Merge branch 'linus' into perf/core

Merge reason: pick up perf fixes

Signed-off-by: Ingo Molnar <mingo@elte.hu>

136 files changed, 1157 insertions, 803 deletions

diff --git a/arch/alpha/kernel/process.c b/arch/alpha/kernel/process.c
index 88e608a..842dba3 100644
--- a/arch/alpha/kernel/process.c
+++ b/arch/alpha/kernel/process.c
@@ -387,8 +387,9 @@ EXPORT_SYMBOL(dump_elf_task_fp);
  * sys_execve() executes a new program.
  */
 asmlinkage int
-do_sys_execve(const char __user *ufilename, char __user * __user *argv,
-	      char __user * __user *envp, struct pt_regs *regs)
+do_sys_execve(const char __user *ufilename,
+	      const char __user *const __user *argv,
+	      const char __user *const __user *envp, struct pt_regs *regs)
 {
 	int error;
 	char *filename;
diff --git a/arch/arm/Makefile b/arch/arm/Makefile
index 99b8200..59c1ce8 100644
--- a/arch/arm/Makefile
+++ b/arch/arm/Makefile
@@ -21,6 +21,9 @@ GZFLAGS		:=-9
 # Explicitly specifiy 32-bit ARM ISA since toolchain default can be -mthumb:
 KBUILD_CFLAGS	+=$(call cc-option,-marm,)
 
+# Never generate .eh_frame
+KBUILD_CFLAGS	+= $(call cc-option,-fno-dwarf2-cfi-asm)
+
 # Do not use arch/arm/defconfig - it's always outdated.
 # Select a platform tht is kept up-to-date
 KBUILD_DEFCONFIG := versatile_defconfig
diff --git a/arch/arm/include/asm/ptrace.h b/arch/arm/include/asm/ptrace.h
index c974be8..7ce15eb 100644
--- a/arch/arm/include/asm/ptrace.h
+++ b/arch/arm/include/asm/ptrace.h
@@ -158,15 +158,24 @@ struct pt_regs {
  */
 static inline int valid_user_regs(struct pt_regs *regs)
 {
-	if (user_mode(regs) && (regs->ARM_cpsr & PSR_I_BIT) == 0) {
-		regs->ARM_cpsr &= ~(PSR_F_BIT | PSR_A_BIT);
-		return 1;
+	unsigned long mode = regs->ARM_cpsr & MODE_MASK;
+
+	/*
+	 * Always clear the F (FIQ) and A (delayed abort) bits
+	 */
+	regs->ARM_cpsr &= ~(PSR_F_BIT | PSR_A_BIT);
+
+	if ((regs->ARM_cpsr & PSR_I_BIT) == 0) {
+		if (mode == USR_MODE)
+			return 1;
+		if (elf_hwcap & HWCAP_26BIT && mode == USR26_MODE)
+			return 1;
 	}
 
 	/*
 	 * Force CPSR to something logical...
 	 */
-	regs->ARM_cpsr &= PSR_f | PSR_s | (PSR_x & ~PSR_A_BIT) | PSR_T_BIT | MODE32_BIT;
+	regs->ARM_cpsr &= PSR_f | PSR_s | PSR_x | PSR_T_BIT | MODE32_BIT;
 	if (!(elf_hwcap & HWCAP_26BIT))
 		regs->ARM_cpsr |= USR_MODE;
 
diff --git a/arch/arm/include/asm/unistd.h b/arch/arm/include/asm/unistd.h
index dd2bf53..d02cfb6 100644
--- a/arch/arm/include/asm/unistd.h
+++ b/arch/arm/include/asm/unistd.h
@@ -392,6 +392,7 @@
 #define __NR_rt_tgsigqueueinfo		(__NR_SYSCALL_BASE+363)
 #define __NR_perf_event_open		(__NR_SYSCALL_BASE+364)
 #define __NR_recvmmsg			(__NR_SYSCALL_BASE+365)
+#define __NR_accept4			(__NR_SYSCALL_BASE+366)
 
 /*
  * The following SWIs are ARM private.
diff --git a/arch/arm/kernel/calls.S b/arch/arm/kernel/calls.S
index 37ae301..afeb71f 100644
--- a/arch/arm/kernel/calls.S
+++ b/arch/arm/kernel/calls.S
@@ -375,6 +375,7 @@
 		CALL(sys_rt_tgsigqueueinfo)
 		CALL(sys_perf_event_open)
 /* 365 */	CALL(sys_recvmmsg)
+		CALL(sys_accept4)
 #ifndef syscalls_counted
 .equ syscalls_padding, ((NR_syscalls + 3) & ~3) - NR_syscalls
 #define syscalls_counted
diff --git a/arch/arm/kernel/kgdb.c b/arch/arm/kernel/kgdb.c
index 778c2f7..d6e8b4d 100644
--- a/arch/arm/kernel/kgdb.c
+++ b/arch/arm/kernel/kgdb.c
@@ -79,7 +79,7 @@ sleeping_thread_to_gdb_regs(unsigned long *gdb_regs, struct task_struct *task)
 		return;
 
 	/* Initialize to zero */
-	for (regno = 0; regno < GDB_MAX_REGS; regno++)
+	for (regno = 0; regno < DBG_MAX_REG_NUM; regno++)
 		gdb_regs[regno] = 0;
 
 	/* Otherwise, we have only some registers from switch_to() */
diff --git a/arch/arm/kernel/sys_arm.c b/arch/arm/kernel/sys_arm.c
index 5b7c541..62e7c61 100644
--- a/arch/arm/kernel/sys_arm.c
+++ b/arch/arm/kernel/sys_arm.c
@@ -62,8 +62,9 @@ asmlinkage int sys_vfork(struct pt_regs *regs)
 /* sys_execve() executes a new program.
  * This is called indirectly via a small wrapper
  */
-asmlinkage int sys_execve(const char __user *filenamei, char __user * __user *argv,
-			  char __user * __user *envp, struct pt_regs *regs)
+asmlinkage int sys_execve(const char __user *filenamei,
+			  const char __user *const __user *argv,
+			  const char __user *const __user *envp, struct pt_regs *regs)
 {
 	int error;
 	char * filename;
@@ -78,14 +79,17 @@ out:
 	return error;
 }
 
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	struct pt_regs regs;
 	int ret;
 
 	memset(&regs, 0, sizeof(struct pt_regs));
-	ret = do_execve(filename, (char __user * __user *)argv,
-			(char __user * __user *)envp, &regs);
+	ret = do_execve(filename,
+			(const char __user *const __user *)argv,
+			(const char __user *const __user *)envp, &regs);
 	if (ret < 0)
 		goto out;
 
diff --git a/arch/arm/mach-imx/mach-cpuimx27.c b/arch/arm/mach-imx/mach-cpuimx27.c
index 575ff1a..339150a 100644
--- a/arch/arm/mach-imx/mach-cpuimx27.c
+++ b/arch/arm/mach-imx/mach-cpuimx27.c
@@ -279,13 +279,13 @@ static void __init eukrea_cpuimx27_init(void)
 #if defined(CONFIG_USB_ULPI)
 	if (otg_mode_host) {
 		otg_pdata.otg = otg_ulpi_create(&mxc_ulpi_access_ops,
-				USB_OTG_DRV_VBUS | USB_OTG_DRV_VBUS_EXT);
+				ULPI_OTG_DRVVBUS | ULPI_OTG_DRVVBUS_EXT);
 
 		mxc_register_device(&mxc_otg_host, &otg_pdata);
 	}
 
 	usbh2_pdata.otg = otg_ulpi_create(&mxc_ulpi_access_ops,
-				USB_OTG_DRV_VBUS | USB_OTG_DRV_VBUS_EXT);
+				ULPI_OTG_DRVVBUS | ULPI_OTG_DRVVBUS_EXT);
 
 	mxc_register_device(&mxc_usbh2, &usbh2_pdata);
 #endif
diff --git a/arch/arm/mach-imx/mach-pca100.c b/arch/arm/mach-imx/mach-pca100.c
index a389d11..23c9e1f 100644
--- a/arch/arm/mach-imx/mach-pca100.c
+++ b/arch/arm/mach-imx/mach-pca100.c
@@ -419,13 +419,13 @@ static void __init pca100_init(void)
 #if defined(CONFIG_USB_ULPI)
 	if (otg_mode_host) {
 		otg_pdata.otg = otg_ulpi_create(&mxc_ulpi_access_ops,
-				USB_OTG_DRV_VBUS | USB_OTG_DRV_VBUS_EXT);
+				ULPI_OTG_DRVVBUS | ULPI_OTG_DRVVBUS_EXT);
 
 		mxc_register_device(&mxc_otg_host, &otg_pdata);
 	}
 
 	usbh2_pdata.otg = otg_ulpi_create(&mxc_ulpi_access_ops,
-				USB_OTG_DRV_VBUS | USB_OTG_DRV_VBUS_EXT);
+				ULPI_OTG_DRVVBUS | ULPI_OTG_DRVVBUS_EXT);
 
 	mxc_register_device(&mxc_usbh2, &usbh2_pdata);
 #endif
diff --git a/arch/arm/mach-mx25/mach-cpuimx25.c b/arch/arm/mach-mx25/mach-cpuimx25.c
index 56b2e26..a5f0174 100644
--- a/arch/arm/mach-mx25/mach-cpuimx25.c
+++ b/arch/arm/mach-mx25/mach-cpuimx25.c
@@ -138,7 +138,7 @@ static void __init eukrea_cpuimx25_init(void)
 #if defined(CONFIG_USB_ULPI)
 	if (otg_mode_host) {
 		otg_pdata.otg = otg_ulpi_create(&mxc_ulpi_access_ops,
-				USB_OTG_DRV_VBUS | USB_OTG_DRV_VBUS_EXT);
+				ULPI_OTG_DRVVBUS | ULPI_OTG_DRVVBUS_EXT);
 
 		mxc_register_device(&mxc_otg, &otg_pdata);
 	}
diff --git a/arch/arm/mach-mx3/mach-cpuimx35.c b/arch/arm/mach-mx3/mach-cpuimx35.c
index 63f970f..9770a6a 100644
--- a/arch/arm/mach-mx3/mach-cpuimx35.c
+++ b/arch/arm/mach-mx3/mach-cpuimx35.c
@@ -192,7 +192,7 @@ static void __init mxc_board_init(void)
 #if defined(CONFIG_USB_ULPI)
 	if (otg_mode_host) {
 		otg_pdata.otg = otg_ulpi_create(&mxc_ulpi_access_ops,
-				USB_OTG_DRV_VBUS | USB_OTG_DRV_VBUS_EXT);
+				ULPI_OTG_DRVVBUS | ULPI_OTG_DRVVBUS_EXT);
 
 		mxc_register_device(&mxc_otg_host, &otg_pdata);
 	}
diff --git a/arch/arm/plat-samsung/dev-hsmmc.c b/arch/arm/plat-samsung/dev-hsmmc.c
index b0f93f1..9d2be09 100644
--- a/arch/arm/plat-samsung/dev-hsmmc.c
+++ b/arch/arm/plat-samsung/dev-hsmmc.c
@@ -70,4 +70,6 @@ void s3c_sdhci0_set_platdata(struct s3c_sdhci_platdata *pd)
 		set->cfg_gpio = pd->cfg_gpio;
 	if (pd->cfg_card)
 		set->cfg_card = pd->cfg_card;
+	if (pd->host_caps)
+		set->host_caps = pd->host_caps;
 }
diff --git a/arch/arm/plat-samsung/dev-hsmmc1.c b/arch/arm/plat-samsung/dev-hsmmc1.c
index 1504fd8..a6c8295 100644
--- a/arch/arm/plat-samsung/dev-hsmmc1.c
+++ b/arch/arm/plat-samsung/dev-hsmmc1.c
@@ -70,4 +70,6 @@ void s3c_sdhci1_set_platdata(struct s3c_sdhci_platdata *pd)
 		set->cfg_gpio = pd->cfg_gpio;
 	if (pd->cfg_card)
 		set->cfg_card = pd->cfg_card;
+	if (pd->host_caps)
+		set->host_caps = pd->host_caps;
 }
diff --git a/arch/arm/plat-samsung/dev-hsmmc2.c b/arch/arm/plat-samsung/dev-hsmmc2.c
index b28ef17..cb0d714 100644
--- a/arch/arm/plat-samsung/dev-hsmmc2.c
+++ b/arch/arm/plat-samsung/dev-hsmmc2.c
@@ -71,4 +71,6 @@ void s3c_sdhci2_set_platdata(struct s3c_sdhci_platdata *pd)
 		set->cfg_gpio = pd->cfg_gpio;
 	if (pd->cfg_card)
 		set->cfg_card = pd->cfg_card;
+	if (pd->host_caps)
+		set->host_caps = pd->host_caps;
 }
diff --git a/arch/avr32/kernel/process.c b/arch/avr32/kernel/process.c
index e5daddf..9c46aaa 100644
--- a/arch/avr32/kernel/process.c
+++ b/arch/avr32/kernel/process.c
@@ -384,8 +384,9 @@ asmlinkage int sys_vfork(struct pt_regs *regs)
 }
 
 asmlinkage int sys_execve(const char __user *ufilename,
-			  char __user *__user *uargv,
-			  char __user *__user *uenvp, struct pt_regs *regs)
+			  const char __user *const __user *uargv,
+			  const char __user *const __user *uenvp,
+			  struct pt_regs *regs)
 {
 	int error;
 	char *filename;
diff --git a/arch/avr32/kernel/sys_avr32.c b/arch/avr32/kernel/sys_avr32.c
index 459349b..62635a0 100644
--- a/arch/avr32/kernel/sys_avr32.c
+++ b/arch/avr32/kernel/sys_avr32.c
@@ -7,7 +7,9 @@
  */
 #include <linux/unistd.h>
 
-int kernel_execve(const char *file, char **argv, char **envp)
+int kernel_execve(const char *file,
+		  const char *const *argv,
+		  const char *const *envp)
 {
 	register long scno asm("r8") = __NR_execve;
 	register long sc1 asm("r12") = (long)file;
diff --git a/arch/blackfin/include/asm/bitops.h b/arch/blackfin/include/asm/bitops.h
index d5872cd..3f7ef4d 100644
--- a/arch/blackfin/include/asm/bitops.h
+++ b/arch/blackfin/include/asm/bitops.h
@@ -22,7 +22,9 @@
 
 #include <asm-generic/bitops/sched.h>
 #include <asm-generic/bitops/ffs.h>
+#include <asm-generic/bitops/const_hweight.h>
 #include <asm-generic/bitops/lock.h>
+
 #include <asm-generic/bitops/ext2-non-atomic.h>
 #include <asm-generic/bitops/ext2-atomic.h>
 #include <asm-generic/bitops/minix.h>
@@ -115,7 +117,7 @@ static inline int test_and_change_bit(int nr, volatile unsigned long *addr)
  * of bits set) of a N-bit word
  */
 
-static inline unsigned int hweight32(unsigned int w)
+static inline unsigned int __arch_hweight32(unsigned int w)
 {
 	unsigned int res;
 
@@ -125,19 +127,20 @@ static inline unsigned int hweight32(unsigned int w)
 	return res;
 }
 
-static inline unsigned int hweight64(__u64 w)
+static inline unsigned int __arch_hweight64(__u64 w)
 {
-	return hweight32((unsigned int)(w >> 32)) + hweight32((unsigned int)w);
+	return __arch_hweight32((unsigned int)(w >> 32)) +
+	       __arch_hweight32((unsigned int)w);
 }
 
-static inline unsigned int hweight16(unsigned int w)
+static inline unsigned int __arch_hweight16(unsigned int w)
 {
-	return hweight32(w & 0xffff);
+	return __arch_hweight32(w & 0xffff);
 }
 
-static inline unsigned int hweight8(unsigned int w)
+static inline unsigned int __arch_hweight8(unsigned int w)
 {
-	return hweight32(w & 0xff);
+	return __arch_hweight32(w & 0xff);
 }
 
 #endif				/* _BLACKFIN_BITOPS_H */
diff --git a/arch/blackfin/include/asm/unistd.h b/arch/blackfin/include/asm/unistd.h
index 22886cb..14fcd25 100644
--- a/arch/blackfin/include/asm/unistd.h
+++ b/arch/blackfin/include/asm/unistd.h
@@ -389,8 +389,11 @@
 #define __NR_rt_tgsigqueueinfo	368
 #define __NR_perf_event_open	369
 #define __NR_recvmmsg		370
+#define __NR_fanotify_init	371
+#define __NR_fanotify_mark	372
+#define __NR_prlimit64		373
 
-#define __NR_syscall		371
+#define __NR_syscall		374
 #define NR_syscalls		__NR_syscall
 
 /* Old optional stuff no one actually uses */
diff --git a/arch/blackfin/kernel/process.c b/arch/blackfin/kernel/process.c
index a566f61..01f98cb 100644
--- a/arch/blackfin/kernel/process.c
+++ b/arch/blackfin/kernel/process.c
@@ -209,7 +209,9 @@ copy_thread(unsigned long clone_flags,
 /*
  * sys_execve() executes a new program.
  */
-asmlinkage int sys_execve(const char __user *name, char __user * __user *argv, char __user * __user *envp)
+asmlinkage int sys_execve(const char __user *name,
+			  const char __user *const __user *argv,
+			  const char __user *const __user *envp)
 {
 	int error;
 	char *filename;
diff --git a/arch/blackfin/mach-common/entry.S b/arch/blackfin/mach-common/entry.S
index a5847f5..af1bffa 100644
--- a/arch/blackfin/mach-common/entry.S
+++ b/arch/blackfin/mach-common/entry.S
@@ -1628,6 +1628,9 @@ ENTRY(_sys_call_table)
 	.long _sys_rt_tgsigqueueinfo
 	.long _sys_perf_event_open
 	.long _sys_recvmmsg		/* 370 */
+	.long _sys_fanotify_init
+	.long _sys_fanotify_mark
+	.long _sys_prlimit64
 
