Merge branch 'for-linus' of git://git390.osdl.marist.edu/pub/scm/linux-2.6

* 'for-linus' of git://git390.osdl.marist.edu/pub/scm/linux-2.6: (85 commits)
  [S390] provide documentation for hvc_iucv kernel parameter.
  [S390] convert ctcm printks to dev_xxx and pr_xxx macros.
  [S390] convert zfcp printks to pr_xxx macros.
  [S390] convert vmlogrdr printks to pr_xxx macros.
  [S390] convert zfcp dumper printks to pr_xxx macros.
  [S390] convert cpu related printks to pr_xxx macros.
  [S390] convert qeth printks to dev_xxx and pr_xxx macros.
  [S390] convert sclp printks to pr_xxx macros.
  [S390] convert iucv printks to dev_xxx and pr_xxx macros.
  [S390] convert ap_bus printks to pr_xxx macros.
  [S390] convert dcssblk and extmem printks messages to pr_xxx macros.
  [S390] convert monwriter printks to pr_xxx macros.
  [S390] convert s390 debug feature printks to pr_xxx macros.
  [S390] convert monreader printks to pr_xxx macros.
  [S390] convert appldata printks to pr_xxx macros.
  [S390] convert setup printks to pr_xxx macros.
  [S390] convert hypfs printks to pr_xxx macros.
  [S390] convert time printks to pr_xxx macros.
  [S390] convert cpacf printks to pr_xxx macros.
  [S390] convert cio printks to pr_xxx macros.
  ...

Documentation/kernel-parameters.txt

@@ -823,6 +823,9 @@ and is between 256 and 4096 characters. It is defined in the file
 
 	hlt		[BUGS=ARM,SH]
 
+	hvc_iucv=	[S390] Number of z/VM IUCV Hypervisor console (HVC)
+			       back-ends. Valid parameters: 0..8
+
 	i8042.debug	[HW] Toggle i8042 debug mode
 	i8042.direct	[HW] Put keyboard port into non-translated mode
 	i8042.dumbkbd	[HW] Pretend that controller can only read data from
@@ -2292,6 +2295,14 @@ and is between 256 and 4096 characters. It is defined in the file
 			See comment before function dc390_setup() in
 			drivers/scsi/tmscsim.c.
 
+	topology=	[S390]
+			Format: {off | on}
+			Specify if the kernel should make use of the cpu
+			topology informations if the hardware supports these.
+			The scheduler will make use of these informations and
+			e.g. base its process migration decisions on it.
+			Default is off.
+
 	tp720=		[HW,PS2]
 
 	trix=		[HW,OSS] MediaTrix AudioTrix Pro

arch/powerpc/include/asm/elf.h

@@ -267,7 +267,7 @@ extern int ucache_bsize;
 #define ARCH_HAS_SETUP_ADDITIONAL_PAGES
 struct linux_binprm;
 extern int arch_setup_additional_pages(struct linux_binprm *bprm,
-				       int executable_stack);
+				       int uses_interp);
 #define VDSO_AUX_ENT(a,b) NEW_AUX_ENT(a,b);
 
 #endif /* __KERNEL__ */

arch/powerpc/kernel/vdso.c

@@ -184,8 +184,7 @@ static void dump_vdso_pages(struct vm_area_struct * vma)
  * This is called from binfmt_elf, we create the special vma for the
  * vDSO and insert it into the mm struct tree
  */
-int arch_setup_additional_pages(struct linux_binprm *bprm,
-				int executable_stack)
+int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 {
 	struct mm_struct *mm = current->mm;
 	struct page **vdso_pagelist;

arch/s390/Kconfig

@@ -43,6 +43,9 @@ config GENERIC_HWEIGHT
 config GENERIC_TIME
 	def_bool y
 
+config GENERIC_TIME_VSYSCALL
+	def_bool y
+
 config GENERIC_CLOCKEVENTS
 	def_bool y
 
@@ -66,10 +69,15 @@ config PGSTE
 	bool
 	default y if KVM
 
+config VIRT_CPU_ACCOUNTING
+	def_bool y
+
 mainmenu "Linux Kernel Configuration"
 
 config S390
 	def_bool y
+	select USE_GENERIC_SMP_HELPERS if SMP
+	select HAVE_FUNCTION_TRACER
 	select HAVE_OPROFILE
 	select HAVE_KPROBES
 	select HAVE_KRETPROBES
@@ -225,6 +233,14 @@ config MARCH_Z9_109
 	  Class (z9 BC). The kernel will be slightly faster but will not
 	  work on older machines such as the z990, z890, z900, and z800.
 
+config MARCH_Z10
+	bool "IBM System z10"
+	help
+	  Select this to enable optimizations for IBM System z10. The
+	  kernel will be slightly faster but will not work on older
+	  machines such as the z990, z890, z900, z800, z9-109, z9-ec
+	  and z9-bc.
+
 endchoice
 
 config PACK_STACK
@@ -343,16 +359,6 @@ config QDIO
 
 	  If unsure, say Y.
 
-config QDIO_DEBUG
-	bool "Extended debugging information"
-	depends on QDIO
-	help
-	  Say Y here to get extended debugging output in
-	    /sys/kernel/debug/s390dbf/qdio...
-	  Warning: this option reduces the performance of the QDIO module.
-
-	  If unsure, say N.
-
 config CHSC_SCH
 	tristate "Support for CHSC subchannels"
 	help
@@ -466,22 +472,9 @@ config PAGE_STATES
 	  hypervisor. The ESSA instruction is used to do the states
 	  changes between a page that has content and the unused state.
 
-config VIRT_TIMER
-	bool "Virtual CPU timer support"
-	help
-	  This provides a kernel interface for virtual CPU timers.
-	  Default is disabled.
-
-config VIRT_CPU_ACCOUNTING
-	bool "Base user process accounting on virtual cpu timer"
-	depends on VIRT_TIMER
-	help
-	  Select this option to use CPU timer deltas to do user
-	  process accounting.
-
 config APPLDATA_BASE
 	bool "Linux - VM Monitor Stream, base infrastructure"
-	depends on PROC_FS && VIRT_TIMER=y
+	depends on PROC_FS
 	help
 	  This provides a kernel interface for creating and updating z/VM APPLDATA
 	  monitor records. The monitor records are updated at certain time

arch/s390/Makefile

@@ -34,6 +34,7 @@ cflags-$(CONFIG_MARCH_G5)   += $(call cc-option,-march=g5)
 cflags-$(CONFIG_MARCH_Z900) += $(call cc-option,-march=z900)
 cflags-$(CONFIG_MARCH_Z990) += $(call cc-option,-march=z990)
 cflags-$(CONFIG_MARCH_Z9_109) += $(call cc-option,-march=z9-109)
+cflags-$(CONFIG_MARCH_Z10) += $(call cc-option,-march=z10)
 
 #KBUILD_IMAGE is necessary for make rpm
 KBUILD_IMAGE	:=arch/s390/boot/image

arch/s390/appldata/appldata.h

@@ -26,10 +26,6 @@
 #define CTL_APPLDATA_NET_SUM	2125
 #define CTL_APPLDATA_PROC	2126
 
-#define P_INFO(x...)	printk(KERN_INFO MY_PRINT_NAME " info: " x)
-#define P_ERROR(x...)	printk(KERN_ERR MY_PRINT_NAME " error: " x)
-#define P_WARNING(x...)	printk(KERN_WARNING MY_PRINT_NAME " status: " x)
-
 struct appldata_ops {
 	struct list_head list;
 	struct ctl_table_header *sysctl_header;

arch/s390/appldata/appldata_base.c

@@ -10,6 +10,9 @@
  * Author: Gerald Schaefer <gerald.schaefer@de.ibm.com>
  */
 
+#define KMSG_COMPONENT	"appldata"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/slab.h>
@@ -32,7 +35,6 @@
 #include "appldata.h"
 
 
-#define MY_PRINT_NAME	"appldata"		/* for debug messages, etc. */
 #define APPLDATA_CPU_INTERVAL	10000		/* default (CPU) time for
 						   sampling interval in
 						   milliseconds */
@@ -390,8 +392,8 @@ appldata_generic_handler(ctl_table *ctl, int write, struct file *filp,
 					(unsigned long) ops->data, ops->size,
 					ops->mod_lvl);
 		if (rc != 0) {
-			P_ERROR("START DIAG 0xDC for %s failed, "
-				"return code: %d\n", ops->name, rc);
+			pr_err("Starting the data collection for %s "
+			       "failed with rc=%d\n", ops->name, rc);
 			module_put(ops->owner);
 		} else
 			ops->active = 1;
@@ -401,8 +403,8 @@ appldata_generic_handler(ctl_table *ctl, int write, struct file *filp,
 				(unsigned long) ops->data, ops->size,
 				ops->mod_lvl);
 		if (rc != 0)
-			P_ERROR("STOP DIAG 0xDC for %s failed, "
-				"return code: %d\n", ops->name, rc);
+			pr_err("Stopping the data collection for %s "
+			       "failed with rc=%d\n", ops->name, rc);
 		module_put(ops->owner);
 	}
 	spin_unlock(&appldata_ops_lock);

arch/s390/appldata/appldata_os.c

@@ -9,6 +9,9 @@
  * Author: Gerald Schaefer <gerald.schaefer@de.ibm.com>
  */
 
+#define KMSG_COMPONENT	"appldata"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/slab.h>
@@ -22,7 +25,6 @@
 #include "appldata.h"
 
 
-#define MY_PRINT_NAME	"appldata_os"		/* for debug messages, etc. */
 #define LOAD_INT(x) ((x) >> FSHIFT)
 #define LOAD_FRAC(x) LOAD_INT(((x) & (FIXED_1-1)) * 100)
 
@@ -143,21 +145,16 @@ static void appldata_get_os_data(void *data)
 					   (unsigned long) ops.data, new_size,
 					   ops.mod_lvl);
 			if (rc != 0)
-				P_ERROR("os: START NEW DIAG 0xDC failed, "
-					"return code: %d, new size = %i\n", rc,
-					new_size);
+				pr_err("Starting a new OS data collection "
+				       "failed with rc=%d\n", rc);
 
 			rc = appldata_diag(APPLDATA_RECORD_OS_ID,
 					   APPLDATA_STOP_REC,
 					   (unsigned long) ops.data, ops.size,
 					   ops.mod_lvl);
 			if (rc != 0)
-				P_ERROR("os: STOP OLD DIAG 0xDC failed, "
-					"return code: %d, old size = %i\n", rc,
-					ops.size);
-			else
-				P_INFO("os: old record size = %i stopped\n",
-					ops.size);
+				pr_err("Stopping a faulty OS data "
+				       "collection failed with rc=%d\n", rc);
 		}
 		ops.size = new_size;
 	}
@@ -178,8 +175,8 @@ static int __init appldata_os_init(void)
 	max_size = sizeof(struct appldata_os_data) +
 		   (NR_CPUS * sizeof(struct appldata_os_per_cpu));
 	if (max_size > APPLDATA_MAX_REC_SIZE) {
-		P_ERROR("Max. size of OS record = %i, bigger than maximum "
-			"record size (%i)\n", max_size, APPLDATA_MAX_REC_SIZE);
+		pr_err("Maximum OS record size %i exceeds the maximum "
+		       "record size %i\n", max_size, APPLDATA_MAX_REC_SIZE);
 		rc = -ENOMEM;
 		goto out;
 	}

arch/s390/crypto/aes_s390.c

@@ -17,6 +17,9 @@
  *
  */
 
+#define KMSG_COMPONENT "aes_s390"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <crypto/aes.h>
 #include <crypto/algapi.h>
 #include <linux/err.h>
@@ -169,7 +172,8 @@ static int fallback_init_cip(struct crypto_tfm *tfm)
 			CRYPTO_ALG_ASYNC | CRYPTO_ALG_NEED_FALLBACK);
 
 	if (IS_ERR(sctx->fallback.cip)) {
-		printk(KERN_ERR "Error allocating fallback algo %s\n", name);
+		pr_err("Allocating AES fallback algorithm %s failed\n",
+		       name);
 		return PTR_ERR(sctx->fallback.blk);
 	}
 
@@ -349,7 +353,8 @@ static int fallback_init_blk(struct crypto_tfm *tfm)
 			CRYPTO_ALG_ASYNC | CRYPTO_ALG_NEED_FALLBACK);
 
 	if (IS_ERR(sctx->fallback.blk)) {
-		printk(KERN_ERR "Error allocating fallback algo %s\n", name);
+		pr_err("Allocating AES fallback algorithm %s failed\n",
+		       name);
 		return PTR_ERR(sctx->fallback.blk);
 	}
 
@@ -515,9 +520,8 @@ static int __init aes_s390_init(void)
 
 	/* z9 109 and z9 BC/EC only support 128 bit key length */
 	if (keylen_flag == AES_KEYLEN_128)
-		printk(KERN_INFO
-		       "aes_s390: hardware acceleration only available for "
-		       "128 bit keys\n");
+		pr_info("AES hardware acceleration is only available for"
+			" 128-bit keys\n");
 
 	ret = crypto_register_alg(&aes_alg);
 	if (ret)

arch/s390/hypfs/hypfs_diag.c

@@ -3,10 +3,13 @@
  *    Hypervisor filesystem for Linux on s390. Diag 204 and 224
  *    implementation.
  *
- *    Copyright (C) IBM Corp. 2006
+ *    Copyright IBM Corp. 2006, 2008
  *    Author(s): Michael Holzheu <holzheu@de.ibm.com>
  */
 
+#define KMSG_COMPONENT "hypfs"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/types.h>
 #include <linux/errno.h>
 #include <linux/string.h>
@@ -527,13 +530,14 @@ __init int hypfs_diag_init(void)
 	int rc;
 
 	if (diag204_probe()) {
-		printk(KERN_ERR "hypfs: diag 204 not working.");
+		pr_err("The hardware system does not support hypfs\n");
 		return -ENODATA;
 	}
 	rc = diag224_get_name_table();
 	if (rc) {
 		diag204_free_buffer();
-		printk(KERN_ERR "hypfs: could not get name table.\n");
+		pr_err("The hardware system does not provide all "
+		       "functions required by hypfs\n");
 	}
 	return rc;
 }

arch/s390/hypfs/inode.c

@@ -2,10 +2,13 @@
  *  arch/s390/hypfs/inode.c
  *    Hypervisor filesystem for Linux on s390.
  *
- *    Copyright (C) IBM Corp. 2006
+ *    Copyright IBM Corp. 2006, 2008
  *    Author(s): Michael Holzheu <holzheu@de.ibm.com>
  */
 
+#define KMSG_COMPONENT "hypfs"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/types.h>
 #include <linux/errno.h>
 #include <linux/fs.h>
@@ -200,7 +203,7 @@ static ssize_t hypfs_aio_write(struct kiocb *iocb, const struct iovec *iov,
 	else
 		rc = hypfs_diag_create_files(sb, sb->s_root);
 	if (rc) {
-		printk(KERN_ERR "hypfs: Update failed\n");
+		pr_err("Updating the hypfs tree failed\n");
 		hypfs_delete_tree(sb->s_root);
 		goto out;
 	}
@@ -252,8 +255,7 @@ static int hypfs_parse_options(char *options, struct super_block *sb)
 			break;
 		case opt_err:
 		default:
-			printk(KERN_ERR "hypfs: Unrecognized mount option "
-			       "\"%s\" or missing value\n", str);
+			pr_err("%s is not a valid mount option\n", str);
 			return -EINVAL;
 		}
 	}
@@ -317,7 +319,7 @@ static int hypfs_fill_super(struct super_block *sb, void *data, int silent)
 	}
 	hypfs_update_update(sb);
 	sb->s_root = root_dentry;
-	printk(KERN_INFO "hypfs: Hypervisor filesystem mounted\n");
+	pr_info("Hypervisor filesystem mounted\n");
 	return 0;
 
 err_tree:
@@ -513,7 +515,7 @@ fail_sysfs:
 	if (!MACHINE_IS_VM)
 		hypfs_diag_exit();
 fail_diag:
-	printk(KERN_ERR "hypfs: Initialization failed with rc = %i.\n", rc);
+	pr_err("Initialization of hypfs failed with rc=%i\n", rc);
 	return rc;
 }
 

arch/s390/include/asm/auxvec.h

@@ -1,4 +1,6 @@
 #ifndef __ASMS390_AUXVEC_H
 #define __ASMS390_AUXVEC_H
 
+#define AT_SYSINFO_EHDR		33
+
 #endif

arch/s390/include/asm/bug.h

@@ -47,7 +47,10 @@
 
 #endif /* CONFIG_DEBUG_BUGVERBOSE */
 
-#define BUG()	__EMIT_BUG(0)
+#define BUG() do {					\
+	__EMIT_BUG(0);					\
+	for (;;);					\
+} while (0)
 
 #define WARN_ON(x) ({					\
 	int __ret_warn_on = !!(x);			\

arch/s390/include/asm/byteorder.h

@@ -11,32 +11,39 @@
 
 #include <asm/types.h>
 
-#ifdef __GNUC__
+#define __BIG_ENDIAN
+
+#ifndef __s390x__
+# define __SWAB_64_THRU_32__
+#endif
 
 #ifdef __s390x__
-static inline __u64 ___arch__swab64p(const __u64 *x)
+static inline __u64 __arch_swab64p(const __u64 *x)
 {
 	__u64 result;
 
 	asm volatile("lrvg %0,%1" : "=d" (result) : "m" (*x));
 	return result;
 }
+#define __arch_swab64p __arch_swab64p
 
-static inline __u64 ___arch__swab64(__u64 x)
+static inline __u64 __arch_swab64(__u64 x)
 {
 	__u64 result;
 
 	asm volatile("lrvgr %0,%1" : "=d" (result) : "d" (x));
 	return result;
 }
+#define __arch_swab64 __arch_swab64
 
-static inline void ___arch__swab64s(__u64 *x)
+static inline void __arch_swab64s(__u64 *x)
 {
-	*x = ___arch__swab64p(x);
+	*x = __arch_swab64p(x);
 }
+#define __arch_swab64s __arch_swab64s
 #endif /* __s390x__ */
 
-static inline __u32 ___arch__swab32p(const __u32 *x)
+static inline __u32 __arch_swab32p(const __u32 *x)
 {
 	__u32 result;
 	
@@ -53,25 +60,20 @@ static inline __u32 ___arch__swab32p(const __u32 *x)
 #endif /* __s390x__ */
 	return result;
 }
+#define __arch_swab32p __arch_swab32p
 
-static inline __u32 ___arch__swab32(__u32 x)
+#ifdef __s390x__
+static inline __u32 __arch_swab32(__u32 x)
 {
-#ifndef __s390x__
-	return ___arch__swab32p(&x);
-#else /* __s390x__ */
 	__u32 result;
 	
 	asm volatile("lrvr  %0,%1" : "=d" (result) : "d" (x));
 	return result;
-#endif /* __s390x__ */
-}
-
-static __inline__ void ___arch__swab32s(__u32 *x)
-{
-	*x = ___arch__swab32p(x);
 }
+#define __arch_swab32 __arch_swab32
+#endif /* __s390x__ */
 
-static __inline__ __u16 ___arch__swab16p(const __u16 *x)
+static inline __u16 __arch_swab16p(const __u16 *x)
 {
 	__u16 result;
 	
@@ -86,40 +88,8 @@ static __inline__ __u16 ___arch__swab16p(const __u16 *x)
 #endif /* __s390x__ */
 	return result;
 }
+#define __arch_swab16p __arch_swab16p
 
-static __inline__ __u16 ___arch__swab16(__u16 x)
-{
-	return ___arch__swab16p(&x);
-}
-
-static __inline__ void ___arch__swab16s(__u16 *x)
-{
-	*x = ___arch__swab16p(x);
-}
-
-#ifdef __s390x__
-#define __arch__swab64(x) ___arch__swab64(x)
-#define __arch__swab64p(x) ___arch__swab64p(x)
-#define __arch__swab64s(x) ___arch__swab64s(x)
-#endif /* __s390x__ */
-#define __arch__swab32(x) ___arch__swab32(x)
-#define __arch__swab16(x) ___arch__swab16(x)
-#define __arch__swab32p(x) ___arch__swab32p(x)
-#define __arch__swab16p(x) ___arch__swab16p(x)
-#define __arch__swab32s(x) ___arch__swab32s(x)
-#define __arch__swab16s(x) ___arch__swab16s(x)
-
-#ifndef __s390x__
-#if !defined(__STRICT_ANSI__) || defined(__KERNEL__)
-#  define __BYTEORDER_HAS_U64__
-#  define __SWAB_64_THRU_32__
-#endif
-#else /* __s390x__ */
-#define __BYTEORDER_HAS_U64__
-#endif /* __s390x__ */
-
-#endif /* __GNUC__ */
-
-#include <linux/byteorder/big_endian.h>
+#include <linux/byteorder.h>
 
 #endif /* _S390_BYTEORDER_H */

arch/s390/include/asm/elf.h

@@ -120,6 +120,10 @@ typedef s390_compat_regs compat_elf_gregset_t;
 #include <asm/system.h>		/* for save_access_regs */
 #include <asm/mmu_context.h>
 
+#include <asm/vdso.h>
+
+extern unsigned int vdso_enabled;
+
 /*
  * This is used to ensure we don't load something for the wrong architecture.
  */
@@ -191,4 +195,16 @@ do {							\
 	current->mm->context.noexec == 0;		\
 })
 
+#define ARCH_DLINFO							    \
+do {									    \
+	if (vdso_enabled)						    \
+		NEW_AUX_ENT(AT_SYSINFO_EHDR,				    \
+			    (unsigned long)current->mm->context.vdso_base); \
+} while (0)
+
+struct linux_binprm;
+
+#define ARCH_HAS_SETUP_ADDITIONAL_PAGES 1
+int arch_setup_additional_pages(struct linux_binprm *, int);
+
 #endif

arch/s390/include/asm/fcx.h

@@ -248,8 +248,8 @@ struct dcw {
 #define TCCB_MAX_SIZE		(sizeof(struct tccb_tcah) + \
 				 TCCB_MAX_DCW * sizeof(struct dcw) + \
 				 sizeof(struct tccb_tcat))
-#define TCCB_SAC_DEFAULT	0xf901
-#define TCCB_SAC_INTRG		0xf902
+#define TCCB_SAC_DEFAULT	0x1ffe
+#define TCCB_SAC_INTRG		0x1fff
 
 /**
  * struct tccb_tcah - Transport-Command-Area Header (TCAH)

arch/s390/include/asm/ftrace.h

@@ -0,0 +1,8 @@
+#ifndef _ASM_S390_FTRACE_H
+#define _ASM_S390_FTRACE_H
+
+#ifndef __ASSEMBLY__
+extern void _mcount(void);
+#endif
+
+#endif /* _ASM_S390_FTRACE_H */

arch/s390/include/asm/isc.h

@@ -17,6 +17,7 @@
 #define CHSC_SCH_ISC 7			/* CHSC subchannels */
 /* Adapter interrupts. */
 #define QDIO_AIRQ_ISC IO_SCH_ISC	/* I/O subchannel in qdio mode */
+#define AP_ISC 6			/* adjunct processor (crypto) devices */
 
 /* Functions for registration of I/O interruption subclasses */
 void isc_register(unsigned int isc);

arch/s390/include/asm/mmu.h

@@ -6,6 +6,7 @@ typedef struct {
 	struct list_head pgtable_list;
 	unsigned long asce_bits;
 	unsigned long asce_limit;
+	unsigned long vdso_base;
 	int noexec;
 	int has_pgste;	 /* The mmu context has extended page tables */
 	int alloc_pgste; /* cloned contexts will have extended page tables */

arch/s390/include/asm/page.h

@@ -152,4 +152,6 @@ void arch_alloc_page(struct page *page, int order);
 #include <asm-generic/memory_model.h>
 #include <asm-generic/page.h>
 
+#define __HAVE_ARCH_GATE_AREA 1
+
 #endif /* _S390_PAGE_H */

arch/s390/include/asm/pgalloc.h

@@ -28,6 +28,8 @@ void disable_noexec(struct mm_struct *, struct task_struct *);
 
 static inline void clear_table(unsigned long *s, unsigned long val, size_t n)
 {
+	typedef struct { char _[n]; } addrtype;
+
 	*s = val;
 	n = (n / 256) - 1;
 	asm volatile(
@@ -39,7 +41,8 @@ static inline void clear_table(unsigned long *s, unsigned long val, size_t n)
 		"0:	mvc	256(256,%0),0(%0)\n"
 		"	la	%0,256(%0)\n"
 		"	brct	%1,0b\n"
-		: "+a" (s), "+d" (n));
+		: "+a" (s), "+d" (n), "=m" (*(addrtype *) s)
+		: "m" (*(addrtype *) s));
 }
 
 static inline void crst_table_init(unsigned long *crst, unsigned long entry)

arch/s390/include/asm/processor.h

@@ -13,6 +13,7 @@
 #ifndef __ASM_S390_PROCESSOR_H
 #define __ASM_S390_PROCESSOR_H
 
+#include <linux/linkage.h>
 #include <asm/ptrace.h>
 
 #ifdef __KERNEL__
@@ -258,7 +259,7 @@ static inline void enabled_wait(void)
  * Function to drop a processor into disabled wait state
  */
 
-static inline void disabled_wait(unsigned long code)
+static inline void ATTRIB_NORET disabled_wait(unsigned long code)
 {
         unsigned long ctl_buf;
         psw_t dw_psw;
@@ -322,6 +323,7 @@ static inline void disabled_wait(unsigned long code)
 		: "=m" (ctl_buf)
 		: "a" (&dw_psw), "a" (&ctl_buf), "m" (dw_psw) : "cc", "0");
 #endif /* __s390x__ */
+	while (1);
 }
 
 /*

arch/s390/include/asm/qdio.h

@@ -373,16 +373,16 @@ struct qdio_initialize {
 #define QDIO_FLAG_SYNC_OUTPUT		0x02
 #define QDIO_FLAG_PCI_OUT		0x10
 
-extern int qdio_initialize(struct qdio_initialize *init_data);
-extern int qdio_allocate(struct qdio_initialize *init_data);
-extern int qdio_establish(struct qdio_initialize *init_data);
+extern int qdio_initialize(struct qdio_initialize *);
+extern int qdio_allocate(struct qdio_initialize *);
+extern int qdio_establish(struct qdio_initialize *);
 extern int qdio_activate(struct ccw_device *);
 
-extern int do_QDIO(struct ccw_device*, unsigned int flags,
-		   int q_nr, int qidx, int count);
-extern int qdio_cleanup(struct ccw_device*, int how);
-extern int qdio_shutdown(struct ccw_device*, int how);
+extern int do_QDIO(struct ccw_device *cdev, unsigned int callflags,
+		   int q_nr, int bufnr, int count);
+extern int qdio_cleanup(struct ccw_device*, int);
+extern int qdio_shutdown(struct ccw_device*, int);
 extern int qdio_free(struct ccw_device *);
-extern struct qdio_ssqd_desc *qdio_get_ssqd_desc(struct ccw_device *cdev);
+extern int qdio_get_ssqd_desc(struct ccw_device *dev, struct qdio_ssqd_desc*);
 
 #endif /* __QDIO_H__ */

arch/s390/include/asm/sigp.h

@@ -61,6 +61,7 @@ typedef enum
 {
 	ec_schedule=0,
 	ec_call_function,
+	ec_call_function_single,
 	ec_bit_last
 } ec_bit_sig;
 

arch/s390/include/asm/smp.h

@@ -91,8 +91,9 @@ extern int __cpu_up (unsigned int cpu);
 extern struct mutex smp_cpu_state_mutex;
 extern int smp_cpu_polarization[];
 
-extern int smp_call_function_mask(cpumask_t mask, void (*func)(void *),
-	void *info, int wait);
+extern void arch_send_call_function_single_ipi(int cpu);
+extern void arch_send_call_function_ipi(cpumask_t mask);
+
 #endif
 
 #ifndef CONFIG_SMP

arch/s390/include/asm/sysinfo.h

@@ -118,4 +118,15 @@ static inline int stsi(void *sysinfo, int fc, int sel1, int sel2)
 	return r0;
 }
 
+/*
+ * Service level reporting interface.
+ */
+struct service_level {
+	struct list_head list;
+	void (*seq_print)(struct seq_file *, struct service_level *);
+};
+
+int register_service_level(struct service_level *);
+int unregister_service_level(struct service_level *);
+
 #endif /* __ASM_S390_SYSINFO_H */

arch/s390/include/asm/system.h

@@ -12,6 +12,7 @@
 #define __ASM_SYSTEM_H
 
 #include <linux/kernel.h>
+#include <linux/errno.h>
 #include <asm/types.h>
 #include <asm/ptrace.h>
 #include <asm/setup.h>
@@ -98,13 +99,9 @@ static inline void restore_access_regs(unsigned int *acrs)
 	prev = __switch_to(prev,next);					     \
 } while (0)
 
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 extern void account_vtime(struct task_struct *);
 extern void account_tick_vtime(struct task_struct *);
 extern void account_system_vtime(struct task_struct *);
-#else
-#define account_vtime(x) do { /* empty */ } while (0)
-#endif
 
 #ifdef CONFIG_PFAULT
 extern void pfault_irq_init(void);
@@ -413,8 +410,6 @@ __set_psw_mask(unsigned long mask)
 #define local_mcck_enable()  __set_psw_mask(psw_kernel_bits)
 #define local_mcck_disable() __set_psw_mask(psw_kernel_bits & ~PSW_MASK_MCHECK)
 
-int stfle(unsigned long long *list, int doublewords);
-
 #ifdef CONFIG_SMP
 
 extern void smp_ctl_set_bit(int cr, int bit);
@@ -438,6 +433,23 @@ static inline unsigned int stfl(void)
 	return S390_lowcore.stfl_fac_list;
 }
 
+static inline int __stfle(unsigned long long *list, int doublewords)
+{
+	typedef struct { unsigned long long _[doublewords]; } addrtype;
+	register unsigned long __nr asm("0") = doublewords - 1;
+
+	asm volatile(".insn s,0xb2b00000,%0" /* stfle */
+		     : "=m" (*(addrtype *) list), "+d" (__nr) : : "cc");
+	return __nr + 1;
+}
+
+static inline int stfle(unsigned long long *list, int doublewords)
+{
+	if (!(stfl() & (1UL << 24)))
+		return -EOPNOTSUPP;
+	return __stfle(list, doublewords);
+}
+
 static inline unsigned short stap(void)
 {
 	unsigned short cpu_address;

arch/s390/include/asm/timer.h

@@ -48,18 +48,9 @@ extern int del_virt_timer(struct vtimer_list *timer);
 extern void init_cpu_vtimer(void);
 extern void vtime_init(void);
 
-#ifdef CONFIG_VIRT_TIMER
-
 extern void vtime_start_cpu_timer(void);
 extern void vtime_stop_cpu_timer(void);
 
-#else
-
-static inline void vtime_start_cpu_timer(void) { }
-static inline void vtime_stop_cpu_timer(void) { }
-
-#endif /* CONFIG_VIRT_TIMER */
-
 #endif /* __KERNEL__ */
 
 #endif /* _ASM_S390_TIMER_H */

arch/s390/include/asm/vdso.h

@@ -0,0 +1,39 @@
+#ifndef __S390_VDSO_H__
+#define __S390_VDSO_H__
+
+#ifdef __KERNEL__
+
+/* Default link addresses for the vDSOs */
+#define VDSO32_LBASE	0
+#define VDSO64_LBASE	0
+
+#define VDSO_VERSION_STRING	LINUX_2.6.26
+
+#ifndef __ASSEMBLY__
+
+/*
+ * Note about this structure:
+ *
+ * NEVER USE THIS IN USERSPACE CODE DIRECTLY. The layout of this
+ * structure is supposed to be known only to the function in the vdso
+ * itself and may change without notice.
+ */
+
+struct vdso_data {
+	__u64 tb_update_count;		/* Timebase atomicity ctr	0x00 */
+	__u64 xtime_tod_stamp;		/* TOD clock for xtime		0x08 */
+	__u64 xtime_clock_sec;		/* Kernel time			0x10 */
+	__u64 xtime_clock_nsec;		/*				0x18 */
+	__u64 wtom_clock_sec;		/* Wall to monotonic clock	0x20 */
+	__u64 wtom_clock_nsec;		/*				0x28 */
+	__u32 tz_minuteswest;		/* Minutes west of Greenwich	0x30 */
+	__u32 tz_dsttime;		/* Type of dst correction	0x34 */
+};
+
+extern struct vdso_data *vdso_data;
+
+#endif /* __ASSEMBLY__ */
+
+#endif /* __KERNEL__ */
+
+#endif /* __S390_VDSO_H__ */

arch/s390/kernel/Makefile

@@ -2,6 +2,11 @@
 # Makefile for the linux kernel.
 #
 
+ifdef CONFIG_FUNCTION_TRACER
+# Do not trace early boot code
+CFLAGS_REMOVE_early.o = -pg
+endif
+
 #
 # Passing null pointers is ok for smp code, since we access the lowcore here.
 #
@@ -12,9 +17,10 @@ CFLAGS_smp.o	:= -Wno-nonnull
 #
 CFLAGS_ptrace.o		+= -DUTS_MACHINE='"$(UTS_MACHINE)"'
 
-obj-y	:=  bitmap.o traps.o time.o process.o base.o early.o \
-            setup.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o \
-	    s390_ext.o debug.o irq.o ipl.o dis.o diag.o mem_detect.o
+obj-y	:=  bitmap.o traps.o time.o process.o base.o early.o setup.o \
+	    processor.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o \
+	    s390_ext.o debug.o irq.o ipl.o dis.o diag.o mem_detect.o \
+	    vdso.o vtime.o
 
 obj-y	+= $(if $(CONFIG_64BIT),entry64.o,entry.o)
 obj-y	+= $(if $(CONFIG_64BIT),reipl64.o,reipl.o)
@@ -30,12 +36,16 @@ obj-$(CONFIG_COMPAT)		+= compat_linux.o compat_signal.o \
 					compat_wrapper.o compat_exec_domain.o \
 					$(compat-obj-y)
 
-obj-$(CONFIG_VIRT_TIMER)	+= vtime.o
 obj-$(CONFIG_STACKTRACE)	+= stacktrace.o
 obj-$(CONFIG_KPROBES)		+= kprobes.o
+obj-$(CONFIG_FUNCTION_TRACER)	+= mcount.o
 
 # Kexec part
 S390_KEXEC_OBJS := machine_kexec.o crash.o
 S390_KEXEC_OBJS += $(if $(CONFIG_64BIT),relocate_kernel64.o,relocate_kernel.o)
 obj-$(CONFIG_KEXEC) += $(S390_KEXEC_OBJS)
 
+# vdso
+obj-$(CONFIG_64BIT)		+= vdso64/
+obj-$(CONFIG_32BIT)		+= vdso32/
+obj-$(CONFIG_COMPAT)		+= vdso32/

arch/s390/kernel/asm-offsets.c

@@ -6,6 +6,7 @@
 
 #include <linux/sched.h>
 #include <linux/kbuild.h>
+#include <asm/vdso.h>
 
 int main(void)
 {
@@ -38,5 +39,19 @@ int main(void)
 	DEFINE(__SF_BACKCHAIN, offsetof(struct stack_frame, back_chain));
 	DEFINE(__SF_GPRS, offsetof(struct stack_frame, gprs));
 	DEFINE(__SF_EMPTY, offsetof(struct stack_frame, empty1));
+	BLANK();
+	/* timeval/timezone offsets for use by vdso */
+	DEFINE(__VDSO_UPD_COUNT, offsetof(struct vdso_data, tb_update_count));
+	DEFINE(__VDSO_XTIME_STAMP, offsetof(struct vdso_data, xtime_tod_stamp));
+	DEFINE(__VDSO_XTIME_SEC, offsetof(struct vdso_data, xtime_clock_sec));
+	DEFINE(__VDSO_XTIME_NSEC, offsetof(struct vdso_data, xtime_clock_nsec));
+	DEFINE(__VDSO_WTOM_SEC, offsetof(struct vdso_data, wtom_clock_sec));
+	DEFINE(__VDSO_WTOM_NSEC, offsetof(struct vdso_data, wtom_clock_nsec));
+	DEFINE(__VDSO_TIMEZONE, offsetof(struct vdso_data, tz_minuteswest));
+	/* constants used by the vdso */
+	DEFINE(CLOCK_REALTIME, CLOCK_REALTIME);
+	DEFINE(CLOCK_MONOTONIC, CLOCK_MONOTONIC);
+	DEFINE(CLOCK_REALTIME_RES, MONOTONIC_RES_NSEC);
+
 	return 0;
 }

arch/s390/kernel/cpcmd.c

@@ -7,6 +7,9 @@
  *               Christian Borntraeger (cborntra@de.ibm.com),
  */
 
+#define KMSG_COMPONENT "cpcmd"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/slab.h>
@@ -104,8 +107,8 @@ int cpcmd(const char *cmd, char *response, int rlen, int *response_code)
 			(((unsigned long)response + rlen) >> 31)) {
 		lowbuf = kmalloc(rlen, GFP_KERNEL | GFP_DMA);
 		if (!lowbuf) {
-			printk(KERN_WARNING
-				"cpcmd: could not allocate response buffer\n");
+			pr_warning("The cpcmd kernel function failed to "
+				   "allocate a response buffer\n");
 			return -ENOMEM;
 		}
 		spin_lock_irqsave(&cpcmd_lock, flags);

arch/s390/kernel/debug.c

@@ -10,6 +10,9 @@
  *    Bugreports to: <Linux390@de.ibm.com>
  */
 
+#define KMSG_COMPONENT "s390dbf"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/stddef.h>
 #include <linux/kernel.h>
 #include <linux/errno.h>
@@ -388,7 +391,7 @@ debug_info_copy(debug_info_t* in, int mode)
 		debug_info_free(rc);
 	} while (1);
 
