etc */
+#endif
+
 struct symbol {
 	struct rb_node	rb_node;
 	u64		start;