 	.rept NR_syscalls-(.-_sys_call_table)/4
 	.long _sys_ni_syscall
diff --git a/arch/cris/arch-v10/kernel/process.c b/arch/cris/arch-v10/kernel/process.c
index 93f0f64..9a57db6 100644
--- a/arch/cris/arch-v10/kernel/process.c
+++ b/arch/cris/arch-v10/kernel/process.c
@@ -204,7 +204,9 @@ asmlinkage int sys_vfork(long r10, long r11, long r12, long r13, long mof, long
 /*
  * sys_execve() executes a new program.
  */
-asmlinkage int sys_execve(const char *fname, char **argv, char **envp,
+asmlinkage int sys_execve(const char *fname,
+			  const char *const *argv,
+			  const char *const *envp,
 			  long r13, long mof, long srp, 
 			  struct pt_regs *regs)
 {
diff --git a/arch/cris/arch-v32/kernel/process.c b/arch/cris/arch-v32/kernel/process.c
index 2661a95..562f847 100644
--- a/arch/cris/arch-v32/kernel/process.c
+++ b/arch/cris/arch-v32/kernel/process.c
@@ -218,8 +218,10 @@ sys_vfork(long r10, long r11, long r12, long r13, long mof, long srp,
 
 /* sys_execve() executes a new program. */
 asmlinkage int
-sys_execve(const char *fname, char **argv, char **envp, long r13, long mof, long srp,
-	struct pt_regs *regs)
+sys_execve(const char *fname,
+	   const char *const *argv,
+	   const char *const *envp, long r13, long mof, long srp,
+	   struct pt_regs *regs)
 {
 	int error;
 	char *filename;
diff --git a/arch/frv/kernel/process.c b/arch/frv/kernel/process.c
index 428931c..2b63b01 100644
--- a/arch/frv/kernel/process.c
+++ b/arch/frv/kernel/process.c
@@ -250,8 +250,9 @@ int copy_thread(unsigned long clone_flags,
 /*
  * sys_execve() executes a new program.
  */
-asmlinkage int sys_execve(const char __user *name, char __user * __user *argv,
-			  char __user * __user *envp)
+asmlinkage int sys_execve(const char __user *name,
+			  const char __user *const __user *argv,
+			  const char __user *const __user *envp)
 {
 	int error;
 	char * filename;
diff --git a/arch/h8300/kernel/process.c b/arch/h8300/kernel/process.c
index 8b7b78d..9747813 100644
--- a/arch/h8300/kernel/process.c
+++ b/arch/h8300/kernel/process.c
@@ -212,7 +212,10 @@ int copy_thread(unsigned long clone_flags,
 /*
  * sys_execve() executes a new program.
  */
-asmlinkage int sys_execve(const char *name, char **argv, char **envp,int dummy,...)
+asmlinkage int sys_execve(const char *name,
+			  const char *const *argv,
+			  const char *const *envp,
+			  int dummy, ...)
 {
 	int error;
 	char * filename;
diff --git a/arch/h8300/kernel/sys_h8300.c b/arch/h8300/kernel/sys_h8300.c
index f9b3f44..dc1ac02 100644
--- a/arch/h8300/kernel/sys_h8300.c
+++ b/arch/h8300/kernel/sys_h8300.c
@@ -51,7 +51,9 @@ asmlinkage void syscall_print(void *dummy,...)
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	register long res __asm__("er0");
 	register char *const *_c __asm__("er3") = envp;
diff --git a/arch/ia64/include/asm/unistd.h b/arch/ia64/include/asm/unistd.h
index 87f1bd1..954d398 100644
--- a/arch/ia64/include/asm/unistd.h
+++ b/arch/ia64/include/asm/unistd.h
@@ -356,8 +356,6 @@ asmlinkage unsigned long sys_mmap2(
 				int fd, long pgoff);
 struct pt_regs;
 struct sigaction;
-long sys_execve(const char __user *filename, char __user * __user *argv,
-			   char __user * __user *envp, struct pt_regs *regs);
 asmlinkage long sys_ia64_pipe(void);
 asmlinkage long sys_rt_sigaction(int sig,
 				 const struct sigaction __user *act,
diff --git a/arch/ia64/kernel/process.c b/arch/ia64/kernel/process.c
index a879c03..16f1c7b 100644
--- a/arch/ia64/kernel/process.c
+++ b/arch/ia64/kernel/process.c
@@ -633,7 +633,9 @@ dump_fpu (struct pt_regs *pt, elf_fpregset_t dst)
 }
 
 long
-sys_execve (const char __user *filename, char __user * __user *argv, char __user * __user *envp,
+sys_execve (const char __user *filename,
+	    const char __user *const __user *argv,
+	    const char __user *const __user *envp,
 	    struct pt_regs *regs)
 {
 	char *fname;
diff --git a/arch/m32r/kernel/process.c b/arch/m32r/kernel/process.c
index 8665a4d..422bea9 100644
--- a/arch/m32r/kernel/process.c
+++ b/arch/m32r/kernel/process.c
@@ -289,8 +289,8 @@ asmlinkage int sys_vfork(unsigned long r0, unsigned long r1, unsigned long r2,
  * sys_execve() executes a new program.
  */
 asmlinkage int sys_execve(const char __user *ufilename,
-			  char __user * __user *uargv,
-			  char __user * __user *uenvp,
+			  const char __user *const __user *uargv,
+			  const char __user *const __user *uenvp,
 			  unsigned long r3, unsigned long r4, unsigned long r5,
 			  unsigned long r6, struct pt_regs regs)
 {
diff --git a/arch/m32r/kernel/sys_m32r.c b/arch/m32r/kernel/sys_m32r.c
index 0a00f46..d841fb6 100644
--- a/arch/m32r/kernel/sys_m32r.c
+++ b/arch/m32r/kernel/sys_m32r.c
@@ -93,7 +93,9 @@ asmlinkage int sys_cachectl(char *addr, int nbytes, int op)
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	register long __scno __asm__ ("r7") = __NR_execve;
 	register long __arg3 __asm__ ("r2") = (long)(envp);
diff --git a/arch/m68k/include/asm/ide.h b/arch/m68k/include/asm/ide.h
index 3958726..492fee8 100644
--- a/arch/m68k/include/asm/ide.h
+++ b/arch/m68k/include/asm/ide.h
@@ -1,6 +1,4 @@
 /*
- *  linux/include/asm-m68k/ide.h
- *
  *  Copyright (C) 1994-1996  Linus Torvalds & authors
  */
 
@@ -34,6 +32,8 @@
 #include <asm/io.h>
 #include <asm/irq.h>
 
+#ifdef CONFIG_MMU
+
 /*
  * Get rid of defs from io.h - ide has its private and conflicting versions
  * Since so far no single m68k platform uses ISA/PCI I/O space for IDE, we
@@ -53,5 +53,14 @@
 #define __ide_mm_outsw(port, addr, n)	raw_outsw((u16 *)port, addr, n)
 #define __ide_mm_outsl(port, addr, n)	raw_outsl((u32 *)port, addr, n)
 
+#else
+
+#define __ide_mm_insw(port, addr, n)	io_insw((unsigned int)port, addr, n)
+#define __ide_mm_insl(port, addr, n)	io_insl((unsigned int)port, addr, n)
+#define __ide_mm_outsw(port, addr, n)	io_outsw((unsigned int)port, addr, n)
+#define __ide_mm_outsl(port, addr, n)	io_outsl((unsigned int)port, addr, n)
+
+#endif /* CONFIG_MMU */
+
 #endif /* __KERNEL__ */
 #endif /* _M68K_IDE_H */
diff --git a/arch/m68k/kernel/process.c b/arch/m68k/kernel/process.c
index 221d0b7..18732ab 100644
--- a/arch/m68k/kernel/process.c
+++ b/arch/m68k/kernel/process.c
@@ -315,7 +315,9 @@ EXPORT_SYMBOL(dump_fpu);
 /*
  * sys_execve() executes a new program.
  */
-asmlinkage int sys_execve(const char __user *name, char __user * __user *argv, char __user * __user *envp)
+asmlinkage int sys_execve(const char __user *name,
+			  const char __user *const __user *argv,
+			  const char __user *const __user *envp)
 {
 	int error;
 	char * filename;
diff --git a/arch/m68k/kernel/sys_m68k.c b/arch/m68k/kernel/sys_m68k.c
index 7789669..2f431ec 100644
--- a/arch/m68k/kernel/sys_m68k.c
+++ b/arch/m68k/kernel/sys_m68k.c
@@ -459,7 +459,9 @@ asmlinkage int sys_getpagesize(void)
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	register long __res asm ("%d0") = __NR_execve;
 	register long __a asm ("%d1") = (long)(filename);
diff --git a/arch/m68knommu/kernel/process.c b/arch/m68knommu/kernel/process.c
index 6350f68..6d33905 100644
--- a/arch/m68knommu/kernel/process.c
+++ b/arch/m68knommu/kernel/process.c
@@ -316,14 +316,14 @@ void dump(struct pt_regs *fp)
 		fp->d0, fp->d1, fp->d2, fp->d3);
 	printk(KERN_EMERG "d4: %08lx    d5: %08lx    a0: %08lx    a1: %08lx\n",
 		fp->d4, fp->d5, fp->a0, fp->a1);
-	printk(KERN_EMERG "\nUSP: %08x   TRAPFRAME: %08x\n",
-		(unsigned int) rdusp(), (unsigned int) fp);
+	printk(KERN_EMERG "\nUSP: %08x   TRAPFRAME: %p\n",
+		(unsigned int) rdusp(), fp);
 
 	printk(KERN_EMERG "\nCODE:");
 	tp = ((unsigned char *) fp->pc) - 0x20;
 	for (sp = (unsigned long *) tp, i = 0; (i < 0x40);  i += 4) {
 		if ((i % 0x10) == 0)
-			printk(KERN_EMERG "%08x: ", (int) (tp + i));
+			printk(KERN_EMERG "%p: ", tp + i);
 		printk("%08x ", (int) *sp++);
 	}
 	printk(KERN_EMERG "\n");
@@ -332,7 +332,7 @@ void dump(struct pt_regs *fp)
 	tp = ((unsigned char *) fp) - 0x40;
 	for (sp = (unsigned long *) tp, i = 0; (i < 0xc0); i += 4) {
 		if ((i % 0x10) == 0)
-			printk(KERN_EMERG "%08x: ", (int) (tp + i));
+			printk(KERN_EMERG "%p: ", tp + i);
 		printk("%08x ", (int) *sp++);
 	}
 	printk(KERN_EMERG "\n");
@@ -341,7 +341,7 @@ void dump(struct pt_regs *fp)
 	tp = (unsigned char *) (rdusp() - 0x10);
 	for (sp = (unsigned long *) tp, i = 0; (i < 0x80); i += 4) {
 		if ((i % 0x10) == 0)
-			printk(KERN_EMERG "%08x: ", (int) (tp + i));
+			printk(KERN_EMERG "%p: ", tp + i);
 		printk("%08x ", (int) *sp++);
 	}
 	printk(KERN_EMERG "\n");
@@ -350,7 +350,9 @@ void dump(struct pt_regs *fp)
 /*
  * sys_execve() executes a new program.
  */
-asmlinkage int sys_execve(const char *name, char **argv, char **envp)
+asmlinkage int sys_execve(const char *name,
+			  const char *const *argv,
+			  const char *const *envp)
 {
 	int error;
 	char * filename;
diff --git a/arch/m68knommu/kernel/sys_m68k.c b/arch/m68knommu/kernel/sys_m68k.c
index d65e9c4..68488ae 100644
--- a/arch/m68knommu/kernel/sys_m68k.c
+++ b/arch/m68knommu/kernel/sys_m68k.c
@@ -44,7 +44,9 @@ asmlinkage int sys_getpagesize(void)
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	register long __res asm ("%d0") = __NR_execve;
 	register long __a asm ("%d1") = (long)(filename);
diff --git a/arch/microblaze/kernel/prom_parse.c b/arch/microblaze/kernel/prom_parse.c
index d33ba17..99d9b61 100644
--- a/arch/microblaze/kernel/prom_parse.c
+++ b/arch/microblaze/kernel/prom_parse.c
@@ -73,7 +73,7 @@ int of_irq_map_pci(struct pci_dev *pdev, struct of_irq *out_irq)
 		/* We can only get here if we hit a P2P bridge with no node,
 		 * let's do standard swizzling and try again
 		 */
-		lspec = of_irq_pci_swizzle(PCI_SLOT(pdev->devfn), lspec);
+		lspec = pci_swizzle_interrupt_pin(pdev, lspec);
 		pdev = ppdev;
 	}
 
diff --git a/arch/microblaze/kernel/sys_microblaze.c b/arch/microblaze/kernel/sys_microblaze.c
index 6abab6e..2250fe9 100644
--- a/arch/microblaze/kernel/sys_microblaze.c
+++ b/arch/microblaze/kernel/sys_microblaze.c
@@ -47,8 +47,10 @@ asmlinkage long microblaze_clone(int flags, unsigned long stack, struct pt_regs
 	return do_fork(flags, stack, regs, 0, NULL, NULL);
 }
 
-asmlinkage long microblaze_execve(const char __user *filenamei, char __user *__user *argv,
-			char __user *__user *envp, struct pt_regs *regs)
+asmlinkage long microblaze_execve(const char __user *filenamei,
+				  const char __user *const __user *argv,
+				  const char __user *const __user *envp,
+				  struct pt_regs *regs)
 {
 	int error;
 	char *filename;
@@ -77,7 +79,9 @@ asmlinkage long sys_mmap(unsigned long addr, unsigned long len,
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	register const char *__a __asm__("r5") = filename;
 	register const void *__b __asm__("r6") = argv;
diff --git a/arch/microblaze/pci/pci-common.c b/arch/microblaze/pci/pci-common.c
index 23be25f..55ef532 100644
--- a/arch/microblaze/pci/pci-common.c
+++ b/arch/microblaze/pci/pci-common.c
@@ -27,10 +27,11 @@
 #include <linux/irq.h>
 #include <linux/vmalloc.h>
 #include <linux/slab.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
 
 #include <asm/processor.h>
 #include <asm/io.h>
-#include <asm/prom.h>
 #include <asm/pci-bridge.h>
 #include <asm/byteorder.h>
 
@@ -1077,7 +1078,7 @@ void __devinit pcibios_setup_bus_devices(struct pci_bus *bus)
 		struct dev_archdata *sd = &dev->dev.archdata;
 
 		/* Setup OF node pointer in archdata */
-		sd->of_node = pci_device_to_OF_node(dev);
+		dev->dev.of_node = pci_device_to_OF_node(dev);
 
 		/* Fixup NUMA node as it may not be setup yet by the generic
 		 * code and is needed by the DMA init
diff --git a/arch/microblaze/pci/xilinx_pci.c b/arch/microblaze/pci/xilinx_pci.c
index 7869a41..0687a42 100644
--- a/arch/microblaze/pci/xilinx_pci.c
+++ b/arch/microblaze/pci/xilinx_pci.c
@@ -16,6 +16,7 @@
 
 #include <linux/ioport.h>
 #include <linux/of.h>
+#include <linux/of_address.h>
 #include <linux/pci.h>
 #include <asm/io.h>
 
diff --git a/arch/mips/kernel/syscall.c b/arch/mips/kernel/syscall.c
index bddce0b..1dc6edf 100644
--- a/arch/mips/kernel/syscall.c
+++ b/arch/mips/kernel/syscall.c
@@ -258,8 +258,10 @@ asmlinkage int sys_execve(nabi_no_regargs struct pt_regs regs)
 	error = PTR_ERR(filename);
 	if (IS_ERR(filename))
 		goto out;
-	error = do_execve(filename, (char __user *__user *) (long)regs.regs[5],
-	                  (char __user *__user *) (long)regs.regs[6], &regs);
+	error = do_execve(filename,
+			  (const char __user *const __user *) (long)regs.regs[5],
+	                  (const char __user *const __user *) (long)regs.regs[6],
+			  &regs);
 	putname(filename);
 
 out:
@@ -436,7 +438,9 @@ asmlinkage void bad_stack(void)
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	register unsigned long __a0 asm("$4") = (unsigned long) filename;
 	register unsigned long __a1 asm("$5") = (unsigned long) argv;
diff --git a/arch/mn10300/kernel/process.c b/arch/mn10300/kernel/process.c
index 762eb32..f48373e 100644
--- a/arch/mn10300/kernel/process.c
+++ b/arch/mn10300/kernel/process.c
@@ -269,8 +269,8 @@ asmlinkage long sys_vfork(void)
 }
 
 asmlinkage long sys_execve(const char __user *name,
-			   char __user * __user *argv,
-			   char __user * __user *envp)
+			   const char __user *const __user *argv,
+			   const char __user *const __user *envp)
 {
 	char *filename;
 	int error;
diff --git a/arch/mn10300/mm/dma-alloc.c b/arch/mn10300/mm/dma-alloc.c
index 4e34880..159acb0 100644
--- a/arch/mn10300/mm/dma-alloc.c
+++ b/arch/mn10300/mm/dma-alloc.c
@@ -25,7 +25,8 @@ void *dma_alloc_coherent(struct device *dev, size_t size,
 	unsigned long addr;
 	void *ret;
 
-	printk("dma_alloc_coherent(%s,%zu,,%x)\n", dev_name(dev), size, gfp);
+	pr_debug("dma_alloc_coherent(%s,%zu,%x)\n",
+		 dev ? dev_name(dev) : "?", size, gfp);
 
 	if (0xbe000000 - pci_sram_allocated >= size) {
 		size = (size + 255) & ~255;
diff --git a/arch/parisc/hpux/fs.c b/arch/parisc/hpux/fs.c
index 1444875..0dc8543 100644
--- a/arch/parisc/hpux/fs.c
+++ b/arch/parisc/hpux/fs.c
@@ -41,8 +41,10 @@ int hpux_execve(struct pt_regs *regs)
 	if (IS_ERR(filename))
 		goto out;
 