-        if(!rc || (mode == NO_AREAS))
+	if (mode == NO_AREAS)
                 goto out;
 
         for(i = 0; i < in->nr_areas; i++){
@@ -693,8 +696,8 @@ debug_info_t *debug_register_mode(const char *name, int pages_per_area,
 	/* Since debugfs currently does not support uid/gid other than root, */
 	/* we do not allow gid/uid != 0 until we get support for that. */
 	if ((uid != 0) || (gid != 0))
-		printk(KERN_WARNING "debug: Warning - Currently only uid/gid "
-		       "= 0 are supported. Using root as owner now!");
+		pr_warning("Root becomes the owner of all s390dbf files "
+			   "in sysfs\n");
 	if (!initialized)
 		BUG();
 	mutex_lock(&debug_mutex);
@@ -709,7 +712,7 @@ debug_info_t *debug_register_mode(const char *name, int pages_per_area,
 	debug_register_view(rc, &debug_pages_view);
 out:
         if (!rc){
-		printk(KERN_ERR "debug: debug_register failed for %s\n",name);
+		pr_err("Registering debug feature %s failed\n", name);
         }
 	mutex_unlock(&debug_mutex);
 	return rc;
@@ -763,8 +766,8 @@ debug_set_size(debug_info_t* id, int nr_areas, int pages_per_area)
 	if(pages_per_area > 0){
 		new_areas = debug_areas_alloc(pages_per_area, nr_areas);
 		if(!new_areas) {
-			printk(KERN_WARNING "debug: could not allocate memory "\
-					 "for pagenumber: %i\n",pages_per_area);
+			pr_info("Allocating memory for %i pages failed\n",
+				pages_per_area);
 			rc = -ENOMEM;
 			goto out;
 		}
@@ -780,8 +783,7 @@ debug_set_size(debug_info_t* id, int nr_areas, int pages_per_area)
 	memset(id->active_entries,0,sizeof(int)*id->nr_areas);
 	memset(id->active_pages, 0, sizeof(int)*id->nr_areas);
 	spin_unlock_irqrestore(&id->lock,flags);
-	printk(KERN_INFO "debug: %s: set new size (%i pages)\n"\
-			 ,id->name, pages_per_area);
+	pr_info("%s: set new size (%i pages)\n" ,id->name, pages_per_area);
 out:
 	return rc;
 }
@@ -800,10 +802,9 @@ debug_set_level(debug_info_t* id, int new_level)
 	spin_lock_irqsave(&id->lock,flags);
         if(new_level == DEBUG_OFF_LEVEL){
                 id->level = DEBUG_OFF_LEVEL;
-                printk(KERN_INFO "debug: %s: switched off\n",id->name);
+		pr_info("%s: switched off\n",id->name);
         } else if ((new_level > DEBUG_MAX_LEVEL) || (new_level < 0)) {
-                printk(KERN_INFO
-                        "debug: %s: level %i is out of range (%i - %i)\n",
+		pr_info("%s: level %i is out of range (%i - %i)\n",
                         id->name, new_level, 0, DEBUG_MAX_LEVEL);
         } else {
                 id->level = new_level;
@@ -1108,8 +1109,8 @@ debug_register_view(debug_info_t * id, struct debug_view *view)
 	pde = debugfs_create_file(view->name, mode, id->debugfs_root_entry,
 				id , &debug_file_ops);
 	if (!pde){
-		printk(KERN_WARNING "debug: debugfs_create_file() failed!"\
-			" Cannot register view %s/%s\n", id->name,view->name);
+		pr_err("Registering view %s/%s failed due to out of "
+		       "memory\n", id->name,view->name);
 		rc = -1;
 		goto out;
 	}
@@ -1119,10 +1120,8 @@ debug_register_view(debug_info_t * id, struct debug_view *view)
 			break;
 	}
 	if (i == DEBUG_MAX_VIEWS) {
-		printk(KERN_WARNING "debug: cannot register view %s/%s\n",
-			id->name,view->name);
-		printk(KERN_WARNING 
-			"debug: maximum number of views reached (%i)!\n", i);
+		pr_err("Registering view %s/%s would exceed the maximum "
+		       "number of views %i\n", id->name, view->name, i);
 		debugfs_remove(pde);
 		rc = -1;
 	} else {
@@ -1303,7 +1302,8 @@ debug_input_level_fn(debug_info_t * id, struct debug_view *view,
 		new_level = debug_get_uint(str);
 	}
 	if(new_level < 0) {
-		printk(KERN_INFO "debug: level `%s` is not valid\n", str);
+		pr_warning("%s is not a valid level for a debug "
+			   "feature\n", str);
 		rc = -EINVAL;
 	} else {
 		debug_set_level(id, new_level);
@@ -1380,7 +1380,8 @@ debug_input_flush_fn(debug_info_t * id, struct debug_view *view,
                 goto out;
         }
 
-        printk(KERN_INFO "debug: area `%c` is not valid\n", input_buf[0]);
+	pr_info("Flushing debug data failed because %c is not a valid "
+		 "area\n", input_buf[0]);
 
 out:
         *offset += user_len;

arch/s390/kernel/entry.S

@@ -109,13 +109,6 @@ STACK_SIZE  = 1 << STACK_SHIFT
  *    R15 - kernel stack pointer
  */
 
-	.macro	STORE_TIMER lc_offset
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
-	stpt	\lc_offset
-#endif
-	.endm
-
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	.macro	UPDATE_VTIME lc_from,lc_to,lc_sum
 	lm	%r10,%r11,\lc_from
 	sl	%r10,\lc_to
@@ -128,7 +121,6 @@ STACK_SIZE  = 1 << STACK_SHIFT
 	al	%r10,BASED(.Lc_1)
 1:	stm	%r10,%r11,\lc_sum
 	.endm
-#endif
 
 	.macro	SAVE_ALL_BASE savearea
 	stm	%r12,%r15,\savearea
@@ -198,7 +190,7 @@ STACK_SIZE  = 1 << STACK_SHIFT
 	ni	\psworg+1,0xfd		# clear wait state bit
 	.endif
 	lm	%r0,%r15,SP_R0(%r15)	# load gprs 0-15 of user
-	STORE_TIMER __LC_EXIT_TIMER
+	stpt	__LC_EXIT_TIMER
 	lpsw	\psworg			# back to caller
 	.endm
 
@@ -247,20 +239,18 @@ __critical_start:
 
 	.globl	system_call
 system_call:
-	STORE_TIMER __LC_SYNC_ENTER_TIMER
+	stpt	__LC_SYNC_ENTER_TIMER
 sysc_saveall:
 	SAVE_ALL_BASE __LC_SAVE_AREA
 	SAVE_ALL_SVC __LC_SVC_OLD_PSW,__LC_SAVE_AREA
 	CREATE_STACK_FRAME __LC_SVC_OLD_PSW,__LC_SAVE_AREA
 	lh	%r7,0x8a	  # get svc number from lowcore
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 sysc_vtime:
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_SYNC_ENTER_TIMER,__LC_USER_TIMER
 sysc_stime:
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 sysc_update:
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER
-#endif
 sysc_do_svc:
 	l	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	ltr	%r7,%r7			# test for svc 0
@@ -436,7 +426,7 @@ ret_from_fork:
 	basr	%r14,%r1
 	TRACE_IRQS_ON
 	stosm	__SF_EMPTY(%r15),0x03	# reenable interrupts
-	b	BASED(sysc_return)
+	b	BASED(sysc_tracenogo)
 
 #
 # kernel_execve function needs to deal with pt_regs that is not
@@ -490,20 +480,18 @@ pgm_check_handler:
  * we just ignore the PER event (FIXME: is there anything we have to do
  * for LPSW?).
  */
-	STORE_TIMER __LC_SYNC_ENTER_TIMER
+	stpt	__LC_SYNC_ENTER_TIMER
 	SAVE_ALL_BASE __LC_SAVE_AREA
 	tm	__LC_PGM_INT_CODE+1,0x80 # check whether we got a per exception
 	bnz	BASED(pgm_per)		# got per exception -> special case
 	SAVE_ALL_SYNC __LC_PGM_OLD_PSW,__LC_SAVE_AREA
 	CREATE_STACK_FRAME __LC_PGM_OLD_PSW,__LC_SAVE_AREA
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	tm	SP_PSW+1(%r15),0x01	# interrupting from user ?
 	bz	BASED(pgm_no_vtime)
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_SYNC_ENTER_TIMER,__LC_USER_TIMER
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER
 pgm_no_vtime:
-#endif
 	l	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	TRACE_IRQS_OFF
 	l	%r3,__LC_PGM_ILC	# load program interruption code
@@ -536,14 +524,12 @@ pgm_per:
 pgm_per_std:
 	SAVE_ALL_SYNC __LC_PGM_OLD_PSW,__LC_SAVE_AREA
 	CREATE_STACK_FRAME __LC_PGM_OLD_PSW,__LC_SAVE_AREA
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	tm	SP_PSW+1(%r15),0x01	# interrupting from user ?
 	bz	BASED(pgm_no_vtime2)
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_SYNC_ENTER_TIMER,__LC_USER_TIMER
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER
 pgm_no_vtime2:
-#endif
 	l	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	TRACE_IRQS_OFF
 	l	%r1,__TI_task(%r9)
@@ -565,11 +551,9 @@ pgm_no_vtime2:
 pgm_svcper:
 	SAVE_ALL_SYNC __LC_SVC_OLD_PSW,__LC_SAVE_AREA
 	CREATE_STACK_FRAME __LC_SVC_OLD_PSW,__LC_SAVE_AREA
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_SYNC_ENTER_TIMER,__LC_USER_TIMER
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER
-#endif
 	lh	%r7,0x8a		# get svc number from lowcore
 	l	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	TRACE_IRQS_OFF
@@ -599,19 +583,17 @@ kernel_per:
 
 	.globl io_int_handler
 io_int_handler:
-	STORE_TIMER __LC_ASYNC_ENTER_TIMER
+	stpt	__LC_ASYNC_ENTER_TIMER
 	stck	__LC_INT_CLOCK
 	SAVE_ALL_BASE __LC_SAVE_AREA+16
 	SAVE_ALL_ASYNC __LC_IO_OLD_PSW,__LC_SAVE_AREA+16
 	CREATE_STACK_FRAME __LC_IO_OLD_PSW,__LC_SAVE_AREA+16
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	tm	SP_PSW+1(%r15),0x01	# interrupting from user ?
 	bz	BASED(io_no_vtime)
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_ASYNC_ENTER_TIMER,__LC_USER_TIMER
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_ASYNC_ENTER_TIMER
 io_no_vtime:
-#endif
 	l	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	TRACE_IRQS_OFF
 	l	%r1,BASED(.Ldo_IRQ)	# load address of do_IRQ
@@ -741,19 +723,17 @@ io_notify_resume:
 
 	.globl	ext_int_handler
 ext_int_handler:
-	STORE_TIMER __LC_ASYNC_ENTER_TIMER
+	stpt	__LC_ASYNC_ENTER_TIMER
 	stck	__LC_INT_CLOCK
 	SAVE_ALL_BASE __LC_SAVE_AREA+16
 	SAVE_ALL_ASYNC __LC_EXT_OLD_PSW,__LC_SAVE_AREA+16
 	CREATE_STACK_FRAME __LC_EXT_OLD_PSW,__LC_SAVE_AREA+16
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	tm	SP_PSW+1(%r15),0x01	# interrupting from user ?
 	bz	BASED(ext_no_vtime)
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_ASYNC_ENTER_TIMER,__LC_USER_TIMER
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_ASYNC_ENTER_TIMER
 ext_no_vtime:
-#endif
 	l	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	TRACE_IRQS_OFF
 	la	%r2,SP_PTREGS(%r15)	# address of register-save area
@@ -776,7 +756,6 @@ mcck_int_handler:
 	la	%r12,__LC_MCK_OLD_PSW
 	tm	__LC_MCCK_CODE,0x80	# system damage?
 	bo	BASED(mcck_int_main)	# yes -> rest of mcck code invalid
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	mvc	__LC_SAVE_AREA+52(8),__LC_ASYNC_ENTER_TIMER
 	mvc	__LC_ASYNC_ENTER_TIMER(8),__LC_CPU_TIMER_SAVE_AREA
 	tm	__LC_MCCK_CODE+5,0x02	# stored cpu timer value valid?
@@ -793,9 +772,7 @@ mcck_int_handler:
 	la	%r14,__LC_LAST_UPDATE_TIMER
 0:	spt	0(%r14)
 	mvc	__LC_ASYNC_ENTER_TIMER(8),0(%r14)
-1:
-#endif
-	tm	__LC_MCCK_CODE+2,0x09	# mwp + ia of old psw valid?
+1:	tm	__LC_MCCK_CODE+2,0x09	# mwp + ia of old psw valid?
 	bno	BASED(mcck_int_main)	# no -> skip cleanup critical
 	tm	__LC_MCK_OLD_PSW+1,0x01	# test problem state bit
 	bnz	BASED(mcck_int_main)	# from user -> load async stack
@@ -812,7 +789,6 @@ mcck_int_main:
 	be	BASED(0f)
 	l	%r15,__LC_PANIC_STACK	# load panic stack
 0:	CREATE_STACK_FRAME __LC_MCK_OLD_PSW,__LC_SAVE_AREA+32
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	tm	__LC_MCCK_CODE+2,0x08	# mwp of old psw valid?
 	bno	BASED(mcck_no_vtime)	# no -> skip cleanup critical
 	tm	SP_PSW+1(%r15),0x01	# interrupting from user ?
@@ -821,7 +797,6 @@ mcck_int_main:
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_ASYNC_ENTER_TIMER
 mcck_no_vtime:
-#endif
 	l	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	la	%r2,SP_PTREGS(%r15)	# load pt_regs
 	l	%r1,BASED(.Ls390_mcck)
@@ -843,16 +818,13 @@ mcck_no_vtime:
 mcck_return:
 	mvc	__LC_RETURN_MCCK_PSW(8),SP_PSW(%r15) # move return PSW
 	ni	__LC_RETURN_MCCK_PSW+1,0xfd # clear wait state bit
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	mvc	__LC_ASYNC_ENTER_TIMER(8),__LC_SAVE_AREA+52
 	tm	__LC_RETURN_MCCK_PSW+1,0x01 # returning to user ?
 	bno	BASED(0f)
 	lm	%r0,%r15,SP_R0(%r15)	# load gprs 0-15
 	stpt	__LC_EXIT_TIMER
 	lpsw	__LC_RETURN_MCCK_PSW	# back to caller
-0:
-#endif
-	lm	%r0,%r15,SP_R0(%r15)	# load gprs 0-15
+0:	lm	%r0,%r15,SP_R0(%r15)	# load gprs 0-15
 	lpsw	__LC_RETURN_MCCK_PSW	# back to caller
 
 	RESTORE_ALL __LC_RETURN_MCCK_PSW,0
@@ -976,13 +948,11 @@ cleanup_system_call:
 	b	BASED(1f)
 0:	la	%r12,__LC_SAVE_AREA+32
 1:
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	clc	__LC_RETURN_PSW+4(4),BASED(cleanup_system_call_insn+4)
 	bh	BASED(0f)
 	mvc	__LC_SYNC_ENTER_TIMER(8),__LC_ASYNC_ENTER_TIMER
 0:	clc	__LC_RETURN_PSW+4(4),BASED(cleanup_system_call_insn+8)
 	bhe	BASED(cleanup_vtime)
-#endif
 	clc	__LC_RETURN_PSW+4(4),BASED(cleanup_system_call_insn)
 	bh	BASED(0f)
 	mvc	__LC_SAVE_AREA(16),0(%r12)
@@ -993,7 +963,6 @@ cleanup_system_call:
 	l	%r12,__LC_SAVE_AREA+48	# argh
 	st	%r15,12(%r12)
 	lh	%r7,0x8a
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 cleanup_vtime:
 	clc	__LC_RETURN_PSW+4(4),BASED(cleanup_system_call_insn+12)
 	bhe	BASED(cleanup_stime)
@@ -1004,18 +973,15 @@ cleanup_stime:
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 cleanup_update:
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER
-#endif
 	mvc	__LC_RETURN_PSW+4(4),BASED(cleanup_table_system_call+4)
 	la	%r12,__LC_RETURN_PSW
 	br	%r14
 cleanup_system_call_insn:
 	.long	sysc_saveall + 0x80000000
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	.long	system_call + 0x80000000
 	.long	sysc_vtime + 0x80000000
 	.long	sysc_stime + 0x80000000
 	.long	sysc_update + 0x80000000
-#endif
 
 cleanup_sysc_return:
 	mvc	__LC_RETURN_PSW(4),0(%r12)
@@ -1026,11 +992,9 @@ cleanup_sysc_return:
 cleanup_sysc_leave:
 	clc	4(4,%r12),BASED(cleanup_sysc_leave_insn)
 	be	BASED(2f)
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	mvc	__LC_EXIT_TIMER(8),__LC_ASYNC_ENTER_TIMER
 	clc	4(4,%r12),BASED(cleanup_sysc_leave_insn+4)
 	be	BASED(2f)
-#endif
 	mvc	__LC_RETURN_PSW(8),SP_PSW(%r15)
 	c	%r12,BASED(.Lmck_old_psw)
 	bne	BASED(0f)
@@ -1043,9 +1007,7 @@ cleanup_sysc_leave:
 	br	%r14
 cleanup_sysc_leave_insn:
 	.long	sysc_done - 4 + 0x80000000
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	.long	sysc_done - 8 + 0x80000000
-#endif
 
 cleanup_io_return:
 	mvc	__LC_RETURN_PSW(4),0(%r12)
@@ -1056,11 +1018,9 @@ cleanup_io_return:
 cleanup_io_leave:
 	clc	4(4,%r12),BASED(cleanup_io_leave_insn)
 	be	BASED(2f)
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	mvc	__LC_EXIT_TIMER(8),__LC_ASYNC_ENTER_TIMER
 	clc	4(4,%r12),BASED(cleanup_io_leave_insn+4)
 	be	BASED(2f)
-#endif
 	mvc	__LC_RETURN_PSW(8),SP_PSW(%r15)
 	c	%r12,BASED(.Lmck_old_psw)
 	bne	BASED(0f)
@@ -1073,9 +1033,7 @@ cleanup_io_leave:
 	br	%r14
 cleanup_io_leave_insn:
 	.long	io_done - 4 + 0x80000000
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	.long	io_done - 8 + 0x80000000
-#endif
 
 /*
  * Integer constants

arch/s390/kernel/entry64.S

@@ -96,20 +96,12 @@ _TIF_WORK_INT = (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_NEED_RESCHED | \
 #define LOCKDEP_SYS_EXIT
 #endif
 
-	.macro	STORE_TIMER lc_offset
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
-	stpt	\lc_offset
-#endif
-	.endm
-
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	.macro	UPDATE_VTIME lc_from,lc_to,lc_sum
 	lg	%r10,\lc_from
 	slg	%r10,\lc_to
 	alg	%r10,\lc_sum
 	stg	%r10,\lc_sum
 	.endm
-#endif
 
 /*
  * Register usage in interrupt handlers:
@@ -186,7 +178,7 @@ _TIF_WORK_INT = (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_NEED_RESCHED | \
 	ni	\psworg+1,0xfd		# clear wait state bit
 	.endif
 	lmg	%r0,%r15,SP_R0(%r15)	# load gprs 0-15 of user
-	STORE_TIMER __LC_EXIT_TIMER
+	stpt	__LC_EXIT_TIMER
 	lpswe	\psworg			# back to caller
 	.endm
 
@@ -233,20 +225,18 @@ __critical_start:
 
 	.globl	system_call
 system_call:
-	STORE_TIMER __LC_SYNC_ENTER_TIMER
+	stpt	__LC_SYNC_ENTER_TIMER
 sysc_saveall:
 	SAVE_ALL_BASE __LC_SAVE_AREA
 	SAVE_ALL_SVC __LC_SVC_OLD_PSW,__LC_SAVE_AREA
 	CREATE_STACK_FRAME __LC_SVC_OLD_PSW,__LC_SAVE_AREA
 	llgh	%r7,__LC_SVC_INT_CODE	# get svc number from lowcore
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 sysc_vtime:
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_SYNC_ENTER_TIMER,__LC_USER_TIMER
 sysc_stime:
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 sysc_update:
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER
-#endif
 sysc_do_svc:
 	lg	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	ltgr	%r7,%r7		# test for svc 0
@@ -417,7 +407,7 @@ ret_from_fork:
 0:	brasl	%r14,schedule_tail
 	TRACE_IRQS_ON
 	stosm	24(%r15),0x03		# reenable interrupts
-	j	sysc_return
+	j	sysc_tracenogo
 
 #
 # kernel_execve function needs to deal with pt_regs that is not
@@ -469,20 +459,18 @@ pgm_check_handler:
  * we just ignore the PER event (FIXME: is there anything we have to do
  * for LPSW?).
  */
-	STORE_TIMER __LC_SYNC_ENTER_TIMER
+	stpt	__LC_SYNC_ENTER_TIMER
 	SAVE_ALL_BASE __LC_SAVE_AREA
 	tm	__LC_PGM_INT_CODE+1,0x80 # check whether we got a per exception
 	jnz	pgm_per 		 # got per exception -> special case
 	SAVE_ALL_SYNC __LC_PGM_OLD_PSW,__LC_SAVE_AREA
 	CREATE_STACK_FRAME __LC_PGM_OLD_PSW,__LC_SAVE_AREA
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	tm	SP_PSW+1(%r15),0x01	# interrupting from user ?
 	jz	pgm_no_vtime
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_SYNC_ENTER_TIMER,__LC_USER_TIMER
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER
 pgm_no_vtime:
-#endif
 	lg	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	mvc	SP_ARGS(8,%r15),__LC_LAST_BREAK
 	TRACE_IRQS_OFF
@@ -516,14 +504,12 @@ pgm_per:
 pgm_per_std:
 	SAVE_ALL_SYNC __LC_PGM_OLD_PSW,__LC_SAVE_AREA
 	CREATE_STACK_FRAME __LC_PGM_OLD_PSW,__LC_SAVE_AREA
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	tm	SP_PSW+1(%r15),0x01	# interrupting from user ?
 	jz	pgm_no_vtime2
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_SYNC_ENTER_TIMER,__LC_USER_TIMER
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER
 pgm_no_vtime2:
-#endif
 	lg	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	TRACE_IRQS_OFF
 	lg	%r1,__TI_task(%r9)
@@ -545,11 +531,9 @@ pgm_no_vtime2:
 pgm_svcper:
 	SAVE_ALL_SYNC __LC_SVC_OLD_PSW,__LC_SAVE_AREA
 	CREATE_STACK_FRAME __LC_SVC_OLD_PSW,__LC_SAVE_AREA
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_SYNC_ENTER_TIMER,__LC_USER_TIMER
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER
-#endif
 	llgh	%r7,__LC_SVC_INT_CODE	# get svc number from lowcore
 	lg	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	lg	%r1,__TI_task(%r9)
@@ -575,19 +559,17 @@ kernel_per:
  */
 	.globl io_int_handler
 io_int_handler:
-	STORE_TIMER __LC_ASYNC_ENTER_TIMER
+	stpt	__LC_ASYNC_ENTER_TIMER
 	stck	__LC_INT_CLOCK
 	SAVE_ALL_BASE __LC_SAVE_AREA+32
 	SAVE_ALL_ASYNC __LC_IO_OLD_PSW,__LC_SAVE_AREA+32
 	CREATE_STACK_FRAME __LC_IO_OLD_PSW,__LC_SAVE_AREA+32
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	tm	SP_PSW+1(%r15),0x01	# interrupting from user ?
 	jz	io_no_vtime
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_ASYNC_ENTER_TIMER,__LC_USER_TIMER
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_ASYNC_ENTER_TIMER
 io_no_vtime:
-#endif
 	lg	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	TRACE_IRQS_OFF
 	la	%r2,SP_PTREGS(%r15)	# address of register-save area
@@ -739,19 +721,17 @@ io_notify_resume:
  */
 	.globl	ext_int_handler
 ext_int_handler:
-	STORE_TIMER __LC_ASYNC_ENTER_TIMER
+	stpt	__LC_ASYNC_ENTER_TIMER
 	stck	__LC_INT_CLOCK
 	SAVE_ALL_BASE __LC_SAVE_AREA+32
 	SAVE_ALL_ASYNC __LC_EXT_OLD_PSW,__LC_SAVE_AREA+32
 	CREATE_STACK_FRAME __LC_EXT_OLD_PSW,__LC_SAVE_AREA+32
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	tm	SP_PSW+1(%r15),0x01	# interrupting from user ?
 	jz	ext_no_vtime
 	UPDATE_VTIME __LC_EXIT_TIMER,__LC_ASYNC_ENTER_TIMER,__LC_USER_TIMER
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_ASYNC_ENTER_TIMER
 ext_no_vtime:
-#endif
 	lg	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	TRACE_IRQS_OFF
 	la	%r2,SP_PTREGS(%r15)	# address of register-save area
@@ -773,7 +753,6 @@ mcck_int_handler:
 	la	%r12,__LC_MCK_OLD_PSW
 	tm	__LC_MCCK_CODE,0x80	# system damage?
 	jo	mcck_int_main		# yes -> rest of mcck code invalid
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	la	%r14,4095
 	mvc	__LC_SAVE_AREA+104(8),__LC_ASYNC_ENTER_TIMER
 	mvc	__LC_ASYNC_ENTER_TIMER(8),__LC_CPU_TIMER_SAVE_AREA-4095(%r14)
@@ -791,9 +770,7 @@ mcck_int_handler:
 	la	%r14,__LC_LAST_UPDATE_TIMER
 0:	spt	0(%r14)
 	mvc	__LC_ASYNC_ENTER_TIMER(8),0(%r14)
-1:
-#endif
-	tm	__LC_MCCK_CODE+2,0x09	# mwp + ia of old psw valid?
+1:	tm	__LC_MCCK_CODE+2,0x09	# mwp + ia of old psw valid?
 	jno	mcck_int_main		# no -> skip cleanup critical
 	tm	__LC_MCK_OLD_PSW+1,0x01 # test problem state bit
 	jnz	mcck_int_main		# from user -> load kernel stack
@@ -809,7 +786,6 @@ mcck_int_main:
 	jz	0f
 	lg	%r15,__LC_PANIC_STACK	# load panic stack
 0:	CREATE_STACK_FRAME __LC_MCK_OLD_PSW,__LC_SAVE_AREA+64
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	tm	__LC_MCCK_CODE+2,0x08	# mwp of old psw valid?
 	jno	mcck_no_vtime		# no -> no timer update
 	tm	SP_PSW+1(%r15),0x01	# interrupting from user ?
@@ -818,7 +794,6 @@ mcck_int_main:
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_ASYNC_ENTER_TIMER
 mcck_no_vtime:
-#endif
 	lg	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	la	%r2,SP_PTREGS(%r15)	# load pt_regs
 	brasl	%r14,s390_do_machine_check
@@ -839,14 +814,11 @@ mcck_return:
 	mvc	__LC_RETURN_MCCK_PSW(16),SP_PSW(%r15) # move return PSW
 	ni	__LC_RETURN_MCCK_PSW+1,0xfd # clear wait state bit
 	lmg	%r0,%r15,SP_R0(%r15)	# load gprs 0-15
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	mvc	__LC_ASYNC_ENTER_TIMER(8),__LC_SAVE_AREA+104
 	tm	__LC_RETURN_MCCK_PSW+1,0x01 # returning to user ?
 	jno	0f
 	stpt	__LC_EXIT_TIMER
-0:
-#endif
-	lpswe	__LC_RETURN_MCCK_PSW	# back to caller
+0:	lpswe	__LC_RETURN_MCCK_PSW	# back to caller
 
 /*
  * Restart interruption handler, kick starter for additional CPUs
@@ -964,13 +936,11 @@ cleanup_system_call:
 	j	1f
 0:	la	%r12,__LC_SAVE_AREA+64
 1:
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	clc	__LC_RETURN_PSW+8(8),BASED(cleanup_system_call_insn+8)
 	jh	0f
 	mvc	__LC_SYNC_ENTER_TIMER(8),__LC_ASYNC_ENTER_TIMER
 0:	clc	__LC_RETURN_PSW+8(8),BASED(cleanup_system_call_insn+16)
 	jhe	cleanup_vtime
-#endif
 	clc	__LC_RETURN_PSW+8(8),BASED(cleanup_system_call_insn)
 	jh	0f
 	mvc	__LC_SAVE_AREA(32),0(%r12)
@@ -981,7 +951,6 @@ cleanup_system_call:
 	lg	%r12,__LC_SAVE_AREA+96	# argh
 	stg	%r15,24(%r12)
 	llgh	%r7,__LC_SVC_INT_CODE
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 cleanup_vtime:
 	clc	__LC_RETURN_PSW+8(8),BASED(cleanup_system_call_insn+24)
 	jhe	cleanup_stime
@@ -992,18 +961,15 @@ cleanup_stime:
 	UPDATE_VTIME __LC_LAST_UPDATE_TIMER,__LC_EXIT_TIMER,__LC_SYSTEM_TIMER
 cleanup_update:
 	mvc	__LC_LAST_UPDATE_TIMER(8),__LC_SYNC_ENTER_TIMER
-#endif
 	mvc	__LC_RETURN_PSW+8(8),BASED(cleanup_table_system_call+8)
 	la	%r12,__LC_RETURN_PSW
 	br	%r14
 cleanup_system_call_insn:
 	.quad	sysc_saveall
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	.quad	system_call
 	.quad	sysc_vtime
 	.quad	sysc_stime
 	.quad	sysc_update
-#endif
 
 cleanup_sysc_return:
 	mvc	__LC_RETURN_PSW(8),0(%r12)
@@ -1014,11 +980,9 @@ cleanup_sysc_return:
 cleanup_sysc_leave:
 	clc	8(8,%r12),BASED(cleanup_sysc_leave_insn)
 	je	2f
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	mvc	__LC_EXIT_TIMER(8),__LC_ASYNC_ENTER_TIMER
 	clc	8(8,%r12),BASED(cleanup_sysc_leave_insn+8)
 	je	2f
-#endif
 	mvc	__LC_RETURN_PSW(16),SP_PSW(%r15)
 	cghi	%r12,__LC_MCK_OLD_PSW
 	jne	0f
@@ -1031,9 +995,7 @@ cleanup_sysc_leave:
 	br	%r14
 cleanup_sysc_leave_insn:
 	.quad	sysc_done - 4
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	.quad	sysc_done - 8
-#endif
 
 cleanup_io_return:
 	mvc	__LC_RETURN_PSW(8),0(%r12)
@@ -1044,11 +1006,9 @@ cleanup_io_return:
 cleanup_io_leave:
 	clc	8(8,%r12),BASED(cleanup_io_leave_insn)
 	je	2f
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	mvc	__LC_EXIT_TIMER(8),__LC_ASYNC_ENTER_TIMER
 	clc	8(8,%r12),BASED(cleanup_io_leave_insn+8)
 	je	2f
-#endif
 	mvc	__LC_RETURN_PSW(16),SP_PSW(%r15)
 	cghi	%r12,__LC_MCK_OLD_PSW
 	jne	0f
@@ -1061,9 +1021,7 @@ cleanup_io_leave:
 	br	%r14
 cleanup_io_leave_insn:
 	.quad	io_done - 4
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	.quad	io_done - 8
-#endif
 
 /*
  * Integer constants

arch/s390/kernel/head.S

@@ -461,6 +461,55 @@ start:
 	.byte 0xf0,0xf1,0xf2,0xf3,0xf4,0xf5,0xf6,0xf7
 	.byte 0xf8,0xf9,0xfa,0xfb,0xfc,0xfd,0xfe,0xff
 
+#
+# startup-code at 0x10000, running in absolute addressing mode
+# this is called either by the ipl loader or directly by PSW restart
+# or linload or SALIPL
+#
+	.org	0x10000
+startup:basr	%r13,0			# get base
+.LPG0:
+
+#ifndef CONFIG_MARCH_G5
+	# check processor version against MARCH_{G5,Z900,Z990,Z9_109,Z10}
+	stidp	__LC_CPUID		# store cpuid
+	lhi	%r0,(3f-2f) / 2
+	la	%r1,2f-.LPG0(%r13)
+0:	clc	__LC_CPUID+4(2),0(%r1)
+	jne	3f
+	lpsw	1f-.LPG0(13)		# machine type not good enough, crash
+	.align 16
+1:	.long	0x000a0000,0x00000000
+2:
+#if defined(CONFIG_MARCH_Z10)
+	.short 0x9672, 0x2064, 0x2066, 0x2084, 0x2086, 0x2094, 0x2096
+#elif defined(CONFIG_MARCH_Z9_109)
+	.short 0x9672, 0x2064, 0x2066, 0x2084, 0x2086
+#elif defined(CONFIG_MARCH_Z990)
+	.short 0x9672, 0x2064, 0x2066
+#elif defined(CONFIG_MARCH_Z900)
+	.short 0x9672
+#endif
+3:	la	%r1,2(%r1)
+	brct	%r0,0b
+#endif
+
+	l	%r13,0f-.LPG0(%r13)
+	b	0(%r13)
+0:	.long	startup_continue
+
+#
+# params at 10400 (setup.h)
+#
+	.org	PARMAREA
+	.long	0,0			# IPL_DEVICE
+	.long	0,0			# INITRD_START
+	.long	0,0			# INITRD_SIZE
+
+	.org	COMMAND_LINE
+	.byte	"root=/dev/ram0 ro"
+	.byte	0
+
 #ifdef CONFIG_64BIT
 #include "head64.S"
 #else

arch/s390/kernel/head31.S

@@ -10,34 +10,13 @@
  *
  */
 
-#
-# startup-code at 0x10000, running in absolute addressing mode
-# this is called either by the ipl loader or directly by PSW restart
-# or linload or SALIPL
-#
-	.org	0x10000
-startup:basr	%r13,0			# get base
-.LPG0:	l	%r13,0f-.LPG0(%r13)
-	b	0(%r13)
-0:	.long	startup_continue
-
-#
-# params at 10400 (setup.h)
-#
-	.org	PARMAREA
-	.long	0,0			# IPL_DEVICE
-	.long	0,0			# INITRD_START
-	.long	0,0			# INITRD_SIZE
-
-	.org	COMMAND_LINE
-	.byte	"root=/dev/ram0 ro"
-	.byte	0
-
 	.org	0x11000
 
 startup_continue:
 	basr	%r13,0			# get base
-.LPG1:	mvi	__LC_AR_MODE_ID,0	# set ESA flag (mode 0)
+.LPG1:
+
+	mvi	__LC_AR_MODE_ID,0	# set ESA flag (mode 0)
 	lctl	%c0,%c15,.Lctl-.LPG1(%r13) # load control registers
 	l	%r12,.Lparmaddr-.LPG1(%r13) # pointer to parameter area
 					# move IPL device to lowcore
@@ -50,7 +29,6 @@ startup_continue:
 	ahi	%r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union+THREAD_SIZE
 	st	%r15,__LC_KERNEL_STACK	# set end of kernel stack
 	ahi	%r15,-96
-	xc	__SF_BACKCHAIN(4,%r15),__SF_BACKCHAIN(%r15) # clear backchain
 #
 # Save ipl parameters, clear bss memory, initialize storage key for kernel pages,
 # and create a kernel NSS if the SAVESYS= parm is defined

arch/s390/kernel/head64.S

@@ -10,29 +10,6 @@
  *
  */
 
-#
-# startup-code at 0x10000, running in absolute addressing mode
-# this is called either by the ipl loader or directly by PSW restart
-# or linload or SALIPL
-#
-	.org	0x10000
-startup:basr	%r13,0			# get base
-.LPG0:	l	%r13,0f-.LPG0(%r13)
-	b	0(%r13)
-0:	.long	startup_continue
-
-#
-# params at 10400 (setup.h)
-#
-	.org	PARMAREA
-	.quad	0			# IPL_DEVICE
-	.quad	0			# INITRD_START
-	.quad	0			# INITRD_SIZE
-
-	.org	COMMAND_LINE
-	.byte	"root=/dev/ram0 ro"
-	.byte	0
-
 	.org	0x11000
 
 startup_continue:
@@ -119,7 +96,6 @@ startup_continue:
 	aghi	%r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union + THREAD_SIZE
 	stg	%r15,__LC_KERNEL_STACK	# set end of kernel stack
 	aghi	%r15,-160
-	xc	__SF_BACKCHAIN(4,%r15),__SF_BACKCHAIN(%r15) # clear backchain
 #
 # Save ipl parameters, clear bss memory, initialize storage key for kernel pages,
 # and create a kernel NSS if the SAVESYS= parm is defined

arch/s390/kernel/mcount.S

@@ -0,0 +1,56 @@
+/*
+ * Copyright IBM Corp. 2008
+ *
+ *   Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>,
+ *
+ */
+
+#ifndef CONFIG_64BIT
+.globl _mcount
+_mcount:
+	stm	%r0,%r5,8(%r15)
+	st	%r14,56(%r15)
+	lr	%r1,%r15
+	ahi	%r15,-96
+	l	%r3,100(%r15)
+	la	%r2,0(%r14)
+	st	%r1,0(%r15)
+	la	%r3,0(%r3)
+	bras	%r14,0f
+	.long	ftrace_trace_function
+0:	l	%r14,0(%r14)
+	l	%r14,0(%r14)
+	basr	%r14,%r14
+	ahi	%r15,96
+	lm	%r0,%r5,8(%r15)
+	l	%r14,56(%r15)
+	br	%r14
+
+.globl ftrace_stub
+ftrace_stub:
+	br	%r14
+
+#else /* CONFIG_64BIT */
+
+.globl _mcount
+_mcount:
+	stmg	%r0,%r5,16(%r15)
+	stg	%r14,112(%r15)
+	lgr	%r1,%r15
+	aghi	%r15,-160
+	stg	%r1,0(%r15)
+	lgr	%r2,%r14
+	lg	%r3,168(%r15)
+	larl	%r14,ftrace_trace_function
+	lg	%r14,0(%r14)
+	basr	%r14,%r14
+	aghi	%r15,160
+	lmg	%r0,%r5,16(%r15)
+	lg	%r14,112(%r15)
+	br	%r14
+
+.globl ftrace_stub
+ftrace_stub:
+	br	%r14
+
+#endif /* CONFIG_64BIT */

arch/s390/kernel/processor.c

@@ -0,0 +1,98 @@
+/*
+ *  arch/s390/kernel/processor.c
+ *
+ *  Copyright IBM Corp. 2008
+ *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
+ */
+
+#define KMSG_COMPONENT "cpu"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/smp.h>
+#include <linux/seq_file.h>
+#include <linux/delay.h>
+
+#include <asm/elf.h>
+#include <asm/lowcore.h>
+#include <asm/param.h>
+
+void __cpuinit print_cpu_info(struct cpuinfo_S390 *cpuinfo)
+{
+	pr_info("Processor %d started, address %d, identification %06X\n",
+		cpuinfo->cpu_nr, cpuinfo->cpu_addr, cpuinfo->cpu_id.ident);
+}
+
+/*
+ * show_cpuinfo - Get information on one CPU for use by procfs.
+ */
+
+static int show_cpuinfo(struct seq_file *m, void *v)
+{
+	static const char *hwcap_str[8] = {
+		"esan3", "zarch", "stfle", "msa", "ldisp", "eimm", "dfp",
+		"edat"
+	};
+       struct cpuinfo_S390 *cpuinfo;
+       unsigned long n = (unsigned long) v - 1;
+       int i;
+
+       s390_adjust_jiffies();
+       preempt_disable();
+       if (!n) {
+	       seq_printf(m, "vendor_id       : IBM/S390\n"
+			  "# processors    : %i\n"
+			  "bogomips per cpu: %lu.%02lu\n",
+			  num_online_cpus(), loops_per_jiffy/(500000/HZ),
+			  (loops_per_jiffy/(5000/HZ))%100);
+	       seq_puts(m, "features\t: ");
+	       for (i = 0; i < 8; i++)
+		       if (hwcap_str[i] && (elf_hwcap & (1UL << i)))
+			       seq_printf(m, "%s ", hwcap_str[i]);
+	       seq_puts(m, "\n");
+       }
+
+       if (cpu_online(n)) {
+#ifdef CONFIG_SMP
+	       if (smp_processor_id() == n)
+		       cpuinfo = &S390_lowcore.cpu_data;
+	       else
+		       cpuinfo = &lowcore_ptr[n]->cpu_data;
+#else
+	       cpuinfo = &S390_lowcore.cpu_data;
+#endif
+	       seq_printf(m, "processor %li: "
+			  "version = %02X,  "
+			  "identification = %06X,  "
+			  "machine = %04X\n",
+			  n, cpuinfo->cpu_id.version,
+			  cpuinfo->cpu_id.ident,
+			  cpuinfo->cpu_id.machine);
+       }
+       preempt_enable();
+       return 0;
+}
+
+static void *c_start(struct seq_file *m, loff_t *pos)
+{
+	return *pos < NR_CPUS ? (void *)((unsigned long) *pos + 1) : NULL;
+}
+
+static void *c_next(struct seq_file *m, void *v, loff_t *pos)
+{
+	++*pos;
+	return c_start(m, pos);
+}
+
+static void c_stop(struct seq_file *m, void *v)
+{
+}
+
+const struct seq_operations cpuinfo_op = {
+	.start	= c_start,
+	.next	= c_next,
+	.stop	= c_stop,
+	.show	= show_cpuinfo,
+};
+

arch/s390/kernel/ptrace.c

@@ -204,7 +204,6 @@ static unsigned long __peek_user(struct task_struct *child, addr_t addr)
 static int
 peek_user(struct task_struct *child, addr_t addr, addr_t data)
 {
-	struct user *dummy = NULL;
 	addr_t tmp, mask;
 