-	error = do_execve(filename, (char __user * __user *) regs->gr[25],
-		(char __user * __user *) regs->gr[24], regs);
+	error = do_execve(filename,
+			  (const char __user *const __user *) regs->gr[25],
+			  (const char __user *const __user *) regs->gr[24],
+			  regs);
 
 	putname(filename);
 
diff --git a/arch/parisc/kernel/process.c b/arch/parisc/kernel/process.c
index 76332da..4b4b918 100644
--- a/arch/parisc/kernel/process.c
+++ b/arch/parisc/kernel/process.c
@@ -348,17 +348,22 @@ asmlinkage int sys_execve(struct pt_regs *regs)
 	error = PTR_ERR(filename);
 	if (IS_ERR(filename))
 		goto out;
-	error = do_execve(filename, (char __user * __user *) regs->gr[25],
-		(char __user * __user *) regs->gr[24], regs);
+	error = do_execve(filename,
+			  (const char __user *const __user *) regs->gr[25],
+			  (const char __user *const __user *) regs->gr[24],
+			  regs);
 	putname(filename);
 out:
 
 	return error;
 }
 
-extern int __execve(const char *filename, char *const argv[],
-		char *const envp[], struct task_struct *task);
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+extern int __execve(const char *filename,
+		    const char *const argv[],
+		    const char *const envp[], struct task_struct *task);
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	return __execve(filename, argv, envp, current);
 }
diff --git a/arch/powerpc/Makefile b/arch/powerpc/Makefile
index e3ea151..b7212b6 100644
--- a/arch/powerpc/Makefile
+++ b/arch/powerpc/Makefile
@@ -164,7 +164,7 @@ drivers-$(CONFIG_OPROFILE)	+= arch/powerpc/oprofile/
 all: zImage
 
 # With make 3.82 we cannot mix normal and wildcard targets
-BOOT_TARGETS1 := zImage zImage.initrd uImaged
+BOOT_TARGETS1 := zImage zImage.initrd uImage
 BOOT_TARGETS2 := zImage% dtbImage% treeImage.% cuImage.% simpleImage.%
 
 PHONY += $(BOOT_TARGETS1) $(BOOT_TARGETS2)
diff --git a/arch/powerpc/boot/dts/canyonlands.dts b/arch/powerpc/boot/dts/canyonlands.dts
index 5806ef0..a303703 100644
--- a/arch/powerpc/boot/dts/canyonlands.dts
+++ b/arch/powerpc/boot/dts/canyonlands.dts
@@ -163,6 +163,14 @@
 			interrupts = <0x1e 4>;
 		};
 
+		SATA0: sata@bffd1000 {
+			compatible = "amcc,sata-460ex";
+			reg = <4 0xbffd1000 0x800 4 0xbffd0800 0x400>;
+			interrupt-parent = <&UIC3>;
+			interrupts = <0x0 0x4       /* SATA */
+				      0x5 0x4>;     /* AHBDMA */
+		};
+
 		POB0: opb {
 			compatible = "ibm,opb-460ex", "ibm,opb";
 			#address-cells = <1>;
diff --git a/arch/powerpc/include/asm/mmu-hash64.h b/arch/powerpc/include/asm/mmu-hash64.h
index 0e398cf..acac35d 100644
--- a/arch/powerpc/include/asm/mmu-hash64.h
+++ b/arch/powerpc/include/asm/mmu-hash64.h
@@ -433,7 +433,7 @@ typedef struct {
  * with.  However gcc is not clever enough to compute the
  * modulus (2^n-1) without a second multiply.
  */
-#define vsid_scrample(protovsid, size) \
+#define vsid_scramble(protovsid, size) \
 	((((protovsid) * VSID_MULTIPLIER_##size) % VSID_MODULUS_##size))
 
 #else /* 1 */
diff --git a/arch/powerpc/include/asm/reg.h b/arch/powerpc/include/asm/reg.h
index d8be016..ff0005eec 100644
--- a/arch/powerpc/include/asm/reg.h
+++ b/arch/powerpc/include/asm/reg.h
@@ -951,7 +951,14 @@
 #ifdef CONFIG_PPC64
 
 extern void ppc64_runlatch_on(void);
-extern void ppc64_runlatch_off(void);
+extern void __ppc64_runlatch_off(void);
+
+#define ppc64_runlatch_off()					\
+	do {							\
+		if (cpu_has_feature(CPU_FTR_CTRL) &&		\
+		    test_thread_flag(TIF_RUNLATCH))		\
+			__ppc64_runlatch_off();			\
+	} while (0)
 
 extern unsigned long scom970_read(unsigned int address);
 extern void scom970_write(unsigned int address, unsigned long value);
diff --git a/arch/powerpc/include/asm/rwsem.h b/arch/powerpc/include/asm/rwsem.h
index 24cd928..8447d89 100644
--- a/arch/powerpc/include/asm/rwsem.h
+++ b/arch/powerpc/include/asm/rwsem.h
@@ -21,15 +21,20 @@
 /*
  * the semaphore definition
  */
-struct rw_semaphore {
-	/* XXX this should be able to be an atomic_t  -- paulus */
-	signed int		count;
-#define RWSEM_UNLOCKED_VALUE		0x00000000
-#define RWSEM_ACTIVE_BIAS		0x00000001
-#define RWSEM_ACTIVE_MASK		0x0000ffff
-#define RWSEM_WAITING_BIAS		(-0x00010000)
+#ifdef CONFIG_PPC64
+# define RWSEM_ACTIVE_MASK		0xffffffffL
+#else
+# define RWSEM_ACTIVE_MASK		0x0000ffffL
+#endif
+
+#define RWSEM_UNLOCKED_VALUE		0x00000000L
+#define RWSEM_ACTIVE_BIAS		0x00000001L
+#define RWSEM_WAITING_BIAS		(-RWSEM_ACTIVE_MASK-1)
 #define RWSEM_ACTIVE_READ_BIAS		RWSEM_ACTIVE_BIAS
 #define RWSEM_ACTIVE_WRITE_BIAS		(RWSEM_WAITING_BIAS + RWSEM_ACTIVE_BIAS)
+
+struct rw_semaphore {
+	long			count;
 	spinlock_t		wait_lock;
 	struct list_head	wait_list;
 #ifdef CONFIG_DEBUG_LOCK_ALLOC
@@ -43,9 +48,13 @@ struct rw_semaphore {
 # define __RWSEM_DEP_MAP_INIT(lockname)
 #endif
 
-#define __RWSEM_INITIALIZER(name) \
-	{ RWSEM_UNLOCKED_VALUE, __SPIN_LOCK_UNLOCKED((name).wait_lock), \
-	  LIST_HEAD_INIT((name).wait_list) __RWSEM_DEP_MAP_INIT(name) }
+#define __RWSEM_INITIALIZER(name)				\
+{								\
+	RWSEM_UNLOCKED_VALUE,					\
+	__SPIN_LOCK_UNLOCKED((name).wait_lock),			\
+	LIST_HEAD_INIT((name).wait_list)			\
+	__RWSEM_DEP_MAP_INIT(name)				\
+}
 
 #define DECLARE_RWSEM(name)		\
 	struct rw_semaphore name = __RWSEM_INITIALIZER(name)
@@ -70,13 +79,13 @@ extern void __init_rwsem(struct rw_semaphore *sem, const char *name,
  */
 static inline void __down_read(struct rw_semaphore *sem)
 {
-	if (unlikely(atomic_inc_return((atomic_t *)(&sem->count)) <= 0))
+	if (unlikely(atomic_long_inc_return((atomic_long_t *)&sem->count) <= 0))
 		rwsem_down_read_failed(sem);
 }
 
 static inline int __down_read_trylock(struct rw_semaphore *sem)
 {
-	int tmp;
+	long tmp;
 
 	while ((tmp = sem->count) >= 0) {
 		if (tmp == cmpxchg(&sem->count, tmp,
@@ -92,10 +101,10 @@ static inline int __down_read_trylock(struct rw_semaphore *sem)
  */
 static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
 {
-	int tmp;
+	long tmp;
 
-	tmp = atomic_add_return(RWSEM_ACTIVE_WRITE_BIAS,
-				(atomic_t *)(&sem->count));
+	tmp = atomic_long_add_return(RWSEM_ACTIVE_WRITE_BIAS,
+				     (atomic_long_t *)&sem->count);
 	if (unlikely(tmp != RWSEM_ACTIVE_WRITE_BIAS))
 		rwsem_down_write_failed(sem);
 }
@@ -107,7 +116,7 @@ static inline void __down_write(struct rw_semaphore *sem)
 
 static inline int __down_write_trylock(struct rw_semaphore *sem)
 {
-	int tmp;
+	long tmp;
 
 	tmp = cmpxchg(&sem->count, RWSEM_UNLOCKED_VALUE,
 		      RWSEM_ACTIVE_WRITE_BIAS);
@@ -119,9 +128,9 @@ static inline int __down_write_trylock(struct rw_semaphore *sem)
  */
 static inline void __up_read(struct rw_semaphore *sem)
 {
-	int tmp;
+	long tmp;
 
-	tmp = atomic_dec_return((atomic_t *)(&sem->count));
+	tmp = atomic_long_dec_return((atomic_long_t *)&sem->count);
 	if (unlikely(tmp < -1 && (tmp & RWSEM_ACTIVE_MASK) == 0))
 		rwsem_wake(sem);
 }
@@ -131,17 +140,17 @@ static inline void __up_read(struct rw_semaphore *sem)
  */
 static inline void __up_write(struct rw_semaphore *sem)
 {
-	if (unlikely(atomic_sub_return(RWSEM_ACTIVE_WRITE_BIAS,
-			      (atomic_t *)(&sem->count)) < 0))
+	if (unlikely(atomic_long_sub_return(RWSEM_ACTIVE_WRITE_BIAS,
+				 (atomic_long_t *)&sem->count) < 0))
 		rwsem_wake(sem);
 }
 
 /*
  * implement atomic add functionality
  */
-static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem)
+static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
 {
-	atomic_add(delta, (atomic_t *)(&sem->count));
+	atomic_long_add(delta, (atomic_long_t *)&sem->count);
 }
 
 /*
@@ -149,9 +158,10 @@ static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem)
  */
 static inline void __downgrade_write(struct rw_semaphore *sem)
 {
-	int tmp;
+	long tmp;
 
-	tmp = atomic_add_return(-RWSEM_WAITING_BIAS, (atomic_t *)(&sem->count));
+	tmp = atomic_long_add_return(-RWSEM_WAITING_BIAS,
+				     (atomic_long_t *)&sem->count);
 	if (tmp < 0)
 		rwsem_downgrade_wake(sem);
 }
@@ -159,14 +169,14 @@ static inline void __downgrade_write(struct rw_semaphore *sem)
 /*
  * implement exchange and add functionality
  */
-static inline int rwsem_atomic_update(int delta, struct rw_semaphore *sem)
+static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem)
 {
-	return atomic_add_return(delta, (atomic_t *)(&sem->count));
+	return atomic_long_add_return(delta, (atomic_long_t *)&sem->count);
 }
 
 static inline int rwsem_is_locked(struct rw_semaphore *sem)
 {
-	return (sem->count != 0);
+	return sem->count != 0;
 }
 
 #endif	/* __KERNEL__ */
diff --git a/arch/powerpc/include/asm/systbl.h b/arch/powerpc/include/asm/systbl.h
index a5ee345..3d21266 100644
--- a/arch/powerpc/include/asm/systbl.h
+++ b/arch/powerpc/include/asm/systbl.h
@@ -326,3 +326,6 @@ SYSCALL_SPU(perf_event_open)
 COMPAT_SYS_SPU(preadv)
 COMPAT_SYS_SPU(pwritev)
 COMPAT_SYS(rt_tgsigqueueinfo)
+SYSCALL(fanotify_init)
+COMPAT_SYS(fanotify_mark)
+SYSCALL_SPU(prlimit64)
diff --git a/arch/powerpc/include/asm/unistd.h b/arch/powerpc/include/asm/unistd.h
index f0a1026..597e6f9 100644
--- a/arch/powerpc/include/asm/unistd.h
+++ b/arch/powerpc/include/asm/unistd.h
@@ -345,10 +345,13 @@
 #define __NR_preadv		320
 #define __NR_pwritev		321
 #define __NR_rt_tgsigqueueinfo	322
+#define __NR_fanotify_init	323
+#define __NR_fanotify_mark	324
+#define __NR_prlimit64		325
 
 #ifdef __KERNEL__
 
-#define __NR_syscalls		323
+#define __NR_syscalls		326
 
 #define __NR__exit __NR_exit
 #define NR_syscalls	__NR_syscalls
diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c
index 65e2b4e..1f9123f 100644
--- a/arch/powerpc/kernel/cputable.c
+++ b/arch/powerpc/kernel/cputable.c
@@ -1826,7 +1826,6 @@ static struct cpu_spec __initdata cpu_specs[] = {
 		.cpu_features		= CPU_FTRS_47X,
 		.cpu_user_features	= COMMON_USER_BOOKE |
 			PPC_FEATURE_HAS_FPU,
-		.cpu_user_features	= COMMON_USER_BOOKE,
 		.mmu_features		= MMU_FTR_TYPE_47x |
 			MMU_FTR_USE_TLBIVAX_BCAST | MMU_FTR_LOCK_BCAST_INVAL,
 		.icache_bsize		= 32,
diff --git a/arch/powerpc/kernel/crash.c b/arch/powerpc/kernel/crash.c
index 417f7b0..4457382 100644
--- a/arch/powerpc/kernel/crash.c
+++ b/arch/powerpc/kernel/crash.c
@@ -402,6 +402,18 @@ void default_machine_crash_shutdown(struct pt_regs *regs)
 	 */
 	hard_irq_disable();
 
+	/*
+	 * Make a note of crashing cpu. Will be used in machine_kexec
+	 * such that another IPI will not be sent.
+	 */
+	crashing_cpu = smp_processor_id();
+	crash_save_cpu(regs, crashing_cpu);
+	crash_kexec_prepare_cpus(crashing_cpu);
+	cpu_set(crashing_cpu, cpus_in_crash);
+#if defined(CONFIG_PPC_STD_MMU_64) && defined(CONFIG_SMP)
+	crash_kexec_wait_realmode(crashing_cpu);
+#endif
+
 	for_each_irq(i) {
 		struct irq_desc *desc = irq_to_desc(i);
 
@@ -438,18 +450,8 @@ void default_machine_crash_shutdown(struct pt_regs *regs)
 	crash_shutdown_cpu = -1;
 	__debugger_fault_handler = old_handler;
 
-	/*
-	 * Make a note of crashing cpu. Will be used in machine_kexec
-	 * such that another IPI will not be sent.
-	 */
-	crashing_cpu = smp_processor_id();
-	crash_save_cpu(regs, crashing_cpu);
-	crash_kexec_prepare_cpus(crashing_cpu);
-	cpu_set(crashing_cpu, cpus_in_crash);
 	crash_kexec_stop_spus();
-#if defined(CONFIG_PPC_STD_MMU_64) && defined(CONFIG_SMP)
-	crash_kexec_wait_realmode(crashing_cpu);
-#endif
+
 	if (ppc_md.kexec_cpu_down)
 		ppc_md.kexec_cpu_down(1, 0);
 }
diff --git a/arch/powerpc/kernel/head_44x.S b/arch/powerpc/kernel/head_44x.S
index 5ab484e..562305b 100644
--- a/arch/powerpc/kernel/head_44x.S
+++ b/arch/powerpc/kernel/head_44x.S
@@ -113,6 +113,10 @@ _ENTRY(_start);
 	stw	r5, 0(r4)	/* Save abatron_pteptrs at a fixed location */
 	stw	r6, 0(r5)
 
+	/* Clear the Machine Check Syndrome Register */
+	li	r0,0
+	mtspr	SPRN_MCSR,r0
+
 	/* Let's move on */
 	lis	r4,start_kernel@h
 	ori	r4,r4,start_kernel@l
diff --git a/arch/powerpc/kernel/head_64.S b/arch/powerpc/kernel/head_64.S
index 844a44b..4d6681d 100644
--- a/arch/powerpc/kernel/head_64.S
+++ b/arch/powerpc/kernel/head_64.S
@@ -572,9 +572,6 @@ __secondary_start:
 	/* Set thread priority to MEDIUM */
 	HMT_MEDIUM
 
-	/* Do early setup for that CPU (stab, slb, hash table pointer) */
-	bl	.early_setup_secondary
-
 	/* Initialize the kernel stack.  Just a repeat for iSeries.	 */
 	LOAD_REG_ADDR(r3, current_set)
 	sldi	r28,r24,3		/* get current_set[cpu#]	 */
@@ -582,6 +579,9 @@ __secondary_start:
 	addi	r1,r1,THREAD_SIZE-STACK_FRAME_OVERHEAD
 	std	r1,PACAKSAVE(r13)
 
+	/* Do early setup for that CPU (stab, slb, hash table pointer) */
+	bl	.early_setup_secondary
+
 	/* Clear backchain so we get nice backtraces */
 	li	r7,0
 	mtlr	r7
diff --git a/arch/powerpc/kernel/idle.c b/arch/powerpc/kernel/idle.c
index 049dda6..39a2baa 100644
--- a/arch/powerpc/kernel/idle.c
+++ b/arch/powerpc/kernel/idle.c
@@ -94,9 +94,9 @@ void cpu_idle(void)
 		HMT_medium();
 		ppc64_runlatch_on();
 		tick_nohz_restart_sched_tick();
+		preempt_enable_no_resched();
 		if (cpu_should_die())
 			cpu_die();
-		preempt_enable_no_resched();
 		schedule();
 		preempt_disable();
 	}
diff --git a/arch/powerpc/kernel/irq.c b/arch/powerpc/kernel/irq.c
index d3ce67c..4a65386 100644
--- a/arch/powerpc/kernel/irq.c
+++ b/arch/powerpc/kernel/irq.c
@@ -67,6 +67,7 @@
 #include <asm/machdep.h>
 #include <asm/udbg.h>
 #include <asm/dbell.h>
+#include <asm/smp.h>
 
 #ifdef CONFIG_PPC64
 #include <asm/paca.h>
@@ -446,22 +447,23 @@ struct thread_info *mcheckirq_ctx[NR_CPUS] __read_mostly;
 void exc_lvl_ctx_init(void)
 {
 	struct thread_info *tp;
-	int i;
+	int i, hw_cpu;
 
 	for_each_possible_cpu(i) {
-		memset((void *)critirq_ctx[i], 0, THREAD_SIZE);
-		tp = critirq_ctx[i];
+		hw_cpu = get_hard_smp_processor_id(i);
+		memset((void *)critirq_ctx[hw_cpu], 0, THREAD_SIZE);
+		tp = critirq_ctx[hw_cpu];
 		tp->cpu = i;
 		tp->preempt_count = 0;
 
 #ifdef CONFIG_BOOKE
-		memset((void *)dbgirq_ctx[i], 0, THREAD_SIZE);
-		tp = dbgirq_ctx[i];
+		memset((void *)dbgirq_ctx[hw_cpu], 0, THREAD_SIZE);
+		tp = dbgirq_ctx[hw_cpu];
 		tp->cpu = i;
 		tp->preempt_count = 0;
 
-		memset((void *)mcheckirq_ctx[i], 0, THREAD_SIZE);
-		tp = mcheckirq_ctx[i];
+		memset((void *)mcheckirq_ctx[hw_cpu], 0, THREAD_SIZE);
+		tp = mcheckirq_ctx[hw_cpu];
 		tp->cpu = i;
 		tp->preempt_count = HARDIRQ_OFFSET;
 #endif
diff --git a/arch/powerpc/kernel/pci_of_scan.c b/arch/powerpc/kernel/pci_of_scan.c
index 6ddb795f..e751506 100644
--- a/arch/powerpc/kernel/pci_of_scan.c
+++ b/arch/powerpc/kernel/pci_of_scan.c
@@ -336,7 +336,7 @@ static void __devinit __of_scan_bus(struct device_node *node,
 		if (dev->hdr_type == PCI_HEADER_TYPE_BRIDGE ||
 		    dev->hdr_type == PCI_HEADER_TYPE_CARDBUS) {
 			struct device_node *child = pci_device_to_OF_node(dev);
-			if (dev)
+			if (child)
 				of_scan_pci_bridge(child, dev);
 		}
 	}
diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c
index feacfb7..b1c648a 100644
--- a/arch/powerpc/kernel/process.c
+++ b/arch/powerpc/kernel/process.c
@@ -728,7 +728,7 @@ int copy_thread(unsigned long clone_flags, unsigned long usp,
 		p->thread.regs = childregs;
 		if (clone_flags & CLONE_SETTLS) {
 #ifdef CONFIG_PPC64
-			if (!test_thread_flag(TIF_32BIT))
+			if (!is_32bit_task())
 				childregs->gpr[13] = childregs->gpr[6];
 			else
 #endif
@@ -823,7 +823,7 @@ void start_thread(struct pt_regs *regs, unsigned long start, unsigned long sp)
 	regs->nip = start;
 	regs->msr = MSR_USER;
 #else
-	if (!test_thread_flag(TIF_32BIT)) {
+	if (!is_32bit_task()) {
 		unsigned long entry, toc;
 
 		/* start is a relocated pointer to the function descriptor for
@@ -995,7 +995,7 @@ int sys_clone(unsigned long clone_flags, unsigned long usp,
 	if (usp == 0)
 		usp = regs->gpr[1];	/* stack pointer for child */
 #ifdef CONFIG_PPC64
-	if (test_thread_flag(TIF_32BIT)) {
+	if (is_32bit_task()) {
 		parent_tidp = TRUNC_PTR(parent_tidp);
 		child_tidp = TRUNC_PTR(child_tidp);
 	}
@@ -1034,8 +1034,9 @@ int sys_execve(unsigned long a0, unsigned long a1, unsigned long a2,
 	flush_fp_to_thread(current);
 	flush_altivec_to_thread(current);
 	flush_spe_to_thread(current);
-	error = do_execve(filename, (char __user * __user *) a1,
-			  (char __user * __user *) a2, regs);
+	error = do_execve(filename,
+			  (const char __user *const __user *) a1,
+			  (const char __user *const __user *) a2, regs);
 	putname(filename);
 out:
 	return error;
@@ -1198,19 +1199,17 @@ void ppc64_runlatch_on(void)
 	}
 }
 
-void ppc64_runlatch_off(void)
+void __ppc64_runlatch_off(void)
 {
 	unsigned long ctrl;
 
-	if (cpu_has_feature(CPU_FTR_CTRL) && test_thread_flag(TIF_RUNLATCH)) {
-		HMT_medium();
+	HMT_medium();
 
-		clear_thread_flag(TIF_RUNLATCH);
+	clear_thread_flag(TIF_RUNLATCH);
 
-		ctrl = mfspr(SPRN_CTRLF);
-		ctrl &= ~CTRL_RUNLATCH;
-		mtspr(SPRN_CTRLT, ctrl);
-	}
+	ctrl = mfspr(SPRN_CTRLF);
+	ctrl &= ~CTRL_RUNLATCH;
+	mtspr(SPRN_CTRLT, ctrl);
 }
 #endif
 
diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c
index a10ffc8..93666f9 100644
--- a/arch/powerpc/kernel/setup_32.c
+++ b/arch/powerpc/kernel/setup_32.c
@@ -258,17 +258,18 @@ static void __init irqstack_early_init(void)
 #if defined(CONFIG_BOOKE) || defined(CONFIG_40x)
 static void __init exc_lvl_early_init(void)
 {
-	unsigned int i;
+	unsigned int i, hw_cpu;
 
 	/* interrupt stacks must be in lowmem, we get that for free on ppc32
 	 * as the memblock is limited to lowmem by MEMBLOCK_REAL_LIMIT */
 	for_each_possible_cpu(i) {
-		critirq_ctx[i] = (struct thread_info *)
+		hw_cpu = get_hard_smp_processor_id(i);
+		critirq_ctx[hw_cpu] = (struct thread_info *)
 			__va(memblock_alloc(THREAD_SIZE, THREAD_SIZE));
 #ifdef CONFIG_BOOKE
-		dbgirq_ctx[i] = (struct thread_info *)
+		dbgirq_ctx[hw_cpu] = (struct thread_info *)
 			__va(memblock_alloc(THREAD_SIZE, THREAD_SIZE));
-		mcheckirq_ctx[i] = (struct thread_info *)
+		mcheckirq_ctx[hw_cpu] = (struct thread_info *)
 			__va(memblock_alloc(THREAD_SIZE, THREAD_SIZE));
 #endif
 	}
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index 1bee4b6..e72690e 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -95,7 +95,7 @@ int ucache_bsize;
 
 #ifdef CONFIG_SMP
 
-static int smt_enabled_cmdline;
+static char *smt_enabled_cmdline;
 
 /* Look for ibm,smt-enabled OF option */
 static void check_smt_enabled(void)
@@ -103,37 +103,46 @@ static void check_smt_enabled(void)
 	struct device_node *dn;
 	const char *smt_option;
 
-	/* Allow the command line to overrule the OF option */
-	if (smt_enabled_cmdline)
-		return;
-
-	dn = of_find_node_by_path("/options");
-
-	if (dn) {
-		smt_option = of_get_property(dn, "ibm,smt-enabled", NULL);
+	/* Default to enabling all threads */
+	smt_enabled_at_boot = threads_per_core;
 
-                if (smt_option) {
-			if (!strcmp(smt_option, "on"))
-				smt_enabled_at_boot = 1;
-			else if (!strcmp(smt_option, "off"))
-				smt_enabled_at_boot = 0;
-                }
-        }
+	/* Allow the command line to overrule the OF option */
+	if (smt_enabled_cmdline) {
+		if (!strcmp(smt_enabled_cmdline, "on"))
+			smt_enabled_at_boot = threads_per_core;
+		else if (!strcmp(smt_enabled_cmdline, "off"))
+			smt_enabled_at_boot = 0;
+		else {
+			long smt;
+			int rc;
+
+			rc = strict_strtol(smt_enabled_cmdline, 10, &smt);
+			if (!rc)
+				smt_enabled_at_boot =
+					min(threads_per_core, (int)smt);
+		}
+	} else {
+		dn = of_find_node_by_path("/options");
+		if (dn) {
+			smt_option = of_get_property(dn, "ibm,smt-enabled",
+						     NULL);
+
+			if (smt_option) {
+				if (!strcmp(smt_option, "on"))
+					smt_enabled_at_boot = threads_per_core;
+				else if (!strcmp(smt_option, "off"))
+					smt_enabled_at_boot = 0;
+			}
+
+			of_node_put(dn);
+		}
+	}
 }
 
 /* Look for smt-enabled= cmdline option */
 static int __init early_smt_enabled(char *p)
 {
-	smt_enabled_cmdline = 1;
-
-	if (!p)
-		return 0;
-
-	if (!strcmp(p, "on") || !strcmp(p, "1"))
-		smt_enabled_at_boot = 1;
-	else if (!strcmp(p, "off") || !strcmp(p, "0"))
-		smt_enabled_at_boot = 0;
-
+	smt_enabled_cmdline = p;
 	return 0;
 }
 early_param("smt-enabled", early_smt_enabled);
@@ -380,8 +389,8 @@ void __init setup_system(void)
 	 */
 	xmon_setup();
 
-	check_smt_enabled();
 	smp_setup_cpu_maps();
+	check_smt_enabled();
 
 #ifdef CONFIG_SMP
 	/* Release secondary cpus out of their spinloops at 0x60 now that
diff --git a/arch/powerpc/kernel/smp.c b/arch/powerpc/kernel/smp.c
index a61b3dd..0008bc5 100644
--- a/arch/powerpc/kernel/smp.c
+++ b/arch/powerpc/kernel/smp.c
@@ -427,11 +427,11 @@ int __cpuinit __cpu_up(unsigned int cpu)
 #endif
 
 	if (!cpu_callin_map[cpu]) {
-		printk("Processor %u is stuck.\n", cpu);
+		printk(KERN_ERR "Processor %u is stuck.\n", cpu);
 		return -ENOENT;
 	}
 
-	printk("Processor %u found.\n", cpu);
+	DBG("Processor %u found.\n", cpu);
 
 	if (smp_ops->give_timebase)
 		smp_ops->give_timebase();
diff --git a/arch/powerpc/kernel/sys_ppc32.c b/arch/powerpc/kernel/sys_ppc32.c
index 20fd701..b1b6043 100644
--- a/arch/powerpc/kernel/sys_ppc32.c
+++ b/arch/powerpc/kernel/sys_ppc32.c
@@ -616,3 +616,11 @@ asmlinkage long compat_sys_sync_file_range2(int fd, unsigned int flags,
 
 	return sys_sync_file_range(fd, offset, nbytes, flags);
 }
+
+asmlinkage long compat_sys_fanotify_mark(int fanotify_fd, unsigned int flags,
+					 unsigned mask_hi, unsigned mask_lo,
+					 int dfd, const char __user *pathname)
+{
+	u64 mask = ((u64)mask_hi << 32) | mask_lo;
+	return sys_fanotify_mark(fanotify_fd, flags, mask, dfd, pathname);
+}
diff --git a/arch/powerpc/kernel/vio.c b/arch/powerpc/kernel/vio.c
index 00b9436..fa3469d 100644
--- a/arch/powerpc/kernel/vio.c
+++ b/arch/powerpc/kernel/vio.c
@@ -1059,7 +1059,7 @@ static struct iommu_table *vio_build_iommu_table(struct vio_dev *dev)
 	if (!dma_window)
 		return NULL;
 
-	tbl = kmalloc(sizeof(*tbl), GFP_KERNEL);
+	tbl = kzalloc(sizeof(*tbl), GFP_KERNEL);
 	if (tbl == NULL)
 		return NULL;
 
@@ -1072,6 +1072,7 @@ static struct iommu_table *vio_build_iommu_table(struct vio_dev *dev)
 	tbl->it_offset = offset >> IOMMU_PAGE_SHIFT;
 	tbl->it_busno = 0;
 	tbl->it_type = TCE_VB;
+	tbl->it_blocksize = 16;
 
 	return iommu_init_table(tbl, -1);
 }
diff --git a/arch/powerpc/mm/init_64.c b/arch/powerpc/mm/init_64.c
index 71f1415..ace85fa 100644
--- a/arch/powerpc/mm/init_64.c
+++ b/arch/powerpc/mm/init_64.c
@@ -79,7 +79,9 @@
 #endif /* CONFIG_PPC_STD_MMU_64 */
 
 phys_addr_t memstart_addr = ~0;
+EXPORT_SYMBOL_GPL(memstart_addr);
 phys_addr_t kernstart_addr;
+EXPORT_SYMBOL_GPL(kernstart_addr);
 
 void free_initmem(void)
 {
diff --git a/arch/powerpc/mm/tlb_nohash_low.S b/arch/powerpc/mm/tlb_nohash_low.S
index cfa7682..b9d9fed 100644
--- a/arch/powerpc/mm/tlb_nohash_low.S
+++ b/arch/powerpc/mm/tlb_nohash_low.S
@@ -200,6 +200,7 @@ _GLOBAL(_tlbivax_bcast)
 	rlwimi	r5,r4,0,16,31
 	wrteei	0
 	mtspr	SPRN_MMUCR,r5
+	isync
 /*	tlbivax	0,r3 - use .long to avoid binutils deps */
 	.long 0x7c000624 | (r3 << 11)
 	isync
diff --git a/arch/powerpc/platforms/Kconfig b/arch/powerpc/platforms/Kconfig
index d1663db..81c9208 100644
--- a/arch/powerpc/platforms/Kconfig
+++ b/arch/powerpc/platforms/Kconfig
@@ -106,8 +106,7 @@ config MMIO_NVRAM
 
 config MPIC_U3_HT_IRQS
 	bool
-	depends on PPC_MAPLE
-	default y
+	default n
 
 config MPIC_BROKEN_REGREAD
 	bool
diff --git a/arch/powerpc/platforms/cell/iommu.c b/arch/powerpc/platforms/cell/iommu.c
index 58b13ce..26a0671 100644
--- a/arch/powerpc/platforms/cell/iommu.c
+++ b/arch/powerpc/platforms/cell/iommu.c
@@ -477,7 +477,7 @@ cell_iommu_setup_window(struct cbe_iommu *iommu, struct device_node *np,
 
 	ioid = cell_iommu_get_ioid(np);
 
-	window = kmalloc_node(sizeof(*window), GFP_KERNEL, iommu->nid);
+	window = kzalloc_node(sizeof(*window), GFP_KERNEL, iommu->nid);
 	BUG_ON(window == NULL);
 
 	window->offset = offset;
diff --git a/arch/powerpc/platforms/iseries/iommu.c b/arch/powerpc/platforms/iseries/iommu.c
index ce61cea..d8b7633 100644
--- a/arch/powerpc/platforms/iseries/iommu.c
+++ b/arch/powerpc/platforms/iseries/iommu.c
@@ -184,7 +184,7 @@ static void pci_dma_dev_setup_iseries(struct pci_dev *pdev)
 
 	BUG_ON(lsn == NULL);
 
-	tbl = kmalloc(sizeof(struct iommu_table), GFP_KERNEL);
+	tbl = kzalloc(sizeof(struct iommu_table), GFP_KERNEL);
 
 	iommu_table_getparms_iSeries(pdn->busno, *lsn, 0, tbl);
 
diff --git a/arch/powerpc/platforms/powermac/feature.c b/arch/powerpc/platforms/powermac/feature.c
index 39df6ab..df42399 100644
--- a/arch/powerpc/platforms/powermac/feature.c
+++ b/arch/powerpc/platforms/powermac/feature.c
@@ -2873,12 +2873,11 @@ set_initial_features(void)
 
 		/* Switch airport off */
 		for_each_node_by_name(np, "radio") {
-			if (np && np->parent == macio_chips[0].of_node) {
+			if (np->parent == macio_chips[0].of_node) {
 				macio_chips[0].flags |= MACIO_FLAG_AIRPORT_ON;
 				core99_airport_enable(np, 0, 0);
 			}
 		}
-		of_node_put(np);
 	}
 
 	/* On all machines that support sound PM, switch sound off */
diff --git a/arch/powerpc/platforms/powermac/pci.c b/arch/powerpc/platforms/powermac/pci.c
index ab2027c..3bc075c 100644
--- a/arch/powerpc/platforms/powermac/pci.c
+++ b/arch/powerpc/platforms/powermac/pci.c
@@ -1155,13 +1155,11 @@ void __init pmac_pcibios_after_init(void)
 			pmac_call_feature(PMAC_FTR_1394_CABLE_POWER, nd, 0, 0);
 		}
 	}
-	of_node_put(nd);
 	for_each_node_by_name(nd, "ethernet") {
 		if (nd->parent && of_device_is_compatible(nd, "gmac")
 		    && of_device_is_compatible(nd->parent, "uni-north"))
 			pmac_call_feature(PMAC_FTR_GMAC_ENABLE, nd, 0, 0);
 	}
-	of_node_put(nd);
 }
 
 void pmac_pci_fixup_cardbus(struct pci_dev* dev)
diff --git a/arch/powerpc/platforms/pseries/iommu.c b/arch/powerpc/platforms/pseries/iommu.c
index 395848e..a77bcae 100644
--- a/arch/powerpc/platforms/pseries/iommu.c
+++ b/arch/powerpc/platforms/pseries/iommu.c
@@ -403,7 +403,7 @@ static void pci_dma_bus_setup_pSeries(struct pci_bus *bus)
 	pci->phb->dma_window_size = 0x8000000ul;
 	pci->phb->dma_window_base_cur = 0x8000000ul;
 