 	/*
@@ -213,8 +212,8 @@ peek_user(struct task_struct *child, addr_t addr, addr_t data)
 	 */
 	mask = __ADDR_MASK;
 #ifdef CONFIG_64BIT
-	if (addr >= (addr_t) &dummy->regs.acrs &&
-	    addr < (addr_t) &dummy->regs.orig_gpr2)
+	if (addr >= (addr_t) &((struct user *) NULL)->regs.acrs &&
+	    addr < (addr_t) &((struct user *) NULL)->regs.orig_gpr2)
 		mask = 3;
 #endif
 	if ((addr & mask) || addr > sizeof(struct user) - __ADDR_MASK)
@@ -312,7 +311,6 @@ static int __poke_user(struct task_struct *child, addr_t addr, addr_t data)
 static int
 poke_user(struct task_struct *child, addr_t addr, addr_t data)
 {
-	struct user *dummy = NULL;
 	addr_t mask;
 
 	/*
@@ -321,8 +319,8 @@ poke_user(struct task_struct *child, addr_t addr, addr_t data)
 	 */
 	mask = __ADDR_MASK;
 #ifdef CONFIG_64BIT
-	if (addr >= (addr_t) &dummy->regs.acrs &&
-	    addr < (addr_t) &dummy->regs.orig_gpr2)
+	if (addr >= (addr_t) &((struct user *) NULL)->regs.acrs &&
+	    addr < (addr_t) &((struct user *) NULL)->regs.orig_gpr2)
 		mask = 3;
 #endif
 	if ((addr & mask) || addr > sizeof(struct user) - __ADDR_MASK)

arch/s390/kernel/s390_ksyms.c

@@ -14,6 +14,7 @@
 #include <asm/delay.h>
 #include <asm/pgalloc.h>
 #include <asm/setup.h>
+#include <asm/ftrace.h>
 #ifdef CONFIG_IP_MULTICAST
 #include <net/arp.h>
 #endif
@@ -43,3 +44,7 @@ EXPORT_SYMBOL(csum_fold);
 EXPORT_SYMBOL(console_mode);
 EXPORT_SYMBOL(console_devno);
 EXPORT_SYMBOL(console_irq);
+
+#ifdef CONFIG_FUNCTION_TRACER
+EXPORT_SYMBOL(_mcount);
+#endif

arch/s390/kernel/setup.c

@@ -14,6 +14,9 @@
  * This file handles the architecture-dependent parts of initialization
  */
 
+#define KMSG_COMPONENT "setup"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/errno.h>
 #include <linux/module.h>
 #include <linux/sched.h>
@@ -32,7 +35,6 @@
 #include <linux/bootmem.h>
 #include <linux/root_dev.h>
 #include <linux/console.h>
-#include <linux/seq_file.h>
 #include <linux/kernel_stat.h>
 #include <linux/device.h>
 #include <linux/notifier.h>
@@ -291,8 +293,8 @@ unsigned int switch_amode = 0;
 #endif
 EXPORT_SYMBOL_GPL(switch_amode);
 
-static void set_amode_and_uaccess(unsigned long user_amode,
-				  unsigned long user32_amode)
+static int set_amode_and_uaccess(unsigned long user_amode,
+				 unsigned long user32_amode)
 {
 	psw_user_bits = PSW_BASE_BITS | PSW_MASK_DAT | user_amode |
 			PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK |
@@ -309,11 +311,11 @@ static void set_amode_and_uaccess(unsigned long user_amode,
 			  PSW_MASK_MCHECK | PSW_DEFAULT_KEY;
 
 	if (MACHINE_HAS_MVCOS) {
-		printk("mvcos available.\n");
 		memcpy(&uaccess, &uaccess_mvcos_switch, sizeof(uaccess));
+		return 1;
 	} else {
-		printk("mvcos not available.\n");
 		memcpy(&uaccess, &uaccess_pt, sizeof(uaccess));
+		return 0;
 	}
 }
 
@@ -328,9 +330,10 @@ static int __init early_parse_switch_amode(char *p)
 early_param("switch_amode", early_parse_switch_amode);
 
 #else /* CONFIG_S390_SWITCH_AMODE */
-static inline void set_amode_and_uaccess(unsigned long user_amode,
-					 unsigned long user32_amode)
+static inline int set_amode_and_uaccess(unsigned long user_amode,
+					unsigned long user32_amode)
 {
+	return 0;
 }
 #endif /* CONFIG_S390_SWITCH_AMODE */
 
@@ -355,11 +358,20 @@ early_param("noexec", early_parse_noexec);
 static void setup_addressing_mode(void)
 {
 	if (s390_noexec) {
-		printk("S390 execute protection active, ");
-		set_amode_and_uaccess(PSW_ASC_SECONDARY, PSW32_ASC_SECONDARY);
+		if (set_amode_and_uaccess(PSW_ASC_SECONDARY,
+					  PSW32_ASC_SECONDARY))
+			pr_info("Execute protection active, "
+				"mvcos available\n");
+		else
+			pr_info("Execute protection active, "
+				"mvcos not available\n");
 	} else if (switch_amode) {
-		printk("S390 address spaces switched, ");
-		set_amode_and_uaccess(PSW_ASC_PRIMARY, PSW32_ASC_PRIMARY);
+		if (set_amode_and_uaccess(PSW_ASC_PRIMARY, PSW32_ASC_PRIMARY))
+			pr_info("Address spaces switched, "
+				"mvcos available\n");
+		else
+			pr_info("Address spaces switched, "
+				"mvcos not available\n");
 	}
 #ifdef CONFIG_TRACE_IRQFLAGS
 	sysc_restore_trace_psw.mask = psw_kernel_bits & ~PSW_MASK_MCHECK;
@@ -572,15 +584,15 @@ setup_memory(void)
 			start = PFN_PHYS(start_pfn) + bmap_size + PAGE_SIZE;
 
 			if (start + INITRD_SIZE > memory_end) {
-				printk("initrd extends beyond end of memory "
-				       "(0x%08lx > 0x%08lx)\n"
+				pr_err("initrd extends beyond end of "
+				       "memory (0x%08lx > 0x%08lx) "
 				       "disabling initrd\n",
 				       start + INITRD_SIZE, memory_end);
 				INITRD_START = INITRD_SIZE = 0;
 			} else {
-				printk("Moving initrd (0x%08lx -> 0x%08lx, "
-				       "size: %ld)\n",
-				       INITRD_START, start, INITRD_SIZE);
+				pr_info("Moving initrd (0x%08lx -> "
+					"0x%08lx, size: %ld)\n",
+					INITRD_START, start, INITRD_SIZE);
 				memmove((void *) start, (void *) INITRD_START,
 					INITRD_SIZE);
 				INITRD_START = start;
@@ -642,8 +654,9 @@ setup_memory(void)
 			initrd_start = INITRD_START;
 			initrd_end = initrd_start + INITRD_SIZE;
 		} else {
-			printk("initrd extends beyond end of memory "
-			       "(0x%08lx > 0x%08lx)\ndisabling initrd\n",
+			pr_err("initrd extends beyond end of "
+			       "memory (0x%08lx > 0x%08lx) "
+			       "disabling initrd\n",
 			       initrd_start + INITRD_SIZE, memory_end);
 			initrd_start = initrd_end = 0;
 		}
@@ -651,23 +664,6 @@ setup_memory(void)
 #endif
 }
 
-static int __init __stfle(unsigned long long *list, int doublewords)
-{
-	typedef struct { unsigned long long _[doublewords]; } addrtype;
-	register unsigned long __nr asm("0") = doublewords - 1;
-
-	asm volatile(".insn s,0xb2b00000,%0" /* stfle */
-		     : "=m" (*(addrtype *) list), "+d" (__nr) : : "cc");
-	return __nr + 1;
-}
-
-int __init stfle(unsigned long long *list, int doublewords)
-{
-	if (!(stfl() & (1UL << 24)))
-		return -EOPNOTSUPP;
-	return __stfle(list, doublewords);
-}
-
 /*
  * Setup hardware capabilities.
  */
@@ -739,8 +735,13 @@ static void __init setup_hwcaps(void)
 		strcpy(elf_platform, "z990");
 		break;
 	case 0x2094:
+	case 0x2096:
 		strcpy(elf_platform, "z9-109");
 		break;
+	case 0x2097:
+	case 0x2098:
+		strcpy(elf_platform, "z10");
+		break;
 	}
 }
 
@@ -752,25 +753,34 @@ static void __init setup_hwcaps(void)
 void __init
 setup_arch(char **cmdline_p)
 {
+	/* set up preferred console */
+	add_preferred_console("ttyS", 0, NULL);
+
         /*
          * print what head.S has found out about the machine
          */
 #ifndef CONFIG_64BIT
-	printk((MACHINE_IS_VM) ?
-	       "We are running under VM (31 bit mode)\n" :
-	       "We are running native (31 bit mode)\n");
-	printk((MACHINE_HAS_IEEE) ?
-	       "This machine has an IEEE fpu\n" :
-	       "This machine has no IEEE fpu\n");
+	if (MACHINE_IS_VM)
+		pr_info("Linux is running as a z/VM "
+			"guest operating system in 31-bit mode\n");
+	else
+		pr_info("Linux is running natively in 31-bit mode\n");
+	if (MACHINE_HAS_IEEE)
+		pr_info("The hardware system has IEEE compatible "
+			"floating point units\n");
+	else
+		pr_info("The hardware system has no IEEE compatible "
+			"floating point units\n");
 #else /* CONFIG_64BIT */
 	if (MACHINE_IS_VM)
-		printk("We are running under VM (64 bit mode)\n");
+		pr_info("Linux is running as a z/VM "
+			"guest operating system in 64-bit mode\n");
 	else if (MACHINE_IS_KVM) {
-		printk("We are running under KVM (64 bit mode)\n");
+		pr_info("Linux is running under KVM in 64-bit mode\n");
 		add_preferred_console("hvc", 0, NULL);
 		s390_virtio_console_init();
 	} else
-		printk("We are running native (64 bit mode)\n");
+		pr_info("Linux is running natively in 64-bit mode\n");
 #endif /* CONFIG_64BIT */
 
 	/* Have one command line that is parsed and saved in /proc/cmdline */
@@ -818,90 +828,3 @@ setup_arch(char **cmdline_p)
 	/* Setup zfcpdump support */
 	setup_zfcpdump(console_devno);
 }
-
-void __cpuinit print_cpu_info(struct cpuinfo_S390 *cpuinfo)
-{
-   printk(KERN_INFO "cpu %d "
-#ifdef CONFIG_SMP
-           "phys_idx=%d "
-#endif
-           "vers=%02X ident=%06X machine=%04X unused=%04X\n",
-           cpuinfo->cpu_nr,
-#ifdef CONFIG_SMP
-           cpuinfo->cpu_addr,
-#endif
-           cpuinfo->cpu_id.version,
-           cpuinfo->cpu_id.ident,
-           cpuinfo->cpu_id.machine,
-           cpuinfo->cpu_id.unused);
-}
-
-/*
- * show_cpuinfo - Get information on one CPU for use by procfs.
- */
-
-static int show_cpuinfo(struct seq_file *m, void *v)
-{
-	static const char *hwcap_str[8] = {
-		"esan3", "zarch", "stfle", "msa", "ldisp", "eimm", "dfp",
-		"edat"
-	};
-        struct cpuinfo_S390 *cpuinfo;
-	unsigned long n = (unsigned long) v - 1;
-	int i;
-
-	s390_adjust_jiffies();
-	preempt_disable();
-	if (!n) {
-		seq_printf(m, "vendor_id       : IBM/S390\n"
-			       "# processors    : %i\n"
-			       "bogomips per cpu: %lu.%02lu\n",
-			       num_online_cpus(), loops_per_jiffy/(500000/HZ),
-			       (loops_per_jiffy/(5000/HZ))%100);
-		seq_puts(m, "features\t: ");
-		for (i = 0; i < 8; i++)
-			if (hwcap_str[i] && (elf_hwcap & (1UL << i)))
-				seq_printf(m, "%s ", hwcap_str[i]);
-		seq_puts(m, "\n");
-	}
-
-	if (cpu_online(n)) {
-#ifdef CONFIG_SMP
-		if (smp_processor_id() == n)
-			cpuinfo = &S390_lowcore.cpu_data;
-		else
-			cpuinfo = &lowcore_ptr[n]->cpu_data;
-#else
-		cpuinfo = &S390_lowcore.cpu_data;
-#endif
-		seq_printf(m, "processor %li: "
-			       "version = %02X,  "
-			       "identification = %06X,  "
-			       "machine = %04X\n",
-			       n, cpuinfo->cpu_id.version,
-			       cpuinfo->cpu_id.ident,
-			       cpuinfo->cpu_id.machine);
-	}
-	preempt_enable();
-        return 0;
-}
-
-static void *c_start(struct seq_file *m, loff_t *pos)
-{
-	return *pos < NR_CPUS ? (void *)((unsigned long) *pos + 1) : NULL;
-}
-static void *c_next(struct seq_file *m, void *v, loff_t *pos)
-{
-	++*pos;
-	return c_start(m, pos);
-}
-static void c_stop(struct seq_file *m, void *v)
-{
-}
-const struct seq_operations cpuinfo_op = {
-	.start	= c_start,
-	.next	= c_next,
-	.stop	= c_stop,
-	.show	= show_cpuinfo,
-};
-

arch/s390/kernel/smp.c

@@ -20,6 +20,9 @@
  * cpu_number_map in other architectures.
  */
 
+#define KMSG_COMPONENT "cpu"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/mm.h>
@@ -77,159 +80,6 @@ static DEFINE_PER_CPU(struct cpu, cpu_devices);
 
 static void smp_ext_bitcall(int, ec_bit_sig);
 
-/*
- * Structure and data for __smp_call_function_map(). This is designed to
- * minimise static memory requirements. It also looks cleaner.
- */
-static DEFINE_SPINLOCK(call_lock);
-
-struct call_data_struct {
-	void (*func) (void *info);
-	void *info;
-	cpumask_t started;
-	cpumask_t finished;
-	int wait;
-};
-
-static struct call_data_struct *call_data;
-
-/*
- * 'Call function' interrupt callback
- */
-static void do_call_function(void)
-{
-	void (*func) (void *info) = call_data->func;
-	void *info = call_data->info;
-	int wait = call_data->wait;
-
-	cpu_set(smp_processor_id(), call_data->started);
-	(*func)(info);
-	if (wait)
-		cpu_set(smp_processor_id(), call_data->finished);;
-}
-
-static void __smp_call_function_map(void (*func) (void *info), void *info,
-				    int wait, cpumask_t map)
-{
-	struct call_data_struct data;
-	int cpu, local = 0;
-
-	/*
-	 * Can deadlock when interrupts are disabled or if in wrong context.
-	 */
-	WARN_ON(irqs_disabled() || in_irq());
-
-	/*
-	 * Check for local function call. We have to have the same call order
-	 * as in on_each_cpu() because of machine_restart_smp().
-	 */
-	if (cpu_isset(smp_processor_id(), map)) {
-		local = 1;
-		cpu_clear(smp_processor_id(), map);
-	}
-
-	cpus_and(map, map, cpu_online_map);
-	if (cpus_empty(map))
-		goto out;
-
-	data.func = func;
-	data.info = info;
-	data.started = CPU_MASK_NONE;
-	data.wait = wait;
-	if (wait)
-		data.finished = CPU_MASK_NONE;
-
-	call_data = &data;
-
-	for_each_cpu_mask(cpu, map)
-		smp_ext_bitcall(cpu, ec_call_function);
-
-	/* Wait for response */
-	while (!cpus_equal(map, data.started))
-		cpu_relax();
-	if (wait)
-		while (!cpus_equal(map, data.finished))
-			cpu_relax();
-out:
-	if (local) {
-		local_irq_disable();
-		func(info);
-		local_irq_enable();
-	}
-}
-
-/*
- * smp_call_function:
- * @func: the function to run; this must be fast and non-blocking
- * @info: an arbitrary pointer to pass to the function
- * @wait: if true, wait (atomically) until function has completed on other CPUs
- *
- * Run a function on all other CPUs.
- *
- * You must not call this function with disabled interrupts, from a
- * hardware interrupt handler or from a bottom half.
- */
-int smp_call_function(void (*func) (void *info), void *info, int wait)
-{
-	cpumask_t map;
-
-	spin_lock(&call_lock);
-	map = cpu_online_map;
-	cpu_clear(smp_processor_id(), map);
-	__smp_call_function_map(func, info, wait, map);
-	spin_unlock(&call_lock);
-	return 0;
-}
-EXPORT_SYMBOL(smp_call_function);
-
-/*
- * smp_call_function_single:
- * @cpu: the CPU where func should run
- * @func: the function to run; this must be fast and non-blocking
- * @info: an arbitrary pointer to pass to the function
- * @wait: if true, wait (atomically) until function has completed on other CPUs
- *
- * Run a function on one processor.
- *
- * You must not call this function with disabled interrupts, from a
- * hardware interrupt handler or from a bottom half.
- */
-int smp_call_function_single(int cpu, void (*func) (void *info), void *info,
-			     int wait)
-{
-	spin_lock(&call_lock);
-	__smp_call_function_map(func, info, wait, cpumask_of_cpu(cpu));
-	spin_unlock(&call_lock);
-	return 0;
-}
-EXPORT_SYMBOL(smp_call_function_single);
-
-/**
- * smp_call_function_mask(): Run a function on a set of other CPUs.
- * @mask: The set of cpus to run on.  Must not include the current cpu.
- * @func: The function to run. This must be fast and non-blocking.
- * @info: An arbitrary pointer to pass to the function.
- * @wait: If true, wait (atomically) until function has completed on other CPUs.
- *
- * Returns 0 on success, else a negative status code.
- *
- * If @wait is true, then returns once @func has returned; otherwise
- * it returns just before the target cpu calls @func.
- *
- * You must not call this function with disabled interrupts or from a
- * hardware interrupt handler or from a bottom half handler.
- */
-int smp_call_function_mask(cpumask_t mask, void (*func)(void *), void *info,
-			   int wait)
-{
-	spin_lock(&call_lock);
-	cpu_clear(smp_processor_id(), mask);
-	__smp_call_function_map(func, info, wait, mask);
-	spin_unlock(&call_lock);
-	return 0;
-}
-EXPORT_SYMBOL(smp_call_function_mask);
-
 void smp_send_stop(void)
 {
 	int cpu, rc;
@@ -271,7 +121,10 @@ static void do_ext_call_interrupt(__u16 code)
 	bits = xchg(&S390_lowcore.ext_call_fast, 0);
 
 	if (test_bit(ec_call_function, &bits))
-		do_call_function();
+		generic_smp_call_function_interrupt();
+
+	if (test_bit(ec_call_function_single, &bits))
+		generic_smp_call_function_single_interrupt();
 }
 
 /*
@@ -288,6 +141,19 @@ static void smp_ext_bitcall(int cpu, ec_bit_sig sig)
 		udelay(10);
 }
 
+void arch_send_call_function_ipi(cpumask_t mask)
+{
+	int cpu;
+
+	for_each_cpu_mask(cpu, mask)
+		smp_ext_bitcall(cpu, ec_call_function);
+}
+
+void arch_send_call_function_single_ipi(int cpu)
+{
+	smp_ext_bitcall(cpu, ec_call_function_single);
+}
+
 #ifndef CONFIG_64BIT
 /*
  * this function sends a 'purge tlb' signal to another CPU.
@@ -388,8 +254,8 @@ static void __init smp_get_save_area(unsigned int cpu, unsigned int phy_cpu)
 	if (ipl_info.type != IPL_TYPE_FCP_DUMP)
 		return;
 	if (cpu >= NR_CPUS) {
-		printk(KERN_WARNING "Registers for cpu %i not saved since dump "
-		       "kernel was compiled with NR_CPUS=%i\n", cpu, NR_CPUS);
+		pr_warning("CPU %i exceeds the maximum %i and is excluded from "
+			   "the dump\n", cpu, NR_CPUS - 1);
 		return;
 	}
 	zfcpdump_save_areas[cpu] = kmalloc(sizeof(union save_area), GFP_KERNEL);
@@ -562,7 +428,7 @@ static void __init smp_detect_cpus(void)
 	}
 out:
 	kfree(info);
-	printk(KERN_INFO "CPUs: %d configured, %d standby\n", c_cpus, s_cpus);
+	pr_info("%d configured CPUs, %d standby CPUs\n", c_cpus, s_cpus);
 	get_online_cpus();
 	__smp_rescan_cpus();
 	put_online_cpus();
@@ -578,19 +444,17 @@ int __cpuinit start_secondary(void *cpuvoid)
 	preempt_disable();
 	/* Enable TOD clock interrupts on the secondary cpu. */
 	init_cpu_timer();
-#ifdef CONFIG_VIRT_TIMER
 	/* Enable cpu timer interrupts on the secondary cpu. */
 	init_cpu_vtimer();
-#endif
 	/* Enable pfault pseudo page faults on this cpu. */
 	pfault_init();
 
 	/* call cpu notifiers */
 	notify_cpu_starting(smp_processor_id());
 	/* Mark this cpu as online */
-	spin_lock(&call_lock);
+	ipi_call_lock();
 	cpu_set(smp_processor_id(), cpu_online_map);
-	spin_unlock(&call_lock);
+	ipi_call_unlock();
 	/* Switch on interrupts */
 	local_irq_enable();
 	/* Print info about this processor */
@@ -639,18 +503,15 @@ static int __cpuinit smp_alloc_lowcore(int cpu)
 
 		save_area = get_zeroed_page(GFP_KERNEL);
 		if (!save_area)
-			goto out_save_area;
+			goto out;
 		lowcore->extended_save_area_addr = (u32) save_area;
 	}
 #endif
 	lowcore_ptr[cpu] = lowcore;
 	return 0;
 
-#ifndef CONFIG_64BIT
-out_save_area:
-	free_page(panic_stack);
-#endif
 out:
+	free_page(panic_stack);
 	free_pages(async_stack, ASYNC_ORDER);
 	free_pages((unsigned long) lowcore, lc_order);
 	return -ENOMEM;
@@ -690,12 +551,8 @@ int __cpuinit __cpu_up(unsigned int cpu)
 
 	ccode = signal_processor_p((__u32)(unsigned long)(lowcore_ptr[cpu]),
 				   cpu, sigp_set_prefix);
-	if (ccode) {
-		printk("sigp_set_prefix failed for cpu %d "
-		       "with condition code %d\n",
-		       (int) cpu, (int) ccode);
+	if (ccode)
 		return -EIO;
-	}
 
 	idle = current_set[cpu];
 	cpu_lowcore = lowcore_ptr[cpu];
@@ -778,7 +635,7 @@ void __cpu_die(unsigned int cpu)
 	while (!smp_cpu_not_running(cpu))
 		cpu_relax();
 	smp_free_lowcore(cpu);
-	printk(KERN_INFO "Processor %d spun down\n", cpu);
+	pr_info("Processor %d stopped\n", cpu);
 }
 
 void cpu_die(void)

arch/s390/kernel/time.c

@@ -12,6 +12,9 @@
  *    Copyright (C) 1991, 1992, 1995  Linus Torvalds
  */
 
+#define KMSG_COMPONENT "time"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/errno.h>
 #include <linux/module.h>
 #include <linux/sched.h>
@@ -20,6 +23,8 @@
 #include <linux/string.h>
 #include <linux/mm.h>
 #include <linux/interrupt.h>
+#include <linux/cpu.h>
+#include <linux/stop_machine.h>
 #include <linux/time.h>
 #include <linux/sysdev.h>
 #include <linux/delay.h>
@@ -36,6 +41,7 @@
 #include <asm/delay.h>
 #include <asm/s390_ext.h>
 #include <asm/div64.h>
+#include <asm/vdso.h>
 #include <asm/irq.h>
 #include <asm/irq_regs.h>
 #include <asm/timer.h>
@@ -223,6 +229,36 @@ static struct clocksource clocksource_tod = {
 };
 
 
+void update_vsyscall(struct timespec *wall_time, struct clocksource *clock)
+{
+	if (clock != &clocksource_tod)
+		return;
+
+	/* Make userspace gettimeofday spin until we're done. */
+	++vdso_data->tb_update_count;
+	smp_wmb();
+	vdso_data->xtime_tod_stamp = clock->cycle_last;
+	vdso_data->xtime_clock_sec = xtime.tv_sec;
+	vdso_data->xtime_clock_nsec = xtime.tv_nsec;
+	vdso_data->wtom_clock_sec = wall_to_monotonic.tv_sec;
+	vdso_data->wtom_clock_nsec = wall_to_monotonic.tv_nsec;
+	smp_wmb();
+	++vdso_data->tb_update_count;
+}
+
+extern struct timezone sys_tz;
+
+void update_vsyscall_tz(void)
+{
+	/* Make userspace gettimeofday spin until we're done. */
+	++vdso_data->tb_update_count;
+	smp_wmb();
+	vdso_data->tz_minuteswest = sys_tz.tz_minuteswest;
+	vdso_data->tz_dsttime = sys_tz.tz_dsttime;
+	smp_wmb();
+	++vdso_data->tb_update_count;
+}
+
 /*
  * Initialize the TOD clock and the CPU timer of
  * the boot cpu.
@@ -253,10 +289,8 @@ void __init time_init(void)
 
 	/* Enable TOD clock interrupts on the boot cpu. */
 	init_cpu_timer();
-
-#ifdef CONFIG_VIRT_TIMER
+	/* Enable cpu timer interrupts on the boot cpu. */
 	vtime_init();
-#endif
 }
 
 /*
@@ -288,8 +322,8 @@ static unsigned long long adjust_time(unsigned long long old,
 	}
 	sched_clock_base_cc += delta;
 	if (adjust.offset != 0) {
-		printk(KERN_NOTICE "etr: time adjusted by %li micro-seconds\n",
-		       adjust.offset);
+		pr_notice("The ETR interface has adjusted the clock "
+			  "by %li microseconds\n", adjust.offset);
 		adjust.modes = ADJ_OFFSET_SINGLESHOT;
 		do_adjtimex(&adjust);
 	}
@@ -360,6 +394,15 @@ static void enable_sync_clock(void)
 	atomic_set_mask(0x80000000, sw_ptr);
 }
 
+/* Single threaded workqueue used for etr and stp sync events */
+static struct workqueue_struct *time_sync_wq;
+
+static void __init time_init_wq(void)
+{
+	if (!time_sync_wq)
+		time_sync_wq = create_singlethread_workqueue("timesync");
+}
+
 /*
  * External Time Reference (ETR) code.
  */
@@ -425,6 +468,7 @@ static struct timer_list etr_timer;
 
 static void etr_timeout(unsigned long dummy);
 static void etr_work_fn(struct work_struct *work);
+static DEFINE_MUTEX(etr_work_mutex);
 static DECLARE_WORK(etr_work, etr_work_fn);
 
 /*
@@ -440,8 +484,8 @@ static void etr_reset(void)
 		etr_tolec = get_clock();
 		set_bit(CLOCK_SYNC_HAS_ETR, &clock_sync_flags);
 	} else if (etr_port0_online || etr_port1_online) {
-		printk(KERN_WARNING "Running on non ETR capable "
-		       "machine, only local mode available.\n");
+		pr_warning("The real or virtual hardware system does "
+			   "not provide an ETR interface\n");
 		etr_port0_online = etr_port1_online = 0;
 	}
 }
@@ -452,17 +496,18 @@ static int __init etr_init(void)
 
 	if (!test_bit(CLOCK_SYNC_HAS_ETR, &clock_sync_flags))
 		return 0;
+	time_init_wq();
 	/* Check if this machine has the steai instruction. */
 	if (etr_steai(&aib, ETR_STEAI_STEPPING_PORT) == 0)
 		etr_steai_available = 1;
 	setup_timer(&etr_timer, etr_timeout, 0UL);
 	if (etr_port0_online) {
 		set_bit(ETR_EVENT_PORT0_CHANGE, &etr_events);
-		schedule_work(&etr_work);
+		queue_work(time_sync_wq, &etr_work);
 	}
 	if (etr_port1_online) {
 		set_bit(ETR_EVENT_PORT1_CHANGE, &etr_events);
-		schedule_work(&etr_work);
+		queue_work(time_sync_wq, &etr_work);
 	}
 	return 0;
 }
@@ -489,7 +534,7 @@ void etr_switch_to_local(void)
 	if (test_bit(CLOCK_SYNC_ETR, &clock_sync_flags))
 		disable_sync_clock(NULL);
 	set_bit(ETR_EVENT_SWITCH_LOCAL, &etr_events);
-	schedule_work(&etr_work);
+	queue_work(time_sync_wq, &etr_work);
 }
 
 /*
@@ -505,7 +550,7 @@ void etr_sync_check(void)
 	if (test_bit(CLOCK_SYNC_ETR, &clock_sync_flags))
 		disable_sync_clock(NULL);
 	set_bit(ETR_EVENT_SYNC_CHECK, &etr_events);
-	schedule_work(&etr_work);
+	queue_work(time_sync_wq, &etr_work);
 }
 
 /*
@@ -529,13 +574,13 @@ static void etr_timing_alert(struct etr_irq_parm *intparm)
 		 * Both ports are not up-to-date now.
 		 */
 		set_bit(ETR_EVENT_PORT_ALERT, &etr_events);
-	schedule_work(&etr_work);
+	queue_work(time_sync_wq, &etr_work);
 }
 
 static void etr_timeout(unsigned long dummy)
 {
 	set_bit(ETR_EVENT_UPDATE, &etr_events);
-	schedule_work(&etr_work);
+	queue_work(time_sync_wq, &etr_work);
 }
 
 /*
@@ -642,14 +687,16 @@ static int etr_aib_follows(struct etr_aib *a1, struct etr_aib *a2, int p)
 }
 
 struct clock_sync_data {
+	atomic_t cpus;
 	int in_sync;
 	unsigned long long fixup_cc;
+	int etr_port;
+	struct etr_aib *etr_aib;
 };
 
-static void clock_sync_cpu_start(void *dummy)
+static void clock_sync_cpu(struct clock_sync_data *sync)
 {
-	struct clock_sync_data *sync = dummy;
-
+	atomic_dec(&sync->cpus);
 	enable_sync_clock();
 	/*
 	 * This looks like a busy wait loop but it isn't. etr_sync_cpus
@@ -675,39 +722,35 @@ static void clock_sync_cpu_start(void *dummy)
 	fixup_clock_comparator(sync->fixup_cc);
 }
 
-static void clock_sync_cpu_end(void *dummy)
-{
-}
-
 /*
  * Sync the TOD clock using the port refered to by aibp. This port
  * has to be enabled and the other port has to be disabled. The
  * last eacr update has to be more than 1.6 seconds in the past.
  */
-static int etr_sync_clock(struct etr_aib *aib, int port)
+static int etr_sync_clock(void *data)
 {
-	struct etr_aib *sync_port;
-	struct clock_sync_data etr_sync;
+	static int first;
 	unsigned long long clock, old_clock, delay, delta;
-	int follows;
+	struct clock_sync_data *etr_sync;
+	struct etr_aib *sync_port, *aib;
+	int port;
 	int rc;
 
-	/* Check if the current aib is adjacent to the sync port aib. */
-	sync_port = (port == 0) ? &etr_port0 : &etr_port1;
-	follows = etr_aib_follows(sync_port, aib, port);
-	memcpy(sync_port, aib, sizeof(*aib));
-	if (!follows)
-		return -EAGAIN;
+	etr_sync = data;
 
-	/*
-	 * Catch all other cpus and make them wait until we have
-	 * successfully synced the clock. smp_call_function will
-	 * return after all other cpus are in etr_sync_cpu_start.
-	 */
-	memset(&etr_sync, 0, sizeof(etr_sync));
-	preempt_disable();
-	smp_call_function(clock_sync_cpu_start, &etr_sync, 0);
-	local_irq_disable();
+	if (xchg(&first, 1) == 1) {
+		/* Slave */
+		clock_sync_cpu(etr_sync);
+		return 0;
+	}
+
+	/* Wait until all other cpus entered the sync function. */
+	while (atomic_read(&etr_sync->cpus) != 0)
+		cpu_relax();
+
+	port = etr_sync->etr_port;
+	aib = etr_sync->etr_aib;
+	sync_port = (port == 0) ? &etr_port0 : &etr_port1;
 	enable_sync_clock();
 
 	/* Set clock to next OTE. */
@@ -724,16 +767,16 @@ static int etr_sync_clock(struct etr_aib *aib, int port)
 		delay = (unsigned long long)
 			(aib->edf2.etv - sync_port->edf2.etv) << 32;
 		delta = adjust_time(old_clock, clock, delay);
-		etr_sync.fixup_cc = delta;
+		etr_sync->fixup_cc = delta;
 		fixup_clock_comparator(delta);
 		/* Verify that the clock is properly set. */
 		if (!etr_aib_follows(sync_port, aib, port)) {
 			/* Didn't work. */
 			disable_sync_clock(NULL);
-			etr_sync.in_sync = -EAGAIN;
+			etr_sync->in_sync = -EAGAIN;
 			rc = -EAGAIN;
 		} else {
-			etr_sync.in_sync = 1;
+			etr_sync->in_sync = 1;
 			rc = 0;
 		}
 	} else {
@@ -741,12 +784,33 @@ static int etr_sync_clock(struct etr_aib *aib, int port)
 		__ctl_clear_bit(0, 29);
 		__ctl_clear_bit(14, 21);
 		disable_sync_clock(NULL);
-		etr_sync.in_sync = -EAGAIN;
+		etr_sync->in_sync = -EAGAIN;
 		rc = -EAGAIN;
 	}
-	local_irq_enable();
-	smp_call_function(clock_sync_cpu_end, NULL, 0);
-	preempt_enable();
+	xchg(&first, 0);
+	return rc;
+}
+
+static int etr_sync_clock_stop(struct etr_aib *aib, int port)
+{
+	struct clock_sync_data etr_sync;
+	struct etr_aib *sync_port;
+	int follows;
+	int rc;
+
+	/* Check if the current aib is adjacent to the sync port aib. */
+	sync_port = (port == 0) ? &etr_port0 : &etr_port1;
+	follows = etr_aib_follows(sync_port, aib, port);
+	memcpy(sync_port, aib, sizeof(*aib));
+	if (!follows)
+		return -EAGAIN;
+	memset(&etr_sync, 0, sizeof(etr_sync));
+	etr_sync.etr_aib = aib;
+	etr_sync.etr_port = port;
+	get_online_cpus();
+	atomic_set(&etr_sync.cpus, num_online_cpus() - 1);
+	rc = stop_machine(etr_sync_clock, &etr_sync, &cpu_online_map);
+	put_online_cpus();
 	return rc;
 }
 
@@ -903,7 +967,7 @@ static void etr_update_eacr(struct etr_eacr eacr)
 }
 
 /*
- * ETR tasklet. In this function you'll find the main logic. In
+ * ETR work. In this function you'll find the main logic. In
  * particular this is the only function that calls etr_update_eacr(),
  * it "controls" the etr control register.
  */
@@ -914,6 +978,9 @@ static void etr_work_fn(struct work_struct *work)
 	struct etr_aib aib;
 	int sync_port;
 
+	/* prevent multiple execution. */
+	mutex_lock(&etr_work_mutex);
+
 	/* Create working copy of etr_eacr. */
 	eacr = etr_eacr;
 
@@ -929,7 +996,7 @@ static void etr_work_fn(struct work_struct *work)
 		del_timer_sync(&etr_timer);
 		etr_update_eacr(eacr);
 		clear_bit(CLOCK_SYNC_ETR, &clock_sync_flags);
-		return;
+		goto out_unlock;
 	}
 
 	/* Store aib to get the current ETR status word. */
@@ -1016,7 +1083,7 @@ static void etr_work_fn(struct work_struct *work)
 	    eacr.es || sync_port < 0) {
 		etr_update_eacr(eacr);
 		etr_set_tolec_timeout(now);
-		return;
+		goto out_unlock;
 	}
 
 	/*
@@ -1036,7 +1103,7 @@ static void etr_work_fn(struct work_struct *work)
 	etr_update_eacr(eacr);
 	set_bit(CLOCK_SYNC_ETR, &clock_sync_flags);
 	if (now < etr_tolec + (1600000 << 12) ||
-	    etr_sync_clock(&aib, sync_port) != 0) {
+	    etr_sync_clock_stop(&aib, sync_port) != 0) {
 		/* Sync failed. Try again in 1/2 second. */
 		eacr.es = 0;
 		etr_update_eacr(eacr);
@@ -1044,6 +1111,8 @@ static void etr_work_fn(struct work_struct *work)
 		etr_set_sync_timeout();
 	} else
 		etr_set_tolec_timeout(now);
+out_unlock:
+	mutex_unlock(&etr_work_mutex);
 }
 
 /*
@@ -1125,13 +1194,13 @@ static ssize_t etr_online_store(struct sys_device *dev,
 			return count;	/* Nothing to do. */
 		etr_port0_online = value;
 		set_bit(ETR_EVENT_PORT0_CHANGE, &etr_events);
-		schedule_work(&etr_work);
+		queue_work(time_sync_wq, &etr_work);
 	} else {
 		if (etr_port1_online == value)
 			return count;	/* Nothing to do. */
 		etr_port1_online = value;
 		set_bit(ETR_EVENT_PORT1_CHANGE, &etr_events);
-		schedule_work(&etr_work);
+		queue_work(time_sync_wq, &etr_work);
 	}
 	return count;
 }
@@ -1332,6 +1401,7 @@ static struct stp_sstpi stp_info;
 static void *stp_page;
 
 static void stp_work_fn(struct work_struct *work);
+static DEFINE_MUTEX(stp_work_mutex);
 static DECLARE_WORK(stp_work, stp_work_fn);
 
 static int __init early_parse_stp(char *p)
@@ -1356,7 +1426,8 @@ static void __init stp_reset(void)
 	if (rc == 0)
 		set_bit(CLOCK_SYNC_HAS_STP, &clock_sync_flags);
 	else if (stp_online) {
-		printk(KERN_WARNING "Running on non STP capable machine.\n");
+		pr_warning("The real or virtual hardware system does "
+			   "not provide an STP interface\n");
 		free_bootmem((unsigned long) stp_page, PAGE_SIZE);
 		stp_page = NULL;
 		stp_online = 0;
@@ -1365,8 +1436,12 @@ static void __init stp_reset(void)
 
 static int __init stp_init(void)
 {
-	if (test_bit(CLOCK_SYNC_HAS_STP, &clock_sync_flags) && stp_online)
-		schedule_work(&stp_work);
+	if (!test_bit(CLOCK_SYNC_HAS_STP, &clock_sync_flags))
+		return 0;
+	time_init_wq();
+	if (!stp_online)
+		return 0;
+	queue_work(time_sync_wq, &stp_work);
 	return 0;
 }
 
@@ -1383,7 +1458,7 @@ arch_initcall(stp_init);
 static void stp_timing_alert(struct stp_irq_parm *intparm)
 {
 	if (intparm->tsc || intparm->lac || intparm->tcpc)
-		schedule_work(&stp_work);
+		queue_work(time_sync_wq, &stp_work);
 }
 