-	tbl = kmalloc_node(sizeof(struct iommu_table), GFP_KERNEL,
+	tbl = kzalloc_node(sizeof(struct iommu_table), GFP_KERNEL,
 			   pci->phb->node);
 
 	iommu_table_setparms(pci->phb, dn, tbl);
@@ -448,7 +448,7 @@ static void pci_dma_bus_setup_pSeriesLP(struct pci_bus *bus)
 		 pdn->full_name, ppci->iommu_table);
 
 	if (!ppci->iommu_table) {
-		tbl = kmalloc_node(sizeof(struct iommu_table), GFP_KERNEL,
+		tbl = kzalloc_node(sizeof(struct iommu_table), GFP_KERNEL,
 				   ppci->phb->node);
 		iommu_table_setparms_lpar(ppci->phb, pdn, tbl, dma_window,
 			bus->number);
@@ -478,7 +478,7 @@ static void pci_dma_dev_setup_pSeries(struct pci_dev *dev)
 		struct pci_controller *phb = PCI_DN(dn)->phb;
 
 		pr_debug(" --> first child, no bridge. Allocating iommu table.\n");
-		tbl = kmalloc_node(sizeof(struct iommu_table), GFP_KERNEL,
+		tbl = kzalloc_node(sizeof(struct iommu_table), GFP_KERNEL,
 				   phb->node);
 		iommu_table_setparms(phb, dn, tbl);
 		PCI_DN(dn)->iommu_table = iommu_init_table(tbl, phb->node);
@@ -544,7 +544,7 @@ static void pci_dma_dev_setup_pSeriesLP(struct pci_dev *dev)
 
 	pci = PCI_DN(pdn);
 	if (!pci->iommu_table) {
-		tbl = kmalloc_node(sizeof(struct iommu_table), GFP_KERNEL,
+		tbl = kzalloc_node(sizeof(struct iommu_table), GFP_KERNEL,
 				   pci->phb->node);
 		iommu_table_setparms_lpar(pci->phb, pdn, tbl, dma_window,
 			pci->phb->bus->number);
diff --git a/arch/powerpc/platforms/pseries/smp.c b/arch/powerpc/platforms/pseries/smp.c
index 3b1bf61..0317cce 100644
--- a/arch/powerpc/platforms/pseries/smp.c
+++ b/arch/powerpc/platforms/pseries/smp.c
@@ -182,10 +182,13 @@ static int smp_pSeries_cpu_bootable(unsigned int nr)
 	/* Special case - we inhibit secondary thread startup
 	 * during boot if the user requests it.
 	 */
-	if (system_state < SYSTEM_RUNNING &&
-	    cpu_has_feature(CPU_FTR_SMT) &&
-	    !smt_enabled_at_boot && cpu_thread_in_core(nr) != 0)
-		return 0;
+	if (system_state < SYSTEM_RUNNING && cpu_has_feature(CPU_FTR_SMT)) {
+		if (!smt_enabled_at_boot && cpu_thread_in_core(nr) != 0)
+			return 0;
+		if (smt_enabled_at_boot
+		    && cpu_thread_in_core(nr) >= smt_enabled_at_boot)
+			return 0;
+	}
 
 	return 1;
 }
diff --git a/arch/powerpc/platforms/pseries/xics.c b/arch/powerpc/platforms/pseries/xics.c
index 5b22b07..93834b0 100644
--- a/arch/powerpc/platforms/pseries/xics.c
+++ b/arch/powerpc/platforms/pseries/xics.c
@@ -928,8 +928,10 @@ void xics_migrate_irqs_away(void)
 		if (xics_status[0] != hw_cpu)
 			goto unlock;
 
-		printk(KERN_WARNING "IRQ %u affinity broken off cpu %u\n",
-		       virq, cpu);
+		/* This is expected during cpu offline. */
+		if (cpu_online(cpu))
+			printk(KERN_WARNING "IRQ %u affinity broken off cpu %u\n",
+			       virq, cpu);
 
 		/* Reset affinity to all cpus */
 		cpumask_setall(irq_to_desc(virq)->affinity);
diff --git a/arch/s390/include/asm/hugetlb.h b/arch/s390/include/asm/hugetlb.h
index 670a1d1..bb8343d 100644
--- a/arch/s390/include/asm/hugetlb.h
+++ b/arch/s390/include/asm/hugetlb.h
@@ -97,6 +97,7 @@ static inline pte_t huge_ptep_get_and_clear(struct mm_struct *mm,
 {
 	pte_t pte = huge_ptep_get(ptep);
 
+	mm->context.flush_mm = 1;
 	pmd_clear((pmd_t *) ptep);
 	return pte;
 }
@@ -167,7 +168,8 @@ static inline void huge_ptep_invalidate(struct mm_struct *mm,
 ({									\
 	pte_t __pte = huge_ptep_get(__ptep);				\
 	if (pte_write(__pte)) {						\
-		if (atomic_read(&(__mm)->mm_users) > 1 ||		\
+		(__mm)->context.flush_mm = 1;				\
+		if (atomic_read(&(__mm)->context.attach_count) > 1 ||	\
 		    (__mm) != current->active_mm)			\
 			huge_ptep_invalidate(__mm, __addr, __ptep);	\
 		set_huge_pte_at(__mm, __addr, __ptep,			\
diff --git a/arch/s390/include/asm/mmu.h b/arch/s390/include/asm/mmu.h
index 99e3409..78522cde 100644
--- a/arch/s390/include/asm/mmu.h
+++ b/arch/s390/include/asm/mmu.h
@@ -2,6 +2,8 @@
 #define __MMU_H
 
 typedef struct {
+	atomic_t attach_count;
+	unsigned int flush_mm;
 	spinlock_t list_lock;
 	struct list_head crst_list;
 	struct list_head pgtable_list;
diff --git a/arch/s390/include/asm/mmu_context.h b/arch/s390/include/asm/mmu_context.h
index 976e273..a6f0e7c 100644
--- a/arch/s390/include/asm/mmu_context.h
+++ b/arch/s390/include/asm/mmu_context.h
@@ -11,11 +11,14 @@
 
 #include <asm/pgalloc.h>
 #include <asm/uaccess.h>
+#include <asm/tlbflush.h>
 #include <asm-generic/mm_hooks.h>
 
 static inline int init_new_context(struct task_struct *tsk,
 				   struct mm_struct *mm)
 {
+	atomic_set(&mm->context.attach_count, 0);
+	mm->context.flush_mm = 0;
 	mm->context.asce_bits = _ASCE_TABLE_LENGTH | _ASCE_USER_BITS;
 #ifdef CONFIG_64BIT
 	mm->context.asce_bits |= _ASCE_TYPE_REGION3;
@@ -76,6 +79,12 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
 {
 	cpumask_set_cpu(smp_processor_id(), mm_cpumask(next));
 	update_mm(next, tsk);
+	atomic_dec(&prev->context.attach_count);
+	WARN_ON(atomic_read(&prev->context.attach_count) < 0);
+	atomic_inc(&next->context.attach_count);
+	/* Check for TLBs not flushed yet */
+	if (next->context.flush_mm)
+		__tlb_flush_mm(next);
 }
 
 #define enter_lazy_tlb(mm,tsk)	do { } while (0)
diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h
index 89a504c..3157441 100644
--- a/arch/s390/include/asm/pgtable.h
+++ b/arch/s390/include/asm/pgtable.h
@@ -880,7 +880,8 @@ static inline void ptep_invalidate(struct mm_struct *mm,
 #define ptep_get_and_clear(__mm, __address, __ptep)			\
 ({									\
 	pte_t __pte = *(__ptep);					\
-	if (atomic_read(&(__mm)->mm_users) > 1 ||			\
+	(__mm)->context.flush_mm = 1;					\
+	if (atomic_read(&(__mm)->context.attach_count) > 1 ||		\
 	    (__mm) != current->active_mm)				\
 		ptep_invalidate(__mm, __address, __ptep);		\
 	else								\
@@ -923,7 +924,8 @@ static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm,
 ({									\
 	pte_t __pte = *(__ptep);					\
 	if (pte_write(__pte)) {						\
-		if (atomic_read(&(__mm)->mm_users) > 1 ||		\
+		(__mm)->context.flush_mm = 1;				\
+		if (atomic_read(&(__mm)->context.attach_count) > 1 ||	\
 		    (__mm) != current->active_mm)			\
 			ptep_invalidate(__mm, __addr, __ptep);		\
 		set_pte_at(__mm, __addr, __ptep, pte_wrprotect(__pte));	\
diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h
index 81150b0..fd1c00d 100644
--- a/arch/s390/include/asm/tlb.h
+++ b/arch/s390/include/asm/tlb.h
@@ -50,8 +50,7 @@ static inline struct mmu_gather *tlb_gather_mmu(struct mm_struct *mm,
 	struct mmu_gather *tlb = &get_cpu_var(mmu_gathers);
 
 	tlb->mm = mm;
-	tlb->fullmm = full_mm_flush || (num_online_cpus() == 1) ||
-		(atomic_read(&mm->mm_users) <= 1 && mm == current->active_mm);
+	tlb->fullmm = full_mm_flush;
 	tlb->nr_ptes = 0;
 	tlb->nr_pxds = TLB_NR_PTRS;
 	if (tlb->fullmm)
diff --git a/arch/s390/include/asm/tlbflush.h b/arch/s390/include/asm/tlbflush.h
index 304cffa..29d5d6d 100644
--- a/arch/s390/include/asm/tlbflush.h
+++ b/arch/s390/include/asm/tlbflush.h
@@ -94,8 +94,12 @@ static inline void __tlb_flush_mm(struct mm_struct * mm)
 
 static inline void __tlb_flush_mm_cond(struct mm_struct * mm)
 {
-	if (atomic_read(&mm->mm_users) <= 1 && mm == current->active_mm)
+	spin_lock(&mm->page_table_lock);
+	if (mm->context.flush_mm) {
 		__tlb_flush_mm(mm);
+		mm->context.flush_mm = 0;
+	}
+	spin_unlock(&mm->page_table_lock);
 }
 
 /*
diff --git a/arch/s390/kernel/entry.h b/arch/s390/kernel/entry.h
index 403fb43..ff579b6 100644
--- a/arch/s390/kernel/entry.h
+++ b/arch/s390/kernel/entry.h
@@ -42,8 +42,8 @@ long sys_clone(unsigned long newsp, unsigned long clone_flags,
 	       int __user *parent_tidptr, int __user *child_tidptr);
 long sys_vfork(void);
 void execve_tail(void);
-long sys_execve(const char __user *name, char __user * __user *argv,
-		char __user * __user *envp);
+long sys_execve(const char __user *name, const char __user *const __user *argv,
+		const char __user *const __user *envp);
 long sys_sigsuspend(int history0, int history1, old_sigset_t mask);
 long sys_sigaction(int sig, const struct old_sigaction __user *act,
 		   struct old_sigaction __user *oact);
diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c
index 7eafaf2..d3a2d1c 100644
--- a/arch/s390/kernel/process.c
+++ b/arch/s390/kernel/process.c
@@ -267,8 +267,9 @@ asmlinkage void execve_tail(void)
 /*
  * sys_execve() executes a new program.
  */
-SYSCALL_DEFINE3(execve, const char __user *, name, char __user * __user *, argv,
-		char __user * __user *, envp)
+SYSCALL_DEFINE3(execve, const char __user *, name,
+		const char __user *const __user *, argv,
+		const char __user *const __user *, envp)
 {
 	struct pt_regs *regs = task_pt_regs(current);
 	char *filename;
diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index 541053e..8127ebd 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -583,6 +583,7 @@ int __cpuinit __cpu_up(unsigned int cpu)
 	sf->gprs[9] = (unsigned long) sf;
 	cpu_lowcore->save_area[15] = (unsigned long) sf;
 	__ctl_store(cpu_lowcore->cregs_save_area, 0, 15);
+	atomic_inc(&init_mm.context.attach_count);
 	asm volatile(
 		"	stam	0,15,0(%0)"
 		: : "a" (&cpu_lowcore->access_regs_save_area) : "memory");
@@ -659,6 +660,7 @@ void __cpu_die(unsigned int cpu)
 	while (sigp_p(0, cpu, sigp_set_prefix) == sigp_busy)
 		udelay(10);
 	smp_free_lowcore(cpu);
+	atomic_dec(&init_mm.context.attach_count);
 	pr_info("Processor %d stopped\n", cpu);
 }
 
diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c
index acc91c7..30eb6d02 100644
--- a/arch/s390/mm/init.c
+++ b/arch/s390/mm/init.c
@@ -74,6 +74,8 @@ void __init paging_init(void)
 	__ctl_load(S390_lowcore.kernel_asce, 13, 13);
 	__raw_local_irq_ssm(ssm_mask);
 
+	atomic_set(&init_mm.context.attach_count, 1);
+
 	sparse_memory_present_with_active_regions(MAX_NUMNODES);
 	sparse_init();
 	memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
diff --git a/arch/score/kernel/sys_score.c b/arch/score/kernel/sys_score.c
index 651096f..e478bf9 100644
--- a/arch/score/kernel/sys_score.c
+++ b/arch/score/kernel/sys_score.c
@@ -99,8 +99,10 @@ score_execve(struct pt_regs *regs)
 	if (IS_ERR(filename))
 		return error;
 
-	error = do_execve(filename, (char __user *__user*)regs->regs[5],
-			  (char __user *__user *) regs->regs[6], regs);
+	error = do_execve(filename,
+			  (const char __user *const __user *)regs->regs[5],
+			  (const char __user *const __user *)regs->regs[6],
+			  regs);
 
 	putname(filename);
 	return error;
@@ -110,7 +112,9 @@ score_execve(struct pt_regs *regs)
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	register unsigned long __r4 asm("r4") = (unsigned long) filename;
 	register unsigned long __r5 asm("r5") = (unsigned long) argv;
diff --git a/arch/sh/kernel/process_32.c b/arch/sh/kernel/process_32.c
index 0529819..762a139 100644
--- a/arch/sh/kernel/process_32.c
+++ b/arch/sh/kernel/process_32.c
@@ -296,9 +296,10 @@ asmlinkage int sys_vfork(unsigned long r4, unsigned long r5,
 /*
  * sys_execve() executes a new program.
  */
-asmlinkage int sys_execve(char __user *ufilename, char __user * __user *uargv,
-			  char __user * __user *uenvp, unsigned long r7,
-			  struct pt_regs __regs)
+asmlinkage int sys_execve(const char __user *ufilename,
+			  const char __user *const __user *uargv,
+			  const char __user *const __user *uenvp,
+			  unsigned long r7, struct pt_regs __regs)
 {
 	struct pt_regs *regs = RELOC_HIDE(&__regs, 0);
 	int error;
diff --git a/arch/sh/kernel/process_64.c b/arch/sh/kernel/process_64.c
index 68d128d..210c1ca 100644
--- a/arch/sh/kernel/process_64.c
+++ b/arch/sh/kernel/process_64.c
@@ -497,8 +497,8 @@ asmlinkage int sys_execve(const char *ufilename, char **uargv,
 		goto out;
 
 	error = do_execve(filename,
-			  (char __user * __user *)uargv,
-			  (char __user * __user *)uenvp,
+			  (const char __user *const __user *)uargv,
+			  (const char __user *const __user *)uenvp,
 			  pregs);
 	putname(filename);
 out:
diff --git a/arch/sh/kernel/sys_sh32.c b/arch/sh/kernel/sys_sh32.c
index eb68bfd..f56b6fe5 100644
--- a/arch/sh/kernel/sys_sh32.c
+++ b/arch/sh/kernel/sys_sh32.c
@@ -71,7 +71,9 @@ asmlinkage int sys_fadvise64_64_wrapper(int fd, u32 offset0, u32 offset1,
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	register long __sc0 __asm__ ("r3") = __NR_execve;
 	register long __sc4 __asm__ ("r4") = (long) filename;
diff --git a/arch/sh/kernel/sys_sh64.c b/arch/sh/kernel/sys_sh64.c
index 2872357..c5a38c4 100644
--- a/arch/sh/kernel/sys_sh64.c
+++ b/arch/sh/kernel/sys_sh64.c
@@ -33,7 +33,9 @@
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	register unsigned long __sc0 __asm__ ("r9") = ((0x13 << 16) | __NR_execve);
 	register unsigned long __sc2 __asm__ ("r2") = (unsigned long) filename;
diff --git a/arch/sparc/include/asm/atomic_64.h b/arch/sparc/include/asm/atomic_64.h
index 2050ca0..bdb2ff8 100644
--- a/arch/sparc/include/asm/atomic_64.h
+++ b/arch/sparc/include/asm/atomic_64.h
@@ -20,14 +20,14 @@
 #define atomic64_set(v, i)	(((v)->counter) = i)
 
 extern void atomic_add(int, atomic_t *);
-extern void atomic64_add(int, atomic64_t *);
+extern void atomic64_add(long, atomic64_t *);
 extern void atomic_sub(int, atomic_t *);
-extern void atomic64_sub(int, atomic64_t *);
+extern void atomic64_sub(long, atomic64_t *);
 
 extern int atomic_add_ret(int, atomic_t *);
-extern int atomic64_add_ret(int, atomic64_t *);
+extern long atomic64_add_ret(long, atomic64_t *);
 extern int atomic_sub_ret(int, atomic_t *);
-extern int atomic64_sub_ret(int, atomic64_t *);
+extern long atomic64_sub_ret(long, atomic64_t *);
 
 #define atomic_dec_return(v) atomic_sub_ret(1, v)
 #define atomic64_dec_return(v) atomic64_sub_ret(1, v)
@@ -91,7 +91,7 @@ static inline int atomic_add_unless(atomic_t *v, int a, int u)
 	((__typeof__((v)->counter))cmpxchg(&((v)->counter), (o), (n)))
 #define atomic64_xchg(v, new) (xchg(&((v)->counter), new))
 
-static inline int atomic64_add_unless(atomic64_t *v, long a, long u)
+static inline long atomic64_add_unless(atomic64_t *v, long a, long u)
 {
 	long c, old;
 	c = atomic64_read(v);
diff --git a/arch/sparc/include/asm/backoff.h b/arch/sparc/include/asm/backoff.h
index fa1fdf6..db3af0d 100644
--- a/arch/sparc/include/asm/backoff.h
+++ b/arch/sparc/include/asm/backoff.h
@@ -8,6 +8,9 @@
 #define BACKOFF_SETUP(reg)	\
 	mov	1, reg
 