 /*
@@ -1397,7 +1472,7 @@ void stp_sync_check(void)
 	if (!test_bit(CLOCK_SYNC_STP, &clock_sync_flags))
 		return;
 	disable_sync_clock(NULL);
-	schedule_work(&stp_work);
+	queue_work(time_sync_wq, &stp_work);
 }
 
 /*
@@ -1411,46 +1486,34 @@ void stp_island_check(void)
 	if (!test_bit(CLOCK_SYNC_STP, &clock_sync_flags))
 		return;
 	disable_sync_clock(NULL);
-	schedule_work(&stp_work);
+	queue_work(time_sync_wq, &stp_work);
 }
 
-/*
- * STP tasklet. Check for the STP state and take over the clock
- * synchronization if the STP clock source is usable.
- */
-static void stp_work_fn(struct work_struct *work)
+
+static int stp_sync_clock(void *data)
 {
-	struct clock_sync_data stp_sync;
+	static int first;
 	unsigned long long old_clock, delta;
+	struct clock_sync_data *stp_sync;
 	int rc;
 
-	if (!stp_online) {
-		chsc_sstpc(stp_page, STP_OP_CTRL, 0x0000);
-		return;
-	}
+	stp_sync = data;
 
-	rc = chsc_sstpc(stp_page, STP_OP_CTRL, 0xb0e0);
-	if (rc)
-		return;
+	if (xchg(&first, 1) == 1) {
+		/* Slave */
+		clock_sync_cpu(stp_sync);
+		return 0;
+	}
 
-	rc = chsc_sstpi(stp_page, &stp_info, sizeof(struct stp_sstpi));
-	if (rc || stp_info.c == 0)
-		return;
+	/* Wait until all other cpus entered the sync function. */
+	while (atomic_read(&stp_sync->cpus) != 0)
+		cpu_relax();
 
-	/*
-	 * Catch all other cpus and make them wait until we have
-	 * successfully synced the clock. smp_call_function will
-	 * return after all other cpus are in clock_sync_cpu_start.
-	 */
-	memset(&stp_sync, 0, sizeof(stp_sync));
-	preempt_disable();
-	smp_call_function(clock_sync_cpu_start, &stp_sync, 0);
-	local_irq_disable();
 	enable_sync_clock();
 
 	set_bit(CLOCK_SYNC_STP, &clock_sync_flags);
 	if (test_and_clear_bit(CLOCK_SYNC_ETR, &clock_sync_flags))
-		schedule_work(&etr_work);
+		queue_work(time_sync_wq, &etr_work);
 
 	rc = 0;
 	if (stp_info.todoff[0] || stp_info.todoff[1] ||
@@ -1469,16 +1532,49 @@ static void stp_work_fn(struct work_struct *work)
 	}
 	if (rc) {
 		disable_sync_clock(NULL);
-		stp_sync.in_sync = -EAGAIN;
+		stp_sync->in_sync = -EAGAIN;
 		clear_bit(CLOCK_SYNC_STP, &clock_sync_flags);
 		if (etr_port0_online || etr_port1_online)
-			schedule_work(&etr_work);
+			queue_work(time_sync_wq, &etr_work);
 	} else
-		stp_sync.in_sync = 1;
+		stp_sync->in_sync = 1;
+	xchg(&first, 0);
+	return 0;
+}
+
+/*
+ * STP work. Check for the STP state and take over the clock
+ * synchronization if the STP clock source is usable.
+ */
+static void stp_work_fn(struct work_struct *work)
+{
+	struct clock_sync_data stp_sync;
+	int rc;
+
+	/* prevent multiple execution. */
+	mutex_lock(&stp_work_mutex);
+
+	if (!stp_online) {
+		chsc_sstpc(stp_page, STP_OP_CTRL, 0x0000);
+		goto out_unlock;
+	}
+
+	rc = chsc_sstpc(stp_page, STP_OP_CTRL, 0xb0e0);
+	if (rc)
+		goto out_unlock;
+
+	rc = chsc_sstpi(stp_page, &stp_info, sizeof(struct stp_sstpi));
+	if (rc || stp_info.c == 0)
+		goto out_unlock;
+
+	memset(&stp_sync, 0, sizeof(stp_sync));
+	get_online_cpus();
+	atomic_set(&stp_sync.cpus, num_online_cpus() - 1);
+	stop_machine(stp_sync_clock, &stp_sync, &cpu_online_map);
+	put_online_cpus();
 
-	local_irq_enable();
-	smp_call_function(clock_sync_cpu_end, NULL, 0);
-	preempt_enable();
+out_unlock:
+	mutex_unlock(&stp_work_mutex);
 }
 
 /*
@@ -1587,7 +1683,7 @@ static ssize_t stp_online_store(struct sysdev_class *class,
 	if (!test_bit(CLOCK_SYNC_HAS_STP, &clock_sync_flags))
 		return -EOPNOTSUPP;
 	stp_online = value;
-	schedule_work(&stp_work);
+	queue_work(time_sync_wq, &stp_work);
 	return count;
 }
 

arch/s390/kernel/topology.c

@@ -3,6 +3,9 @@
  *    Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>
  */
 
+#define KMSG_COMPONENT "cpu"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/kernel.h>
 #include <linux/mm.h>
 #include <linux/init.h>
@@ -12,6 +15,7 @@
 #include <linux/workqueue.h>
 #include <linux/cpu.h>
 #include <linux/smp.h>
+#include <linux/cpuset.h>
 #include <asm/delay.h>
 #include <asm/s390_ext.h>
 #include <asm/sysinfo.h>
@@ -57,11 +61,11 @@ struct core_info {
 	cpumask_t mask;
 };
 
+static int topology_enabled;
 static void topology_work_fn(struct work_struct *work);
 static struct tl_info *tl_info;
 static struct core_info core_info;
 static int machine_has_topology;
-static int machine_has_topology_irq;
 static struct timer_list topology_timer;
 static void set_topology_timer(void);
 static DECLARE_WORK(topology_work, topology_work_fn);
@@ -77,8 +81,8 @@ cpumask_t cpu_coregroup_map(unsigned int cpu)
 	cpumask_t mask;
 
 	cpus_clear(mask);
-	if (!machine_has_topology)
-		return cpu_present_map;
+	if (!topology_enabled || !machine_has_topology)
+		return cpu_possible_map;
 	spin_lock_irqsave(&topology_lock, flags);
 	while (core) {
 		if (cpu_isset(cpu, core->mask)) {
@@ -168,7 +172,7 @@ static void topology_update_polarization_simple(void)
 	int cpu;
 
 	mutex_lock(&smp_cpu_state_mutex);
-	for_each_present_cpu(cpu)
+	for_each_possible_cpu(cpu)
 		smp_cpu_polarization[cpu] = POLARIZATION_HRZ;
 	mutex_unlock(&smp_cpu_state_mutex);
 }
@@ -199,7 +203,7 @@ int topology_set_cpu_management(int fc)
 		rc = ptf(PTF_HORIZONTAL);
 	if (rc)
 		return -EBUSY;
-	for_each_present_cpu(cpu)
+	for_each_possible_cpu(cpu)
 		smp_cpu_polarization[cpu] = POLARIZATION_UNKNWN;
 	return rc;
 }
@@ -208,7 +212,7 @@ static void update_cpu_core_map(void)
 {
 	int cpu;
 
-	for_each_present_cpu(cpu)
+	for_each_possible_cpu(cpu)
 		cpu_core_map[cpu] = cpu_coregroup_map(cpu);
 }
 
@@ -235,7 +239,7 @@ int arch_update_cpu_topology(void)
 
 static void topology_work_fn(struct work_struct *work)
 {
-	arch_reinit_sched_domains();
+	rebuild_sched_domains();
 }
 
 void topology_schedule_update(void)
@@ -258,10 +262,14 @@ static void set_topology_timer(void)
 	add_timer(&topology_timer);
 }
 
-static void topology_interrupt(__u16 code)
+static int __init early_parse_topology(char *p)
 {
-	schedule_work(&topology_work);
+	if (strncmp(p, "on", 2))
+		return 0;
+	topology_enabled = 1;
+	return 0;
 }
+early_param("topology", early_parse_topology);
 
 static int __init init_topology_update(void)
 {
@@ -273,14 +281,7 @@ static int __init init_topology_update(void)
 		goto out;
 	}
 	init_timer_deferrable(&topology_timer);
-	if (machine_has_topology_irq) {
-		rc = register_external_interrupt(0x2005, topology_interrupt);
-		if (rc)
-			goto out;
-		ctl_set_bit(0, 8);
-	}
-	else
-		set_topology_timer();
+	set_topology_timer();
 out:
 	update_cpu_core_map();
 	return rc;
@@ -301,9 +302,6 @@ void __init s390_init_cpu_topology(void)
 		return;
 	machine_has_topology = 1;
 
-	if (facility_bits & (1ULL << 51))
-		machine_has_topology_irq = 1;
-
 	tl_info = alloc_bootmem_pages(PAGE_SIZE);
 	info = tl_info;
 	stsi(info, 15, 1, 2);
@@ -312,7 +310,7 @@ void __init s390_init_cpu_topology(void)
 	for (i = 0; i < info->mnest - 2; i++)
 		nr_cores *= info->mag[NR_MAG - 3 - i];
 
-	printk(KERN_INFO "CPU topology:");
+	pr_info("The CPU configuration topology of the machine is:");
 	for (i = 0; i < NR_MAG; i++)
 		printk(" %d", info->mag[i]);
 	printk(" / %d\n", info->mnest);
@@ -327,5 +325,4 @@ void __init s390_init_cpu_topology(void)
 	return;
 error:
 	machine_has_topology = 0;
-	machine_has_topology_irq = 0;
 }

arch/s390/kernel/vdso.c

@@ -0,0 +1,234 @@
+/*
+ * vdso setup for s390
+ *
+ *  Copyright IBM Corp. 2008
+ *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License (version 2 only)
+ * as published by the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+#include <linux/errno.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/smp.h>
+#include <linux/stddef.h>
+#include <linux/unistd.h>
+#include <linux/slab.h>
+#include <linux/user.h>
+#include <linux/elf.h>
+#include <linux/security.h>
+#include <linux/bootmem.h>
+
+#include <asm/pgtable.h>
+#include <asm/system.h>
+#include <asm/processor.h>
+#include <asm/mmu.h>
+#include <asm/mmu_context.h>
+#include <asm/sections.h>
+#include <asm/vdso.h>
+
+/* Max supported size for symbol names */
+#define MAX_SYMNAME	64
+
+#if defined(CONFIG_32BIT) || defined(CONFIG_COMPAT)
+extern char vdso32_start, vdso32_end;
+static void *vdso32_kbase = &vdso32_start;
+static unsigned int vdso32_pages;
+static struct page **vdso32_pagelist;
+#endif
+
+#ifdef CONFIG_64BIT
+extern char vdso64_start, vdso64_end;
+static void *vdso64_kbase = &vdso64_start;
+static unsigned int vdso64_pages;
+static struct page **vdso64_pagelist;
+#endif /* CONFIG_64BIT */
+
+/*
+ * Should the kernel map a VDSO page into processes and pass its
+ * address down to glibc upon exec()?
+ */
+unsigned int __read_mostly vdso_enabled = 1;
+
+static int __init vdso_setup(char *s)
+{
+	vdso_enabled = simple_strtoul(s, NULL, 0);
+	return 1;
+}
+__setup("vdso=", vdso_setup);
+
+/*
+ * The vdso data page
+ */
+static union {
+	struct vdso_data	data;
+	u8			page[PAGE_SIZE];
+} vdso_data_store __attribute__((__section__(".data.page_aligned")));
+struct vdso_data *vdso_data = &vdso_data_store.data;
+
+/*
+ * This is called from binfmt_elf, we create the special vma for the
+ * vDSO and insert it into the mm struct tree
+ */
+int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
+{
+	struct mm_struct *mm = current->mm;
+	struct page **vdso_pagelist;
+	unsigned long vdso_pages;
+	unsigned long vdso_base;
+	int rc;
+
+	if (!vdso_enabled)
+		return 0;
+	/*
+	 * Only map the vdso for dynamically linked elf binaries.
+	 */
+	if (!uses_interp)
+		return 0;
+
+	vdso_base = mm->mmap_base;
+#ifdef CONFIG_64BIT
+	vdso_pagelist = vdso64_pagelist;
+	vdso_pages = vdso64_pages;
+#ifdef CONFIG_COMPAT
+	if (test_thread_flag(TIF_31BIT)) {
+		vdso_pagelist = vdso32_pagelist;
+		vdso_pages = vdso32_pages;
+	}
+#endif
+#else
+	vdso_pagelist = vdso32_pagelist;
+	vdso_pages = vdso32_pages;
+#endif
+
+	/*
+	 * vDSO has a problem and was disabled, just don't "enable" it for
+	 * the process
+	 */
+	if (vdso_pages == 0)
+		return 0;
+
+	current->mm->context.vdso_base = 0;
+
+	/*
+	 * pick a base address for the vDSO in process space. We try to put
+	 * it at vdso_base which is the "natural" base for it, but we might
+	 * fail and end up putting it elsewhere.
+	 */
+	down_write(&mm->mmap_sem);
+	vdso_base = get_unmapped_area(NULL, vdso_base,
+				      vdso_pages << PAGE_SHIFT, 0, 0);
+	if (IS_ERR_VALUE(vdso_base)) {
+		rc = vdso_base;
+		goto out_up;
+	}
+
+	/*
+	 * our vma flags don't have VM_WRITE so by default, the process
+	 * isn't allowed to write those pages.
+	 * gdb can break that with ptrace interface, and thus trigger COW
+	 * on those pages but it's then your responsibility to never do that
+	 * on the "data" page of the vDSO or you'll stop getting kernel
+	 * updates and your nice userland gettimeofday will be totally dead.
+	 * It's fine to use that for setting breakpoints in the vDSO code
+	 * pages though
+	 *
+	 * Make sure the vDSO gets into every core dump.
+	 * Dumping its contents makes post-mortem fully interpretable later
+	 * without matching up the same kernel and hardware config to see
+	 * what PC values meant.
+	 */
+	rc = install_special_mapping(mm, vdso_base, vdso_pages << PAGE_SHIFT,
+				     VM_READ|VM_EXEC|
+				     VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC|
+				     VM_ALWAYSDUMP,
+				     vdso_pagelist);
+	if (rc)
+		goto out_up;
+
+	/* Put vDSO base into mm struct */
+	current->mm->context.vdso_base = vdso_base;
+
+	up_write(&mm->mmap_sem);
+	return 0;
+
+out_up:
+	up_write(&mm->mmap_sem);
+	return rc;
+}
+
+const char *arch_vma_name(struct vm_area_struct *vma)
+{
+	if (vma->vm_mm && vma->vm_start == vma->vm_mm->context.vdso_base)
+		return "[vdso]";
+	return NULL;
+}
+
+static int __init vdso_init(void)
+{
+	int i;
+
+#if defined(CONFIG_32BIT) || defined(CONFIG_COMPAT)
+	/* Calculate the size of the 32 bit vDSO */
+	vdso32_pages = ((&vdso32_end - &vdso32_start
+			 + PAGE_SIZE - 1) >> PAGE_SHIFT) + 1;
+
+	/* Make sure pages are in the correct state */
+	vdso32_pagelist = kzalloc(sizeof(struct page *) * (vdso32_pages + 1),
+				  GFP_KERNEL);
+	BUG_ON(vdso32_pagelist == NULL);
+	for (i = 0; i < vdso32_pages - 1; i++) {
+		struct page *pg = virt_to_page(vdso32_kbase + i*PAGE_SIZE);
+		ClearPageReserved(pg);
+		get_page(pg);
+		vdso32_pagelist[i] = pg;
+	}
+	vdso32_pagelist[vdso32_pages - 1] = virt_to_page(vdso_data);
+	vdso32_pagelist[vdso32_pages] = NULL;
+#endif
+
+#ifdef CONFIG_64BIT
+	/* Calculate the size of the 64 bit vDSO */
+	vdso64_pages = ((&vdso64_end - &vdso64_start
+			 + PAGE_SIZE - 1) >> PAGE_SHIFT) + 1;
+
+	/* Make sure pages are in the correct state */
+	vdso64_pagelist = kzalloc(sizeof(struct page *) * (vdso64_pages + 1),
+				  GFP_KERNEL);
+	BUG_ON(vdso64_pagelist == NULL);
+	for (i = 0; i < vdso64_pages - 1; i++) {
+		struct page *pg = virt_to_page(vdso64_kbase + i*PAGE_SIZE);
+		ClearPageReserved(pg);
+		get_page(pg);
+		vdso64_pagelist[i] = pg;
+	}
+	vdso64_pagelist[vdso64_pages - 1] = virt_to_page(vdso_data);
+	vdso64_pagelist[vdso64_pages] = NULL;
+#endif /* CONFIG_64BIT */
+
+	get_page(virt_to_page(vdso_data));
+
+	smp_wmb();
+
+	return 0;
+}
+arch_initcall(vdso_init);
+
+int in_gate_area_no_task(unsigned long addr)
+{
+	return 0;
+}
+
+int in_gate_area(struct task_struct *task, unsigned long addr)
+{
+	return 0;
+}
+
+struct vm_area_struct *get_gate_vma(struct task_struct *tsk)
+{
+	return NULL;
+}

arch/s390/kernel/vdso32/Makefile

@@ -0,0 +1,55 @@
+# List of files in the vdso, has to be asm only for now
+
+obj-vdso32 = gettimeofday.o clock_getres.o clock_gettime.o note.o
+
+# Build rules
+
+targets := $(obj-vdso32) vdso32.so vdso32.so.dbg
+obj-vdso32 := $(addprefix $(obj)/, $(obj-vdso32))
+
+KBUILD_AFLAGS_31 := $(filter-out -m64,$(KBUILD_AFLAGS))
+KBUILD_AFLAGS_31 += -m31 -s
+
+KBUILD_CFLAGS_31 := $(filter-out -m64,$(KBUILD_CFLAGS))
+KBUILD_CFLAGS_31 += -m31 -fPIC -shared -fno-common -fno-builtin
+KBUILD_CFLAGS_31 += -nostdlib -Wl,-soname=linux-vdso32.so.1 \
+			$(call ld-option, -Wl$(comma)--hash-style=sysv)
+
+$(targets:%=$(obj)/%.dbg): KBUILD_CFLAGS = $(KBUILD_CFLAGS_31)
+$(targets:%=$(obj)/%.dbg): KBUILD_AFLAGS = $(KBUILD_AFLAGS_31)
+
+obj-y += vdso32_wrapper.o
+extra-y += vdso32.lds
+CPPFLAGS_vdso32.lds += -P -C -U$(ARCH)
+
+# Force dependency (incbin is bad)
+$(obj)/vdso32_wrapper.o : $(obj)/vdso32.so
+
+# link rule for the .so file, .lds has to be first
+$(obj)/vdso32.so.dbg: $(src)/vdso32.lds $(obj-vdso32)
+	$(call if_changed,vdso32ld)
+
+# strip rule for the .so file
+$(obj)/%.so: OBJCOPYFLAGS := -S
+$(obj)/%.so: $(obj)/%.so.dbg FORCE
+	$(call if_changed,objcopy)
+
+# assembly rules for the .S files
+$(obj-vdso32): %.o: %.S
+	$(call if_changed_dep,vdso32as)
+
+# actual build commands
+quiet_cmd_vdso32ld = VDSO32L $@
+      cmd_vdso32ld = $(CC) $(c_flags) -Wl,-T $^ -o $@
+quiet_cmd_vdso32as = VDSO32A $@
+      cmd_vdso32as = $(CC) $(a_flags) -c -o $@ $<
+
+# install commands for the unstripped file
+quiet_cmd_vdso_install = INSTALL $@
+      cmd_vdso_install = cp $(obj)/$@.dbg $(MODLIB)/vdso/$@
+
+vdso32.so: $(obj)/vdso32.so.dbg
+	@mkdir -p $(MODLIB)/vdso
+	$(call cmd,vdso_install)
+
+vdso_install: vdso32.so

arch/s390/kernel/vdso32/clock_getres.S

@@ -0,0 +1,39 @@
+/*
+ * Userland implementation of clock_getres() for 32 bits processes in a
+ * s390 kernel for use in the vDSO
+ *
+ *  Copyright IBM Corp. 2008
+ *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License (version 2 only)
+ * as published by the Free Software Foundation.
+ */
+#include <asm/vdso.h>
+#include <asm/asm-offsets.h>
+#include <asm/unistd.h>
+
+	.text
+	.align 4
+	.globl __kernel_clock_getres
+	.type  __kernel_clock_getres,@function
+__kernel_clock_getres:
+	.cfi_startproc
+	chi	%r2,CLOCK_REALTIME
+	je	0f
+	chi	%r2,CLOCK_MONOTONIC
+	jne	3f
+0:	ltr	%r3,%r3
+	jz	2f				/* res == NULL */
+	basr	%r1,0
+1:	l	%r0,4f-1b(%r1)
+	xc	0(4,%r3),0(%r3)			/* set tp->tv_sec to zero */
+	st	%r0,4(%r3)			/* store tp->tv_usec */
+2:	lhi	%r2,0
+	br	%r14
+3:	lhi	%r1,__NR_clock_getres		/* fallback to svc */
+	svc	0
+	br	%r14
+4:	.long	CLOCK_REALTIME_RES
+	.cfi_endproc
+	.size	__kernel_clock_getres,.-__kernel_clock_getres

arch/s390/kernel/vdso32/clock_gettime.S

@@ -0,0 +1,128 @@
+/*
+ * Userland implementation of clock_gettime() for 32 bits processes in a
+ * s390 kernel for use in the vDSO
+ *
+ *  Copyright IBM Corp. 2008
+ *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License (version 2 only)
+ * as published by the Free Software Foundation.
+ */
+#include <asm/vdso.h>
+#include <asm/asm-offsets.h>
+#include <asm/unistd.h>
+
+	.text
+	.align 4
+	.globl __kernel_clock_gettime
+	.type  __kernel_clock_gettime,@function
+__kernel_clock_gettime:
+	.cfi_startproc
+	basr	%r5,0
+0:	al	%r5,21f-0b(%r5)			/* get &_vdso_data */
+	chi	%r2,CLOCK_REALTIME
+	je	10f
+	chi	%r2,CLOCK_MONOTONIC
+	jne	19f
+
+	/* CLOCK_MONOTONIC */
+	ltr	%r3,%r3
+	jz	9f				/* tp == NULL */
+1:	l	%r4,__VDSO_UPD_COUNT+4(%r5)	/* load update counter */
+	tml	%r4,0x0001			/* pending update ? loop */
+	jnz	1b
+	stck	24(%r15)			/* Store TOD clock */
+	lm	%r0,%r1,24(%r15)
+	s	%r0,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
+	sl	%r1,__VDSO_XTIME_STAMP+4(%r5)
+	brc	3,2f
+	ahi	%r0,-1
+2:	mhi	%r0,1000			/* cyc2ns(clock,cycle_delta) */
+	lr	%r2,%r0
+	lhi	%r0,1000
+	ltr	%r1,%r1
+	mr	%r0,%r0
+	jnm	3f
+	ahi	%r0,1000
+3:	alr	%r0,%r2
+	srdl	%r0,12
+	al	%r0,__VDSO_XTIME_NSEC(%r5)	/*  + xtime */
+	al	%r1,__VDSO_XTIME_NSEC+4(%r5)
+	brc	12,4f
+	ahi	%r0,1
+4:	l	%r2,__VDSO_XTIME_SEC+4(%r5)
+	al	%r0,__VDSO_WTOM_NSEC(%r5)	/*  + wall_to_monotonic */
+	al	%r1,__VDSO_WTOM_NSEC+4(%r5)
+	brc	12,5f
+	ahi	%r0,1
+5:	al	%r2,__VDSO_WTOM_SEC+4(%r5)
+	cl	%r4,__VDSO_UPD_COUNT+4(%r5)	/* check update counter */
+	jne	1b
+	basr	%r5,0
+6:	ltr	%r0,%r0
+	jnz	7f
+	cl	%r1,20f-6b(%r5)
+	jl	8f
+7:	ahi	%r2,1
+	sl	%r1,20f-6b(%r5)
+	brc	3,6b
+	ahi	%r0,-1
+	j	6b
+8:	st	%r2,0(%r3)			/* store tp->tv_sec */
+	st	%r1,4(%r3)			/* store tp->tv_nsec */
+9:	lhi	%r2,0
+	br	%r14
+
+	/* CLOCK_REALTIME */
+10:	ltr	%r3,%r3				/* tp == NULL */
+	jz	18f
+11:	l	%r4,__VDSO_UPD_COUNT+4(%r5)	/* load update counter */
+	tml	%r4,0x0001			/* pending update ? loop */
+	jnz	11b
+	stck	24(%r15)			/* Store TOD clock */
+	lm	%r0,%r1,24(%r15)
+	s	%r0,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
+	sl	%r1,__VDSO_XTIME_STAMP+4(%r5)
+	brc	3,12f
+	ahi	%r0,-1
+12:	mhi	%r0,1000			/* cyc2ns(clock,cycle_delta) */
+	lr	%r2,%r0
+	lhi	%r0,1000
+	ltr	%r1,%r1
+	mr	%r0,%r0
+	jnm	13f
+	ahi	%r0,1000
+13:	alr	%r0,%r2
+	srdl	%r0,12
+	al	%r0,__VDSO_XTIME_NSEC(%r5)	/*  + xtime */
+	al	%r1,__VDSO_XTIME_NSEC+4(%r5)
+	brc	12,14f
+	ahi	%r0,1
+14:	l	%r2,__VDSO_XTIME_SEC+4(%r5)
+	cl	%r4,__VDSO_UPD_COUNT+4(%r5)	/* check update counter */
+	jne	11b
+	basr	%r5,0
+15:	ltr	%r0,%r0
+	jnz	16f
+	cl	%r1,20f-15b(%r5)
+	jl	17f
+16:	ahi	%r2,1
+	sl	%r1,20f-15b(%r5)
+	brc	3,15b
+	ahi	%r0,-1
+	j	15b
+17:	st	%r2,0(%r3)			/* store tp->tv_sec */
+	st	%r1,4(%r3)			/* store tp->tv_nsec */
+18:	lhi	%r2,0
+	br	%r14
+
+	/* Fallback to system call */
+19:	lhi	%r1,__NR_clock_gettime
+	svc	0
+	br	%r14
+
+20:	.long	1000000000
+21:	.long	_vdso_data - 0b
+	.cfi_endproc
+	.size	__kernel_clock_gettime,.-__kernel_clock_gettime

arch/s390/kernel/vdso32/gettimeofday.S

@@ -0,0 +1,82 @@
+/*
+ * Userland implementation of gettimeofday() for 32 bits processes in a
+ * s390 kernel for use in the vDSO
+ *
+ *  Copyright IBM Corp. 2008
+ *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License (version 2 only)
+ * as published by the Free Software Foundation.
+ */
+#include <asm/vdso.h>
+#include <asm/asm-offsets.h>
+#include <asm/unistd.h>
+
+#include <asm/vdso.h>
+#include <asm/asm-offsets.h>
+#include <asm/unistd.h>
+
+	.text
+	.align 4
+	.globl __kernel_gettimeofday
+	.type  __kernel_gettimeofday,@function
+__kernel_gettimeofday:
+	.cfi_startproc
+	basr	%r5,0
+0:	al	%r5,13f-0b(%r5)			/* get &_vdso_data */
+1:	ltr	%r3,%r3				/* check if tz is NULL */
+	je	2f
+	mvc	0(8,%r3),__VDSO_TIMEZONE(%r5)
+2:	ltr	%r2,%r2				/* check if tv is NULL */
+	je	10f
+	l	%r4,__VDSO_UPD_COUNT+4(%r5)	/* load update counter */
+	tml	%r4,0x0001			/* pending update ? loop */
+	jnz	1b
+	stck	24(%r15)			/* Store TOD clock */
+	lm	%r0,%r1,24(%r15)
+	s	%r0,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
+	sl	%r1,__VDSO_XTIME_STAMP+4(%r5)
+	brc	3,3f
+	ahi	%r0,-1
+3:	mhi	%r0,1000			/* cyc2ns(clock,cycle_delta) */
+	st	%r0,24(%r15)
+	lhi	%r0,1000
+	ltr	%r1,%r1
+	mr	%r0,%r0
+	jnm	4f
+	ahi	%r0,1000
+4:	al	%r0,24(%r15)
+	srdl	%r0,12
+	al	%r0,__VDSO_XTIME_NSEC(%r5)	/*  + xtime */
+	al	%r1,__VDSO_XTIME_NSEC+4(%r5)
+	brc	12,5f
+	ahi	%r0,1
+5:	mvc	24(4,%r15),__VDSO_XTIME_SEC+4(%r5)
+	cl	%r4,__VDSO_UPD_COUNT+4(%r5)	/* check update counter */
+	jne	1b
+	l	%r4,24(%r15)			/* get tv_sec from stack */
+	basr	%r5,0
+6:	ltr	%r0,%r0
+	jnz	7f
+	cl	%r1,11f-6b(%r5)
+	jl	8f
+7:	ahi	%r4,1
+	sl	%r1,11f-6b(%r5)
+	brc	3,6b
+	ahi	%r0,-1
+	j	6b
+8:	st	%r4,0(%r2)			/* store tv->tv_sec */
+	ltr	%r1,%r1
+	m	%r0,12f-6b(%r5)
+	jnm	9f
+	al	%r0,12f-6b(%r5)
+9:	srl	%r0,6
+	st	%r0,4(%r2)			/* store tv->tv_usec */
+10:	slr	%r2,%r2
+	br	%r14
+11:	.long	1000000000
+12:	.long	274877907
+13:	.long	_vdso_data - 0b
+	.cfi_endproc
+	.size	__kernel_gettimeofday,.-__kernel_gettimeofday

arch/s390/kernel/vdso32/note.S

@@ -0,0 +1,12 @@
+/*
+ * This supplies .note.* sections to go into the PT_NOTE inside the vDSO text.
+ * Here we can supply some information useful to userland.
+ */
+
+#include <linux/uts.h>
+#include <linux/version.h>
+#include <linux/elfnote.h>
+
+ELFNOTE_START(Linux, 0, "a")
+	.long LINUX_VERSION_CODE
+ELFNOTE_END

arch/s390/kernel/vdso32/vdso32.lds.S

@@ -0,0 +1,138 @@
+/*
+ * This is the infamous ld script for the 32 bits vdso
+ * library
+ */
+#include <asm/vdso.h>
+
+OUTPUT_FORMAT("elf32-s390", "elf32-s390", "elf32-s390")
+OUTPUT_ARCH(s390:31-bit)
+ENTRY(_start)
+
+SECTIONS
+{
+	. = VDSO32_LBASE + SIZEOF_HEADERS;
+
+	.hash		: { *(.hash) }			:text
+	.gnu.hash	: { *(.gnu.hash) }
+	.dynsym		: { *(.dynsym) }
+	.dynstr		: { *(.dynstr) }
+	.gnu.version	: { *(.gnu.version) }
+	.gnu.version_d	: { *(.gnu.version_d) }
+	.gnu.version_r	: { *(.gnu.version_r) }
+
+	.note		: { *(.note.*) }		:text	:note
+
+	. = ALIGN(16);
+	.text		: {
+		*(.text .stub .text.* .gnu.linkonce.t.*)
+	} :text
+	PROVIDE(__etext = .);
+	PROVIDE(_etext = .);
+	PROVIDE(etext = .);
+
+	/*
+	 * Other stuff is appended to the text segment:
+	 */
+	.rodata		: { *(.rodata .rodata.* .gnu.linkonce.r.*) }
+	.rodata1	: { *(.rodata1) }
+
+	.dynamic	: { *(.dynamic) }		:text	:dynamic
+
+	.eh_frame_hdr	: { *(.eh_frame_hdr) }		:text	:eh_frame_hdr
+	.eh_frame	: { KEEP (*(.eh_frame)) }	:text
+	.gcc_except_table : { *(.gcc_except_table .gcc_except_table.*) }
+
+	.rela.dyn ALIGN(8) : { *(.rela.dyn) }
+	.got ALIGN(8)	: { *(.got .toc) }
+
+	_end = .;
+	PROVIDE(end = .);
+
+	/*
+	 * Stabs debugging sections are here too.
+	 */
+	.stab	       0 : { *(.stab) }
+	.stabstr       0 : { *(.stabstr) }
+	.stab.excl     0 : { *(.stab.excl) }
+	.stab.exclstr  0 : { *(.stab.exclstr) }
+	.stab.index    0 : { *(.stab.index) }
+	.stab.indexstr 0 : { *(.stab.indexstr) }
+	.comment       0 : { *(.comment) }
+
+	/*
+	 * DWARF debug sections.
+	 * Symbols in the DWARF debugging sections are relative to the
+	 * beginning of the section so we begin them at 0.
+	 */
+	/* DWARF 1 */
+	.debug		0 : { *(.debug) }
+	.line		0 : { *(.line) }
+	/* GNU DWARF 1 extensions */
+	.debug_srcinfo	0 : { *(.debug_srcinfo) }
+	.debug_sfnames	0 : { *(.debug_sfnames) }
+	/* DWARF 1.1 and DWARF 2 */
+	.debug_aranges	0 : { *(.debug_aranges) }
+	.debug_pubnames 0 : { *(.debug_pubnames) }
+	/* DWARF 2 */
+	.debug_info	0 : { *(.debug_info .gnu.linkonce.wi.*) }
+	.debug_abbrev	0 : { *(.debug_abbrev) }
+	.debug_line	0 : { *(.debug_line) }
+	.debug_frame	0 : { *(.debug_frame) }
+	.debug_str	0 : { *(.debug_str) }
+	.debug_loc	0 : { *(.debug_loc) }
+	.debug_macinfo	0 : { *(.debug_macinfo) }
+	/* SGI/MIPS DWARF 2 extensions */
+	.debug_weaknames 0 : { *(.debug_weaknames) }
+	.debug_funcnames 0 : { *(.debug_funcnames) }
+	.debug_typenames 0 : { *(.debug_typenames) }
+	.debug_varnames  0 : { *(.debug_varnames) }
+	/* DWARF 3 */
+	.debug_pubtypes 0 : { *(.debug_pubtypes) }
+	.debug_ranges	0 : { *(.debug_ranges) }
+	.gnu.attributes 0 : { KEEP (*(.gnu.attributes)) }
+
+	. = ALIGN(4096);
+	PROVIDE(_vdso_data = .);
+
+	/DISCARD/	: {
+		*(.note.GNU-stack)
+		*(.branch_lt)
+		*(.data .data.* .gnu.linkonce.d.* .sdata*)
+		*(.bss .sbss .dynbss .dynsbss)
+	}
+}
+
+/*
+ * Very old versions of ld do not recognize this name token; use the constant.
+ */
+#define PT_GNU_EH_FRAME	0x6474e550
+
+/*
+ * We must supply the ELF program headers explicitly to get just one
+ * PT_LOAD segment, and set the flags explicitly to make segments read-only.
+ */
+PHDRS
+{
+	text		PT_LOAD FILEHDR PHDRS FLAGS(5);	/* PF_R|PF_X */
+	dynamic		PT_DYNAMIC FLAGS(4);		/* PF_R */
+	note		PT_NOTE FLAGS(4);		/* PF_R */
+	eh_frame_hdr	PT_GNU_EH_FRAME;
+}
+
+/*
+ * This controls what symbols we export from the DSO.
+ */
+VERSION
+{
+	VDSO_VERSION_STRING {
+	global:
+		/*
+		 * Has to be there for the kernel to find
+		 */
+		__kernel_gettimeofday;
+		__kernel_clock_gettime;
+		__kernel_clock_getres;
+
+	local: *;
+	};
+}

arch/s390/kernel/vdso32/vdso32_wrapper.S

@@ -0,0 +1,13 @@
+#include <linux/init.h>
+#include <asm/page.h>
+
+	.section ".data.page_aligned"
+
+	.globl vdso32_start, vdso32_end
+	.balign PAGE_SIZE
+vdso32_start:
+	.incbin "arch/s390/kernel/vdso32/vdso32.so"
+	.balign PAGE_SIZE
+vdso32_end:
+
+	.previous

arch/s390/kernel/vdso64/Makefile

@@ -0,0 +1,55 @@
+# List of files in the vdso, has to be asm only for now
+
+obj-vdso64 = gettimeofday.o clock_getres.o clock_gettime.o note.o
+
+# Build rules
+
+targets := $(obj-vdso64) vdso64.so vdso64.so.dbg
+obj-vdso64 := $(addprefix $(obj)/, $(obj-vdso64))
+
+KBUILD_AFLAGS_64 := $(filter-out -m64,$(KBUILD_AFLAGS))
+KBUILD_AFLAGS_64 += -m64 -s
+
+KBUILD_CFLAGS_64 := $(filter-out -m64,$(KBUILD_CFLAGS))
+KBUILD_CFLAGS_64 += -m64 -fPIC -shared -fno-common -fno-builtin
+KBUILD_CFLAGS_64 += -nostdlib -Wl,-soname=linux-vdso64.so.1 \
+			$(call ld-option, -Wl$(comma)--hash-style=sysv)
+
+$(targets:%=$(obj)/%.dbg): KBUILD_CFLAGS = $(KBUILD_CFLAGS_64)
+$(targets:%=$(obj)/%.dbg): KBUILD_AFLAGS = $(KBUILD_AFLAGS_64)
+
+obj-y += vdso64_wrapper.o
+extra-y += vdso64.lds
+CPPFLAGS_vdso64.lds += -P -C -U$(ARCH)
+
+# Force dependency (incbin is bad)
+$(obj)/vdso64_wrapper.o : $(obj)/vdso64.so
+
+# link rule for the .so file, .lds has to be first
+$(obj)/vdso64.so.dbg: $(src)/vdso64.lds $(obj-vdso64)
+	$(call if_changed,vdso64ld)
+
+# strip rule for the .so file
+$(obj)/%.so: OBJCOPYFLAGS := -S
+$(obj)/%.so: $(obj)/%.so.dbg FORCE
+	$(call if_changed,objcopy)
+
+# assembly rules for the .S files
+$(obj-vdso64): %.o: %.S
+	$(call if_changed_dep,vdso64as)
+
+# actual build commands
+quiet_cmd_vdso64ld = VDSO64L $@
+      cmd_vdso64ld = $(CC) $(c_flags) -Wl,-T $^ -o $@
+quiet_cmd_vdso64as = VDSO64A $@
+      cmd_vdso64as = $(CC) $(a_flags) -c -o $@ $<
+
+# install commands for the unstripped file
+quiet_cmd_vdso_install = INSTALL $@
+      cmd_vdso_install = cp $(obj)/$@.dbg $(MODLIB)/vdso/$@
+
+vdso64.so: $(obj)/vdso64.so.dbg
+	@mkdir -p $(MODLIB)/vdso
+	$(call cmd,vdso_install)
+
+vdso_install: vdso64.so