+#define BACKOFF_LABEL(spin_label, continue_label) \
+	spin_label
+
 #define BACKOFF_SPIN(reg, tmp, label)	\
 	mov	reg, tmp; \
 88:	brnz,pt	tmp, 88b; \
@@ -22,9 +25,11 @@
 #else
 
 #define BACKOFF_SETUP(reg)
-#define BACKOFF_SPIN(reg, tmp, label) \
-	ba,pt	%xcc, label; \
-	 nop;
+
+#define BACKOFF_LABEL(spin_label, continue_label) \
+	continue_label
+
+#define BACKOFF_SPIN(reg, tmp, label)
 
 #endif
 
diff --git a/arch/sparc/include/asm/fb.h b/arch/sparc/include/asm/fb.h
index e834880..2173432 100644
--- a/arch/sparc/include/asm/fb.h
+++ b/arch/sparc/include/asm/fb.h
@@ -1,5 +1,6 @@
 #ifndef _SPARC_FB_H_
 #define _SPARC_FB_H_
+#include <linux/console.h>
 #include <linux/fb.h>
 #include <linux/fs.h>
 #include <asm/page.h>
@@ -18,6 +19,9 @@ static inline int fb_is_primary_device(struct fb_info *info)
 	struct device *dev = info->device;
 	struct device_node *node;
 
+	if (console_set_on_cmdline)
+		return 0;
+
 	node = dev->of_node;
 	if (node &&
 	    node == of_console_device)
diff --git a/arch/sparc/include/asm/oplib_64.h b/arch/sparc/include/asm/oplib_64.h
index a5db031..3e0b2d6 100644
--- a/arch/sparc/include/asm/oplib_64.h
+++ b/arch/sparc/include/asm/oplib_64.h
@@ -185,9 +185,8 @@ extern int prom_getunumber(int syndrome_code,
 			   char *buf, int buflen);
 
 /* Retain physical memory to the caller across soft resets. */
-extern unsigned long prom_retain(const char *name,
-				 unsigned long pa_low, unsigned long pa_high,
-				 long size, long align);
+extern int prom_retain(const char *name, unsigned long size,
+		       unsigned long align, unsigned long *paddr);
 
 /* Load explicit I/D TLB entries into the calling processor. */
 extern long prom_itlb_load(unsigned long index,
@@ -287,26 +286,6 @@ extern void prom_sun4v_guest_soft_state(void);
 extern int prom_ihandle2path(int handle, char *buffer, int bufsize);
 
 /* Client interface level routines. */
-extern long p1275_cmd(const char *, long, ...);
-
-#if 0
-#define P1275_SIZE(x) ((((long)((x) / 32)) << 32) | (x))
-#else
-#define P1275_SIZE(x) x
-#endif
-
-/* We support at most 16 input and 1 output argument */
-#define P1275_ARG_NUMBER		0
-#define P1275_ARG_IN_STRING		1
-#define P1275_ARG_OUT_BUF		2
-#define P1275_ARG_OUT_32B		3
-#define P1275_ARG_IN_FUNCTION		4
-#define P1275_ARG_IN_BUF		5
-#define P1275_ARG_IN_64B		6
-
-#define P1275_IN(x) ((x) & 0xf)
-#define P1275_OUT(x) (((x) << 4) & 0xf0)
-#define P1275_INOUT(i,o) (P1275_IN(i)|P1275_OUT(o))
-#define P1275_ARG(n,x) ((x) << ((n)*3 + 8))
+extern void p1275_cmd_direct(unsigned long *);
 
 #endif /* !(__SPARC64_OPLIB_H) */
diff --git a/arch/sparc/include/asm/rwsem-const.h b/arch/sparc/include/asm/rwsem-const.h
deleted file mode 100644
index a303c9d..0000000
--- a/arch/sparc/include/asm/rwsem-const.h
+++ /dev/null
@@ -1,12 +0,0 @@
-/* rwsem-const.h: RW semaphore counter constants.  */
-#ifndef _SPARC64_RWSEM_CONST_H
-#define _SPARC64_RWSEM_CONST_H
-
-#define RWSEM_UNLOCKED_VALUE		0x00000000
-#define RWSEM_ACTIVE_BIAS		0x00000001
-#define RWSEM_ACTIVE_MASK		0x0000ffff
-#define RWSEM_WAITING_BIAS		0xffff0000
-#define RWSEM_ACTIVE_READ_BIAS		RWSEM_ACTIVE_BIAS
-#define RWSEM_ACTIVE_WRITE_BIAS		(RWSEM_WAITING_BIAS + RWSEM_ACTIVE_BIAS)
-
-#endif /* _SPARC64_RWSEM_CONST_H */
diff --git a/arch/sparc/include/asm/rwsem.h b/arch/sparc/include/asm/rwsem.h
index 6e56210..a2b4302 100644
--- a/arch/sparc/include/asm/rwsem.h
+++ b/arch/sparc/include/asm/rwsem.h
@@ -15,16 +15,21 @@
 
 #include <linux/list.h>
 #include <linux/spinlock.h>
-#include <asm/rwsem-const.h>
 
 struct rwsem_waiter;
 
 struct rw_semaphore {
-	signed int count;
-	spinlock_t		wait_lock;
-	struct list_head	wait_list;
+	signed long			count;
+#define RWSEM_UNLOCKED_VALUE		0x00000000L
+#define RWSEM_ACTIVE_BIAS		0x00000001L
+#define RWSEM_ACTIVE_MASK		0xffffffffL
+#define RWSEM_WAITING_BIAS		(-RWSEM_ACTIVE_MASK-1)
+#define RWSEM_ACTIVE_READ_BIAS		RWSEM_ACTIVE_BIAS
+#define RWSEM_ACTIVE_WRITE_BIAS		(RWSEM_WAITING_BIAS + RWSEM_ACTIVE_BIAS)
+	spinlock_t			wait_lock;
+	struct list_head		wait_list;
 #ifdef CONFIG_DEBUG_LOCK_ALLOC
-	struct lockdep_map	dep_map;
+	struct lockdep_map		dep_map;
 #endif
 };
 
@@ -41,6 +46,11 @@ struct rw_semaphore {
 #define DECLARE_RWSEM(name) \
 	struct rw_semaphore name = __RWSEM_INITIALIZER(name)
 
+extern struct rw_semaphore *rwsem_down_read_failed(struct rw_semaphore *sem);
+extern struct rw_semaphore *rwsem_down_write_failed(struct rw_semaphore *sem);
+extern struct rw_semaphore *rwsem_wake(struct rw_semaphore *sem);
+extern struct rw_semaphore *rwsem_downgrade_wake(struct rw_semaphore *sem);
+
 extern void __init_rwsem(struct rw_semaphore *sem, const char *name,
 			 struct lock_class_key *key);
 
@@ -51,27 +61,103 @@ do {								\
 	__init_rwsem((sem), #sem, &__key);			\
 } while (0)
 
-extern void __down_read(struct rw_semaphore *sem);
-extern int __down_read_trylock(struct rw_semaphore *sem);
-extern void __down_write(struct rw_semaphore *sem);
-extern int __down_write_trylock(struct rw_semaphore *sem);
-extern void __up_read(struct rw_semaphore *sem);
-extern void __up_write(struct rw_semaphore *sem);
-extern void __downgrade_write(struct rw_semaphore *sem);
+/*
+ * lock for reading
+ */
+static inline void __down_read(struct rw_semaphore *sem)
+{
+	if (unlikely(atomic64_inc_return((atomic64_t *)(&sem->count)) <= 0L))
+		rwsem_down_read_failed(sem);
+}
+
+static inline int __down_read_trylock(struct rw_semaphore *sem)
+{
+	long tmp;
+
+	while ((tmp = sem->count) >= 0L) {
+		if (tmp == cmpxchg(&sem->count, tmp,
+				   tmp + RWSEM_ACTIVE_READ_BIAS)) {
+			return 1;
+		}
+	}
+	return 0;
+}
 
+/*
+ * lock for writing
+ */
 static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
 {
-	__down_write(sem);
+	long tmp;
+
+	tmp = atomic64_add_return(RWSEM_ACTIVE_WRITE_BIAS,
+				  (atomic64_t *)(&sem->count));
+	if (unlikely(tmp != RWSEM_ACTIVE_WRITE_BIAS))
+		rwsem_down_write_failed(sem);
 }
 
-static inline int rwsem_atomic_update(int delta, struct rw_semaphore *sem)
+static inline void __down_write(struct rw_semaphore *sem)
 {
-	return atomic_add_return(delta, (atomic_t *)(&sem->count));
+	__down_write_nested(sem, 0);
+}
+
+static inline int __down_write_trylock(struct rw_semaphore *sem)
+{
+	long tmp;
+
+	tmp = cmpxchg(&sem->count, RWSEM_UNLOCKED_VALUE,
+		      RWSEM_ACTIVE_WRITE_BIAS);
+	return tmp == RWSEM_UNLOCKED_VALUE;
 }
 
-static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem)
+/*
+ * unlock after reading
+ */
+static inline void __up_read(struct rw_semaphore *sem)
+{
+	long tmp;
+
+	tmp = atomic64_dec_return((atomic64_t *)(&sem->count));
+	if (unlikely(tmp < -1L && (tmp & RWSEM_ACTIVE_MASK) == 0L))
+		rwsem_wake(sem);
+}
+
+/*
+ * unlock after writing
+ */
+static inline void __up_write(struct rw_semaphore *sem)
+{
+	if (unlikely(atomic64_sub_return(RWSEM_ACTIVE_WRITE_BIAS,
+					 (atomic64_t *)(&sem->count)) < 0L))
+		rwsem_wake(sem);
+}
+
+/*
+ * implement atomic add functionality
+ */
+static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
+{
+	atomic64_add(delta, (atomic64_t *)(&sem->count));
+}
+
+/*
+ * downgrade write lock to read lock
+ */
+static inline void __downgrade_write(struct rw_semaphore *sem)
+{
+	long tmp;
+
+	tmp = atomic64_add_return(-RWSEM_WAITING_BIAS, (atomic64_t *)(&sem->count));
+	if (tmp < 0L)
+		rwsem_downgrade_wake(sem);
+}
+
+/*
+ * implement exchange and add functionality
+ */
+static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem)
 {
-	atomic_add(delta, (atomic_t *)(&sem->count));
+	return atomic64_add_return(delta, (atomic64_t *)(&sem->count));
 }
 
 static inline int rwsem_is_locked(struct rw_semaphore *sem)
diff --git a/arch/sparc/include/asm/system_64.h b/arch/sparc/include/asm/system_64.h
index d24cfe1..e3b65d8 100644
--- a/arch/sparc/include/asm/system_64.h
+++ b/arch/sparc/include/asm/system_64.h
@@ -106,6 +106,7 @@ do {	__asm__ __volatile__("ba,pt	%%xcc, 1f\n\t" \
  */
 #define write_pic(__p)  					\
 	__asm__ __volatile__("ba,pt	%%xcc, 99f\n\t"		\
+			     " nop\n\t"				\
 			     ".align	64\n"			\
 			  "99:wr	%0, 0x0, %%pic\n\t"	\
 			     "rd	%%pic, %%g0" : : "r" (__p))
diff --git a/arch/sparc/include/asm/unistd.h b/arch/sparc/include/asm/unistd.h
index d0b3b01..03eb5a8 100644
--- a/arch/sparc/include/asm/unistd.h
+++ b/arch/sparc/include/asm/unistd.h
@@ -397,8 +397,11 @@
 #define __NR_rt_tgsigqueueinfo	326
 #define __NR_perf_event_open	327
 #define __NR_recvmmsg		328
+#define __NR_fanotify_init	329
+#define __NR_fanotify_mark	330
+#define __NR_prlimit64		331
 
-#define NR_syscalls		329
+#define NR_syscalls		332
 
 #ifdef __32bit_syscall_numbers__
 /* Sparc 32-bit only has the "setresuid32", "getresuid32" variants,
diff --git a/arch/sparc/kernel/process_32.c b/arch/sparc/kernel/process_32.c
index 40e29fc..1752929 100644
--- a/arch/sparc/kernel/process_32.c
+++ b/arch/sparc/kernel/process_32.c
@@ -633,8 +633,10 @@ asmlinkage int sparc_execve(struct pt_regs *regs)
 	if(IS_ERR(filename))
 		goto out;
 	error = do_execve(filename,
-			  (char __user * __user *)regs->u_regs[base + UREG_I1],
-			  (char __user * __user *)regs->u_regs[base + UREG_I2],
+			  (const char __user *const  __user *)
+			  regs->u_regs[base + UREG_I1],
+			  (const char __user *const  __user *)
+			  regs->u_regs[base + UREG_I2],
 			  regs);
 	putname(filename);
 out:
diff --git a/arch/sparc/kernel/process_64.c b/arch/sparc/kernel/process_64.c
index dbe81a3..485f547 100644
--- a/arch/sparc/kernel/process_64.c
+++ b/arch/sparc/kernel/process_64.c
@@ -739,9 +739,9 @@ asmlinkage int sparc_execve(struct pt_regs *regs)
 	if (IS_ERR(filename))
 		goto out;
 	error = do_execve(filename,
-			  (char __user * __user *)
+			  (const char __user *const __user *)
 			  regs->u_regs[base + UREG_I1],
-			  (char __user * __user *)
+			  (const char __user *const __user *)
 			  regs->u_regs[base + UREG_I2], regs);
 	putname(filename);
 	if (!error) {
diff --git a/arch/sparc/kernel/sys32.S b/arch/sparc/kernel/sys32.S
index 46a76ba..44e5faf 100644
--- a/arch/sparc/kernel/sys32.S
+++ b/arch/sparc/kernel/sys32.S
@@ -330,6 +330,15 @@ do_sys_accept4: /* sys_accept4(int, struct sockaddr *, int *, int) */
 	nop
 	nop
 
+	.globl		sys32_fanotify_mark
+sys32_fanotify_mark:
+	sethi		%hi(sys_fanotify_mark), %g1
+	sllx		%o2, 32, %o2
+	or		%o2, %o3, %o2
+	mov		%o4, %o3
+	jmpl		%g1 + %lo(sys_fanotify_mark), %g0
+	 mov		%o5, %o4
+
 	.section	__ex_table,"a"
 	.align		4
 	.word		1b, __retl_efault, 2b, __retl_efault
diff --git a/arch/sparc/kernel/sys_sparc_32.c b/arch/sparc/kernel/sys_sparc_32.c
index ee995b7..5079413 100644
--- a/arch/sparc/kernel/sys_sparc_32.c
+++ b/arch/sparc/kernel/sys_sparc_32.c
@@ -282,7 +282,9 @@ out:
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	long __res;
 	register long __g1 __asm__ ("g1") = __NR_execve;
diff --git a/arch/sparc/kernel/sys_sparc_64.c b/arch/sparc/kernel/sys_sparc_64.c
index 3d435c4..f836f4e 100644
--- a/arch/sparc/kernel/sys_sparc_64.c
+++ b/arch/sparc/kernel/sys_sparc_64.c
@@ -758,7 +758,9 @@ SYSCALL_DEFINE5(rt_sigaction, int, sig, const struct sigaction __user *, act,
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	long __res;
 	register long __g1 __asm__ ("g1") = __NR_execve;
diff --git a/arch/sparc/kernel/systbls_32.S b/arch/sparc/kernel/systbls_32.S
index 801fc8e..ec396e1 100644
--- a/arch/sparc/kernel/systbls_32.S
+++ b/arch/sparc/kernel/systbls_32.S
@@ -82,5 +82,6 @@ sys_call_table:
 /*310*/	.long sys_utimensat, sys_signalfd, sys_timerfd_create, sys_eventfd, sys_fallocate
 /*315*/	.long sys_timerfd_settime, sys_timerfd_gettime, sys_signalfd4, sys_eventfd2, sys_epoll_create1
 /*320*/	.long sys_dup3, sys_pipe2, sys_inotify_init1, sys_accept4, sys_preadv
-/*325*/	.long sys_pwritev, sys_rt_tgsigqueueinfo, sys_perf_event_open, sys_recvmmsg
+/*325*/	.long sys_pwritev, sys_rt_tgsigqueueinfo, sys_perf_event_open, sys_recvmmsg, sys_fanotify_init
+/*330*/	.long sys_fanotify_mark, sys_prlimit64
 
diff --git a/arch/sparc/kernel/systbls_64.S b/arch/sparc/kernel/systbls_64.S
index 9db058d..8cfcaa5 100644
--- a/arch/sparc/kernel/systbls_64.S
+++ b/arch/sparc/kernel/systbls_64.S
@@ -83,7 +83,8 @@ sys_call_table32:
 /*310*/	.word compat_sys_utimensat, compat_sys_signalfd, sys_timerfd_create, sys_eventfd, compat_sys_fallocate
 	.word compat_sys_timerfd_settime, compat_sys_timerfd_gettime, compat_sys_signalfd4, sys_eventfd2, sys_epoll_create1
 /*320*/	.word sys_dup3, sys_pipe2, sys_inotify_init1, sys_accept4, compat_sys_preadv
-	.word compat_sys_pwritev, compat_sys_rt_tgsigqueueinfo, sys_perf_event_open, compat_sys_recvmmsg
+	.word compat_sys_pwritev, compat_sys_rt_tgsigqueueinfo, sys_perf_event_open, compat_sys_recvmmsg, sys_fanotify_init
+/*330*/	.word sys32_fanotify_mark, sys_prlimit64
 