arch/s390/kernel/vdso64/clock_getres.S

@@ -0,0 +1,39 @@
+/*
+ * Userland implementation of clock_getres() for 64 bits processes in a
+ * s390 kernel for use in the vDSO
+ *
+ *  Copyright IBM Corp. 2008
+ *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License (version 2 only)
+ * as published by the Free Software Foundation.
+ */
+#include <asm/vdso.h>
+#include <asm/asm-offsets.h>
+#include <asm/unistd.h>
+
+	.text
+	.align 4
+	.globl __kernel_clock_getres
+	.type  __kernel_clock_getres,@function
+__kernel_clock_getres:
+	.cfi_startproc
+	cghi	%r2,CLOCK_REALTIME
+	je	0f
+	cghi	%r2,CLOCK_MONOTONIC
+	jne	2f
+0:	ltgr	%r3,%r3
+	jz	1f				/* res == NULL */
+	larl	%r1,3f
+	lg	%r0,0(%r1)
+	xc	0(8,%r3),0(%r3)			/* set tp->tv_sec to zero */
+	stg	%r0,8(%r3)			/* store tp->tv_usec */
+1:	lghi	%r2,0
+	br	%r14
+2:	lghi	%r1,__NR_clock_getres		/* fallback to svc */
+	svc	0
+	br	%r14
+3:	.quad	CLOCK_REALTIME_RES
+	.cfi_endproc
+	.size	__kernel_clock_getres,.-__kernel_clock_getres

arch/s390/kernel/vdso64/clock_gettime.S

@@ -0,0 +1,89 @@
+/*
+ * Userland implementation of clock_gettime() for 64 bits processes in a
+ * s390 kernel for use in the vDSO
+ *
+ *  Copyright IBM Corp. 2008
+ *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License (version 2 only)
+ * as published by the Free Software Foundation.
+ */
+#include <asm/vdso.h>
+#include <asm/asm-offsets.h>
+#include <asm/unistd.h>
+
+	.text
+	.align 4
+	.globl __kernel_clock_gettime
+	.type  __kernel_clock_gettime,@function
+__kernel_clock_gettime:
+	.cfi_startproc
+	larl	%r5,_vdso_data
+	cghi	%r2,CLOCK_REALTIME
+	je	4f
+	cghi	%r2,CLOCK_MONOTONIC
+	jne	9f
+
+	/* CLOCK_MONOTONIC */
+	ltgr	%r3,%r3
+	jz	3f				/* tp == NULL */
+0:	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
+	tmll	%r4,0x0001			/* pending update ? loop */
+	jnz	0b
+	stck	48(%r15)			/* Store TOD clock */
+	lg	%r1,48(%r15)
+	sg	%r1,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
+	mghi	%r1,1000
+	srlg	%r1,%r1,12			/* cyc2ns(clock,cycle_delta) */
+	alg	%r1,__VDSO_XTIME_NSEC(%r5)	/*  + xtime */
+	lg	%r0,__VDSO_XTIME_SEC(%r5)
+	alg	%r1,__VDSO_WTOM_NSEC(%r5)	/*  + wall_to_monotonic */
+	alg	%r0,__VDSO_WTOM_SEC(%r5)
+	clg	%r4,__VDSO_UPD_COUNT(%r5)	/* check update counter */
+	jne	0b
+	larl	%r5,10f
+1:	clg	%r1,0(%r5)
+	jl	2f
+	slg	%r1,0(%r5)
+	aghi	%r0,1
+	j	1b
+2:	stg	%r0,0(%r3)			/* store tp->tv_sec */
+	stg	%r1,8(%r3)			/* store tp->tv_nsec */
+3:	lghi	%r2,0
+	br	%r14
+
+	/* CLOCK_REALTIME */
+4:	ltr	%r3,%r3				/* tp == NULL */
+	jz	8f
+5:	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
+	tmll	%r4,0x0001			/* pending update ? loop */
+	jnz	5b
+	stck	48(%r15)			/* Store TOD clock */
+	lg	%r1,48(%r15)
+	sg	%r1,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
+	mghi	%r1,1000
+	srlg	%r1,%r1,12			/* cyc2ns(clock,cycle_delta) */
+	alg	%r1,__VDSO_XTIME_NSEC(%r5)	/*  + xtime */
+	lg	%r0,__VDSO_XTIME_SEC(%r5)
+	clg	%r4,__VDSO_UPD_COUNT(%r5)	/* check update counter */
+	jne	5b
+	larl	%r5,10f
+6:	clg	%r1,0(%r5)
+	jl	7f
+	slg	%r1,0(%r5)
+	aghi	%r0,1
+	j	6b
+7:	stg	%r0,0(%r3)			/* store tp->tv_sec */
+	stg	%r1,8(%r3)			/* store tp->tv_nsec */
+8:	lghi	%r2,0
+	br	%r14
+
+	/* Fallback to system call */
+9:	lghi	%r1,__NR_clock_gettime
+	svc	0
+	br	%r14
+
+10:	.quad	1000000000
+	.cfi_endproc
+	.size	__kernel_clock_gettime,.-__kernel_clock_gettime

arch/s390/kernel/vdso64/gettimeofday.S

@@ -0,0 +1,56 @@
+/*
+ * Userland implementation of gettimeofday() for 64 bits processes in a
+ * s390 kernel for use in the vDSO
+ *
+ *  Copyright IBM Corp. 2008
+ *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License (version 2 only)
+ * as published by the Free Software Foundation.
+ */
+#include <asm/vdso.h>
+#include <asm/asm-offsets.h>
+#include <asm/unistd.h>
+
+	.text
+	.align 4
+	.globl __kernel_gettimeofday
+	.type  __kernel_gettimeofday,@function
+__kernel_gettimeofday:
+	.cfi_startproc
+	larl	%r5,_vdso_data
+0:	ltgr	%r3,%r3				/* check if tz is NULL */
+	je	1f
+	mvc	0(8,%r3),__VDSO_TIMEZONE(%r5)
+1:	ltgr	%r2,%r2				/* check if tv is NULL */
+	je	4f
+	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
+	tmll	%r4,0x0001			/* pending update ? loop */
+	jnz	0b
+	stck	48(%r15)			/* Store TOD clock */
+	lg	%r1,48(%r15)
+	sg	%r1,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
+	mghi	%r1,1000
+	srlg	%r1,%r1,12			/* cyc2ns(clock,cycle_delta) */
+	alg	%r1,__VDSO_XTIME_NSEC(%r5)	/*  + xtime.tv_nsec */
+	lg	%r0,__VDSO_XTIME_SEC(%r5)	/* xtime.tv_sec */
+	clg	%r4,__VDSO_UPD_COUNT(%r5)	/* check update counter */
+	jne	0b
+	larl	%r5,5f
+2:	clg	%r1,0(%r5)
+	jl	3f
+	slg	%r1,0(%r5)
+	aghi	%r0,1
+	j	2b
+3:	stg	%r0,0(%r2)			/* store tv->tv_sec */
+	slgr	%r0,%r0				/* tv_nsec -> tv_usec */
+	ml	%r0,8(%r5)
+	srlg	%r0,%r0,6
+	stg	%r0,8(%r2)			/* store tv->tv_usec */
+4:	lghi	%r2,0
+	br	%r14
+5:	.quad	1000000000
+	.long	274877907
+	.cfi_endproc
+	.size	__kernel_gettimeofday,.-__kernel_gettimeofday

arch/s390/kernel/vdso64/note.S

@@ -0,0 +1,12 @@
+/*
+ * This supplies .note.* sections to go into the PT_NOTE inside the vDSO text.
+ * Here we can supply some information useful to userland.
+ */
+
+#include <linux/uts.h>
+#include <linux/version.h>
+#include <linux/elfnote.h>
+
+ELFNOTE_START(Linux, 0, "a")
+	.long LINUX_VERSION_CODE
+ELFNOTE_END

arch/s390/kernel/vdso64/vdso64.lds.S

@@ -0,0 +1,138 @@
+/*
+ * This is the infamous ld script for the 64 bits vdso
+ * library
+ */
+#include <asm/vdso.h>
+
+OUTPUT_FORMAT("elf64-s390", "elf64-s390", "elf64-s390")
+OUTPUT_ARCH(s390:64-bit)
+ENTRY(_start)
+
+SECTIONS
+{
+	. = VDSO64_LBASE + SIZEOF_HEADERS;
+
+	.hash		: { *(.hash) }			:text
+	.gnu.hash	: { *(.gnu.hash) }
+	.dynsym		: { *(.dynsym) }
+	.dynstr		: { *(.dynstr) }
+	.gnu.version	: { *(.gnu.version) }
+	.gnu.version_d	: { *(.gnu.version_d) }
+	.gnu.version_r	: { *(.gnu.version_r) }
+
+	.note		: { *(.note.*) }		:text	:note
+
+	. = ALIGN(16);
+	.text		: {
+		*(.text .stub .text.* .gnu.linkonce.t.*)
+	} :text
+	PROVIDE(__etext = .);
+	PROVIDE(_etext = .);
+	PROVIDE(etext = .);
+
+	/*
+	 * Other stuff is appended to the text segment:
+	 */
+	.rodata		: { *(.rodata .rodata.* .gnu.linkonce.r.*) }
+	.rodata1	: { *(.rodata1) }
+
+	.dynamic	: { *(.dynamic) }		:text	:dynamic
+
+	.eh_frame_hdr	: { *(.eh_frame_hdr) }		:text	:eh_frame_hdr
+	.eh_frame	: { KEEP (*(.eh_frame)) }	:text
+	.gcc_except_table : { *(.gcc_except_table .gcc_except_table.*) }
+
+	.rela.dyn ALIGN(8) : { *(.rela.dyn) }
+	.got ALIGN(8)	: { *(.got .toc) }
+
+	_end = .;
+	PROVIDE(end = .);
+
+	/*
+	 * Stabs debugging sections are here too.
+	 */
+	.stab	       0 : { *(.stab) }
+	.stabstr       0 : { *(.stabstr) }
+	.stab.excl     0 : { *(.stab.excl) }
+	.stab.exclstr  0 : { *(.stab.exclstr) }
+	.stab.index    0 : { *(.stab.index) }
+	.stab.indexstr 0 : { *(.stab.indexstr) }
+	.comment       0 : { *(.comment) }
+
+	/*
+	 * DWARF debug sections.
+	 * Symbols in the DWARF debugging sections are relative to the
+	 * beginning of the section so we begin them at 0.
+	 */
+	/* DWARF 1 */
+	.debug		0 : { *(.debug) }
+	.line		0 : { *(.line) }
+	/* GNU DWARF 1 extensions */
+	.debug_srcinfo	0 : { *(.debug_srcinfo) }
+	.debug_sfnames	0 : { *(.debug_sfnames) }
+	/* DWARF 1.1 and DWARF 2 */
+	.debug_aranges	0 : { *(.debug_aranges) }
+	.debug_pubnames 0 : { *(.debug_pubnames) }
+	/* DWARF 2 */
+	.debug_info	0 : { *(.debug_info .gnu.linkonce.wi.*) }
+	.debug_abbrev	0 : { *(.debug_abbrev) }
+	.debug_line	0 : { *(.debug_line) }
+	.debug_frame	0 : { *(.debug_frame) }
+	.debug_str	0 : { *(.debug_str) }
+	.debug_loc	0 : { *(.debug_loc) }
+	.debug_macinfo	0 : { *(.debug_macinfo) }
+	/* SGI/MIPS DWARF 2 extensions */
+	.debug_weaknames 0 : { *(.debug_weaknames) }
+	.debug_funcnames 0 : { *(.debug_funcnames) }
+	.debug_typenames 0 : { *(.debug_typenames) }
+	.debug_varnames  0 : { *(.debug_varnames) }
+	/* DWARF 3 */
+	.debug_pubtypes 0 : { *(.debug_pubtypes) }
+	.debug_ranges	0 : { *(.debug_ranges) }
+	.gnu.attributes 0 : { KEEP (*(.gnu.attributes)) }
+
+	. = ALIGN(4096);
+	PROVIDE(_vdso_data = .);
+
+	/DISCARD/	: {
+		*(.note.GNU-stack)
+		*(.branch_lt)
+		*(.data .data.* .gnu.linkonce.d.* .sdata*)
+		*(.bss .sbss .dynbss .dynsbss)
+	}
+}
+
+/*
+ * Very old versions of ld do not recognize this name token; use the constant.
+ */
+#define PT_GNU_EH_FRAME	0x6474e550
+
+/*
+ * We must supply the ELF program headers explicitly to get just one
+ * PT_LOAD segment, and set the flags explicitly to make segments read-only.
+ */
+PHDRS
+{
+	text		PT_LOAD FILEHDR PHDRS FLAGS(5);	/* PF_R|PF_X */
+	dynamic		PT_DYNAMIC FLAGS(4);		/* PF_R */
+	note		PT_NOTE FLAGS(4);		/* PF_R */
+	eh_frame_hdr	PT_GNU_EH_FRAME;
+}
+
+/*
+ * This controls what symbols we export from the DSO.
+ */
+VERSION
+{
+	VDSO_VERSION_STRING {
+	global:
+		/*
+		 * Has to be there for the kernel to find
+		 */
+		__kernel_gettimeofday;
+		__kernel_clock_gettime;
+		__kernel_clock_getres;
+
+	local: *;
+	};
+}

arch/s390/kernel/vdso64/vdso64_wrapper.S

@@ -0,0 +1,13 @@
+#include <linux/init.h>
+#include <asm/page.h>
+
+	.section ".data.page_aligned"
+
+	.globl vdso64_start, vdso64_end
+	.balign PAGE_SIZE
+vdso64_start:
+	.incbin "arch/s390/kernel/vdso64/vdso64.so"
+	.balign PAGE_SIZE
+vdso64_end:
+
+	.previous

arch/s390/kernel/vtime.c

@@ -27,7 +27,6 @@
 static ext_int_info_t ext_int_info_timer;
 static DEFINE_PER_CPU(struct vtimer_queue, virt_cpu_timer);
 
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING
 /*
  * Update process times based on virtual cpu times stored by entry.S
  * to the lowcore fields user_timer, system_timer & steal_clock.
@@ -125,16 +124,6 @@ static inline void set_vtimer(__u64 expires)
 	/* store expire time for this CPU timer */
 	__get_cpu_var(virt_cpu_timer).to_expire = expires;
 }
-#else
-static inline void set_vtimer(__u64 expires)
-{
-	S390_lowcore.last_update_timer = expires;
-	asm volatile ("SPT %0" : : "m" (S390_lowcore.last_update_timer));
-
-	/* store expire time for this CPU timer */
-	__get_cpu_var(virt_cpu_timer).to_expire = expires;
-}
-#endif
 
 void vtime_start_cpu_timer(void)
 {

arch/s390/mm/extmem.c

@@ -7,6 +7,9 @@
  * (C) IBM Corporation 2002-2004
  */
 
+#define KMSG_COMPONENT "extmem"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/kernel.h>
 #include <linux/string.h>
 #include <linux/spinlock.h>
@@ -24,19 +27,6 @@
 #include <asm/cpcmd.h>
 #include <asm/setup.h>
 
-#define DCSS_DEBUG	/* Debug messages on/off */
-
-#define DCSS_NAME "extmem"
-#ifdef DCSS_DEBUG
-#define PRINT_DEBUG(x...)	printk(KERN_DEBUG DCSS_NAME " debug:" x)
-#else
-#define PRINT_DEBUG(x...)   do {} while (0)
-#endif
-#define PRINT_INFO(x...)	printk(KERN_INFO DCSS_NAME " info:" x)
-#define PRINT_WARN(x...)	printk(KERN_WARNING DCSS_NAME " warning:" x)
-#define PRINT_ERR(x...)		printk(KERN_ERR DCSS_NAME " error:" x)
-
-
 #define DCSS_LOADSHR    0x00
 #define DCSS_LOADNSR    0x04
 #define DCSS_PURGESEG   0x08
@@ -286,7 +276,7 @@ query_segment_type (struct dcss_segment *seg)
 		goto out_free;
 	}
 	if (diag_cc > 1) {
-		PRINT_WARN ("segment_type: diag returned error %ld\n", vmrc);
+		pr_warning("Querying a DCSS type failed with rc=%ld\n", vmrc);
 		rc = dcss_diag_translate_rc (vmrc);
 		goto out_free;
 	}
@@ -368,7 +358,6 @@ query_segment_type (struct dcss_segment *seg)
  * -EIO     : could not perform query diagnose
  * -ENOENT  : no such segment
  * -ENOTSUPP: multi-part segment cannot be used with linux
- * -ENOSPC  : segment cannot be used (overlaps with storage)
  * -ENOMEM  : out of memory
  * 0 .. 6   : type of segment as defined in include/asm-s390/extmem.h
  */
@@ -480,9 +469,8 @@ __segment_load (char *name, int do_nonshared, unsigned long *addr, unsigned long
 		goto out_resource;
 	}
 	if (diag_cc > 1) {
-		PRINT_WARN ("segment_load: could not load segment %s - "
-				"diag returned error (%ld)\n",
-				name, end_addr);
+		pr_warning("Loading DCSS %s failed with rc=%ld\n", name,
+			   end_addr);
 		rc = dcss_diag_translate_rc(end_addr);
 		dcss_diag(&purgeseg_scode, seg->dcss_name,
 				&dummy, &dummy);
@@ -496,15 +484,13 @@ __segment_load (char *name, int do_nonshared, unsigned long *addr, unsigned long
 	*addr = seg->start_addr;
 	*end  = seg->end;
 	if (do_nonshared)
-		PRINT_INFO ("segment_load: loaded segment %s range %p .. %p "
-				"type %s in non-shared mode\n", name,
-				(void*)seg->start_addr, (void*)seg->end,
-				segtype_string[seg->vm_segtype]);
+		pr_info("DCSS %s of range %p to %p and type %s loaded as "
+			"exclusive-writable\n", name, (void*) seg->start_addr,
+			(void*) seg->end, segtype_string[seg->vm_segtype]);
 	else {
-		PRINT_INFO ("segment_load: loaded segment %s range %p .. %p "
-				"type %s in shared mode\n", name,
-				(void*)seg->start_addr, (void*)seg->end,
-				segtype_string[seg->vm_segtype]);
+		pr_info("DCSS %s of range %p to %p and type %s loaded in "
+			"shared access mode\n", name, (void*) seg->start_addr,
+			(void*) seg->end, segtype_string[seg->vm_segtype]);
 	}
 	goto out;
  out_resource:
@@ -593,14 +579,14 @@ segment_modify_shared (char *name, int do_nonshared)
 		goto out_unlock;
 	}
 	if (do_nonshared == seg->do_nonshared) {
-		PRINT_INFO ("segment_modify_shared: not reloading segment %s"
-				" - already in requested mode\n",name);
+		pr_info("DCSS %s is already in the requested access "
+			"mode\n", name);
 		rc = 0;
 		goto out_unlock;
 	}
 	if (atomic_read (&seg->ref_count) != 1) {
-		PRINT_WARN ("segment_modify_shared: not reloading segment %s - "
-				"segment is in use by other driver(s)\n",name);
+		pr_warning("DCSS %s is in use and cannot be reloaded\n",
+			   name);
 		rc = -EAGAIN;
 		goto out_unlock;
 	}
@@ -613,8 +599,8 @@ segment_modify_shared (char *name, int do_nonshared)
 			seg->res->flags |= IORESOURCE_READONLY;
 
 	if (request_resource(&iomem_resource, seg->res)) {
-		PRINT_WARN("segment_modify_shared: could not reload segment %s"
-			   " - overlapping resources\n", name);
+		pr_warning("DCSS %s overlaps with used memory resources "
+			   "and cannot be reloaded\n", name);
 		rc = -EBUSY;
 		kfree(seg->res);
 		goto out_del_mem;
@@ -632,9 +618,8 @@ segment_modify_shared (char *name, int do_nonshared)
 		goto out_del_res;
 	}
 	if (diag_cc > 1) {
-		PRINT_WARN ("segment_modify_shared: could not reload segment %s"
-				" - diag returned error (%ld)\n",
-				name, end_addr);
+		pr_warning("Reloading DCSS %s failed with rc=%ld\n", name,
+			   end_addr);
 		rc = dcss_diag_translate_rc(end_addr);
 		goto out_del_res;
 	}
@@ -673,8 +658,7 @@ segment_unload(char *name)
 	mutex_lock(&dcss_lock);
 	seg = segment_by_name (name);
 	if (seg == NULL) {
-		PRINT_ERR ("could not find segment %s in segment_unload, "
-				"please report to linux390@de.ibm.com\n",name);
+		pr_err("Unloading unknown DCSS %s failed\n", name);
 		goto out_unlock;
 	}
 	if (atomic_dec_return(&seg->ref_count) != 0)
@@ -709,8 +693,7 @@ segment_save(char *name)
 	seg = segment_by_name (name);
 
 	if (seg == NULL) {
-		PRINT_ERR("could not find segment %s in segment_save, please "
-			  "report to linux390@de.ibm.com\n", name);
+		pr_err("Saving unknown DCSS %s failed\n", name);
 		goto out;
 	}
 
@@ -727,14 +710,14 @@ segment_save(char *name)
 	response = 0;
 	cpcmd(cmd1, NULL, 0, &response);
 	if (response) {
-		PRINT_ERR("segment_save: DEFSEG failed with response code %i\n",
-			  response);
+		pr_err("Saving a DCSS failed with DEFSEG response code "
+		       "%i\n", response);
 		goto out;
 	}
 	cpcmd(cmd2, NULL, 0, &response);
 	if (response) {
-		PRINT_ERR("segment_save: SAVESEG failed with response code %i\n",
-			  response);
+		pr_err("Saving a DCSS failed with SAVESEG response code "
+		       "%i\n", response);
 		goto out;
 	}
 out:
@@ -749,44 +732,41 @@ void segment_warning(int rc, char *seg_name)
 {
 	switch (rc) {
 	case -ENOENT:
-		PRINT_WARN("cannot load/query segment %s, "
-			   "does not exist\n", seg_name);
+		pr_err("DCSS %s cannot be loaded or queried\n", seg_name);
 		break;
 	case -ENOSYS:
-		PRINT_WARN("cannot load/query segment %s, "
-			   "not running on VM\n", seg_name);
+		pr_err("DCSS %s cannot be loaded or queried without "
+		       "z/VM\n", seg_name);
 		break;
 	case -EIO:
-		PRINT_WARN("cannot load/query segment %s, "
-			   "hardware error\n", seg_name);
+		pr_err("Loading or querying DCSS %s resulted in a "
+		       "hardware error\n", seg_name);
 		break;
 	case -ENOTSUPP:
-		PRINT_WARN("cannot load/query segment %s, "
-			   "is a multi-part segment\n", seg_name);
+		pr_err("DCSS %s has multiple page ranges and cannot be "
+		       "loaded or queried\n", seg_name);
 		break;
 	case -ENOSPC:
-		PRINT_WARN("cannot load/query segment %s, "
-			   "overlaps with storage\n", seg_name);
+		pr_err("DCSS %s overlaps with used storage and cannot "
+		       "be loaded\n", seg_name);
 		break;
 	case -EBUSY:
-		PRINT_WARN("cannot load/query segment %s, "
-			   "overlaps with already loaded dcss\n", seg_name);
+		pr_err("%s needs used memory resources and cannot be "
+		       "loaded or queried\n", seg_name);
 		break;
 	case -EPERM:
-		PRINT_WARN("cannot load/query segment %s, "
-			   "already loaded in incompatible mode\n", seg_name);
+		pr_err("DCSS %s is already loaded in a different access "
+		       "mode\n", seg_name);
 		break;
 	case -ENOMEM:
-		PRINT_WARN("cannot load/query segment %s, "
-			   "out of memory\n", seg_name);
+		pr_err("There is not enough memory to load or query "
+		       "DCSS %s\n", seg_name);
 		break;
 	case -ERANGE:
-		PRINT_WARN("cannot load/query segment %s, "
-			   "exceeds kernel mapping range\n", seg_name);
+		pr_err("DCSS %s exceeds the kernel mapping range (%lu) "
+		       "and cannot be loaded\n", seg_name, VMEM_MAX_PHYS);
 		break;
 	default:
-		PRINT_WARN("cannot load/query segment %s, "
-			   "return value %i\n", seg_name, rc);
 		break;
 	}
 }

arch/sh/include/asm/elf.h

@@ -202,7 +202,7 @@ do {									\
 #define ARCH_HAS_SETUP_ADDITIONAL_PAGES
 struct linux_binprm;
 extern int arch_setup_additional_pages(struct linux_binprm *bprm,
-				       int executable_stack);
+				       int uses_interp);
 
 extern unsigned int vdso_enabled;
 extern void __kernel_vsyscall;

arch/sh/kernel/vsyscall/vsyscall.c

@@ -59,8 +59,7 @@ int __init vsyscall_init(void)
 }
 
 /* Setup a VMA at program startup for the vsyscall page */
-int arch_setup_additional_pages(struct linux_binprm *bprm,
-				int executable_stack)
+int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 {
 	struct mm_struct *mm = current->mm;
 	unsigned long addr;

arch/x86/include/asm/elf.h

@@ -325,7 +325,7 @@ struct linux_binprm;
 
 #define ARCH_HAS_SETUP_ADDITIONAL_PAGES 1
 extern int arch_setup_additional_pages(struct linux_binprm *bprm,
-				       int executable_stack);
+				       int uses_interp);
 
 extern int syscall32_setup_pages(struct linux_binprm *, int exstack);
 #define compat_arch_setup_additional_pages	syscall32_setup_pages

arch/x86/vdso/vdso32-setup.c

@@ -310,7 +310,7 @@ int __init sysenter_setup(void)
 }
 
 /* Setup a VMA at program startup for the vsyscall page */
-int arch_setup_additional_pages(struct linux_binprm *bprm, int exstack)
+int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 {
 	struct mm_struct *mm = current->mm;
 	unsigned long addr;

arch/x86/vdso/vma.c

@@ -98,7 +98,7 @@ static unsigned long vdso_addr(unsigned long start, unsigned len)
 
 /* Setup a VMA at program startup for the vsyscall page.
    Not called for compat tasks */
-int arch_setup_additional_pages(struct linux_binprm *bprm, int exstack)
+int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 {
 	struct mm_struct *mm = current->mm;
 	unsigned long addr;

drivers/char/Kconfig

@@ -622,6 +622,16 @@ config HVC_BEAT
 	help
 	  Toshiba's Cell Reference Set Beat Console device driver
 
+config HVC_IUCV
+	bool "z/VM IUCV Hypervisor console support (VM only)"
+	depends on S390
+	select HVC_DRIVER
+	select IUCV
+	default y
+	help
+	  This driver provides a Hypervisor console (HVC) back-end to access
+	  a Linux (console) terminal via a z/VM IUCV communication path.
+
 config HVC_XEN
 	bool "Xen Hypervisor Console support"
 	depends on XEN

drivers/char/Makefile

@@ -50,6 +50,7 @@ obj-$(CONFIG_HVC_BEAT)		+= hvc_beat.o
 obj-$(CONFIG_HVC_DRIVER)	+= hvc_console.o
 obj-$(CONFIG_HVC_IRQ)		+= hvc_irq.o
 obj-$(CONFIG_HVC_XEN)		+= hvc_xen.o
+obj-$(CONFIG_HVC_IUCV)		+= hvc_iucv.o
 obj-$(CONFIG_VIRTIO_CONSOLE)	+= virtio_console.o
 obj-$(CONFIG_RAW_DRIVER)	+= raw.o
 obj-$(CONFIG_SGI_SNSC)		+= snsc.o snsc_event.o

drivers/char/hvc_iucv.c

@@ -0,0 +1,850 @@
+/*
+ * hvc_iucv.c - z/VM IUCV back-end for the Hypervisor Console (HVC)
+ *
+ * This back-end for HVC provides terminal access via
+ * z/VM IUCV communication paths.
+ *
+ * Copyright IBM Corp. 2008.
+ *
+ * Author(s):	Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
+ */
+#define KMSG_COMPONENT		"hvc_iucv"
+
+#include <linux/types.h>
+#include <asm/ebcdic.h>
+#include <linux/mempool.h>
+#include <linux/module.h>
+#include <linux/tty.h>
+#include <net/iucv/iucv.h>
+
+#include "hvc_console.h"
+
+
+/* HVC backend for z/VM IUCV */
+#define HVC_IUCV_MAGIC		0xc9e4c3e5
+#define MAX_HVC_IUCV_LINES	HVC_ALLOC_TTY_ADAPTERS
+#define MEMPOOL_MIN_NR		(PAGE_SIZE / sizeof(struct iucv_tty_buffer)/4)
+
+/* IUCV TTY message  */
+#define MSG_VERSION		0x02	/* Message version */
+#define MSG_TYPE_ERROR		0x01	/* Error message */
+#define MSG_TYPE_TERMENV	0x02	/* Terminal environment variable */
+#define MSG_TYPE_TERMIOS	0x04	/* Terminal IO struct update */
+#define MSG_TYPE_WINSIZE	0x08	/* Terminal window size update */
+#define MSG_TYPE_DATA		0x10	/* Terminal data */
+
+#define MSG_SIZE(s)		((s) + offsetof(struct iucv_tty_msg, data))
+struct iucv_tty_msg {
+	u8	version;		/* Message version */
+	u8	type;			/* Message type */
+#define MSG_MAX_DATALEN		(~(u16)0)
+	u16	datalen;		/* Payload length */
+	u8	data[];			/* Payload buffer */
+} __attribute__((packed));
+
+enum iucv_state_t {
+	IUCV_DISCONN	= 0,
+	IUCV_CONNECTED	= 1,
+	IUCV_SEVERED	= 2,
+};
+
+enum tty_state_t {
+	TTY_CLOSED	= 0,
+	TTY_OPENED	= 1,
+};
+
+struct hvc_iucv_private {
+	struct hvc_struct	*hvc; /* HVC console struct reference */
+	u8			srv_name[8];	/* IUCV service name (ebcdic) */
+	enum iucv_state_t	iucv_state;	/* IUCV connection status */
+	enum tty_state_t	tty_state;	/* TTY status */
+	struct iucv_path	*path;		/* IUCV path pointer */
+	spinlock_t		lock;		/* hvc_iucv_private lock */
+	struct list_head	tty_outqueue;	/* outgoing IUCV messages */
+	struct list_head	tty_inqueue;	/* incoming IUCV messages */
+};
+
+struct iucv_tty_buffer {
+	struct list_head	list;	/* list pointer */
+	struct iucv_message	msg;	/* store an incoming IUCV message */
+	size_t			offset;	/* data buffer offset */
+	struct iucv_tty_msg	*mbuf;	/* buffer to store input/output data */
+};
+
+/* IUCV callback handler */
+static	int hvc_iucv_path_pending(struct iucv_path *, u8[8], u8[16]);
+static void hvc_iucv_path_severed(struct iucv_path *, u8[16]);
+static void hvc_iucv_msg_pending(struct iucv_path *, struct iucv_message *);
+static void hvc_iucv_msg_complete(struct iucv_path *, struct iucv_message *);
+
+
+/* Kernel module parameters */
+static unsigned long hvc_iucv_devices;
+
+/* Array of allocated hvc iucv tty lines... */
+static struct hvc_iucv_private *hvc_iucv_table[MAX_HVC_IUCV_LINES];
+
+/* Kmem cache and mempool for iucv_tty_buffer elements */
+static struct kmem_cache *hvc_iucv_buffer_cache;
+static mempool_t *hvc_iucv_mempool;
+
+/* IUCV handler callback functions */
+static struct iucv_handler hvc_iucv_handler = {
+	.path_pending  = hvc_iucv_path_pending,
+	.path_severed  = hvc_iucv_path_severed,
+	.message_complete = hvc_iucv_msg_complete,
+	.message_pending  = hvc_iucv_msg_pending,
+};
+
+
+/**
+ * hvc_iucv_get_private() - Return a struct hvc_iucv_private instance.
+ * @num:	The HVC virtual terminal number (vtermno)
+ *
+ * This function returns the struct hvc_iucv_private instance that corresponds
+ * to the HVC virtual terminal number specified as parameter @num.
+ */
+struct hvc_iucv_private *hvc_iucv_get_private(uint32_t num)
+{
+	if ((num < HVC_IUCV_MAGIC) || (num - HVC_IUCV_MAGIC > hvc_iucv_devices))
+		return NULL;
+	return hvc_iucv_table[num - HVC_IUCV_MAGIC];
+}
+
+/**
+ * alloc_tty_buffer() - Returns a new struct iucv_tty_buffer element.
+ * @size:	Size of the internal buffer used to store data.
+ * @flags:	Memory allocation flags passed to mempool.
+ *
+ * This function allocates a new struct iucv_tty_buffer element and, optionally,
+ * allocates an internal data buffer with the specified size @size.
+ * Note: The total message size arises from the internal buffer size and the
+ *	 members of the iucv_tty_msg structure.
+ *
+ * The function returns NULL if memory allocation has failed.
+ */
+static struct iucv_tty_buffer *alloc_tty_buffer(size_t size, gfp_t flags)
+{
+	struct iucv_tty_buffer *bufp;
+
+	bufp = mempool_alloc(hvc_iucv_mempool, flags);
+	if (!bufp)
+		return NULL;
+	memset(bufp, 0, sizeof(struct iucv_tty_buffer));
+
+	if (size > 0) {
+		bufp->msg.length = MSG_SIZE(size);
+		bufp->mbuf = kmalloc(bufp->msg.length, flags);
+		if (!bufp->mbuf) {
+			mempool_free(bufp, hvc_iucv_mempool);
+			return NULL;
+		}
+		bufp->mbuf->version = MSG_VERSION;
+		bufp->mbuf->type    = MSG_TYPE_DATA;
+		bufp->mbuf->datalen = (u16) size;
+	}
+	return bufp;
+}
+
+/**
+ * destroy_tty_buffer() - destroy struct iucv_tty_buffer element.
+ * @bufp:	Pointer to a struct iucv_tty_buffer element, SHALL NOT be NULL.
+ *
+ * The destroy_tty_buffer() function frees the internal data buffer and returns
+ * the struct iucv_tty_buffer element back to the mempool for freeing.
+ */
+static void destroy_tty_buffer(struct iucv_tty_buffer *bufp)
+{
+	kfree(bufp->mbuf);
+	mempool_free(bufp, hvc_iucv_mempool);
+}
+
+/**
+ * destroy_tty_buffer_list() - call destroy_tty_buffer() for each list element.
+ * @list:	List head pointer to a list containing struct iucv_tty_buffer
+ *		elements.
+ *
+ * Calls destroy_tty_buffer() for each struct iucv_tty_buffer element in the
+ * list @list.
+ */
+static void destroy_tty_buffer_list(struct list_head *list)
+{
+	struct iucv_tty_buffer *ent, *next;
+
+	list_for_each_entry_safe(ent, next, list, list) {
+		list_del(&ent->list);
+		destroy_tty_buffer(ent);
+	}
+}
+
+/**
+ * hvc_iucv_write() - Receive IUCV message write data to HVC console buffer.
+ * @priv:		Pointer to hvc_iucv_private structure.
+ * @buf:		HVC console buffer for writing received terminal data.
+ * @count:		HVC console buffer size.
+ * @has_more_data:	Pointer to an int variable.
+ *
+ * The function picks up pending messages from the input queue and receives
+ * the message data that is then written to the specified buffer @buf.
+ * If the buffer size @count is less than the data message size, then the
+ * message is kept on the input queue and @has_more_data is set to 1.
+ * If the message data has been entirely written, the message is removed from
+ * the input queue.
+ *
+ * The function returns the number of bytes written to the terminal, zero if
+ * there are no pending data messages available or if there is no established
+ * IUCV path.
+ * If the IUCV path has been severed, then -EPIPE is returned to cause a
+ * hang up (that is issued by the HVC console layer).
+ */
+static int hvc_iucv_write(struct hvc_iucv_private *priv,
+			  char *buf, int count, int *has_more_data)
+{
+	struct iucv_tty_buffer *rb;
+	int written;
+	int rc;
+
+	/* Immediately return if there is no IUCV connection */
+	if (priv->iucv_state == IUCV_DISCONN)
+		return 0;
+
+	/* If the IUCV path has been severed, return -EPIPE to inform the
+	 * hvc console layer to hang up the tty device. */
+	if (priv->iucv_state == IUCV_SEVERED)
+		return -EPIPE;
+
+	/* check if there are pending messages */
+	if (list_empty(&priv->tty_inqueue))
+		return 0;
+
+	/* receive a iucv message and flip data to the tty (ldisc) */
+	rb = list_first_entry(&priv->tty_inqueue, struct iucv_tty_buffer, list);
+
+	written = 0;
+	if (!rb->mbuf) { /* message not yet received ... */
+		/* allocate mem to store msg data; if no memory is available
+		 * then leave the buffer on the list and re-try later */
+		rb->mbuf = kmalloc(rb->msg.length, GFP_ATOMIC);
+		if (!rb->mbuf)
+			return -ENOMEM;
+
+		rc = __iucv_message_receive(priv->path, &rb->msg, 0,
+					    rb->mbuf, rb->msg.length, NULL);
+		switch (rc) {
+		case 0: /* Successful	    */
+			break;
+		case 2:	/* No message found */
+		case 9: /* Message purged   */
+			break;
+		default:
+			written = -EIO;
+		}
+		/* remove buffer if an error has occured or received data
+		 * is not correct */
+		if (rc || (rb->mbuf->version != MSG_VERSION) ||
+			  (rb->msg.length    != MSG_SIZE(rb->mbuf->datalen)))
+			goto out_remove_buffer;
+	}
+
+	switch (rb->mbuf->type) {
+	case MSG_TYPE_DATA:
+		written = min_t(int, rb->mbuf->datalen - rb->offset, count);
+		memcpy(buf, rb->mbuf->data + rb->offset, written);
+		if (written < (rb->mbuf->datalen - rb->offset)) {
+			rb->offset += written;
+			*has_more_data = 1;
+			goto out_written;
+		}
+		break;
+
+	case MSG_TYPE_WINSIZE:
+		if (rb->mbuf->datalen != sizeof(struct winsize))
+			break;
+		hvc_resize(priv->hvc, *((struct winsize *)rb->mbuf->data));
+		break;
+
+	case MSG_TYPE_ERROR:	/* ignored ... */
+	case MSG_TYPE_TERMENV:	/* ignored ... */
+	case MSG_TYPE_TERMIOS:	/* ignored ... */
+		break;
+	}
+
+out_remove_buffer:
+	list_del(&rb->list);
+	destroy_tty_buffer(rb);
+	*has_more_data = !list_empty(&priv->tty_inqueue);
+
+out_written:
+	return written;
+}
+
+/**
+ * hvc_iucv_get_chars() - HVC get_chars operation.
+ * @vtermno:	HVC virtual terminal number.
+ * @buf:	Pointer to a buffer to store data
+ * @count:	Size of buffer available for writing
+ *
+ * The hvc_console thread calls this method to read characters from
+ * the terminal backend. If an IUCV communication path has been established,
+ * pending IUCV messages are received and data is copied into buffer @buf
+ * up to @count bytes.
+ *
+ * Locking:	The routine gets called under an irqsave() spinlock; and
+ *		the routine locks the struct hvc_iucv_private->lock to call
+ *		helper functions.
+ */
+static int hvc_iucv_get_chars(uint32_t vtermno, char *buf, int count)
+{
+	struct hvc_iucv_private *priv = hvc_iucv_get_private(vtermno);
+	int written;
+	int has_more_data;
+
+	if (count <= 0)
+		return 0;
+
+	if (!priv)
+		return -ENODEV;
+
+	spin_lock(&priv->lock);
+	has_more_data = 0;
+	written = hvc_iucv_write(priv, buf, count, &has_more_data);
+	spin_unlock(&priv->lock);
+
+	/* if there are still messages on the queue... schedule another run */
+	if (has_more_data)
+		hvc_kick();
+
+	return written;
+}
+
+/**
+ * hvc_iucv_send() - Send an IUCV message containing terminal data.
+ * @priv:	Pointer to struct hvc_iucv_private instance.
+ * @buf:	Buffer containing data to send.
+ * @size:	Size of buffer and amount of data to send.
+ *
+ * If an IUCV communication path is established, the function copies the buffer
+ * data to a newly allocated struct iucv_tty_buffer element, sends the data and
+ * puts the element to the outqueue.
+ *
+ * If there is no IUCV communication path established, the function returns 0.
+ * If an existing IUCV communicaton path has been severed, the function returns
+ * -EPIPE (can be passed to HVC layer to cause a tty hangup).
+ */
+static int hvc_iucv_send(struct hvc_iucv_private *priv, const char *buf,
+			 int count)
+{
+	struct iucv_tty_buffer *sb;
+	int rc;
+	u16 len;
+
+	if (priv->iucv_state == IUCV_SEVERED)
+		return -EPIPE;
+
+	if (priv->iucv_state == IUCV_DISCONN)
+		return 0;
+
+	len = min_t(u16, MSG_MAX_DATALEN, count);
+
+	/* allocate internal buffer to store msg data and also compute total
+	 * message length */
+	sb = alloc_tty_buffer(len, GFP_ATOMIC);
+	if (!sb)
+		return -ENOMEM;
+
+	sb->mbuf->datalen = len;
+	memcpy(sb->mbuf->data, buf, len);
+
+	list_add_tail(&sb->list, &priv->tty_outqueue);
+
+	rc = __iucv_message_send(priv->path, &sb->msg, 0, 0,
+				 (void *) sb->mbuf, sb->msg.length);
+	if (rc) {
+		list_del(&sb->list);
+		destroy_tty_buffer(sb);
+		len = 0;
+	}
+
+	return len;
+}
+
+/**
+ * hvc_iucv_put_chars() - HVC put_chars operation.
+ * @vtermno:	HVC virtual terminal number.
+ * @buf:	Pointer to an buffer to read data from
+ * @count:	Size of buffer available for reading
+ *
+ * The hvc_console thread calls this method to write characters from
+ * to the terminal backend.
+ * The function calls hvc_iucv_send() under the lock of the
+ * struct hvc_iucv_private instance that corresponds to the tty @vtermno.
+ *
+ * Locking:	The method gets called under an irqsave() spinlock; and
+ *		locks struct hvc_iucv_private->lock.
+ */
+static int hvc_iucv_put_chars(uint32_t vtermno, const char *buf, int count)
+{
+	struct hvc_iucv_private *priv = hvc_iucv_get_private(vtermno);
+	int sent;
+
+	if (count <= 0)
+		return 0;
+
+	if (!priv)
+		return -ENODEV;
+
+	spin_lock(&priv->lock);
+	sent = hvc_iucv_send(priv, buf, count);
+	spin_unlock(&priv->lock);
+
+	return sent;
+}
+
+/**
+ * hvc_iucv_notifier_add() - HVC notifier for opening a TTY for the first time.
+ * @hp:	Pointer to the HVC device (struct hvc_struct)
+ * @id:	Additional data (originally passed to hvc_alloc): the index of an struct
+ *	hvc_iucv_private instance.
+ *
+ * The function sets the tty state to TTY_OPEN for the struct hvc_iucv_private
+ * instance that is derived from @id. Always returns 0.
+ *
+ * Locking:	struct hvc_iucv_private->lock, spin_lock_bh
+ */
+static int hvc_iucv_notifier_add(struct hvc_struct *hp, int id)
+{
+	struct hvc_iucv_private *priv;
+
+	priv = hvc_iucv_get_private(id);
+	if (!priv)
+		return 0;
+
+	spin_lock_bh(&priv->lock);
+	priv->tty_state = TTY_OPENED;
+	spin_unlock_bh(&priv->lock);
+
+	return 0;
+}
+
+/**
+ * hvc_iucv_cleanup() - Clean up function if the tty portion is finally closed.
+ * @priv:	Pointer to the struct hvc_iucv_private instance.
+ *
+ * The functions severs the established IUCV communication path (if any), and
+ * destroy struct iucv_tty_buffer elements from the in- and outqueue. Finally,
+ * the functions resets the states to TTY_CLOSED and IUCV_DISCONN.
+ */
+static void hvc_iucv_cleanup(struct hvc_iucv_private *priv)
+{
+	destroy_tty_buffer_list(&priv->tty_outqueue);
+	destroy_tty_buffer_list(&priv->tty_inqueue);
+
+	priv->tty_state = TTY_CLOSED;
+	priv->iucv_state = IUCV_DISCONN;
+}
+
+/**
+ * hvc_iucv_notifier_hangup() - HVC notifier for tty hangups.
+ * @hp: Pointer to the HVC device (struct hvc_struct)
+ * @id: Additional data (originally passed to hvc_alloc): the index of an struct
+ *	hvc_iucv_private instance.
+ *
+ * This routine notifies the HVC backend that a tty hangup (carrier loss,
+ * virtual or otherwise) has occured.
+ *
+ * The HVC backend for z/VM IUCV ignores virtual hangups (vhangup()), to keep
+ * an existing IUCV communication path established.
+ * (Background: vhangup() is called from user space (by getty or login) to
+ *		disable writing to the tty by other applications).
+ *
+ * If the tty has been opened (e.g. getty) and an established IUCV path has been
+ * severed (we caused the tty hangup in that case), then the functions invokes
+ * hvc_iucv_cleanup() to clean up.
+ *
+ * Locking:	struct hvc_iucv_private->lock
+ */
+static void hvc_iucv_notifier_hangup(struct hvc_struct *hp, int id)
+{
+	struct hvc_iucv_private *priv;
+
+	priv = hvc_iucv_get_private(id);
+	if (!priv)
+		return;
+
+	spin_lock_bh(&priv->lock);
+	/* NOTE: If the hangup was scheduled by ourself (from the iucv
+	 *	 path_servered callback [IUCV_SEVERED]), then we have to
+	 *	 finally clean up the tty backend structure and set state to
+	 *	 TTY_CLOSED.
+	 *
+	 *	 If the tty was hung up otherwise (e.g. vhangup()), then we
+	 *	 ignore this hangup and keep an established IUCV path open...
+	 *	 (...the reason is that we are not able to connect back to the
+	 *	 client if we disconnect on hang up) */
+	priv->tty_state = TTY_CLOSED;
+
+	if (priv->iucv_state == IUCV_SEVERED)
+		hvc_iucv_cleanup(priv);
+	spin_unlock_bh(&priv->lock);
+}
+
+/**
+ * hvc_iucv_notifier_del() - HVC notifier for closing a TTY for the last time.
+ * @hp:		Pointer to the HVC device (struct hvc_struct)
+ * @id:		Additional data (originally passed to hvc_alloc):
+ *		the index of an struct hvc_iucv_private instance.
+ *
+ * This routine notifies the HVC backend that the last tty device file
+ * descriptor has been closed.
+ * The function calls hvc_iucv_cleanup() to clean up the struct hvc_iucv_private
+ * instance.
+ *
+ * Locking:	struct hvc_iucv_private->lock
+ */
+static void hvc_iucv_notifier_del(struct hvc_struct *hp, int id)
+{
+	struct hvc_iucv_private *priv;
+	struct iucv_path	*path;
+
+	priv = hvc_iucv_get_private(id);
+	if (!priv)
+		return;
+
+	spin_lock_bh(&priv->lock);
+	path = priv->path;		/* save reference to IUCV path */
+	priv->path = NULL;
+	hvc_iucv_cleanup(priv);
+	spin_unlock_bh(&priv->lock);
+
+	/* sever IUCV path outside of priv->lock due to lock ordering of:
+	 * priv->lock <--> iucv_table_lock */
+	if (path) {
+		iucv_path_sever(path, NULL);
+		iucv_path_free(path);
+	}
+}
+
+/**
+ * hvc_iucv_path_pending() - IUCV handler to process a connection request.
+ * @path:	Pending path (struct iucv_path)
+ * @ipvmid:	Originator z/VM system identifier
+ * @ipuser:	User specified data for this path
+ *		(AF_IUCV: port/service name and originator port)
+ *
+ * The function uses the @ipuser data to check to determine if the pending
+ * path belongs to a terminal managed by this HVC backend.
+ * If the check is successful, then an additional check is done to ensure
+ * that a terminal cannot be accessed multiple times (only one connection
+ * to a terminal is allowed). In that particular case, the pending path is
+ * severed. If it is the first connection, the pending path is accepted and
+ * associated to the struct hvc_iucv_private. The iucv state is updated to
+ * reflect that a communication path has been established.
+ *
+ * Returns 0 if the path belongs to a terminal managed by the this HVC backend;
+ * otherwise returns -ENODEV in order to dispatch this path to other handlers.
+ *
+ * Locking:	struct hvc_iucv_private->lock
+ */
+static	int hvc_iucv_path_pending(struct iucv_path *path,
+				  u8 ipvmid[8], u8 ipuser[16])
+{
+	struct hvc_iucv_private *priv;
+	u8 nuser_data[16];
+	int i, rc;
+
+	priv = NULL;
+	for (i = 0; i < hvc_iucv_devices; i++)
+		if (hvc_iucv_table[i] &&
+		    (0 == memcmp(hvc_iucv_table[i]->srv_name, ipuser, 8))) {
+			priv = hvc_iucv_table[i];
+			break;
+		}
+
+	if (!priv)
+		return -ENODEV;
+
+	spin_lock(&priv->lock);
+
+	/* If the terminal is already connected or being severed, then sever
+	 * this path to enforce that there is only ONE established communication
+	 * path per terminal. */
+	if (priv->iucv_state != IUCV_DISCONN) {
+		iucv_path_sever(path, ipuser);
+		iucv_path_free(path);
+		goto out_path_handled;
+	}
+
+	/* accept path */
+	memcpy(nuser_data, ipuser + 8, 8);  /* remote service (for af_iucv) */
+	memcpy(nuser_data + 8, ipuser, 8);  /* local service  (for af_iucv) */
+	path->msglim = 0xffff;		    /* IUCV MSGLIMIT */
+	path->flags &= ~IUCV_IPRMDATA;	    /* TODO: use IUCV_IPRMDATA */
+	rc = iucv_path_accept(path, &hvc_iucv_handler, nuser_data, priv);
+	if (rc) {
+		iucv_path_sever(path, ipuser);
+		iucv_path_free(path);
+		goto out_path_handled;
+	}
+	priv->path = path;
+	priv->iucv_state = IUCV_CONNECTED;
+
+out_path_handled:
+	spin_unlock(&priv->lock);
+	return 0;
+}
+
+/**
+ * hvc_iucv_path_severed() - IUCV handler to process a path sever.
+ * @path:	Pending path (struct iucv_path)
+ * @ipuser:	User specified data for this path
+ *		(AF_IUCV: port/service name and originator port)
+ *
+ * The function also severs the path (as required by the IUCV protocol) and
+ * sets the iucv state to IUCV_SEVERED for the associated struct
+ * hvc_iucv_private instance. Later, the IUCV_SEVERED state triggers a tty
+ * hangup (hvc_iucv_get_chars() / hvc_iucv_write()).
+ *
+ * If tty portion of the HVC is closed then clean up the outqueue in addition.
+ *
+ * Locking:	struct hvc_iucv_private->lock
+ */
+static void hvc_iucv_path_severed(struct iucv_path *path, u8 ipuser[16])
+{
+	struct hvc_iucv_private *priv = path->private;
+
+	spin_lock(&priv->lock);
+	priv->iucv_state = IUCV_SEVERED;
+
+	/* NOTE: If the tty has not yet been opened by a getty program
+	 *	 (e.g. to see console messages), then cleanup the
+	 *	 hvc_iucv_private structure to allow re-connects.
+	 *
+	 *	 If the tty has been opened, the get_chars() callback returns
+	 *	 -EPIPE to signal the hvc console layer to hang up the tty. */
+	priv->path = NULL;
+	if (priv->tty_state == TTY_CLOSED)
+		hvc_iucv_cleanup(priv);
+	spin_unlock(&priv->lock);
+
+	/* finally sever path (outside of priv->lock due to lock ordering) */
+	iucv_path_sever(path, ipuser);
+	iucv_path_free(path);
+}
+
+/**
+ * hvc_iucv_msg_pending() - IUCV handler to process an incoming IUCV message.
+ * @path:	Pending path (struct iucv_path)
+ * @msg:	Pointer to the IUCV message
+ *
+ * The function stores an incoming message on the input queue for later
+ * processing (by hvc_iucv_get_chars() / hvc_iucv_write()).
+ * However, if the tty has not yet been opened, the message is rejected.
+ *
+ * Locking:	struct hvc_iucv_private->lock
+ */
+static void hvc_iucv_msg_pending(struct iucv_path *path,
+				 struct iucv_message *msg)
+{
+	struct hvc_iucv_private *priv = path->private;
+	struct iucv_tty_buffer *rb;
+
+	spin_lock(&priv->lock);
+
+	/* reject messages if tty has not yet been opened */
+	if (priv->tty_state == TTY_CLOSED) {
+		iucv_message_reject(path, msg);
+		goto unlock_return;
+	}
+
+	/* allocate buffer an empty buffer element */
+	rb = alloc_tty_buffer(0, GFP_ATOMIC);
+	if (!rb) {
+		iucv_message_reject(path, msg);
+		goto unlock_return;	/* -ENOMEM */
+	}
+	rb->msg = *msg;
+
+	list_add_tail(&rb->list, &priv->tty_inqueue);
+
+	hvc_kick();	/* wakup hvc console thread */
+
+unlock_return:
+	spin_unlock(&priv->lock);
+}
+
+/**
+ * hvc_iucv_msg_complete() - IUCV handler to process message completion
+ * @path:	Pending path (struct iucv_path)
+ * @msg:	Pointer to the IUCV message
+ *
+ * The function is called upon completion of message delivery and the
+ * message is removed from the outqueue. Additional delivery information
+ * can be found in msg->audit: rejected messages (0x040000 (IPADRJCT)) and
+ * purged messages (0x010000 (IPADPGNR)).
+ *
+ * Locking:	struct hvc_iucv_private->lock
+ */
+static void hvc_iucv_msg_complete(struct iucv_path *path,
+				  struct iucv_message *msg)
+{
+	struct hvc_iucv_private *priv = path->private;
+	struct iucv_tty_buffer	*ent, *next;
+	LIST_HEAD(list_remove);
+
+	spin_lock(&priv->lock);
+	list_for_each_entry_safe(ent, next, &priv->tty_outqueue, list)
+		if (ent->msg.id == msg->id) {
+			list_move(&ent->list, &list_remove);
+			break;
+		}
+	spin_unlock(&priv->lock);
+	destroy_tty_buffer_list(&list_remove);
+}
+
+
+/* HVC operations */
+static struct hv_ops hvc_iucv_ops = {
+	.get_chars = hvc_iucv_get_chars,
+	.put_chars = hvc_iucv_put_chars,
+	.notifier_add = hvc_iucv_notifier_add,
+	.notifier_del = hvc_iucv_notifier_del,
+	.notifier_hangup = hvc_iucv_notifier_hangup,
+};
+
+/**
+ * hvc_iucv_alloc() - Allocates a new struct hvc_iucv_private instance
+ * @id:	hvc_iucv_table index
+ *
+ * This function allocates a new hvc_iucv_private struct and put the
+ * instance into hvc_iucv_table at index @id.
+ * Returns 0 on success; otherwise non-zero.
+ */
+static int __init hvc_iucv_alloc(int id)
+{
+	struct hvc_iucv_private *priv;
+	char name[9];
+	int rc;
+
+	priv = kzalloc(sizeof(struct hvc_iucv_private), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	spin_lock_init(&priv->lock);
+	INIT_LIST_HEAD(&priv->tty_outqueue);
+	INIT_LIST_HEAD(&priv->tty_inqueue);
+
+	/* Finally allocate hvc */
+	priv->hvc = hvc_alloc(HVC_IUCV_MAGIC + id,
+			      HVC_IUCV_MAGIC + id, &hvc_iucv_ops, PAGE_SIZE);
+	if (IS_ERR(priv->hvc)) {
+		rc = PTR_ERR(priv->hvc);
+		kfree(priv);
+		return rc;
+	}
+
+	/* setup iucv related information */
+	snprintf(name, 9, "ihvc%-4d", id);
+	memcpy(priv->srv_name, name, 8);
+	ASCEBC(priv->srv_name, 8);
+
+	hvc_iucv_table[id] = priv;
+	return 0;
+}
+
+/**
+ * hvc_iucv_init() - Initialization of HVC backend for z/VM IUCV
+ */
+static int __init hvc_iucv_init(void)
+{
+	int rc, i;
+
+	if (!MACHINE_IS_VM) {
+		pr_warning("The z/VM IUCV Hypervisor console cannot be "
+			   "used without z/VM.\n");
+		return -ENODEV;
+	}
+
+	if (!hvc_iucv_devices)
+		return -ENODEV;
+
+	if (hvc_iucv_devices > MAX_HVC_IUCV_LINES)
+		return -EINVAL;
+
+	hvc_iucv_buffer_cache = kmem_cache_create(KMSG_COMPONENT,
+					   sizeof(struct iucv_tty_buffer),
+					   0, 0, NULL);
+	if (!hvc_iucv_buffer_cache) {
+		pr_err("Not enough memory for driver initialization "
+			"(rs=%d).\n", 1);
+		return -ENOMEM;
+	}
+
+	hvc_iucv_mempool = mempool_create_slab_pool(MEMPOOL_MIN_NR,
+						    hvc_iucv_buffer_cache);
+	if (!hvc_iucv_mempool) {
+		pr_err("Not enough memory for driver initialization "
+			"(rs=%d).\n", 2);
+		kmem_cache_destroy(hvc_iucv_buffer_cache);
+		return -ENOMEM;
+	}
+
+	/* allocate hvc_iucv_private structs */
+	for (i = 0; i < hvc_iucv_devices; i++) {
+		rc = hvc_iucv_alloc(i);
+		if (rc) {
+			pr_err("Could not create new z/VM IUCV HVC backend "
+				"rc=%d.\n", rc);
+			goto out_error_hvc;
+		}
+	}
+
+	/* register IUCV callback handler */
+	rc = iucv_register(&hvc_iucv_handler, 0);
+	if (rc) {
+		pr_err("Could not register iucv handler (rc=%d).\n", rc);
+		goto out_error_iucv;
+	}
+
+	return 0;
+
+out_error_iucv:
+	iucv_unregister(&hvc_iucv_handler, 0);
+out_error_hvc:
+	for (i = 0; i < hvc_iucv_devices; i++)
+		if (hvc_iucv_table[i]) {
+			if (hvc_iucv_table[i]->hvc)
+				hvc_remove(hvc_iucv_table[i]->hvc);
+			kfree(hvc_iucv_table[i]);
+		}
+	mempool_destroy(hvc_iucv_mempool);
+	kmem_cache_destroy(hvc_iucv_buffer_cache);
+	return rc;
+}
+
+/**
+ * hvc_iucv_console_init() - Early console initialization
+ */
+static	int __init hvc_iucv_console_init(void)
+{
+	if (!MACHINE_IS_VM || !hvc_iucv_devices)
+		return -ENODEV;
+	return hvc_instantiate(HVC_IUCV_MAGIC, 0, &hvc_iucv_ops);
+}
+
+/**
+ * hvc_iucv_config() - Parsing of hvc_iucv=  kernel command line parameter
+ * @val:	Parameter value (numeric)
+ */
+static	int __init hvc_iucv_config(char *val)
+{
+	 return strict_strtoul(val, 10, &hvc_iucv_devices);
+}
+
+
+module_init(hvc_iucv_init);
+console_initcall(hvc_iucv_console_init);
+__setup("hvc_iucv=", hvc_iucv_config);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("HVC back-end for z/VM IUCV.");
+MODULE_AUTHOR("Hendrik Brueckner <brueckner@linux.vnet.ibm.com>");

drivers/s390/block/dasd.c

@@ -1898,15 +1898,19 @@ restart_cb:
 		wait_event(dasd_flush_wq, (cqr->status < DASD_CQR_QUEUED));
 		/* Process finished ERP request. */
 		if (cqr->refers) {
+			spin_lock_bh(&block->queue_lock);
 			__dasd_block_process_erp(block, cqr);
+			spin_unlock_bh(&block->queue_lock);
 			/* restart list_for_xx loop since dasd_process_erp
 			 * might remove multiple elements */
 			goto restart_cb;
 		}
 		/* call the callback function */
+		spin_lock_irq(&block->request_queue_lock);
 		cqr->endclk = get_clock();
 		list_del_init(&cqr->blocklist);
 		__dasd_cleanup_cqr(cqr);
+		spin_unlock_irq(&block->request_queue_lock);
 	}
 	return rc;
 }

drivers/s390/block/dasd_devmap.c

@@ -23,6 +23,7 @@
 
 /* This is ugly... */
 #define PRINTK_HEADER "dasd_devmap:"
+#define DASD_BUS_ID_SIZE 20
 
 #include "dasd_int.h"
 
@@ -41,7 +42,7 @@ EXPORT_SYMBOL_GPL(dasd_page_cache);
  */
 struct dasd_devmap {
 	struct list_head list;
-	char bus_id[BUS_ID_SIZE];
+	char bus_id[DASD_BUS_ID_SIZE];
         unsigned int devindex;
         unsigned short features;
 	struct dasd_device *device;
@@ -94,7 +95,7 @@ dasd_hash_busid(const char *bus_id)
 	int hash, i;
 
 	hash = 0;
-	for (i = 0; (i < BUS_ID_SIZE) && *bus_id; i++, bus_id++)
+	for (i = 0; (i < DASD_BUS_ID_SIZE) && *bus_id; i++, bus_id++)
 		hash += *bus_id;
 	return hash & 0xff;
 }
@@ -301,7 +302,7 @@ dasd_parse_range( char *parsestring ) {
 	int from, from_id0, from_id1;
 	int to, to_id0, to_id1;
 	int features, rc;
-	char bus_id[BUS_ID_SIZE+1], *str;
+	char bus_id[DASD_BUS_ID_SIZE+1], *str;
 
 	str = parsestring;
 	rc = dasd_busid(&str, &from_id0, &from_id1, &from);
@@ -407,14 +408,14 @@ dasd_add_busid(const char *bus_id, int features)
 	devmap = NULL;
 	hash = dasd_hash_busid(bus_id);
 	list_for_each_entry(tmp, &dasd_hashlists[hash], list)
-		if (strncmp(tmp->bus_id, bus_id, BUS_ID_SIZE) == 0) {
+		if (strncmp(tmp->bus_id, bus_id, DASD_BUS_ID_SIZE) == 0) {
 			devmap = tmp;
 			break;
 		}
 	if (!devmap) {
 		/* This bus_id is new. */
 		new->devindex = dasd_max_devindex++;
-		strncpy(new->bus_id, bus_id, BUS_ID_SIZE);
+		strncpy(new->bus_id, bus_id, DASD_BUS_ID_SIZE);
 		new->features = features;
 		new->device = NULL;
 		list_add(&new->list, &dasd_hashlists[hash]);
@@ -439,7 +440,7 @@ dasd_find_busid(const char *bus_id)
 	devmap = ERR_PTR(-ENODEV);
 	hash = dasd_hash_busid(bus_id);
 	list_for_each_entry(tmp, &dasd_hashlists[hash], list) {
-		if (strncmp(tmp->bus_id, bus_id, BUS_ID_SIZE) == 0) {
+		if (strncmp(tmp->bus_id, bus_id, DASD_BUS_ID_SIZE) == 0) {
 			devmap = tmp;
 			break;
 		}
@@ -561,7 +562,7 @@ dasd_create_device(struct ccw_device *cdev)
 	}
 
 	spin_lock_irqsave(get_ccwdev_lock(cdev), flags);
-	cdev->dev.driver_data = device;
+	dev_set_drvdata(&cdev->dev, device);
 	spin_unlock_irqrestore(get_ccwdev_lock(cdev), flags);
 
 	return device;
@@ -597,7 +598,7 @@ dasd_delete_device(struct dasd_device *device)
 
 	/* Disconnect dasd_device structure from ccw_device structure. */
 	spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
-	device->cdev->dev.driver_data = NULL;
+	dev_set_drvdata(&device->cdev->dev, NULL);
 	spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
 
 	/*
@@ -638,7 +639,7 @@ dasd_put_device_wake(struct dasd_device *device)
 struct dasd_device *
 dasd_device_from_cdev_locked(struct ccw_device *cdev)
 {
-	struct dasd_device *device = cdev->dev.driver_data;
+	struct dasd_device *device = dev_get_drvdata(&cdev->dev);
 
 	if (!device)
 		return ERR_PTR(-ENODEV);

drivers/s390/block/dasd_eckd.c

@@ -1496,7 +1496,7 @@ static void dasd_eckd_handle_unsolicited_interrupt(struct dasd_device *device,
 
 
 	/* service information message SIM */
-	if (irb->esw.esw0.erw.cons && (irb->ecw[27] & DASD_SENSE_BIT_0) &&
+	if (irb->esw.esw0.erw.cons && !(irb->ecw[27] & DASD_SENSE_BIT_0) &&
 	    ((irb->ecw[6] & DASD_SIM_SENSE) == DASD_SIM_SENSE)) {
 		dasd_3990_erp_handle_sim(device, irb->ecw);
 		dasd_schedule_device_bh(device);

drivers/s390/block/dasd_proc.c

@@ -180,12 +180,12 @@ dasd_calc_metrics(char *page, char **start, off_t off,
 
 #ifdef CONFIG_DASD_PROFILE
 static char *
-dasd_statistics_array(char *str, unsigned int *array, int shift)
+dasd_statistics_array(char *str, unsigned int *array, int factor)
 {
 	int i;
 
 	for (i = 0; i < 32; i++) {
-		str += sprintf(str, "%7d ", array[i] >> shift);
+		str += sprintf(str, "%7d ", array[i] / factor);
 		if (i == 15)
 			str += sprintf(str, "\n");
 	}
@@ -202,7 +202,7 @@ dasd_statistics_read(char *page, char **start, off_t off,
 #ifdef CONFIG_DASD_PROFILE
 	struct dasd_profile_info_t *prof;
 	char *str;
-	int shift;
+	int factor;
 
 	/* check for active profiling */
 	if (dasd_profile_level == DASD_PROFILE_OFF) {
@@ -214,12 +214,14 @@ dasd_statistics_read(char *page, char **start, off_t off,
 
 	prof = &dasd_global_profile;
 	/* prevent couter 'overflow' on output */
-	for (shift = 0; (prof->dasd_io_reqs >> shift) > 9999999; shift++);
+	for (factor = 1; (prof->dasd_io_reqs / factor) > 9999999;
+	     factor *= 10);
 
 	str = page;
 	str += sprintf(str, "%d dasd I/O requests\n", prof->dasd_io_reqs);
-	str += sprintf(str, "with %d sectors(512B each)\n",
+	str += sprintf(str, "with %u sectors(512B each)\n",
 		       prof->dasd_io_sects);
+	str += sprintf(str, "Scale Factor is  %d\n", factor);
 	str += sprintf(str,
 		       "   __<4	   ___8	   __16	   __32	   __64	   _128	"
 		       "   _256	   _512	   __1k	   __2k	   __4k	   __8k	"
@@ -230,22 +232,22 @@ dasd_statistics_read(char *page, char **start, off_t off,
 		       "   __1G	   __2G	   __4G " "   _>4G\n");
 
 	str += sprintf(str, "Histogram of sizes (512B secs)\n");
-	str = dasd_statistics_array(str, prof->dasd_io_secs, shift);
+	str = dasd_statistics_array(str, prof->dasd_io_secs, factor);
 	str += sprintf(str, "Histogram of I/O times (microseconds)\n");
-	str = dasd_statistics_array(str, prof->dasd_io_times, shift);
+	str = dasd_statistics_array(str, prof->dasd_io_times, factor);
 	str += sprintf(str, "Histogram of I/O times per sector\n");
-	str = dasd_statistics_array(str, prof->dasd_io_timps, shift);
+	str = dasd_statistics_array(str, prof->dasd_io_timps, factor);
 	str += sprintf(str, "Histogram of I/O time till ssch\n");
-	str = dasd_statistics_array(str, prof->dasd_io_time1, shift);
+	str = dasd_statistics_array(str, prof->dasd_io_time1, factor);
 	str += sprintf(str, "Histogram of I/O time between ssch and irq\n");
-	str = dasd_statistics_array(str, prof->dasd_io_time2, shift);
+	str = dasd_statistics_array(str, prof->dasd_io_time2, factor);
 	str += sprintf(str, "Histogram of I/O time between ssch "
 			    "and irq per sector\n");
-	str = dasd_statistics_array(str, prof->dasd_io_time2ps, shift);
+	str = dasd_statistics_array(str, prof->dasd_io_time2ps, factor);
 	str += sprintf(str, "Histogram of I/O time between irq and end\n");
-	str = dasd_statistics_array(str, prof->dasd_io_time3, shift);
+	str = dasd_statistics_array(str, prof->dasd_io_time3, factor);
 	str += sprintf(str, "# of req in chanq at enqueuing (1..32) \n");
-	str = dasd_statistics_array(str, prof->dasd_io_nr_req, shift);
+	str = dasd_statistics_array(str, prof->dasd_io_nr_req, factor);
 	len = str - page;
 #else
 	len = sprintf(page, "Statistics are not activated in this kernel\n");

drivers/s390/block/dcssblk.c

@@ -4,6 +4,9 @@
  * Authors: Carsten Otte, Stefan Weinhuber, Gerald Schaefer
  */
 
+#define KMSG_COMPONENT "dcssblk"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/module.h>
 #include <linux/moduleparam.h>
 #include <linux/ctype.h>
@@ -17,19 +20,10 @@
 #include <linux/interrupt.h>
 #include <asm/s390_rdev.h>
 
-//#define DCSSBLK_DEBUG		/* Debug messages on/off */
 #define DCSSBLK_NAME "dcssblk"
 #define DCSSBLK_MINORS_PER_DISK 1
 #define DCSSBLK_PARM_LEN 400
-
-#ifdef DCSSBLK_DEBUG
-#define PRINT_DEBUG(x...) printk(KERN_DEBUG DCSSBLK_NAME " debug: " x)
-#else
-#define PRINT_DEBUG(x...) do {} while (0)
-#endif
-#define PRINT_INFO(x...)  printk(KERN_INFO DCSSBLK_NAME " info: " x)
-#define PRINT_WARN(x...)  printk(KERN_WARNING DCSSBLK_NAME " warning: " x)
-#define PRINT_ERR(x...)	  printk(KERN_ERR DCSSBLK_NAME " error: " x)
+#define DCSS_BUS_ID_SIZE 20
 
 static int dcssblk_open(struct block_device *bdev, fmode_t mode);
 static int dcssblk_release(struct gendisk *disk, fmode_t mode);
@@ -50,7 +44,7 @@ static struct block_device_operations dcssblk_devops = {
 struct dcssblk_dev_info {
 	struct list_head lh;
 	struct device dev;
-	char segment_name[BUS_ID_SIZE];
+	char segment_name[DCSS_BUS_ID_SIZE];
 	atomic_t use_count;
 	struct gendisk *gd;
 	unsigned long start;
@@ -65,7 +59,7 @@ struct dcssblk_dev_info {
 
 struct segment_info {
 	struct list_head lh;
-	char segment_name[BUS_ID_SIZE];
+	char segment_name[DCSS_BUS_ID_SIZE];
 	unsigned long start;
 	unsigned long end;
 	int segment_type;
@@ -261,10 +255,9 @@ dcssblk_is_continuous(struct dcssblk_dev_info *dev_info)
 	/* check continuity */
 	for (i = 0; i < dev_info->num_of_segments - 1; i++) {
 		if ((sort_list[i].end + 1) != sort_list[i+1].start) {
-			PRINT_ERR("Segment %s is not contiguous with "
-				"segment %s\n",
-				sort_list[i].segment_name,
-				sort_list[i+1].segment_name);
+			pr_err("Adjacent DCSSs %s and %s are not "
+			       "contiguous\n", sort_list[i].segment_name,
+			       sort_list[i+1].segment_name);
 			rc = -EINVAL;
 			goto out;
 		}
@@ -275,10 +268,10 @@ dcssblk_is_continuous(struct dcssblk_dev_info *dev_info)
 				!(sort_list[i+1].segment_type &
 				SEGMENT_EXCLUSIVE) ||
 				(sort_list[i+1].segment_type == SEG_TYPE_ER)) {
-				PRINT_ERR("Segment %s has different type from "
-					"segment %s\n",
-					sort_list[i].segment_name,
-					sort_list[i+1].segment_name);
+				pr_err("DCSS %s and DCSS %s have "
+				       "incompatible types\n",
+				       sort_list[i].segment_name,
+				       sort_list[i+1].segment_name);
 				rc = -EINVAL;
 				goto out;
 			}
@@ -380,8 +373,9 @@ dcssblk_shared_store(struct device *dev, struct device_attribute *attr, const ch
 	} else if (inbuf[0] == '0') {
 		/* reload segments in exclusive mode */
 		if (dev_info->segment_type == SEG_TYPE_SC) {
-			PRINT_ERR("Segment type SC (%s) cannot be loaded in "
-				"non-shared mode\n", dev_info->segment_name);
+			pr_err("DCSS %s is of type SC and cannot be "
+			       "loaded as exclusive-writable\n",
+			       dev_info->segment_name);
 			rc = -EINVAL;
 			goto out;
 		}
@@ -404,9 +398,8 @@ dcssblk_shared_store(struct device *dev, struct device_attribute *attr, const ch
 	goto out;
 
 removeseg:
-	PRINT_ERR("Could not reload segment(s) of the device %s, removing "
-		"segment(s) now!\n",
-		dev_info->segment_name);
+	pr_err("DCSS device %s is removed after a failed access mode "
+	       "change\n", dev_info->segment_name);
 	temp = entry;
 	list_for_each_entry(entry, &dev_info->seg_list, lh) {
 		if (entry != temp)
@@ -454,17 +447,17 @@ dcssblk_save_store(struct device *dev, struct device_attribute *attr, const char
 	if (inbuf[0] == '1') {
 		if (atomic_read(&dev_info->use_count) == 0) {
 			// device is idle => we save immediately
-			PRINT_INFO("Saving segment(s) of the device %s\n",
-				   dev_info->segment_name);
+			pr_info("All DCSSs that map to device %s are "
+				"saved\n", dev_info->segment_name);
 			list_for_each_entry(entry, &dev_info->seg_list, lh) {
 				segment_save(entry->segment_name);
 			}
 		}  else {
 			// device is busy => we save it when it becomes
 			// idle in dcssblk_release
-			PRINT_INFO("Device %s is currently busy, segment(s) "
-				   "will be saved when it becomes idle...\n",
-				   dev_info->segment_name);
+			pr_info("Device %s is in use, its DCSSs will be "
+				"saved when it becomes idle\n",
+				dev_info->segment_name);
 			dev_info->save_pending = 1;
 		}
 	} else if (inbuf[0] == '0') {
@@ -472,9 +465,9 @@ dcssblk_save_store(struct device *dev, struct device_attribute *attr, const char
 			// device is busy & the user wants to undo his save
 			// request
 			dev_info->save_pending = 0;
-			PRINT_INFO("Pending save for segment(s) of the device "
-					"%s deactivated\n",
-					dev_info->segment_name);
+			pr_info("A pending save request for device %s "
+				"has been canceled\n",
+				dev_info->segment_name);
 		}
 	} else {
 		up_write(&dcssblk_devices_sem);
@@ -614,9 +607,8 @@ dcssblk_add_store(struct device *dev, struct device_attribute *attr, const char
 
 	seg_byte_size = (dev_info->end - dev_info->start + 1);
 	set_capacity(dev_info->gd, seg_byte_size >> 9); // size in sectors
-	PRINT_INFO("Loaded segment(s) %s, size = %lu Byte, "
-		   "capacity = %lu (512 Byte) sectors\n", local_buf,
-		   seg_byte_size, seg_byte_size >> 9);
+	pr_info("Loaded %s with total size %lu bytes and capacity %lu "
+		"sectors\n", local_buf, seg_byte_size, seg_byte_size >> 9);
 
 	dev_info->save_pending = 0;
 	dev_info->is_shared = 1;
@@ -744,13 +736,15 @@ dcssblk_remove_store(struct device *dev, struct device_attribute *attr, const ch
 	dev_info = dcssblk_get_device_by_name(local_buf);
 	if (dev_info == NULL) {
 		up_write(&dcssblk_devices_sem);
-		PRINT_WARN("Device %s is not loaded!\n", local_buf);
+		pr_warning("Device %s cannot be removed because it is not a "
+			   "known device\n", local_buf);
 		rc = -ENODEV;
 		goto out_buf;
 	}
 	if (atomic_read(&dev_info->use_count) != 0) {
 		up_write(&dcssblk_devices_sem);
-		PRINT_WARN("Device %s is in use!\n", local_buf);
+		pr_warning("Device %s cannot be removed while it is in "
+			   "use\n", local_buf);
 		rc = -EBUSY;
 		goto out_buf;
 	}
@@ -807,8 +801,8 @@ dcssblk_release(struct gendisk *disk, fmode_t mode)
 	down_write(&dcssblk_devices_sem);
 	if (atomic_dec_and_test(&dev_info->use_count)
 	    && (dev_info->save_pending)) {
-		PRINT_INFO("Device %s became idle and is being saved now\n",
-			    dev_info->segment_name);
+		pr_info("Device %s has become idle and is being saved "
+			"now\n", dev_info->segment_name);
 		list_for_each_entry(entry, &dev_info->seg_list, lh) {
 			segment_save(entry->segment_name);
 		}
@@ -851,7 +845,8 @@ dcssblk_make_request(struct request_queue *q, struct bio *bio)
 		case SEG_TYPE_SC:
 			/* cannot write to these segments */
 			if (bio_data_dir(bio) == WRITE) {
-				PRINT_WARN("rejecting write to ro device %s\n",
+				pr_warning("Writing to %s failed because it "
+					   "is a read-only device\n",
 					   dev_name(&dev_info->dev));
 				goto fail;
 			}

drivers/s390/block/xpram.c

@@ -25,6 +25,9 @@
  *   generic hard disk support to replace ad-hoc partitioning
  */
 
+#define KMSG_COMPONENT "xpram"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/module.h>
 #include <linux/moduleparam.h>
 #include <linux/ctype.h>  /* isdigit, isxdigit */
@@ -42,12 +45,6 @@
 #define XPRAM_DEVS	1	/* one partition */
 #define XPRAM_MAX_DEVS	32	/* maximal number of devices (partitions) */
 
-#define PRINT_DEBUG(x...)	printk(KERN_DEBUG XPRAM_NAME " debug:" x)
-#define PRINT_INFO(x...)	printk(KERN_INFO XPRAM_NAME " info:" x)
-#define PRINT_WARN(x...)	printk(KERN_WARNING XPRAM_NAME " warning:" x)
-#define PRINT_ERR(x...)		printk(KERN_ERR XPRAM_NAME " error:" x)
-
-
 typedef struct {
 	unsigned int	size;		/* size of xpram segment in pages */
 	unsigned int	offset;		/* start page of xpram segment */
@@ -264,7 +261,7 @@ static int __init xpram_setup_sizes(unsigned long pages)
 
 	/* Check number of devices. */
 	if (devs <= 0 || devs > XPRAM_MAX_DEVS) {
-		PRINT_ERR("invalid number %d of devices\n",devs);
+		pr_err("%d is not a valid number of XPRAM devices\n",devs);
 		return -EINVAL;
 	}
 	xpram_devs = devs;
@@ -295,22 +292,22 @@ static int __init xpram_setup_sizes(unsigned long pages)
 			mem_auto_no++;
 	}
 	