 #endif /* CONFIG_COMPAT */
 
@@ -158,4 +159,5 @@ sys_call_table:
 /*310*/	.word sys_utimensat, sys_signalfd, sys_timerfd_create, sys_eventfd, sys_fallocate
 	.word sys_timerfd_settime, sys_timerfd_gettime, sys_signalfd4, sys_eventfd2, sys_epoll_create1
 /*320*/	.word sys_dup3, sys_pipe2, sys_inotify_init1, sys_accept4, sys_preadv
-	.word sys_pwritev, sys_rt_tgsigqueueinfo, sys_perf_event_open, sys_recvmmsg
+	.word sys_pwritev, sys_rt_tgsigqueueinfo, sys_perf_event_open, sys_recvmmsg, sys_fanotify_init
+/*330*/	.word sys_fanotify_mark, sys_prlimit64
diff --git a/arch/sparc/lib/Makefile b/arch/sparc/lib/Makefile
index c4b5e03..846d1c4 100644
--- a/arch/sparc/lib/Makefile
+++ b/arch/sparc/lib/Makefile
@@ -15,7 +15,7 @@ lib-$(CONFIG_SPARC32) += divdi3.o udivdi3.o
 lib-$(CONFIG_SPARC32) += copy_user.o locks.o
 lib-y                 += atomic_$(BITS).o
 lib-$(CONFIG_SPARC32) += lshrdi3.o ashldi3.o
-lib-y                 += rwsem_$(BITS).o
+lib-$(CONFIG_SPARC32) += rwsem_32.o
 lib-$(CONFIG_SPARC32) += muldi3.o bitext.o cmpdi2.o
 
 lib-$(CONFIG_SPARC64) += copy_page.o clear_page.o bzero.o
diff --git a/arch/sparc/lib/atomic_64.S b/arch/sparc/lib/atomic_64.S
index 0268210..59186e0 100644
--- a/arch/sparc/lib/atomic_64.S
+++ b/arch/sparc/lib/atomic_64.S
@@ -21,7 +21,7 @@ atomic_add: /* %o0 = increment, %o1 = atomic_ptr */
 	add	%g1, %o0, %g7
 	cas	[%o1], %g1, %g7
 	cmp	%g1, %g7
-	bne,pn	%icc, 2f
+	bne,pn	%icc, BACKOFF_LABEL(2f, 1b)
 	 nop
 	retl
 	 nop
@@ -36,7 +36,7 @@ atomic_sub: /* %o0 = decrement, %o1 = atomic_ptr */
 	sub	%g1, %o0, %g7
 	cas	[%o1], %g1, %g7
 	cmp	%g1, %g7
-	bne,pn	%icc, 2f
+	bne,pn	%icc, BACKOFF_LABEL(2f, 1b)
 	 nop
 	retl
 	 nop
@@ -51,11 +51,10 @@ atomic_add_ret: /* %o0 = increment, %o1 = atomic_ptr */
 	add	%g1, %o0, %g7
 	cas	[%o1], %g1, %g7
 	cmp	%g1, %g7
-	bne,pn	%icc, 2f
-	 add	%g7, %o0, %g7
-	sra	%g7, 0, %o0
+	bne,pn	%icc, BACKOFF_LABEL(2f, 1b)
+	 add	%g1, %o0, %g1
 	retl
-	 nop
+	 sra	%g1, 0, %o0
 2:	BACKOFF_SPIN(%o2, %o3, 1b)
 	.size	atomic_add_ret, .-atomic_add_ret
 
@@ -67,11 +66,10 @@ atomic_sub_ret: /* %o0 = decrement, %o1 = atomic_ptr */
 	sub	%g1, %o0, %g7
 	cas	[%o1], %g1, %g7
 	cmp	%g1, %g7
-	bne,pn	%icc, 2f
-	 sub	%g7, %o0, %g7
-	sra	%g7, 0, %o0
+	bne,pn	%icc, BACKOFF_LABEL(2f, 1b)
+	 sub	%g1, %o0, %g1
 	retl
-	 nop
+	 sra	%g1, 0, %o0
 2:	BACKOFF_SPIN(%o2, %o3, 1b)
 	.size	atomic_sub_ret, .-atomic_sub_ret
 
@@ -83,7 +81,7 @@ atomic64_add: /* %o0 = increment, %o1 = atomic_ptr */
 	add	%g1, %o0, %g7
 	casx	[%o1], %g1, %g7
 	cmp	%g1, %g7
-	bne,pn	%xcc, 2f
+	bne,pn	%xcc, BACKOFF_LABEL(2f, 1b)
 	 nop
 	retl
 	 nop
@@ -98,7 +96,7 @@ atomic64_sub: /* %o0 = decrement, %o1 = atomic_ptr */
 	sub	%g1, %o0, %g7
 	casx	[%o1], %g1, %g7
 	cmp	%g1, %g7
-	bne,pn	%xcc, 2f
+	bne,pn	%xcc, BACKOFF_LABEL(2f, 1b)
 	 nop
 	retl
 	 nop
@@ -113,11 +111,10 @@ atomic64_add_ret: /* %o0 = increment, %o1 = atomic_ptr */
 	add	%g1, %o0, %g7
 	casx	[%o1], %g1, %g7
 	cmp	%g1, %g7
-	bne,pn	%xcc, 2f
-	 add	%g7, %o0, %g7
-	mov	%g7, %o0
-	retl
+	bne,pn	%xcc, BACKOFF_LABEL(2f, 1b)
 	 nop
+	retl
+	 add	%g1, %o0, %o0
 2:	BACKOFF_SPIN(%o2, %o3, 1b)
 	.size	atomic64_add_ret, .-atomic64_add_ret
 
@@ -129,10 +126,9 @@ atomic64_sub_ret: /* %o0 = decrement, %o1 = atomic_ptr */
 	sub	%g1, %o0, %g7
 	casx	[%o1], %g1, %g7
 	cmp	%g1, %g7
-	bne,pn	%xcc, 2f
-	 sub	%g7, %o0, %g7
-	mov	%g7, %o0
-	retl
+	bne,pn	%xcc, BACKOFF_LABEL(2f, 1b)
 	 nop
+	retl
+	 sub	%g1, %o0, %o0
 2:	BACKOFF_SPIN(%o2, %o3, 1b)
 	.size	atomic64_sub_ret, .-atomic64_sub_ret
diff --git a/arch/sparc/lib/bitops.S b/arch/sparc/lib/bitops.S
index 2b7228c..3dc61d5 100644
--- a/arch/sparc/lib/bitops.S
+++ b/arch/sparc/lib/bitops.S
@@ -22,7 +22,7 @@ test_and_set_bit:	/* %o0=nr, %o1=addr */
 	or	%g7, %o2, %g1
 	casx	[%o1], %g7, %g1
 	cmp	%g7, %g1
-	bne,pn	%xcc, 2f
+	bne,pn	%xcc, BACKOFF_LABEL(2f, 1b)
 	 and	%g7, %o2, %g2
 	clr	%o0
 	movrne	%g2, 1, %o0
@@ -45,7 +45,7 @@ test_and_clear_bit:	/* %o0=nr, %o1=addr */
 	andn	%g7, %o2, %g1
 	casx	[%o1], %g7, %g1
 	cmp	%g7, %g1
-	bne,pn	%xcc, 2f
+	bne,pn	%xcc, BACKOFF_LABEL(2f, 1b)
 	 and	%g7, %o2, %g2
 	clr	%o0
 	movrne	%g2, 1, %o0
@@ -68,7 +68,7 @@ test_and_change_bit:	/* %o0=nr, %o1=addr */
 	xor	%g7, %o2, %g1
 	casx	[%o1], %g7, %g1
 	cmp	%g7, %g1
-	bne,pn	%xcc, 2f
+	bne,pn	%xcc, BACKOFF_LABEL(2f, 1b)
 	 and	%g7, %o2, %g2
 	clr	%o0
 	movrne	%g2, 1, %o0
@@ -91,7 +91,7 @@ set_bit:		/* %o0=nr, %o1=addr */
 	or	%g7, %o2, %g1
 	casx	[%o1], %g7, %g1
 	cmp	%g7, %g1
-	bne,pn	%xcc, 2f
+	bne,pn	%xcc, BACKOFF_LABEL(2f, 1b)
 	 nop
 	retl
 	 nop
@@ -112,7 +112,7 @@ clear_bit:		/* %o0=nr, %o1=addr */
 	andn	%g7, %o2, %g1
 	casx	[%o1], %g7, %g1
 	cmp	%g7, %g1
-	bne,pn	%xcc, 2f
+	bne,pn	%xcc, BACKOFF_LABEL(2f, 1b)
 	 nop
 	retl
 	 nop
@@ -133,7 +133,7 @@ change_bit:		/* %o0=nr, %o1=addr */
 	xor	%g7, %o2, %g1
 	casx	[%o1], %g7, %g1
 	cmp	%g7, %g1
-	bne,pn	%xcc, 2f
+	bne,pn	%xcc, BACKOFF_LABEL(2f, 1b)
 	 nop
 	retl
 	 nop
diff --git a/arch/sparc/lib/rwsem_64.S b/arch/sparc/lib/rwsem_64.S
deleted file mode 100644
index 91a7d29..0000000
--- a/arch/sparc/lib/rwsem_64.S
+++ /dev/null
@@ -1,163 +0,0 @@
-/* rwsem.S: RW semaphore assembler.
- *
- * Written by David S. Miller (davem@redhat.com), 2001.
- * Derived from asm-i386/rwsem.h
- */
-
-#include <asm/rwsem-const.h>
-
-	.section	.sched.text, "ax"
-
-	.globl		__down_read
-__down_read:
-1:	lduw		[%o0], %g1
-	add		%g1, 1, %g7
-	cas		[%o0], %g1, %g7
-	cmp		%g1, %g7
-	bne,pn		%icc, 1b
-	 add		%g7, 1, %g7
-	cmp		%g7, 0
-	bl,pn		%icc, 3f
-	 nop
-2:
-	retl
-	 nop
-3:
-	save		%sp, -192, %sp
-	call		rwsem_down_read_failed
-	 mov		%i0, %o0
-	ret
-	 restore
-	.size		__down_read, .-__down_read
-
-	.globl		__down_read_trylock
-__down_read_trylock:
-1:	lduw		[%o0], %g1
-	add		%g1, 1, %g7
-	cmp		%g7, 0
-	bl,pn		%icc, 2f
-	 mov		0, %o1
-	cas		[%o0], %g1, %g7
-	cmp		%g1, %g7
-	bne,pn		%icc, 1b
-	 mov		1, %o1
-2:	retl
-	 mov		%o1, %o0
-	.size		__down_read_trylock, .-__down_read_trylock
-
-	.globl		__down_write
-__down_write:
-	sethi		%hi(RWSEM_ACTIVE_WRITE_BIAS), %g1
-	or		%g1, %lo(RWSEM_ACTIVE_WRITE_BIAS), %g1
-1:
-	lduw		[%o0], %g3
-	add		%g3, %g1, %g7
-	cas		[%o0], %g3, %g7
-	cmp		%g3, %g7
-	bne,pn		%icc, 1b
-	 cmp		%g7, 0
-	bne,pn		%icc, 3f
-	 nop
-2:	retl
-	 nop
-3:
-	save		%sp, -192, %sp
-	call		rwsem_down_write_failed
-	 mov		%i0, %o0
-	ret
-	 restore
-	.size		__down_write, .-__down_write
-
-	.globl		__down_write_trylock
-__down_write_trylock:
-	sethi		%hi(RWSEM_ACTIVE_WRITE_BIAS), %g1
-	or		%g1, %lo(RWSEM_ACTIVE_WRITE_BIAS), %g1
-1:
-	lduw		[%o0], %g3
-	cmp		%g3, 0
-	bne,pn		%icc, 2f
-	 mov		0, %o1
-	add		%g3, %g1, %g7
-	cas		[%o0], %g3, %g7
-	cmp		%g3, %g7
-	bne,pn		%icc, 1b
-	 mov		1, %o1
-2:	retl
-	 mov		%o1, %o0
-	.size		__down_write_trylock, .-__down_write_trylock
-
-	.globl		__up_read
-__up_read:
-1:
-	lduw		[%o0], %g1
-	sub		%g1, 1, %g7
-	cas		[%o0], %g1, %g7
-	cmp		%g1, %g7
-	bne,pn		%icc, 1b
-	 cmp		%g7, 0
-	bl,pn		%icc, 3f
-	 nop
-2:	retl
-	 nop
-3:	sethi		%hi(RWSEM_ACTIVE_MASK), %g1
-	sub		%g7, 1, %g7
-	or		%g1, %lo(RWSEM_ACTIVE_MASK), %g1
-	andcc		%g7, %g1, %g0
-	bne,pn		%icc, 2b
-	 nop
-	save		%sp, -192, %sp
-	call		rwsem_wake
-	 mov		%i0, %o0
-	ret
-	 restore
-	.size		__up_read, .-__up_read
-
-	.globl		__up_write
-__up_write:
-	sethi		%hi(RWSEM_ACTIVE_WRITE_BIAS), %g1
-	or		%g1, %lo(RWSEM_ACTIVE_WRITE_BIAS), %g1
-1:
-	lduw		[%o0], %g3
-	sub		%g3, %g1, %g7
-	cas		[%o0], %g3, %g7
-	cmp		%g3, %g7
-	bne,pn		%icc, 1b
-	 sub		%g7, %g1, %g7
-	cmp		%g7, 0
-	bl,pn		%icc, 3f
-	 nop
-2:
-	retl
-	 nop
-3:
-	save		%sp, -192, %sp
-	call		rwsem_wake
-	 mov		%i0, %o0
-	ret
-	 restore
-	.size		__up_write, .-__up_write
-
-	.globl		__downgrade_write
-__downgrade_write:
-	sethi		%hi(RWSEM_WAITING_BIAS), %g1
-	or		%g1, %lo(RWSEM_WAITING_BIAS), %g1
-1:
-	lduw		[%o0], %g3
-	sub		%g3, %g1, %g7
-	cas		[%o0], %g3, %g7
-	cmp		%g3, %g7
-	bne,pn		%icc, 1b
-	 sub		%g7, %g1, %g7
-	cmp		%g7, 0
-	bl,pn		%icc, 3f
-	 nop
-2:
-	retl
-	 nop
-3:
-	save		%sp, -192, %sp
-	call		rwsem_downgrade_wake
-	 mov		%i0, %o0
-	ret
-	 restore
-	.size		__downgrade_write, .-__downgrade_write
diff --git a/arch/sparc/prom/cif.S b/arch/sparc/prom/cif.S
index 5f27ad7..9c86b4b 100644
--- a/arch/sparc/prom/cif.S
+++ b/arch/sparc/prom/cif.S
@@ -9,18 +9,18 @@
 #include <asm/thread_info.h>
 
 	.text
-	.globl	prom_cif_interface
-prom_cif_interface:
-	sethi	%hi(p1275buf), %o0
-	or	%o0, %lo(p1275buf), %o0
-	ldx	[%o0 + 0x010], %o1	! prom_cif_stack
-	save	%o1, -192, %sp
-	ldx	[%i0 + 0x008], %l2	! prom_cif_handler
+	.globl	prom_cif_direct
+prom_cif_direct:
+	sethi	%hi(p1275buf), %o1
+	or	%o1, %lo(p1275buf), %o1
+	ldx	[%o1 + 0x0010], %o2	! prom_cif_stack
+	save	%o2, -192, %sp
+	ldx	[%i1 + 0x0008], %l2	! prom_cif_handler
 	mov	%g4, %l0
 	mov	%g5, %l1
 	mov	%g6, %l3
 	call	%l2
-	 add	%i0, 0x018, %o0		! prom_args
+	 mov	%i0, %o0		! prom_args
 	mov	%l0, %g4
 	mov	%l1, %g5
 	mov	%l3, %g6
diff --git a/arch/sparc/prom/console_64.c b/arch/sparc/prom/console_64.c
index f55d58a..10322dc 100644
--- a/arch/sparc/prom/console_64.c
+++ b/arch/sparc/prom/console_64.c
@@ -21,14 +21,22 @@ extern int prom_stdin, prom_stdout;
 inline int
 prom_nbgetchar(void)
 {
+	unsigned long args[7];
 	char inc;
 
-	if (p1275_cmd("read", P1275_ARG(1,P1275_ARG_OUT_BUF)|
-			      P1275_INOUT(3,1),
-			      prom_stdin, &inc, P1275_SIZE(1)) == 1)
+	args[0] = (unsigned long) "read";
+	args[1] = 3;
+	args[2] = 1;
+	args[3] = (unsigned int) prom_stdin;
+	args[4] = (unsigned long) &inc;
+	args[5] = 1;
+	args[6] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	if (args[6] == 1)
 		return inc;
-	else
-		return -1;
+	return -1;
 }
 
 /* Non blocking put character to console device, returns -1 if
@@ -37,12 +45,22 @@ prom_nbgetchar(void)
 inline int
 prom_nbputchar(char c)
 {
+	unsigned long args[7];
 	char outc;
 	
 	outc = c;
-	if (p1275_cmd("write", P1275_ARG(1,P1275_ARG_IN_BUF)|
-			       P1275_INOUT(3,1),
-			       prom_stdout, &outc, P1275_SIZE(1)) == 1)
+
+	args[0] = (unsigned long) "write";
+	args[1] = 3;
+	args[2] = 1;
+	args[3] = (unsigned int) prom_stdout;
+	args[4] = (unsigned long) &outc;
+	args[5] = 1;
+	args[6] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	if (args[6] == 1)
 		return 0;
 	else
 		return -1;
@@ -67,7 +85,15 @@ prom_putchar(char c)
 void
 prom_puts(const char *s, int len)
 {
-	p1275_cmd("write", P1275_ARG(1,P1275_ARG_IN_BUF)|
-			   P1275_INOUT(3,1),
-			   prom_stdout, s, P1275_SIZE(len));
+	unsigned long args[7];
+
+	args[0] = (unsigned long) "write";
+	args[1] = 3;
+	args[2] = 1;
+	args[3] = (unsigned int) prom_stdout;
+	args[4] = (unsigned long) s;
+	args[5] = len;
+	args[6] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
 }
diff --git a/arch/sparc/prom/devops_64.c b/arch/sparc/prom/devops_64.c
index 9dbd803..a017119 100644
--- a/arch/sparc/prom/devops_64.c
+++ b/arch/sparc/prom/devops_64.c
@@ -18,16 +18,32 @@
 int
 prom_devopen(const char *dstr)
 {
-	return p1275_cmd ("open", P1275_ARG(0,P1275_ARG_IN_STRING)|
-				  P1275_INOUT(1,1),
-				  dstr);
+	unsigned long args[5];
+
+	args[0] = (unsigned long) "open";
+	args[1] = 1;
+	args[2] = 1;
+	args[3] = (unsigned long) dstr;
+	args[4] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	return (int) args[4];
 }
 