-	PRINT_INFO("  number of devices (partitions): %d \n", xpram_devs);
+	pr_info("  number of devices (partitions): %d \n", xpram_devs);
 	for (i = 0; i < xpram_devs; i++) {
 		if (xpram_sizes[i])
-			PRINT_INFO("  size of partition %d: %u kB\n",
-				   i, xpram_sizes[i]);
+			pr_info("  size of partition %d: %u kB\n",
+				i, xpram_sizes[i]);
 		else
-			PRINT_INFO("  size of partition %d to be set "
-				   "automatically\n",i);
+			pr_info("  size of partition %d to be set "
+				"automatically\n",i);
 	}
-	PRINT_DEBUG("  memory needed (for sized partitions): %lu kB\n",
-		    mem_needed);
-	PRINT_DEBUG("  partitions to be sized automatically: %d\n",
-		    mem_auto_no);
+	pr_info("  memory needed (for sized partitions): %lu kB\n",
+		mem_needed);
+	pr_info("  partitions to be sized automatically: %d\n",
+		mem_auto_no);
 
 	if (mem_needed > pages * 4) {
-		PRINT_ERR("Not enough expanded memory available\n");
+		pr_err("Not enough expanded memory available\n");
 		return -EINVAL;
 	}
 
@@ -322,8 +319,8 @@ static int __init xpram_setup_sizes(unsigned long pages)
 	 */
 	if (mem_auto_no) {
 		mem_auto = ((pages - mem_needed / 4) / mem_auto_no) * 4;
-		PRINT_INFO("  automatically determined "
-			   "partition size: %lu kB\n", mem_auto);
+		pr_info("  automatically determined "
+			"partition size: %lu kB\n", mem_auto);
 		for (i = 0; i < xpram_devs; i++)
 			if (xpram_sizes[i] == 0)
 				xpram_sizes[i] = mem_auto;
@@ -405,12 +402,12 @@ static int __init xpram_init(void)
 
 	/* Find out size of expanded memory. */
 	if (xpram_present() != 0) {
-		PRINT_WARN("No expanded memory available\n");
+		pr_err("No expanded memory available\n");
 		return -ENODEV;
 	}
 	xpram_pages = xpram_highest_page_index() + 1;
-	PRINT_INFO("  %u pages expanded memory found (%lu KB).\n",
-		   xpram_pages, (unsigned long) xpram_pages*4);
+	pr_info("  %u pages expanded memory found (%lu KB).\n",
+		xpram_pages, (unsigned long) xpram_pages*4);
 	rc = xpram_setup_sizes(xpram_pages);
 	if (rc)
 		return rc;

drivers/s390/char/monreader.c

@@ -7,6 +7,9 @@
  *   Author: Gerald Schaefer <gerald.schaefer@de.ibm.com>
  */
 
+#define KMSG_COMPONENT "monreader"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/module.h>
 #include <linux/moduleparam.h>
 #include <linux/init.h>
@@ -24,19 +27,6 @@
 #include <asm/ebcdic.h>
 #include <asm/extmem.h>
 
-//#define MON_DEBUG			/* Debug messages on/off */
-
-#define MON_NAME "monreader"
-
-#define P_INFO(x...)	printk(KERN_INFO MON_NAME " info: " x)
-#define P_ERROR(x...)	printk(KERN_ERR MON_NAME " error: " x)
-#define P_WARNING(x...)	printk(KERN_WARNING MON_NAME " warning: " x)
-
-#ifdef MON_DEBUG
-#define P_DEBUG(x...)   printk(KERN_DEBUG MON_NAME " debug: " x)
-#else
-#define P_DEBUG(x...)   do {} while (0)
-#endif
 
 #define MON_COLLECT_SAMPLE 0x80
 #define MON_COLLECT_EVENT  0x40
@@ -172,7 +162,7 @@ static int mon_send_reply(struct mon_msg *monmsg,
 	} else
 		monmsg->replied_msglim = 1;
 	if (rc) {
-		P_ERROR("read, IUCV reply failed with rc = %i\n\n", rc);
+		pr_err("Reading monitor data failed with rc=%i\n", rc);
 		return -EIO;
 	}
 	return 0;
@@ -251,7 +241,8 @@ static void mon_iucv_path_severed(struct iucv_path *path, u8 ipuser[16])
 {
 	struct mon_private *monpriv = path->private;
 
-	P_ERROR("IUCV connection severed with rc = 0x%X\n", ipuser[0]);
+	pr_err("z/VM *MONITOR system service disconnected with rc=%i\n",
+	       ipuser[0]);
 	iucv_path_sever(path, NULL);
 	atomic_set(&monpriv->iucv_severed, 1);
 	wake_up(&mon_conn_wait_queue);
@@ -266,8 +257,7 @@ static void mon_iucv_message_pending(struct iucv_path *path,
 	memcpy(&monpriv->msg_array[monpriv->write_index]->msg,
 	       msg, sizeof(*msg));
 	if (atomic_inc_return(&monpriv->msglim_count) == MON_MSGLIM) {
-		P_WARNING("IUCV message pending, message limit (%i) reached\n",
-			  MON_MSGLIM);
+		pr_warning("The read queue for monitor data is full\n");
 		monpriv->msg_array[monpriv->write_index]->msglim_reached = 1;
 	}
 	monpriv->write_index = (monpriv->write_index + 1) % MON_MSGLIM;
@@ -311,8 +301,8 @@ static int mon_open(struct inode *inode, struct file *filp)
 	rc = iucv_path_connect(monpriv->path, &monreader_iucv_handler,
 			       MON_SERVICE, NULL, user_data_connect, monpriv);
 	if (rc) {
-		P_ERROR("iucv connection to *MONITOR failed with "
-			"IPUSER SEVER code = %i\n", rc);
+		pr_err("Connecting to the z/VM *MONITOR system service "
+		       "failed with rc=%i\n", rc);
 		rc = -EIO;
 		goto out_path;
 	}
@@ -353,7 +343,8 @@ static int mon_close(struct inode *inode, struct file *filp)
 	 */
 	rc = iucv_path_sever(monpriv->path, user_data_sever);
 	if (rc)
-		P_ERROR("close, iucv_sever failed with rc = %i\n", rc);
+		pr_warning("Disconnecting the z/VM *MONITOR system service "
+			   "failed with rc=%i\n", rc);
 
 	atomic_set(&monpriv->iucv_severed, 0);
 	atomic_set(&monpriv->iucv_connected, 0);
@@ -469,7 +460,8 @@ static int __init mon_init(void)
 	int rc;
 
 	if (!MACHINE_IS_VM) {
-		P_ERROR("not running under z/VM, driver not loaded\n");
+		pr_err("The z/VM *MONITOR record device driver cannot be "
+		       "loaded without z/VM\n");
 		return -ENODEV;
 	}
 
@@ -478,7 +470,8 @@ static int __init mon_init(void)
 	 */
 	rc = iucv_register(&monreader_iucv_handler, 1);
 	if (rc) {
-		P_ERROR("failed to register with iucv driver\n");
+		pr_err("The z/VM *MONITOR record device driver failed to "
+		       "register with IUCV\n");
 		return rc;
 	}
 
@@ -488,8 +481,8 @@ static int __init mon_init(void)
 		goto out_iucv;
 	}
 	if (rc != SEG_TYPE_SC) {
-		P_ERROR("segment %s has unsupported type, should be SC\n",
-			mon_dcss_name);
+		pr_err("The specified *MONITOR DCSS %s does not have the "
+		       "required type SC\n", mon_dcss_name);
 		rc = -EINVAL;
 		goto out_iucv;
 	}

drivers/s390/char/monwriter.c

@@ -8,6 +8,9 @@
  * Author(s): Melissa Howland <Melissa.Howland@us.ibm.com>
  */
 
+#define KMSG_COMPONENT "monwriter"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/module.h>
 #include <linux/moduleparam.h>
 #include <linux/init.h>
@@ -64,9 +67,9 @@ static int monwrite_diag(struct monwrite_hdr *myhdr, char *buffer, int fcn)
 	rc = appldata_asm(&id, fcn, (void *) buffer, myhdr->datalen);
 	if (rc <= 0)
 		return rc;
+	pr_err("Writing monitor data failed with rc=%i\n", rc);
 	if (rc == 5)
 		return -EPERM;
-	printk("DIAG X'DC' error with return code: %i\n", rc);
 	return -EINVAL;
 }
 

drivers/s390/char/sclp_cmd.c

@@ -6,6 +6,9 @@
  *		 Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
  */
 
+#define KMSG_COMPONENT "sclp_cmd"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/completion.h>
 #include <linux/init.h>
 #include <linux/errno.h>
@@ -16,9 +19,8 @@
 #include <linux/memory.h>
 #include <asm/chpid.h>
 #include <asm/sclp.h>
-#include "sclp.h"
 
-#define TAG	"sclp_cmd: "
+#include "sclp.h"
 
 #define SCLP_CMDW_READ_SCP_INFO		0x00020001
 #define SCLP_CMDW_READ_SCP_INFO_FORCED	0x00120001
@@ -169,8 +171,8 @@ static int do_sync_request(sclp_cmdw_t cmd, void *sccb)
 
 	/* Check response. */
 	if (request->status != SCLP_REQ_DONE) {
-		printk(KERN_WARNING TAG "sync request failed "
-		       "(cmd=0x%08x, status=0x%02x)\n", cmd, request->status);
+		pr_warning("sync request failed (cmd=0x%08x, "
+			   "status=0x%02x)\n", cmd, request->status);
 		rc = -EIO;
 	}
 out:
@@ -224,8 +226,8 @@ int sclp_get_cpu_info(struct sclp_cpu_info *info)
 	if (rc)
 		goto out;
 	if (sccb->header.response_code != 0x0010) {
-		printk(KERN_WARNING TAG "readcpuinfo failed "
-		       "(response=0x%04x)\n", sccb->header.response_code);
+		pr_warning("readcpuinfo failed (response=0x%04x)\n",
+			   sccb->header.response_code);
 		rc = -EIO;
 		goto out;
 	}
@@ -262,8 +264,9 @@ static int do_cpu_configure(sclp_cmdw_t cmd)
 	case 0x0120:
 		break;
 	default:
-		printk(KERN_WARNING TAG "configure cpu failed (cmd=0x%08x, "
-		       "response=0x%04x)\n", cmd, sccb->header.response_code);
+		pr_warning("configure cpu failed (cmd=0x%08x, "
+			   "response=0x%04x)\n", cmd,
+			   sccb->header.response_code);
 		rc = -EIO;
 		break;
 	}
@@ -626,9 +629,9 @@ static int do_chp_configure(sclp_cmdw_t cmd)
 	case 0x0450:
 		break;
 	default:
-		printk(KERN_WARNING TAG "configure channel-path failed "
-		       "(cmd=0x%08x, response=0x%04x)\n", cmd,
-		       sccb->header.response_code);
+		pr_warning("configure channel-path failed "
+			   "(cmd=0x%08x, response=0x%04x)\n", cmd,
+			   sccb->header.response_code);
 		rc = -EIO;
 		break;
 	}
@@ -695,8 +698,8 @@ int sclp_chp_read_info(struct sclp_chp_info *info)
 	if (rc)
 		goto out;
 	if (sccb->header.response_code != 0x0010) {
-		printk(KERN_WARNING TAG "read channel-path info failed "
-		       "(response=0x%04x)\n", sccb->header.response_code);
+		pr_warning("read channel-path info failed "
+			   "(response=0x%04x)\n", sccb->header.response_code);
 		rc = -EIO;
 		goto out;
 	}

drivers/s390/char/sclp_config.c

@@ -5,15 +5,17 @@
  *    Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>
  */
 
+#define KMSG_COMPONENT "sclp_config"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/init.h>
 #include <linux/errno.h>
 #include <linux/cpu.h>
 #include <linux/sysdev.h>
 #include <linux/workqueue.h>
 #include <asm/smp.h>
-#include "sclp.h"
 
-#define TAG	"sclp_config: "
+#include "sclp.h"
 
 struct conf_mgm_data {
 	u8 reserved;
@@ -31,7 +33,7 @@ static void sclp_cpu_capability_notify(struct work_struct *work)
 	int cpu;
 	struct sys_device *sysdev;
 
-	printk(KERN_WARNING TAG "cpu capability changed.\n");
+	pr_warning("cpu capability changed.\n");
 	get_online_cpus();
 	for_each_online_cpu(cpu) {
 		sysdev = get_cpu_sysdev(cpu);
@@ -78,7 +80,7 @@ static int __init sclp_conf_init(void)
 		return rc;
 
 	if (!(sclp_conf_register.sclp_send_mask & EVTYP_CONFMGMDATA_MASK)) {
-		printk(KERN_WARNING TAG "no configuration management.\n");
+		pr_warning("no configuration management.\n");
 		sclp_unregister(&sclp_conf_register);
 		rc = -ENOSYS;
 	}

drivers/s390/char/sclp_cpi_sys.c

@@ -7,6 +7,9 @@
  *		 Michael Ernst <mernst@de.ibm.com>
  */
 
+#define KMSG_COMPONENT "sclp_cpi"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/kernel.h>
 #include <linux/init.h>
 #include <linux/stat.h>
@@ -20,6 +23,7 @@
 #include <linux/completion.h>
 #include <asm/ebcdic.h>
 #include <asm/sclp.h>
+
 #include "sclp.h"
 #include "sclp_rw.h"
 #include "sclp_cpi_sys.h"
@@ -150,16 +154,16 @@ static int cpi_req(void)
 	wait_for_completion(&completion);
 
 	if (req->status != SCLP_REQ_DONE) {
-		printk(KERN_WARNING "cpi: request failed (status=0x%02x)\n",
-			req->status);
+		pr_warning("request failed (status=0x%02x)\n",
+			   req->status);
 		rc = -EIO;
 		goto out_free_req;
 	}
 
 	response = ((struct cpi_sccb *) req->sccb)->header.response_code;
 	if (response != 0x0020) {
-		printk(KERN_WARNING "cpi: failed with "
-			"response code 0x%x\n", response);
+		pr_warning("request failed with response code 0x%x\n",
+			   response);
 		rc = -EIO;
 	}
 

drivers/s390/char/sclp_sdias.c

@@ -5,15 +5,18 @@
  * Author(s): Michael Holzheu
  */
 
+#define KMSG_COMPONENT "sclp_sdias"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/sched.h>
 #include <asm/sclp.h>
 #include <asm/debug.h>
 #include <asm/ipl.h>
+
 #include "sclp.h"
 #include "sclp_rw.h"
 
 #define TRACE(x...) debug_sprintf_event(sdias_dbf, 1, x)
-#define ERROR_MSG(x...) printk ( KERN_ALERT "SDIAS: " x )
 
 #define SDIAS_RETRIES 300
 #define SDIAS_SLEEP_TICKS 50
@@ -131,7 +134,7 @@ int sclp_sdias_blk_count(void)
 
 	rc = sdias_sclp_send(&request);
 	if (rc) {
-		ERROR_MSG("sclp_send failed for get_nr_blocks\n");
+		pr_err("sclp_send failed for get_nr_blocks\n");
 		goto out;
 	}
 	if (sccb.hdr.response_code != 0x0020) {
@@ -145,7 +148,8 @@ int sclp_sdias_blk_count(void)
 			rc = sccb.evbuf.blk_cnt;
 			break;
 		default:
-			ERROR_MSG("SCLP error: %x\n", sccb.evbuf.event_status);
+			pr_err("SCLP error: %x\n",
+			       sccb.evbuf.event_status);
 			rc = -EIO;
 			goto out;
 	}
@@ -201,7 +205,7 @@ int sclp_sdias_copy(void *dest, int start_blk, int nr_blks)
 
 	rc = sdias_sclp_send(&request);
 	if (rc) {
-		ERROR_MSG("sclp_send failed: %x\n", rc);
+		pr_err("sclp_send failed: %x\n", rc);
 		goto out;
 	}
 	if (sccb.hdr.response_code != 0x0020) {
@@ -219,9 +223,9 @@ int sclp_sdias_copy(void *dest, int start_blk, int nr_blks)
 		case EVSTATE_NO_DATA:
 			TRACE("no data\n");
 		default:
-			ERROR_MSG("Error from SCLP while copying hsa. "
-				  "Event status = %x\n",
-				sccb.evbuf.event_status);
+			pr_err("Error from SCLP while copying hsa. "
+			       "Event status = %x\n",
+			       sccb.evbuf.event_status);
 			rc = -EIO;
 	}
 out:

drivers/s390/char/sclp_vt220.c

@@ -583,23 +583,6 @@ sclp_vt220_chars_in_buffer(struct tty_struct *tty)
 	return count;
 }
 
-static void
-__sclp_vt220_flush_buffer(void)
-{
-	unsigned long flags;
-
-	sclp_vt220_emit_current();
-	spin_lock_irqsave(&sclp_vt220_lock, flags);
-	if (timer_pending(&sclp_vt220_timer))
-		del_timer(&sclp_vt220_timer);
-	while (sclp_vt220_outqueue_count > 0) {
-		spin_unlock_irqrestore(&sclp_vt220_lock, flags);
-		sclp_sync_wait();
-		spin_lock_irqsave(&sclp_vt220_lock, flags);
-	}
-	spin_unlock_irqrestore(&sclp_vt220_lock, flags);
-}
-
 /*
  * Pass on all buffers to the hardware. Return only when there are no more
  * buffers pending.
@@ -745,6 +728,22 @@ sclp_vt220_con_device(struct console *c, int *index)
 	return sclp_vt220_driver;
 }
 
+static void __sclp_vt220_flush_buffer(void)
+{
+	unsigned long flags;
+
+	sclp_vt220_emit_current();
+	spin_lock_irqsave(&sclp_vt220_lock, flags);
+	if (timer_pending(&sclp_vt220_timer))
+		del_timer(&sclp_vt220_timer);
+	while (sclp_vt220_outqueue_count > 0) {
+		spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+		sclp_sync_wait();
+		spin_lock_irqsave(&sclp_vt220_lock, flags);
+	}
+	spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+}
+
 static int
 sclp_vt220_notify(struct notifier_block *self,
 			  unsigned long event, void *data)

drivers/s390/char/vmcp.c

@@ -11,12 +11,14 @@
  * The idea of this driver is based on cpint from Neale Ferguson and #CP in CMS
  */
 
+#define KMSG_COMPONENT "vmcp"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/fs.h>
 #include <linux/init.h>
 #include <linux/kernel.h>
 #include <linux/miscdevice.h>
 #include <linux/module.h>
-#include <linux/smp_lock.h>
 #include <asm/cpcmd.h>
 #include <asm/debug.h>
 #include <asm/uaccess.h>
@@ -26,8 +28,6 @@ MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Christian Borntraeger <borntraeger@de.ibm.com>");
 MODULE_DESCRIPTION("z/VM CP interface");
 
-#define PRINTK_HEADER "vmcp: "
-
 static debug_info_t *vmcp_debug;
 
 static int vmcp_open(struct inode *inode, struct file *file)
@@ -41,13 +41,11 @@ static int vmcp_open(struct inode *inode, struct file *file)
 	if (!session)
 		return -ENOMEM;
 
-	lock_kernel();
 	session->bufsize = PAGE_SIZE;
 	session->response = NULL;
 	session->resp_size = 0;
 	mutex_init(&session->mutex);
 	file->private_data = session;
-	unlock_kernel();
 	return nonseekable_open(inode, file);
 }
 
@@ -193,7 +191,8 @@ static int __init vmcp_init(void)
 	int ret;
 
 	if (!MACHINE_IS_VM) {
-		PRINT_WARN("z/VM CP interface is only available under z/VM\n");
+		pr_warning("The z/VM CP interface device driver cannot be "
+			   "loaded without z/VM\n");
 		return -ENODEV;
 	}
 

drivers/s390/char/vmlogrdr.c

@@ -10,6 +10,10 @@
  *		   Stefan Weinhuber <wein@de.ibm.com>
  *
  */
+
+#define KMSG_COMPONENT "vmlogrdr"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/errno.h>
@@ -28,8 +32,6 @@
 #include <linux/smp_lock.h>
 #include <linux/string.h>
 
-
-
 MODULE_AUTHOR
 	("(C) 2004 IBM Corporation by Xenia Tkatschow (xenia@us.ibm.com)\n"
 	 "                            Stefan Weinhuber (wein@de.ibm.com)");
@@ -174,8 +176,7 @@ static void vmlogrdr_iucv_path_severed(struct iucv_path *path, u8 ipuser[16])
 	struct vmlogrdr_priv_t * logptr = path->private;
 	u8 reason = (u8) ipuser[8];
 
-	printk (KERN_ERR "vmlogrdr: connection severed with"
-		" reason %i\n", reason);
+	pr_err("vmlogrdr: connection severed with reason %i\n", reason);
 
 	iucv_path_sever(path, NULL);
 	kfree(path);
@@ -333,8 +334,8 @@ static int vmlogrdr_open (struct inode *inode, struct file *filp)
 	if (logptr->autorecording) {
 		ret = vmlogrdr_recording(logptr,1,logptr->autopurge);
 		if (ret)
-			printk (KERN_WARNING "vmlogrdr: failed to start "
-				"recording automatically\n");
+			pr_warning("vmlogrdr: failed to start "
+				   "recording automatically\n");
 	}
 
 	/* create connection to the system service */
@@ -345,9 +346,9 @@ static int vmlogrdr_open (struct inode *inode, struct file *filp)
 				       logptr->system_service, NULL, NULL,
 				       logptr);
 	if (connect_rc) {
-		printk (KERN_ERR "vmlogrdr: iucv connection to %s "
-			"failed with rc %i \n", logptr->system_service,
-			connect_rc);
+		pr_err("vmlogrdr: iucv connection to %s "
+		       "failed with rc %i \n",
+		       logptr->system_service, connect_rc);
 		goto out_path;
 	}
 
@@ -388,8 +389,8 @@ static int vmlogrdr_release (struct inode *inode, struct file *filp)
 	if (logptr->autorecording) {
 		ret = vmlogrdr_recording(logptr,0,logptr->autopurge);
 		if (ret)
-			printk (KERN_WARNING "vmlogrdr: failed to stop "
-				"recording automatically\n");
+			pr_warning("vmlogrdr: failed to stop "
+				   "recording automatically\n");
 	}
 	logptr->dev_in_use = 0;
 
@@ -823,8 +824,7 @@ static int __init vmlogrdr_init(void)
 	dev_t dev;
 
 	if (! MACHINE_IS_VM) {
-		printk (KERN_ERR "vmlogrdr: not running under VM, "
-				"driver not loaded.\n");
+		pr_err("not running under VM, driver not loaded.\n");
 		return -ENODEV;
 	}
 

drivers/s390/char/vmur.c

@@ -8,6 +8,9 @@
  *	    Frank Munzert <munzert@de.ibm.com>
  */
 
+#define KMSG_COMPONENT "vmur"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/cdev.h>
 #include <linux/smp_lock.h>
 
@@ -40,8 +43,6 @@ MODULE_AUTHOR("IBM Corporation");
 MODULE_DESCRIPTION("s390 z/VM virtual unit record device driver");
 MODULE_LICENSE("GPL");
 
-#define PRINTK_HEADER "vmur: "
-
 static dev_t ur_first_dev_maj_min;
 static struct class *vmur_class;
 static struct debug_info *vmur_dbf;
@@ -987,7 +988,8 @@ static int __init ur_init(void)
 	dev_t dev;
 
 	if (!MACHINE_IS_VM) {
-		PRINT_ERR("%s is only available under z/VM.\n", ur_banner);
+		pr_err("The %s cannot be loaded without z/VM\n",
+		       ur_banner);
 		return -ENODEV;
 	}
 
@@ -1006,7 +1008,8 @@ static int __init ur_init(void)
 
 	rc = alloc_chrdev_region(&dev, 0, NUM_MINORS, "vmur");
 	if (rc) {
-		PRINT_ERR("alloc_chrdev_region failed: err = %d\n", rc);
+		pr_err("Kernel function alloc_chrdev_region failed with "
+		       "error code %d\n", rc);
 		goto fail_unregister_driver;
 	}
 	ur_first_dev_maj_min = MKDEV(MAJOR(dev), 0);
@@ -1016,7 +1019,7 @@ static int __init ur_init(void)
 		rc = PTR_ERR(vmur_class);
 		goto fail_unregister_region;
 	}
-	PRINT_INFO("%s loaded.\n", ur_banner);
+	pr_info("%s loaded.\n", ur_banner);
 	return 0;
 
 fail_unregister_region:
@@ -1034,7 +1037,7 @@ static void __exit ur_exit(void)
 	unregister_chrdev_region(ur_first_dev_maj_min, NUM_MINORS);
 	ccw_driver_unregister(&ur_driver);
 	debug_unregister(vmur_dbf);
-	PRINT_INFO("%s unloaded.\n", ur_banner);
+	pr_info("%s unloaded.\n", ur_banner);
 }
 
 module_init(ur_init);

drivers/s390/char/zcore.c

@@ -9,6 +9,9 @@
  * Author(s): Michael Holzheu
  */
 
+#define KMSG_COMPONENT "zdump"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/init.h>
 #include <linux/miscdevice.h>
 #include <linux/utsname.h>
@@ -24,8 +27,6 @@
 #include "sclp.h"
 
 #define TRACE(x...) debug_sprintf_event(zcore_dbf, 1, x)
-#define MSG(x...) printk( KERN_ALERT x )
-#define ERROR_MSG(x...) printk ( KERN_ALERT "DUMP: " x )
 
 #define TO_USER		0
 #define TO_KERNEL	1
@@ -563,19 +564,19 @@ static int __init sys_info_init(enum arch_id arch)
 
 	switch (arch) {
 	case ARCH_S390X:
-		MSG("DETECTED 'S390X (64 bit) OS'\n");
+		pr_alert("DETECTED 'S390X (64 bit) OS'\n");
 		sys_info.sa_base = SAVE_AREA_BASE_S390X;
 		sys_info.sa_size = sizeof(struct save_area_s390x);
 		set_s390x_lc_mask(&sys_info.lc_mask);
 		break;
 	case ARCH_S390:
-		MSG("DETECTED 'S390 (32 bit) OS'\n");
+		pr_alert("DETECTED 'S390 (32 bit) OS'\n");
 		sys_info.sa_base = SAVE_AREA_BASE_S390;
 		sys_info.sa_size = sizeof(struct save_area_s390);
 		set_s390_lc_mask(&sys_info.lc_mask);
 		break;
 	default:
-		ERROR_MSG("unknown architecture 0x%x.\n",arch);
+		pr_alert("0x%x is an unknown architecture.\n",arch);
 		return -EINVAL;
 	}
 	sys_info.arch = arch;
@@ -674,7 +675,8 @@ static int __init zcore_init(void)
 
 #ifndef __s390x__
 	if (arch == ARCH_S390X) {
-		ERROR_MSG("32 bit dumper can't dump 64 bit system!\n");
+		pr_alert("The 32-bit dump tool cannot be used for a "
+			 "64-bit system\n");
 		rc = -EINVAL;
 		goto fail;
 	}

drivers/s390/cio/blacklist.c

@@ -9,6 +9,9 @@
  *		 Arnd Bergmann (arndb@de.ibm.com)
  */
 
+#define KMSG_COMPONENT "cio"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/init.h>
 #include <linux/vmalloc.h>
 #include <linux/slab.h>
@@ -50,9 +53,10 @@ static int blacklist_range(range_action action, unsigned int from_ssid,
 {
 	if ((from_ssid > to_ssid) || ((from_ssid == to_ssid) && (from > to))) {
 		if (msgtrigger)
-			printk(KERN_WARNING "cio: Invalid cio_ignore range "
-			       "0.%x.%04x-0.%x.%04x\n", from_ssid, from,
-			       to_ssid, to);
+			pr_warning("0.%x.%04x to 0.%x.%04x is not a valid "
+				   "range for cio_ignore\n", from_ssid, from,
+				   to_ssid, to);
+
 		return 1;
 	}
 
@@ -140,8 +144,8 @@ static int parse_busid(char *str, unsigned int *cssid, unsigned int *ssid,
 	rc = 0;
 out:
 	if (rc && msgtrigger)
-		printk(KERN_WARNING "cio: Invalid cio_ignore device '%s'\n",
-		       str);
+		pr_warning("%s is not a valid device for the cio_ignore "
+			   "kernel parameter\n", str);
 
 	return rc;
 }

drivers/s390/cio/ccwgroup.c

@@ -19,6 +19,8 @@
 #include <asm/ccwdev.h>
 #include <asm/ccwgroup.h>
 
+#define CCW_BUS_ID_SIZE		20
+
 /* In Linux 2.4, we had a channel device layer called "chandev"
  * that did all sorts of obscure stuff for networking devices.
  * This is another driver that serves as a replacement for just
@@ -89,15 +91,23 @@ ccwgroup_ungroup_store(struct device *dev, struct device_attribute *attr, const
 
 	gdev = to_ccwgroupdev(dev);
 
-	if (gdev->state != CCWGROUP_OFFLINE)
-		return -EINVAL;
-
+	/* Prevent concurrent online/offline processing and ungrouping. */
+	if (atomic_cmpxchg(&gdev->onoff, 0, 1) != 0)
+		return -EAGAIN;
+	if (gdev->state != CCWGROUP_OFFLINE) {
+		rc = -EINVAL;
+		goto out;
+	}
 	/* Note that we cannot unregister the device from one of its
 	 * attribute methods, so we have to use this roundabout approach.
 	 */
 	rc = device_schedule_callback(dev, ccwgroup_ungroup_callback);
-	if (rc)
-		count = rc;
+out:
+	if (rc) {
+		/* Release onoff "lock" when ungrouping failed. */
+		atomic_set(&gdev->onoff, 0);
+		return rc;
+	}
 	return count;
 }
 
@@ -172,7 +182,7 @@ static int __get_next_bus_id(const char **buf, char *bus_id)
 		len = end - start + 1;
 		end++;
 	}
-	if (len < BUS_ID_SIZE) {
+	if (len < CCW_BUS_ID_SIZE) {
 		strlcpy(bus_id, start, len);
 		rc = 0;
 	} else
@@ -181,7 +191,7 @@ static int __get_next_bus_id(const char **buf, char *bus_id)
 	return rc;
 }
 
-static int __is_valid_bus_id(char bus_id[BUS_ID_SIZE])
+static int __is_valid_bus_id(char bus_id[CCW_BUS_ID_SIZE])
 {
 	int cssid, ssid, devno;
 
@@ -213,7 +223,7 @@ int ccwgroup_create_from_string(struct device *root, unsigned int creator_id,
 {
 	struct ccwgroup_device *gdev;
 	int rc, i;
-	char tmp_bus_id[BUS_ID_SIZE];
+	char tmp_bus_id[CCW_BUS_ID_SIZE];
 	const char *curr_buf;
 
 	gdev = kzalloc(sizeof(*gdev) + num_devices * sizeof(gdev->cdev[0]),

drivers/s390/cio/chsc.c

@@ -8,6 +8,9 @@
  *		 Arnd Bergmann (arndb@de.ibm.com)
  */
 
+#define KMSG_COMPONENT "cio"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/module.h>
 #include <linux/slab.h>
 #include <linux/init.h>
@@ -333,6 +336,7 @@ static void chsc_process_sei_chp_config(struct chsc_sei_area *sei_area)
 	struct chp_config_data *data;
 	struct chp_id chpid;
 	int num;
+	char *events[3] = {"configure", "deconfigure", "cancel deconfigure"};
 
 	CIO_CRW_EVENT(4, "chsc: channel-path-configuration notification\n");
 	if (sei_area->rs != 0)
@@ -343,8 +347,8 @@ static void chsc_process_sei_chp_config(struct chsc_sei_area *sei_area)
 		if (!chp_test_bit(data->map, num))
 			continue;
 		chpid.id = num;
-		printk(KERN_WARNING "cio: processing configure event %d for "
-		       "chpid %x.%02x\n", data->op, chpid.cssid, chpid.id);
+		pr_notice("Processing %s for channel path %x.%02x\n",
+			  events[data->op], chpid.cssid, chpid.id);
 		switch (data->op) {
 		case 0:
 			chp_cfg_schedule(chpid, 1);

drivers/s390/cio/chsc_sch.c

@@ -61,7 +61,7 @@ static void chsc_subchannel_irq(struct subchannel *sch)
 	}
 	private->request = NULL;
 	memcpy(&request->irb, irb, sizeof(*irb));
-	stsch(sch->schid, &sch->schib);
+	cio_update_schib(sch);
 	complete(&request->completion);
 	put_device(&sch->dev);
 }

drivers/s390/cio/cio.c

@@ -9,6 +9,9 @@
  *		 Martin Schwidefsky (schwidefsky@de.ibm.com)
  */
 
+#define KMSG_COMPONENT "cio"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/slab.h>
@@ -104,44 +107,6 @@ cio_get_options (struct subchannel *sch)
        return flags;
 }
 
-/*
- * Use tpi to get a pending interrupt, call the interrupt handler and
- * return a pointer to the subchannel structure.
- */
-static int
-cio_tpi(void)
-{
-	struct tpi_info *tpi_info;
-	struct subchannel *sch;
-	struct irb *irb;
-	int irq_context;
-
-	tpi_info = (struct tpi_info *) __LC_SUBCHANNEL_ID;
-	if (tpi (NULL) != 1)
-		return 0;
-	irb = (struct irb *) __LC_IRB;
-	/* Store interrupt response block to lowcore. */
-	if (tsch (tpi_info->schid, irb) != 0)
-		/* Not status pending or not operational. */
-		return 1;
-	sch = (struct subchannel *)(unsigned long)tpi_info->intparm;
-	if (!sch)
-		return 1;
-	irq_context = in_interrupt();
-	if (!irq_context)
-		local_bh_disable();
-	irq_enter ();
-	spin_lock(sch->lock);
-	memcpy(&sch->schib.scsw, &irb->scsw, sizeof(union scsw));
-	if (sch->driver && sch->driver->irq)
-		sch->driver->irq(sch);
-	spin_unlock(sch->lock);
-	irq_exit ();
-	if (!irq_context)
-		_local_bh_enable();
-	return 1;
-}
-
 static int
 cio_start_handle_notoper(struct subchannel *sch, __u8 lpm)
 {
@@ -152,11 +117,13 @@ cio_start_handle_notoper(struct subchannel *sch, __u8 lpm)
 	else
 		sch->lpm = 0;
 
-	stsch (sch->schid, &sch->schib);
-
 	CIO_MSG_EVENT(2, "cio_start: 'not oper' status for "
 		      "subchannel 0.%x.%04x!\n", sch->schid.ssid,
 		      sch->schid.sch_no);
+
+	if (cio_update_schib(sch))
+		return -ENODEV;
+
 	sprintf(dbf_text, "no%s", dev_name(&sch->dev));
 	CIO_TRACE_EVENT(0, dbf_text);
 	CIO_HEX_EVENT(0, &sch->schib, sizeof (struct schib));
@@ -354,7 +321,8 @@ cio_cancel (struct subchannel *sch)
 	switch (ccode) {
 	case 0:		/* success */
 		/* Update information in scsw. */
-		stsch (sch->schid, &sch->schib);
+		if (cio_update_schib(sch))
+			return -ENODEV;
 		return 0;
 	case 1:		/* status pending */
 		return -EBUSY;
@@ -365,36 +333,93 @@ cio_cancel (struct subchannel *sch)
 	}
 }
 
+
+static void cio_apply_config(struct subchannel *sch, struct schib *schib)
+{
+	schib->pmcw.intparm = sch->config.intparm;
+	schib->pmcw.mbi = sch->config.mbi;
+	schib->pmcw.isc = sch->config.isc;
+	schib->pmcw.ena = sch->config.ena;
+	schib->pmcw.mme = sch->config.mme;
+	schib->pmcw.mp = sch->config.mp;
+	schib->pmcw.csense = sch->config.csense;
+	schib->pmcw.mbfc = sch->config.mbfc;
+	if (sch->config.mbfc)
+		schib->mba = sch->config.mba;
+}
+
+static int cio_check_config(struct subchannel *sch, struct schib *schib)
+{
+	return (schib->pmcw.intparm == sch->config.intparm) &&
+		(schib->pmcw.mbi == sch->config.mbi) &&
+		(schib->pmcw.isc == sch->config.isc) &&
+		(schib->pmcw.ena == sch->config.ena) &&
+		(schib->pmcw.mme == sch->config.mme) &&
+		(schib->pmcw.mp == sch->config.mp) &&
+		(schib->pmcw.csense == sch->config.csense) &&
+		(schib->pmcw.mbfc == sch->config.mbfc) &&
+		(!sch->config.mbfc || (schib->mba == sch->config.mba));
+}
+
 /*
- * Function: cio_modify
- * Issues a "Modify Subchannel" on the specified subchannel
+ * cio_commit_config - apply configuration to the subchannel
  */
-int
-cio_modify (struct subchannel *sch)
+int cio_commit_config(struct subchannel *sch)
 {
-	int ccode, retry, ret;
+	struct schib schib;
+	int ccode, retry, ret = 0;
+
+	if (stsch(sch->schid, &schib) || !css_sch_is_valid(&schib))
+		return -ENODEV;
 
-	ret = 0;
 	for (retry = 0; retry < 5; retry++) {
-		ccode = msch_err (sch->schid, &sch->schib);
-		if (ccode < 0)	/* -EIO if msch gets a program check. */
+		/* copy desired changes to local schib */
+		cio_apply_config(sch, &schib);
+		ccode = msch_err(sch->schid, &schib);
+		if (ccode < 0) /* -EIO if msch gets a program check. */
 			return ccode;
 		switch (ccode) {
 		case 0: /* successfull */
-			return 0;
-		case 1:	/* status pending */
+			if (stsch(sch->schid, &schib) ||
+			    !css_sch_is_valid(&schib))
+				return -ENODEV;
+			if (cio_check_config(sch, &schib)) {
+				/* commit changes from local schib */
+				memcpy(&sch->schib, &schib, sizeof(schib));
+				return 0;
+			}
+			ret = -EAGAIN;
+			break;
+		case 1: /* status pending */
 			return -EBUSY;
-		case 2:	/* busy */
-			udelay (100);	/* allow for recovery */
+		case 2: /* busy */
+			udelay(100); /* allow for recovery */
 			ret = -EBUSY;
 			break;
-		case 3:	/* not operational */
+		case 3: /* not operational */
 			return -ENODEV;
 		}
 	}
 	return ret;
 }
 
+/**
+ * cio_update_schib - Perform stsch and update schib if subchannel is valid.
+ * @sch: subchannel on which to perform stsch
+ * Return zero on success, -ENODEV otherwise.
+ */
+int cio_update_schib(struct subchannel *sch)
+{
+	struct schib schib;
+
+	if (stsch(sch->schid, &schib) || !css_sch_is_valid(&schib))
+		return -ENODEV;
+
+	memcpy(&sch->schib, &schib, sizeof(schib));
+	return 0;
+}
+EXPORT_SYMBOL_GPL(cio_update_schib);
+
 /**
  * cio_enable_subchannel - enable a subchannel.
  * @sch: subchannel to be enabled
@@ -403,7 +428,6 @@ cio_modify (struct subchannel *sch)
 int cio_enable_subchannel(struct subchannel *sch, u32 intparm)
 {
 	char dbf_txt[15];
-	int ccode;
 	int retry;
 	int ret;
 
@@ -412,33 +436,27 @@ int cio_enable_subchannel(struct subchannel *sch, u32 intparm)
 
 	if (sch_is_pseudo_sch(sch))
 		return -EINVAL;
-	ccode = stsch (sch->schid, &sch->schib);
-	if (ccode)
+	if (cio_update_schib(sch))
 		return -ENODEV;
 
-	for (retry = 5, ret = 0; retry > 0; retry--) {
-		sch->schib.pmcw.ena = 1;
-		sch->schib.pmcw.isc = sch->isc;
-		sch->schib.pmcw.intparm = intparm;
-		ret = cio_modify(sch);
-		if (ret == -ENODEV)
-			break;
-		if (ret == -EIO)
+	sch->config.ena = 1;
+	sch->config.isc = sch->isc;
+	sch->config.intparm = intparm;
+
+	for (retry = 0; retry < 3; retry++) {
+		ret = cio_commit_config(sch);
+		if (ret == -EIO) {
 			/*
-			 * Got a program check in cio_modify. Try without
+			 * Got a program check in msch. Try without
 			 * the concurrent sense bit the next time.
 			 */
-			sch->schib.pmcw.csense = 0;
-		if (ret == 0) {
-			stsch (sch->schid, &sch->schib);
-			if (sch->schib.pmcw.ena)
-				break;
-		}
-		if (ret == -EBUSY) {
+			sch->config.csense = 0;
+		} else if (ret == -EBUSY) {
 			struct irb irb;
 			if (tsch(sch->schid, &irb) != 0)
 				break;
-		}
+		} else
+			break;
 	}
 	sprintf (dbf_txt, "ret:%d", ret);
 	CIO_TRACE_EVENT (2, dbf_txt);
@@ -453,8 +471,6 @@ EXPORT_SYMBOL_GPL(cio_enable_subchannel);
 int cio_disable_subchannel(struct subchannel *sch)
 {
 	char dbf_txt[15];
-	int ccode;
-	int retry;
 	int ret;
 
 	CIO_TRACE_EVENT (2, "dissch");
@@ -462,8 +478,7 @@ int cio_disable_subchannel(struct subchannel *sch)
 
 	if (sch_is_pseudo_sch(sch))
 		return 0;
-	ccode = stsch (sch->schid, &sch->schib);
-	if (ccode == 3)		/* Not operational. */
+	if (cio_update_schib(sch))
 		return -ENODEV;
 
 	if (scsw_actl(&sch->schib.scsw) != 0)
@@ -473,24 +488,9 @@ int cio_disable_subchannel(struct subchannel *sch)
 		 */
 		return -EBUSY;
 
-	for (retry = 5, ret = 0; retry > 0; retry--) {
-		sch->schib.pmcw.ena = 0;
-		ret = cio_modify(sch);
-		if (ret == -ENODEV)
-			break;
-		if (ret == -EBUSY)
-			/*
-			 * The subchannel is busy or status pending.
-			 * We'll disable when the next interrupt was delivered
-			 * via the state machine.
-			 */
-			break;
-		if (ret == 0) {
-			stsch (sch->schid, &sch->schib);
-			if (!sch->schib.pmcw.ena)
-				break;
-		}
-	}
+	sch->config.ena = 0;
+	ret = cio_commit_config(sch);
+
 	sprintf (dbf_txt, "ret:%d", ret);
 	CIO_TRACE_EVENT (2, dbf_txt);
 	return ret;
@@ -687,6 +687,43 @@ static char console_sch_name[10] = "0.x.xxxx";
 static struct io_subchannel_private console_priv;
 static int console_subchannel_in_use;
 
+/*
+ * Use tpi to get a pending interrupt, call the interrupt handler and
+ * return a pointer to the subchannel structure.
+ */
+static int cio_tpi(void)
+{
+	struct tpi_info *tpi_info;
+	struct subchannel *sch;
+	struct irb *irb;
+	int irq_context;
+
+	tpi_info = (struct tpi_info *) __LC_SUBCHANNEL_ID;
+	if (tpi(NULL) != 1)
+		return 0;
+	irb = (struct irb *) __LC_IRB;
+	/* Store interrupt response block to lowcore. */
+	if (tsch(tpi_info->schid, irb) != 0)
+		/* Not status pending or not operational. */
+		return 1;
+	sch = (struct subchannel *)(unsigned long)tpi_info->intparm;
+	if (!sch)
+		return 1;
+	irq_context = in_interrupt();
+	if (!irq_context)
+		local_bh_disable();
+	irq_enter();
+	spin_lock(sch->lock);
+	memcpy(&sch->schib.scsw, &irb->scsw, sizeof(union scsw));
+	if (sch->driver && sch->driver->irq)
+		sch->driver->irq(sch);
+	spin_unlock(sch->lock);
+	irq_exit();
+	if (!irq_context)
+		_local_bh_enable();
+	return 1;
+}
+
 void *cio_get_console_priv(void)
 {
 	return &console_priv;
@@ -780,7 +817,7 @@ cio_probe_console(void)
 	sch_no = cio_get_console_sch_no();
 	if (sch_no == -1) {
 		console_subchannel_in_use = 0;
-		printk(KERN_WARNING "cio: No ccw console found!\n");
+		pr_warning("No CCW console was found\n");
 		return ERR_PTR(-ENODEV);
 	}
 	memset(&console_subchannel, 0, sizeof(struct subchannel));
@@ -796,10 +833,9 @@ cio_probe_console(void)
 	 * enable console I/O-interrupt subclass
 	 */
 	isc_register(CONSOLE_ISC);
-	console_subchannel.schib.pmcw.isc = CONSOLE_ISC;
-	console_subchannel.schib.pmcw.intparm =
-		(u32)(addr_t)&console_subchannel;
-	ret = cio_modify(&console_subchannel);
+	console_subchannel.config.isc = CONSOLE_ISC;
+	console_subchannel.config.intparm = (u32)(addr_t)&console_subchannel;
+	ret = cio_commit_config(&console_subchannel);
 	if (ret) {
 		isc_unregister(CONSOLE_ISC);
 		console_subchannel_in_use = 0;
@@ -811,8 +847,8 @@ cio_probe_console(void)
 void
 cio_release_console(void)
 {
-	console_subchannel.schib.pmcw.intparm = 0;
-	cio_modify(&console_subchannel);
+	console_subchannel.config.intparm = 0;
+	cio_commit_config(&console_subchannel);
 	isc_unregister(CONSOLE_ISC);
 	console_subchannel_in_use = 0;
 }
@@ -852,7 +888,8 @@ __disable_subchannel_easy(struct subchannel_id schid, struct schib *schib)
 		cc = msch(schid, schib);
 		if (cc)
 			return (cc==3?-ENODEV:-EBUSY);
-		stsch(schid, schib);
+		if (stsch(schid, schib) || !css_sch_is_valid(schib))
+			return -ENODEV;
 		if (!schib->pmcw.ena)
 			return 0;
 	}

drivers/s390/cio/cio.h

@@ -45,6 +45,19 @@ struct pmcw {
 				/*  ... in an operand exception.       */
 } __attribute__ ((packed));
 
+/* Target SCHIB configuration. */
+struct schib_config {
+	u64 mba;
+	u32 intparm;
+	u16 mbi;
+	u32 isc:3;
+	u32 ena:1;
+	u32 mme:2;
+	u32 mp:1;
+	u32 csense:1;
+	u32 mbfc:1;
+} __attribute__ ((packed));
+
 /*
  * subchannel information block
  */
@@ -82,6 +95,8 @@ struct subchannel {
 	struct device dev;	/* entry in device tree */
 	struct css_driver *driver;
 	void *private; /* private per subchannel type data */
+	struct work_struct work;
+	struct schib_config config;
 } __attribute__ ((aligned(8)));
 
 #define IO_INTERRUPT_TYPE	   0 /* I/O interrupt type */
@@ -100,7 +115,8 @@ extern int cio_start_key (struct subchannel *, struct ccw1 *, __u8, __u8);
 extern int cio_cancel (struct subchannel *);
 extern int cio_set_options (struct subchannel *, int);
 extern int cio_get_options (struct subchannel *);
-extern int cio_modify (struct subchannel *);
+extern int cio_update_schib(struct subchannel *sch);
+extern int cio_commit_config(struct subchannel *sch);
 
 int cio_tm_start_key(struct subchannel *sch, struct tcw *tcw, u8 lpm, u8 key);
 int cio_tm_intrg(struct subchannel *sch);

drivers/s390/cio/cmf.c

@@ -25,6 +25,9 @@
  * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
 
+#define KMSG_COMPONENT "cio"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/bootmem.h>
 #include <linux/device.h>
 #include <linux/init.h>
@@ -185,56 +188,19 @@ static inline void cmf_activate(void *area, unsigned int onoff)
 static int set_schib(struct ccw_device *cdev, u32 mme, int mbfc,
 		     unsigned long address)
 {
-	int ret;
-	int retry;
 	struct subchannel *sch;
-	struct schib *schib;
 
 	sch = to_subchannel(cdev->dev.parent);
-	schib = &sch->schib;
-	/* msch can silently fail, so do it again if necessary */
-	for (retry = 0; retry < 3; retry++) {
-		/* prepare schib */
-		stsch(sch->schid, schib);
-		schib->pmcw.mme  = mme;
-		schib->pmcw.mbfc = mbfc;
-		/* address can be either a block address or a block index */
-		if (mbfc)
-			schib->mba = address;
-		else
-			schib->pmcw.mbi = address;
-
-		/* try to submit it */
-		switch(ret = msch_err(sch->schid, schib)) {
-			case 0:
-				break;
-			case 1:
-			case 2: /* in I/O or status pending */
-				ret = -EBUSY;
-				break;
-			case 3: /* subchannel is no longer valid */
-				ret = -ENODEV;
-				break;
-			default: /* msch caught an exception */
-				ret = -EINVAL;
-				break;
-		}
-		stsch(sch->schid, schib); /* restore the schib */
-
-		if (ret)
-			break;
 
-		/* check if it worked */
-		if (schib->pmcw.mme  == mme &&
-		    schib->pmcw.mbfc == mbfc &&
-		    (mbfc ? (schib->mba == address)
-			  : (schib->pmcw.mbi == address)))
-			return 0;
+	sch->config.mme = mme;
+	sch->config.mbfc = mbfc;
+	/* address can be either a block address or a block index */
+	if (mbfc)
+		sch->config.mba = address;
+	else
+		sch->config.mbi = address;
 
-		ret = -EINVAL;
-	}
-
-	return ret;
+	return cio_commit_config(sch);
 }
 
 struct set_schib_struct {
@@ -338,7 +304,7 @@ static int cmf_copy_block(struct ccw_device *cdev)
 
 	sch = to_subchannel(cdev->dev.parent);
 
-	if (stsch(sch->schid, &sch->schib))
+	if (cio_update_schib(sch))
 		return -ENODEV;
 
 	if (scsw_fctl(&sch->schib.scsw) & SCSW_FCTL_START_FUNC) {
@@ -1359,9 +1325,8 @@ static int __init init_cmf(void)
 	default:
 		return 1;
 	}
-
-	printk(KERN_INFO "cio: Channel measurement facility using %s "
-	       "format (%s)\n", format_string, detect_string);
+	pr_info("Channel measurement facility initialized using format "
+		"%s (mode %s)\n", format_string, detect_string);
 	return 0;
 }
 

drivers/s390/cio/css.c

@@ -6,6 +6,10 @@
  *    Author(s): Arnd Bergmann (arndb@de.ibm.com)
  *		 Cornelia Huck (cornelia.huck@de.ibm.com)
  */
+
+#define KMSG_COMPONENT "cio"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/device.h>
@@ -128,8 +132,8 @@ css_free_subchannel(struct subchannel *sch)
 {
 	if (sch) {
 		/* Reset intparm to zeroes. */
-		sch->schib.pmcw.intparm = 0;
-		cio_modify(sch);
+		sch->config.intparm = 0;
+		cio_commit_config(sch);
 		kfree(sch->lock);
 		kfree(sch);
 	}
@@ -844,8 +848,8 @@ out:
 	s390_unregister_crw_handler(CRW_RSC_CSS);
 	chsc_free_sei_area();
 	kfree(slow_subchannel_set);
-	printk(KERN_WARNING"cio: failed to initialize css driver (%d)!\n",
-	       ret);
+	pr_alert("The CSS device driver initialization failed with "
+		 "errno=%d\n", ret);
 	return ret;
 }
 

drivers/s390/cio/device.c

@@ -376,19 +376,23 @@ int ccw_device_set_offline(struct ccw_device *cdev)
 			dev_fsm_event(cdev, DEV_EVENT_NOTOPER);
 		}
 		spin_unlock_irq(cdev->ccwlock);
+		/* Give up reference from ccw_device_set_online(). */
+		put_device(&cdev->dev);
 		return ret;
 	}
 	spin_unlock_irq(cdev->ccwlock);
-	if (ret == 0)
+	if (ret == 0) {
 		wait_event(cdev->private->wait_q, dev_fsm_final_state(cdev));
-	else {
+		/* Give up reference from ccw_device_set_online(). */
+		put_device(&cdev->dev);
+	} else {
 		CIO_MSG_EVENT(0, "ccw_device_offline returned %d, "
 			      "device 0.%x.%04x\n",
 			      ret, cdev->private->dev_id.ssid,
 			      cdev->private->dev_id.devno);
 		cdev->online = 1;
 	}
- 	return ret;
+	return ret;
 }
 
 /**
@@ -411,6 +415,9 @@ int ccw_device_set_online(struct ccw_device *cdev)
 		return -ENODEV;
 	if (cdev->online || !cdev->drv)
 		return -EINVAL;
+	/* Hold on to an extra reference while device is online. */
+	if (!get_device(&cdev->dev))
+		return -ENODEV;
 
 	spin_lock_irq(cdev->ccwlock);
 	ret = ccw_device_online(cdev);
@@ -422,10 +429,15 @@ int ccw_device_set_online(struct ccw_device *cdev)
 			      "device 0.%x.%04x\n",
 			      ret, cdev->private->dev_id.ssid,
 			      cdev->private->dev_id.devno);
+		/* Give up online reference since onlining failed. */
+		put_device(&cdev->dev);
 		return ret;
 	}
-	if (cdev->private->state != DEV_STATE_ONLINE)
+	if (cdev->private->state != DEV_STATE_ONLINE) {
+		/* Give up online reference since onlining failed. */
+		put_device(&cdev->dev);
 		return -ENODEV;
+	}
 	if (!cdev->drv->set_online || cdev->drv->set_online(cdev) == 0) {
 		cdev->online = 1;
 		return 0;
@@ -440,6 +452,8 @@ int ccw_device_set_online(struct ccw_device *cdev)
 			      "device 0.%x.%04x\n",
 			      ret, cdev->private->dev_id.ssid,
 			      cdev->private->dev_id.devno);
+	/* Give up online reference since onlining failed. */
+	put_device(&cdev->dev);
 	return (ret == 0) ? -ENODEV : ret;
 }
 
@@ -704,6 +718,8 @@ ccw_device_release(struct device *dev)
 	struct ccw_device *cdev;
 
 	cdev = to_ccwdev(dev);
+	/* Release reference of parent subchannel. */
+	put_device(cdev->dev.parent);
 	kfree(cdev->private);
 	kfree(cdev);
 }
@@ -735,8 +751,8 @@ static int io_subchannel_initialize_dev(struct subchannel *sch,
 	/* Do first half of device_register. */
 	device_initialize(&cdev->dev);
 	if (!get_device(&sch->dev)) {
-		if (cdev->dev.release)
-			cdev->dev.release(&cdev->dev);
+		/* Release reference from device_initialize(). */
+		put_device(&cdev->dev);
 		return -ENODEV;
 	}
 	return 0;
@@ -778,37 +794,55 @@ static void sch_attach_disconnected_device(struct subchannel *sch,
 	struct subchannel *other_sch;
 	int ret;
 
-	other_sch = to_subchannel(get_device(cdev->dev.parent));
+	/* Get reference for new parent. */
+	if (!get_device(&sch->dev))
+		return;
+	other_sch = to_subchannel(cdev->dev.parent);
+	/* Note: device_move() changes cdev->dev.parent */
 	ret = device_move(&cdev->dev, &sch->dev);
 	if (ret) {
 		CIO_MSG_EVENT(0, "Moving disconnected device 0.%x.%04x failed "
 			      "(ret=%d)!\n", cdev->private->dev_id.ssid,
 			      cdev->private->dev_id.devno, ret);
-		put_device(&other_sch->dev);
+		/* Put reference for new parent. */
+		put_device(&sch->dev);
 		return;
 	}
 	sch_set_cdev(other_sch, NULL);
 	/* No need to keep a subchannel without ccw device around. */
 	css_sch_device_unregister(other_sch);
-	put_device(&other_sch->dev);
 	sch_attach_device(sch, cdev);
+	/* Put reference for old parent. */
+	put_device(&other_sch->dev);
 }
 
 static void sch_attach_orphaned_device(struct subchannel *sch,
 				       struct ccw_device *cdev)
 {
 	int ret;
+	struct subchannel *pseudo_sch;
 
-	/* Try to move the ccw device to its new subchannel. */
+	/* Get reference for new parent. */
+	if (!get_device(&sch->dev))
+		return;
+	pseudo_sch = to_subchannel(cdev->dev.parent);
+	/*
+	 * Try to move the ccw device to its new subchannel.
+	 * Note: device_move() changes cdev->dev.parent
+	 */
 	ret = device_move(&cdev->dev, &sch->dev);
 	if (ret) {
 		CIO_MSG_EVENT(0, "Moving device 0.%x.%04x from orphanage "
 			      "failed (ret=%d)!\n",
 			      cdev->private->dev_id.ssid,
 			      cdev->private->dev_id.devno, ret);
+		/* Put reference for new parent. */
+		put_device(&sch->dev);
 		return;
 	}
 	sch_attach_device(sch, cdev);
+	/* Put reference on pseudo subchannel. */
+	put_device(&pseudo_sch->dev);
 }
 
 static void sch_create_and_recog_new_device(struct subchannel *sch)
@@ -830,9 +864,11 @@ static void sch_create_and_recog_new_device(struct subchannel *sch)
 		spin_lock_irq(sch->lock);
 		sch_set_cdev(sch, NULL);
 		spin_unlock_irq(sch->lock);
-		if (cdev->dev.release)
-			cdev->dev.release(&cdev->dev);
 		css_sch_device_unregister(sch);
+		/* Put reference from io_subchannel_create_ccwdev(). */
+		put_device(&sch->dev);
+		/* Give up initial reference. */
+		put_device(&cdev->dev);
 	}
 }
 
@@ -854,15 +890,20 @@ void ccw_device_move_to_orphanage(struct work_struct *work)
 	dev_id.devno = sch->schib.pmcw.dev;
 	dev_id.ssid = sch->schid.ssid;
 
+	/* Increase refcount for pseudo subchannel. */
+	get_device(&css->pseudo_subchannel->dev);
 	/*
 	 * Move the orphaned ccw device to the orphanage so the replacing
 	 * ccw device can take its place on the subchannel.
+	 * Note: device_move() changes cdev->dev.parent
 	 */
 	ret = device_move(&cdev->dev, &css->pseudo_subchannel->dev);
 	if (ret) {
 		CIO_MSG_EVENT(0, "Moving device 0.%x.%04x to orphanage failed "
 			      "(ret=%d)!\n", cdev->private->dev_id.ssid,
 			      cdev->private->dev_id.devno, ret);
+		/* Decrease refcount for pseudo subchannel again. */
+		put_device(&css->pseudo_subchannel->dev);
 		return;
 	}
 	cdev->ccwlock = css->pseudo_subchannel->lock;
@@ -875,17 +916,23 @@ void ccw_device_move_to_orphanage(struct work_struct *work)
 	if (replacing_cdev) {
 		sch_attach_disconnected_device(sch, replacing_cdev);
 		/* Release reference from get_disc_ccwdev_by_dev_id() */
-		put_device(&cdev->dev);
+		put_device(&replacing_cdev->dev);
+		/* Release reference of subchannel from old cdev. */
+		put_device(&sch->dev);
 		return;
 	}
 	replacing_cdev = get_orphaned_ccwdev_by_dev_id(css, &dev_id);
 	if (replacing_cdev) {
 		sch_attach_orphaned_device(sch, replacing_cdev);
 		/* Release reference from get_orphaned_ccwdev_by_dev_id() */
-		put_device(&cdev->dev);
+		put_device(&replacing_cdev->dev);
+		/* Release reference of subchannel from old cdev. */
+		put_device(&sch->dev);
 		return;
 	}
 	sch_create_and_recog_new_device(sch);
+	/* Release reference of subchannel from old cdev. */
+	put_device(&sch->dev);
 }
 
 /*
@@ -903,6 +950,14 @@ io_subchannel_register(struct work_struct *work)
 	priv = container_of(work, struct ccw_device_private, kick_work);
 	cdev = priv->cdev;
 	sch = to_subchannel(cdev->dev.parent);
+	/*
+	 * Check if subchannel is still registered. It may have become
+	 * unregistered if a machine check hit us after finishing
+	 * device recognition but before the register work could be
+	 * queued.
+	 */
+	if (!device_is_registered(&sch->dev))
+		goto out_err;
 	css_update_ssd_info(sch);
 	/*
 	 * io_subchannel_register() will also be called after device
@@ -910,7 +965,7 @@ io_subchannel_register(struct work_struct *work)
 	 * be registered). We need to reprobe since we may now have sense id
 	 * information.
 	 */
-	if (klist_node_attached(&cdev->dev.knode_parent)) {
+	if (device_is_registered(&cdev->dev)) {
 		if (!cdev->drv) {
 			ret = device_reprobe(&cdev->dev);
 			if (ret)
@@ -934,22 +989,19 @@ io_subchannel_register(struct work_struct *work)
 		CIO_MSG_EVENT(0, "Could not register ccw dev 0.%x.%04x: %d\n",
 			      cdev->private->dev_id.ssid,
 			      cdev->private->dev_id.devno, ret);
-		put_device(&cdev->dev);
 		spin_lock_irqsave(sch->lock, flags);
 		sch_set_cdev(sch, NULL);
 		spin_unlock_irqrestore(sch->lock, flags);
-		kfree (cdev->private);
-		kfree (cdev);
-		put_device(&sch->dev);
-		if (atomic_dec_and_test(&ccw_device_init_count))
-			wake_up(&ccw_device_init_wq);
-		return;
+		/* Release initial device reference. */
+		put_device(&cdev->dev);
+		goto out_err;
 	}
-	put_device(&cdev->dev);
 out:
 	cdev->private->flags.recog_done = 1;
-	put_device(&sch->dev);
 	wake_up(&cdev->private->wait_q);
+out_err:
+	/* Release reference for workqueue processing. */
+	put_device(&cdev->dev);
 	if (atomic_dec_and_test(&ccw_device_init_count))
 		wake_up(&ccw_device_init_wq);
 }
@@ -968,8 +1020,8 @@ static void ccw_device_call_sch_unregister(struct work_struct *work)
 	sch = to_subchannel(cdev->dev.parent);
 	css_sch_device_unregister(sch);
 	/* Reset intparm to zeroes. */
-	sch->schib.pmcw.intparm = 0;
-	cio_modify(sch);
+	sch->config.intparm = 0;
+	cio_commit_config(sch);
 	/* Release cdev reference for workqueue processing.*/
 	put_device(&cdev->dev);
 	/* Release subchannel reference for local processing. */
@@ -998,8 +1050,6 @@ io_subchannel_recog_done(struct ccw_device *cdev)
 		PREPARE_WORK(&cdev->private->kick_work,
 			     ccw_device_call_sch_unregister);
 		queue_work(slow_path_wq, &cdev->private->kick_work);
-		/* Release subchannel reference for asynchronous recognition. */
-		put_device(&sch->dev);
 		if (atomic_dec_and_test(&ccw_device_init_count))
 			wake_up(&ccw_device_init_wq);
 		break;
@@ -1070,10 +1120,15 @@ static void ccw_device_move_to_sch(struct work_struct *work)
 	priv = container_of(work, struct ccw_device_private, kick_work);
 	sch = priv->sch;
 	cdev = priv->cdev;
-	former_parent = ccw_device_is_orphan(cdev) ?
-		NULL : to_subchannel(get_device(cdev->dev.parent));
+	former_parent = to_subchannel(cdev->dev.parent);
+	/* Get reference for new parent. */
+	if (!get_device(&sch->dev))
+		return;
 	mutex_lock(&sch->reg_mutex);
-	/* Try to move the ccw device to its new subchannel. */
+	/*
+	 * Try to move the ccw device to its new subchannel.
+	 * Note: device_move() changes cdev->dev.parent
+	 */
 	rc = device_move(&cdev->dev, &sch->dev);
 	mutex_unlock(&sch->reg_mutex);
 	if (rc) {
@@ -1083,21 +1138,23 @@ static void ccw_device_move_to_sch(struct work_struct *work)
 			      cdev->private->dev_id.devno, sch->schid.ssid,
 			      sch->schid.sch_no, rc);
 		css_sch_device_unregister(sch);
+		/* Put reference for new parent again. */
+		put_device(&sch->dev);
 		goto out;
 	}
-	if (former_parent) {
+	if (!sch_is_pseudo_sch(former_parent)) {
 		spin_lock_irq(former_parent->lock);
 		sch_set_cdev(former_parent, NULL);
 		spin_unlock_irq(former_parent->lock);
 		css_sch_device_unregister(former_parent);
 		/* Reset intparm to zeroes. */
-		former_parent->schib.pmcw.intparm = 0;
-		cio_modify(former_parent);
+		former_parent->config.intparm = 0;
+		cio_commit_config(former_parent);
 	}
 	sch_attach_device(sch, cdev);
 out:
-	if (former_parent)
-		put_device(&former_parent->dev);
+	/* Put reference for old parent. */
+	put_device(&former_parent->dev);
 	put_device(&cdev->dev);
 }
 
@@ -1113,6 +1170,15 @@ static void io_subchannel_irq(struct subchannel *sch)
 		dev_fsm_event(cdev, DEV_EVENT_INTERRUPT);
 }
 
+void io_subchannel_init_config(struct subchannel *sch)
+{
+	memset(&sch->config, 0, sizeof(sch->config));
+	sch->config.csense = 1;
+	/* Use subchannel mp mode when there is more than 1 installed CHPID. */
+	if ((sch->schib.pmcw.pim & (sch->schib.pmcw.pim - 1)) != 0)
+		sch->config.mp = 1;
+}
+
 static void io_subchannel_init_fields(struct subchannel *sch)
 {
 	if (cio_is_console(sch->schid))
@@ -1127,18 +1193,34 @@ static void io_subchannel_init_fields(struct subchannel *sch)
 		      sch->schib.pmcw.dev, sch->schid.ssid,
 		      sch->schid.sch_no, sch->schib.pmcw.pim,
 		      sch->schib.pmcw.pam, sch->schib.pmcw.pom);
-	/* Initially set up some fields in the pmcw. */
-	sch->schib.pmcw.ena = 0;
-	sch->schib.pmcw.csense = 1;	/* concurrent sense */
-	if ((sch->lpm & (sch->lpm - 1)) != 0)
-		sch->schib.pmcw.mp = 1; /* multipath mode */
-	/* clean up possible residual cmf stuff */
-	sch->schib.pmcw.mme = 0;
-	sch->schib.pmcw.mbfc = 0;
-	sch->schib.pmcw.mbi = 0;
-	sch->schib.mba = 0;
+
+	io_subchannel_init_config(sch);
 }
 
+static void io_subchannel_do_unreg(struct work_struct *work)
+{
+	struct subchannel *sch;
+
+	sch = container_of(work, struct subchannel, work);
+	css_sch_device_unregister(sch);
+	/* Reset intparm to zeroes. */
+	sch->config.intparm = 0;
+	cio_commit_config(sch);
+	put_device(&sch->dev);
+}
+
+/* Schedule unregister if we have no cdev. */
+static void io_subchannel_schedule_removal(struct subchannel *sch)
+{
+	get_device(&sch->dev);
+	INIT_WORK(&sch->work, io_subchannel_do_unreg);
+	queue_work(slow_path_wq, &sch->work);
+}
+
+/*
+ * Note: We always return 0 so that we bind to the device even on error.
+ * This is needed so that our remove function is called on unregister.
+ */
 static int io_subchannel_probe(struct subchannel *sch)
 {
 	struct ccw_device *cdev;
@@ -1168,9 +1250,8 @@ static int io_subchannel_probe(struct subchannel *sch)
 		ccw_device_register(cdev);
 		/*
 		 * Check if the device is already online. If it is
-		 * the reference count needs to be corrected
-		 * (see ccw_device_online and css_init_done for the
-		 * ugly details).
+		 * the reference count needs to be corrected since we
+		 * didn't obtain a reference in ccw_device_set_online.
 		 */
 		if (cdev->private->state != DEV_STATE_NOT_OPER &&
 		    cdev->private->state != DEV_STATE_OFFLINE &&
@@ -1179,23 +1260,24 @@ static int io_subchannel_probe(struct subchannel *sch)
 		return 0;
 	}
 	io_subchannel_init_fields(sch);
-	/*
-	 * First check if a fitting device may be found amongst the
-	 * disconnected devices or in the orphanage.
-	 */
-	dev_id.devno = sch->schib.pmcw.dev;
-	dev_id.ssid = sch->schid.ssid;
+	rc = cio_commit_config(sch);
+	if (rc)
+		goto out_schedule;
 	rc = sysfs_create_group(&sch->dev.kobj,
 				&io_subchannel_attr_group);
 	if (rc)
-		return rc;
+		goto out_schedule;
 	/* Allocate I/O subchannel private data. */
 	sch->private = kzalloc(sizeof(struct io_subchannel_private),
 			       GFP_KERNEL | GFP_DMA);
-	if (!sch->private) {
-		rc = -ENOMEM;
+	if (!sch->private)
 		goto out_err;
-	}
+	/*
+	 * First check if a fitting device may be found amongst the
+	 * disconnected devices or in the orphanage.
+	 */
+	dev_id.devno = sch->schib.pmcw.dev;
+	dev_id.ssid = sch->schid.ssid;
 	cdev = get_disc_ccwdev_by_dev_id(&dev_id, NULL);
 	if (!cdev)
 		cdev = get_orphaned_ccwdev_by_dev_id(to_css(sch->dev.parent),
@@ -1213,24 +1295,21 @@ static int io_subchannel_probe(struct subchannel *sch)
 		return 0;
 	}
 	cdev = io_subchannel_create_ccwdev(sch);
-	if (IS_ERR(cdev)) {
-		rc = PTR_ERR(cdev);
+	if (IS_ERR(cdev))
 		goto out_err;
-	}
 	rc = io_subchannel_recog(cdev, sch);
 	if (rc) {
 		spin_lock_irqsave(sch->lock, flags);
-		sch_set_cdev(sch, NULL);
+		io_subchannel_recog_done(cdev);
 		spin_unlock_irqrestore(sch->lock, flags);
-		if (cdev->dev.release)
-			cdev->dev.release(&cdev->dev);
-		goto out_err;
 	}
 	return 0;
 out_err:
 	kfree(sch->private);
 	sysfs_remove_group(&sch->dev.kobj, &io_subchannel_attr_group);
-	return rc;
+out_schedule:
+	io_subchannel_schedule_removal(sch);
+	return 0;
 }
 
 static int
@@ -1275,10 +1354,7 @@ static void io_subchannel_verify(struct subchannel *sch)
 
 static int check_for_io_on_path(struct subchannel *sch, int mask)
 {
-	int cc;
-
-	cc = stsch(sch->schid, &sch->schib);
-	if (cc)
+	if (cio_update_schib(sch))
 		return 0;
 	if (scsw_actl(&sch->schib.scsw) && sch->schib.pmcw.lpum == mask)
 		return 1;
@@ -1347,15 +1423,13 @@ static int io_subchannel_chp_event(struct subchannel *sch,
 		io_subchannel_verify(sch);
 		break;
 	case CHP_OFFLINE:
-		if (stsch(sch->schid, &sch->schib))
-			return -ENXIO;
-		if (!css_sch_is_valid(&sch->schib))
+		if (cio_update_schib(sch))
 			return -ENODEV;
 		io_subchannel_terminate_path(sch, mask);
 		break;
 	case CHP_ONLINE:
-		if (stsch(sch->schid, &sch->schib))
-			return -ENXIO;
+		if (cio_update_schib(sch))
+			return -ENODEV;
 		sch->lpm |= mask & sch->opm;
 		io_subchannel_verify(sch);
 		break;
@@ -1610,8 +1684,8 @@ static int io_subchannel_sch_event(struct subchannel *sch, int slow)
 		spin_lock_irqsave(sch->lock, flags);
 
 		/* Reset intparm to zeroes. */
-		sch->schib.pmcw.intparm = 0;
-		cio_modify(sch);
+		sch->config.intparm = 0;
+		cio_commit_config(sch);
 		break;
 	case REPROBE:
 		ccw_device_trigger_reprobe(cdev);
@@ -1652,6 +1726,9 @@ static int ccw_device_console_enable(struct ccw_device *cdev,
 	sch->private = cio_get_console_priv();
 	memset(sch->private, 0, sizeof(struct io_subchannel_private));
 	io_subchannel_init_fields(sch);
+	rc = cio_commit_config(sch);
+	if (rc)
+		return rc;
 	sch->driver = &io_subchannel_driver;
 	/* Initialize the ccw_device structure. */
 	cdev->dev.parent= &sch->dev;
@@ -1723,7 +1800,7 @@ __ccwdev_check_busid(struct device *dev, void *id)
 
 	bus_id = id;
 
-	return (strncmp(bus_id, dev_name(dev), BUS_ID_SIZE) == 0);
+	return (strcmp(bus_id, dev_name(dev)) == 0);
 }
 
 
@@ -1806,6 +1883,8 @@ ccw_device_remove (struct device *dev)
 				      "device 0.%x.%04x\n",
 				      ret, cdev->private->dev_id.ssid,
 				      cdev->private->dev_id.devno);
+		/* Give up reference obtained in ccw_device_set_online(). */
+		put_device(&cdev->dev);
 	}
 	ccw_device_set_timeout(cdev, 0);
 	cdev->drv = NULL;

drivers/s390/cio/device.h

@@ -76,6 +76,7 @@ extern wait_queue_head_t ccw_device_init_wq;
 extern atomic_t ccw_device_init_count;
 
 void io_subchannel_recog_done(struct ccw_device *cdev);
+void io_subchannel_init_config(struct subchannel *sch);
 
 int ccw_device_cancel_halt_clear(struct ccw_device *);
 

drivers/s390/cio/device_fsm.c

@@ -140,8 +140,7 @@ ccw_device_cancel_halt_clear(struct ccw_device *cdev)
 	int ret;
 
 	sch = to_subchannel(cdev->dev.parent);
-	ret = stsch(sch->schid, &sch->schib);
-	if (ret || !sch->schib.pmcw.dnv)
+	if (cio_update_schib(sch))
 		return -ENODEV; 
 	if (!sch->schib.pmcw.ena)
 		/* Not operational -> done. */
@@ -245,11 +244,13 @@ ccw_device_recog_done(struct ccw_device *cdev, int state)
 	 * through ssch() and the path information is up to date.
 	 */
 	old_lpm = sch->lpm;
-	stsch(sch->schid, &sch->schib);
-	sch->lpm = sch->schib.pmcw.pam & sch->opm;
+
 	/* Check since device may again have become not operational. */
-	if (!sch->schib.pmcw.dnv)
+	if (cio_update_schib(sch))
 		state = DEV_STATE_NOT_OPER;
+	else
+		sch->lpm = sch->schib.pmcw.pam & sch->opm;
+
 	if (cdev->private->state == DEV_STATE_DISCONNECTED_SENSE_ID)
 		/* Force reprobe on all chpids. */
 		old_lpm = 0;
@@ -399,9 +400,6 @@ ccw_device_done(struct ccw_device *cdev, int state)
 		ccw_device_oper_notify(cdev);
 	}
 	wake_up(&cdev->private->wait_q);
-
-	if (css_init_done && state != DEV_STATE_ONLINE)
-		put_device (&cdev->dev);
 }
 
 static int cmp_pgid(struct pgid *p1, struct pgid *p2)
@@ -552,7 +550,11 @@ ccw_device_verify_done(struct ccw_device *cdev, int err)
 
 	sch = to_subchannel(cdev->dev.parent);
 	/* Update schib - pom may have changed. */
-	stsch(sch->schid, &sch->schib);
+	if (cio_update_schib(sch)) {
+		cdev->private->flags.donotify = 0;
+		ccw_device_done(cdev, DEV_STATE_NOT_OPER);
+		return;
+	}
 	/* Update lpm with verified path mask. */
 	sch->lpm = sch->vpm;
 	/* Repeat path verification? */
@@ -611,8 +613,6 @@ ccw_device_online(struct ccw_device *cdev)
 	    (cdev->private->state != DEV_STATE_BOXED))
 		return -EINVAL;
 	sch = to_subchannel(cdev->dev.parent);
-	if (css_init_done && !get_device(&cdev->dev))
-		return -ENODEV;
 	ret = cio_enable_subchannel(sch, (u32)(addr_t)sch);
 	if (ret != 0) {
 		/* Couldn't enable the subchannel for i/o. Sick device. */
@@ -672,7 +672,7 @@ ccw_device_offline(struct ccw_device *cdev)
 		return 0;
 	}
 	sch = to_subchannel(cdev->dev.parent);
-	if (stsch(sch->schid, &sch->schib) || !sch->schib.pmcw.dnv)
+	if (cio_update_schib(sch))
 		return -ENODEV;
 	if (scsw_actl(&sch->schib.scsw) != 0)
 		return -EBUSY;
@@ -750,7 +750,10 @@ ccw_device_online_verify(struct ccw_device *cdev, enum dev_event dev_event)
 	 * Since we might not just be coming from an interrupt from the
 	 * subchannel we have to update the schib.
 	 */
-	stsch(sch->schid, &sch->schib);
+	if (cio_update_schib(sch)) {
+		ccw_device_verify_done(cdev, -ENODEV);
+		return;
+	}
 
 	if (scsw_actl(&sch->schib.scsw) != 0 ||
 	    (scsw_stctl(&sch->schib.scsw) & SCSW_STCTL_STATUS_PEND) ||
@@ -1016,20 +1019,21 @@ void ccw_device_trigger_reprobe(struct ccw_device *cdev)
 
 	sch = to_subchannel(cdev->dev.parent);
 	/* Update some values. */
-	if (stsch(sch->schid, &sch->schib))
-		return;
-	if (!sch->schib.pmcw.dnv)
+	if (cio_update_schib(sch))
 		return;
 	/*
 	 * The pim, pam, pom values may not be accurate, but they are the best
 	 * we have before performing device selection :/
 	 */
 	sch->lpm = sch->schib.pmcw.pam & sch->opm;
-	/* Re-set some bits in the pmcw that were lost. */
-	sch->schib.pmcw.csense = 1;
-	sch->schib.pmcw.ena = 0;
-	if ((sch->lpm & (sch->lpm - 1)) != 0)
-		sch->schib.pmcw.mp = 1;
+	/*
+	 * Use the initial configuration since we can't be shure that the old
+	 * paths are valid.
+	 */
+	io_subchannel_init_config(sch);
+	if (cio_commit_config(sch))
+		return;
+
 	/* We should also udate ssd info, but this has to wait. */
 	/* Check if this is another device which appeared on the same sch. */
 	if (sch->schib.pmcw.dev != cdev->private->dev_id.devno) {

drivers/s390/cio/device_pgid.c

@@ -504,7 +504,7 @@ ccw_device_verify_start(struct ccw_device *cdev)
 	sch->vpm = 0;
 
 	/* Get current pam. */
-	if (stsch(sch->schid, &sch->schib)) {
+	if (cio_update_schib(sch)) {
 		ccw_device_verify_done(cdev, -ENODEV);
 		return;
 	}