 /* Close the device described by device handle 'dhandle'. */
 int
 prom_devclose(int dhandle)
 {
-	p1275_cmd ("close", P1275_INOUT(1,0), dhandle);
+	unsigned long args[4];
+
+	args[0] = (unsigned long) "close";
+	args[1] = 1;
+	args[2] = 0;
+	args[3] = (unsigned int) dhandle;
+
+	p1275_cmd_direct(args);
+
 	return 0;
 }
 
@@ -37,5 +53,15 @@ prom_devclose(int dhandle)
 void
 prom_seek(int dhandle, unsigned int seekhi, unsigned int seeklo)
 {
-	p1275_cmd ("seek", P1275_INOUT(3,1), dhandle, seekhi, seeklo);
+	unsigned long args[7];
+
+	args[0] = (unsigned long) "seek";
+	args[1] = 3;
+	args[2] = 1;
+	args[3] = (unsigned int) dhandle;
+	args[4] = seekhi;
+	args[5] = seeklo;
+	args[6] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
 }
diff --git a/arch/sparc/prom/misc_64.c b/arch/sparc/prom/misc_64.c
index 39fc6af..6cb1581 100644
--- a/arch/sparc/prom/misc_64.c
+++ b/arch/sparc/prom/misc_64.c
@@ -20,10 +20,17 @@
 
 int prom_service_exists(const char *service_name)
 {
-	int err = p1275_cmd("test", P1275_ARG(0, P1275_ARG_IN_STRING) |
-			    P1275_INOUT(1, 1), service_name);
+	unsigned long args[5];
 
-	if (err)
+	args[0] = (unsigned long) "test";
+	args[1] = 1;
+	args[2] = 1;
+	args[3] = (unsigned long) service_name;
+	args[4] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	if (args[4])
 		return 0;
 	return 1;
 }
@@ -31,30 +38,47 @@ int prom_service_exists(const char *service_name)
 void prom_sun4v_guest_soft_state(void)
 {
 	const char *svc = "SUNW,soft-state-supported";
+	unsigned long args[3];
 
 	if (!prom_service_exists(svc))
 		return;
-	p1275_cmd(svc, P1275_INOUT(0, 0));
+	args[0] = (unsigned long) svc;
+	args[1] = 0;
+	args[2] = 0;
+	p1275_cmd_direct(args);
 }
 
 /* Reset and reboot the machine with the command 'bcommand'. */
 void prom_reboot(const char *bcommand)
 {
+	unsigned long args[4];
+
 #ifdef CONFIG_SUN_LDOMS
 	if (ldom_domaining_enabled)
 		ldom_reboot(bcommand);
 #endif
-	p1275_cmd("boot", P1275_ARG(0, P1275_ARG_IN_STRING) |
-		  P1275_INOUT(1, 0), bcommand);
+	args[0] = (unsigned long) "boot";
+	args[1] = 1;
+	args[2] = 0;
+	args[3] = (unsigned long) bcommand;
+
+	p1275_cmd_direct(args);
 }
 
 /* Forth evaluate the expression contained in 'fstring'. */
 void prom_feval(const char *fstring)
 {
+	unsigned long args[5];
+
 	if (!fstring || fstring[0] == 0)
 		return;
-	p1275_cmd("interpret", P1275_ARG(0, P1275_ARG_IN_STRING) |
-		  P1275_INOUT(1, 1), fstring);
+	args[0] = (unsigned long) "interpret";
+	args[1] = 1;
+	args[2] = 1;
+	args[3] = (unsigned long) fstring;
+	args[4] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
 }
 EXPORT_SYMBOL(prom_feval);
 
@@ -68,6 +92,7 @@ extern void smp_release(void);
  */
 void prom_cmdline(void)
 {
+	unsigned long args[3];
 	unsigned long flags;
 
 	local_irq_save(flags);
@@ -76,7 +101,11 @@ void prom_cmdline(void)
 	smp_capture();
 #endif
 
-	p1275_cmd("enter", P1275_INOUT(0, 0));
+	args[0] = (unsigned long) "enter";
+	args[1] = 0;
+	args[2] = 0;
+
+	p1275_cmd_direct(args);
 
 #ifdef CONFIG_SMP
 	smp_release();
@@ -90,22 +119,32 @@ void prom_cmdline(void)
  */
 void notrace prom_halt(void)
 {
+	unsigned long args[3];
+
 #ifdef CONFIG_SUN_LDOMS
 	if (ldom_domaining_enabled)
 		ldom_power_off();
 #endif
 again:
-	p1275_cmd("exit", P1275_INOUT(0, 0));
+	args[0] = (unsigned long) "exit";
+	args[1] = 0;
+	args[2] = 0;
+	p1275_cmd_direct(args);
 	goto again; /* PROM is out to get me -DaveM */
 }
 
 void prom_halt_power_off(void)
 {
+	unsigned long args[3];
+
 #ifdef CONFIG_SUN_LDOMS
 	if (ldom_domaining_enabled)
 		ldom_power_off();
 #endif
-	p1275_cmd("SUNW,power-off", P1275_INOUT(0, 0));
+	args[0] = (unsigned long) "SUNW,power-off";
+	args[1] = 0;
+	args[2] = 0;
+	p1275_cmd_direct(args);
 
 	/* if nothing else helps, we just halt */
 	prom_halt();
@@ -114,10 +153,15 @@ void prom_halt_power_off(void)
 /* Set prom sync handler to call function 'funcp'. */
 void prom_setcallback(callback_func_t funcp)
 {
+	unsigned long args[5];
 	if (!funcp)
 		return;
-	p1275_cmd("set-callback", P1275_ARG(0, P1275_ARG_IN_FUNCTION) |
-		  P1275_INOUT(1, 1), funcp);
+	args[0] = (unsigned long) "set-callback";
+	args[1] = 1;
+	args[2] = 1;
+	args[3] = (unsigned long) funcp;
+	args[4] = (unsigned long) -1;
+	p1275_cmd_direct(args);
 }
 
 /* Get the idprom and stuff it into buffer 'idbuf'.  Returns the
@@ -173,57 +217,61 @@ static int prom_get_memory_ihandle(void)
 }
 
 /* Load explicit I/D TLB entries. */
+static long tlb_load(const char *type, unsigned long index,
+		     unsigned long tte_data, unsigned long vaddr)
+{
+	unsigned long args[9];
+
+	args[0] = (unsigned long) prom_callmethod_name;
+	args[1] = 5;
+	args[2] = 1;
+	args[3] = (unsigned long) type;
+	args[4] = (unsigned int) prom_get_mmu_ihandle();
+	args[5] = vaddr;
+	args[6] = tte_data;
+	args[7] = index;
+	args[8] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	return (long) args[8];
+}
+
 long prom_itlb_load(unsigned long index,
 		    unsigned long tte_data,
 		    unsigned long vaddr)
 {
-	return p1275_cmd(prom_callmethod_name,
-			 (P1275_ARG(0, P1275_ARG_IN_STRING) |
-			  P1275_ARG(2, P1275_ARG_IN_64B) |
-			  P1275_ARG(3, P1275_ARG_IN_64B) |
-			  P1275_INOUT(5, 1)),
-			 "SUNW,itlb-load",
-			 prom_get_mmu_ihandle(),
-			 /* And then our actual args are pushed backwards. */
-			 vaddr,
-			 tte_data,
-			 index);
+	return tlb_load("SUNW,itlb-load", index, tte_data, vaddr);
 }
 
 long prom_dtlb_load(unsigned long index,
 		    unsigned long tte_data,
 		    unsigned long vaddr)
 {
-	return p1275_cmd(prom_callmethod_name,
-			 (P1275_ARG(0, P1275_ARG_IN_STRING) |
-			  P1275_ARG(2, P1275_ARG_IN_64B) |
-			  P1275_ARG(3, P1275_ARG_IN_64B) |
-			  P1275_INOUT(5, 1)),
-			 "SUNW,dtlb-load",
-			 prom_get_mmu_ihandle(),
-			 /* And then our actual args are pushed backwards. */
-			 vaddr,
-			 tte_data,
-			 index);
+	return tlb_load("SUNW,dtlb-load", index, tte_data, vaddr);
 }
 
 int prom_map(int mode, unsigned long size,
 	     unsigned long vaddr, unsigned long paddr)
 {
-	int ret = p1275_cmd(prom_callmethod_name,
-			    (P1275_ARG(0, P1275_ARG_IN_STRING) |
-			     P1275_ARG(3, P1275_ARG_IN_64B) |
-			     P1275_ARG(4, P1275_ARG_IN_64B) |
-			     P1275_ARG(6, P1275_ARG_IN_64B) |
-			     P1275_INOUT(7, 1)),
-			    prom_map_name,
-			    prom_get_mmu_ihandle(),
-			    mode,
-			    size,
-			    vaddr,
-			    0,
-			    paddr);
-
+	unsigned long args[11];
+	int ret;
+
+	args[0] = (unsigned long) prom_callmethod_name;
+	args[1] = 7;
+	args[2] = 1;
+	args[3] = (unsigned long) prom_map_name;
+	args[4] = (unsigned int) prom_get_mmu_ihandle();
+	args[5] = (unsigned int) mode;
+	args[6] = size;
+	args[7] = vaddr;
+	args[8] = 0;
+	args[9] = paddr;
+	args[10] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	ret = (int) args[10];
 	if (ret == 0)
 		ret = -1;
 	return ret;
@@ -231,40 +279,51 @@ int prom_map(int mode, unsigned long size,
 
 void prom_unmap(unsigned long size, unsigned long vaddr)
 {
-	p1275_cmd(prom_callmethod_name,
-		  (P1275_ARG(0, P1275_ARG_IN_STRING) |
-		   P1275_ARG(2, P1275_ARG_IN_64B) |
-		   P1275_ARG(3, P1275_ARG_IN_64B) |
-		   P1275_INOUT(4, 0)),
-		  prom_unmap_name,
-		  prom_get_mmu_ihandle(),
-		  size,
-		  vaddr);
+	unsigned long args[7];
+
+	args[0] = (unsigned long) prom_callmethod_name;
+	args[1] = 4;
+	args[2] = 0;
+	args[3] = (unsigned long) prom_unmap_name;
+	args[4] = (unsigned int) prom_get_mmu_ihandle();
+	args[5] = size;
+	args[6] = vaddr;
+
+	p1275_cmd_direct(args);
 }
 
 /* Set aside physical memory which is not touched or modified
  * across soft resets.
  */
-unsigned long prom_retain(const char *name,
-			  unsigned long pa_low, unsigned long pa_high,
-			  long size, long align)
+int prom_retain(const char *name, unsigned long size,
+		unsigned long align, unsigned long *paddr)
 {
-	/* XXX I don't think we return multiple values correctly.
-	 * XXX OBP supposedly returns pa_low/pa_high here, how does
-	 * XXX it work?
+	unsigned long args[11];
+
+	args[0] = (unsigned long) prom_callmethod_name;
+	args[1] = 5;
+	args[2] = 3;
+	args[3] = (unsigned long) "SUNW,retain";
+	args[4] = (unsigned int) prom_get_memory_ihandle();
+	args[5] = align;
+	args[6] = size;
+	args[7] = (unsigned long) name;
+	args[8] = (unsigned long) -1;
+	args[9] = (unsigned long) -1;
+	args[10] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	if (args[8])
+		return (int) args[8];
+
+	/* Next we get "phys_high" then "phys_low".  On 64-bit
+	 * the phys_high cell is don't care since the phys_low
+	 * cell has the full value.
 	 */
+	*paddr = args[10];
 
-	/* If align is zero, the pa_low/pa_high args are passed,
-	 * else they are not.
-	 */
-	if (align == 0)
-		return p1275_cmd("SUNW,retain",
-				 (P1275_ARG(0, P1275_ARG_IN_BUF) | P1275_INOUT(5, 2)),
-				 name, pa_low, pa_high, size, align);
-	else
-		return p1275_cmd("SUNW,retain",
-				 (P1275_ARG(0, P1275_ARG_IN_BUF) | P1275_INOUT(3, 2)),
-				 name, size, align);
+	return 0;
 }
 
 /* Get "Unumber" string for the SIMM at the given
@@ -277,62 +336,129 @@ int prom_getunumber(int syndrome_code,
 		    unsigned long phys_addr,
 		    char *buf, int buflen)
 {
-	return p1275_cmd(prom_callmethod_name,
-			 (P1275_ARG(0, P1275_ARG_IN_STRING)	|
-			  P1275_ARG(3, P1275_ARG_OUT_BUF)	|
-			  P1275_ARG(6, P1275_ARG_IN_64B)	|
-			  P1275_INOUT(8, 2)),
-			 "SUNW,get-unumber", prom_get_memory_ihandle(),
-			 buflen, buf, P1275_SIZE(buflen),
-			 0, phys_addr, syndrome_code);
+	unsigned long args[12];
+
+	args[0] = (unsigned long) prom_callmethod_name;
+	args[1] = 7;
+	args[2] = 2;
+	args[3] = (unsigned long) "SUNW,get-unumber";
+	args[4] = (unsigned int) prom_get_memory_ihandle();
+	args[5] = buflen;
+	args[6] = (unsigned long) buf;
+	args[7] = 0;
+	args[8] = phys_addr;
+	args[9] = (unsigned int) syndrome_code;
+	args[10] = (unsigned long) -1;
+	args[11] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	return (int) args[10];
 }
 
 /* Power management extensions. */
 void prom_sleepself(void)
 {
-	p1275_cmd("SUNW,sleep-self", P1275_INOUT(0, 0));
+	unsigned long args[3];
+
+	args[0] = (unsigned long) "SUNW,sleep-self";
+	args[1] = 0;
+	args[2] = 0;
+	p1275_cmd_direct(args);
 }
 
 int prom_sleepsystem(void)
 {
-	return p1275_cmd("SUNW,sleep-system", P1275_INOUT(0, 1));
+	unsigned long args[4];
+
+	args[0] = (unsigned long) "SUNW,sleep-system";
+	args[1] = 0;
+	args[2] = 1;
+	args[3] = (unsigned long) -1;
+	p1275_cmd_direct(args);
+
+	return (int) args[3];
 }
 
 int prom_wakeupsystem(void)
 {
-	return p1275_cmd("SUNW,wakeup-system", P1275_INOUT(0, 1));
+	unsigned long args[4];
+
+	args[0] = (unsigned long) "SUNW,wakeup-system";
+	args[1] = 0;
+	args[2] = 1;
+	args[3] = (unsigned long) -1;
+	p1275_cmd_direct(args);
+
+	return (int) args[3];
 }
 
 #ifdef CONFIG_SMP
 void prom_startcpu(int cpunode, unsigned long pc, unsigned long arg)
 {
-	p1275_cmd("SUNW,start-cpu", P1275_INOUT(3, 0), cpunode, pc, arg);
+	unsigned long args[6];
+
+	args[0] = (unsigned long) "SUNW,start-cpu";
+	args[1] = 3;
+	args[2] = 0;
+	args[3] = (unsigned int) cpunode;
+	args[4] = pc;
+	args[5] = arg;
+	p1275_cmd_direct(args);
 }
 
 void prom_startcpu_cpuid(int cpuid, unsigned long pc, unsigned long arg)
 {
-	p1275_cmd("SUNW,start-cpu-by-cpuid", P1275_INOUT(3, 0),
-		  cpuid, pc, arg);
+	unsigned long args[6];
+
+	args[0] = (unsigned long) "SUNW,start-cpu-by-cpuid";
+	args[1] = 3;
+	args[2] = 0;
+	args[3] = (unsigned int) cpuid;
+	args[4] = pc;
+	args[5] = arg;
+	p1275_cmd_direct(args);
 }
 
 void prom_stopcpu_cpuid(int cpuid)
 {
-	p1275_cmd("SUNW,stop-cpu-by-cpuid", P1275_INOUT(1, 0),
-		  cpuid);
+	unsigned long args[4];
+
+	args[0] = (unsigned long) "SUNW,stop-cpu-by-cpuid";
+	args[1] = 1;
+	args[2] = 0;
+	args[3] = (unsigned int) cpuid;
+	p1275_cmd_direct(args);
 }
 
 void prom_stopself(void)
 {
-	p1275_cmd("SUNW,stop-self", P1275_INOUT(0, 0));
+	unsigned long args[3];
+
+	args[0] = (unsigned long) "SUNW,stop-self";
+	args[1] = 0;
+	args[2] = 0;
+	p1275_cmd_direct(args);
 }
 
 void prom_idleself(void)
 {
-	p1275_cmd("SUNW,idle-self", P1275_INOUT(0, 0));
+	unsigned long args[3];
+
+	args[0] = (unsigned long) "SUNW,idle-self";
+	args[1] = 0;
+	args[2] = 0;
+	p1275_cmd_direct(args);
 }
 
 void prom_resumecpu(int cpunode)
 {
-	p1275_cmd("SUNW,resume-cpu", P1275_INOUT(1, 0), cpunode);
+	unsigned long args[4];
+
+	args[0] = (unsigned long) "SUNW,resume-cpu";
+	args[1] = 1;
+	args[2] = 0;
+	args[3] = (unsigned int) cpunode;
+	p1275_cmd_direct(args);
 }
 #endif
diff --git a/arch/sparc/prom/p1275.c b/arch/sparc/prom/p1275.c
index 2d8b70d..fa6e4e2 100644
--- a/arch/sparc/prom/p1275.c
+++ b/arch/sparc/prom/p1275.c
@@ -22,13 +22,11 @@ struct {
 	long prom_callback;			/* 0x00 */
 	void (*prom_cif_handler)(long *);	/* 0x08 */
 	unsigned long prom_cif_stack;		/* 0x10 */
-	unsigned long prom_args [23];		/* 0x18 */
-	char prom_buffer [3000];
 } p1275buf;
 
 extern void prom_world(int);
 
-extern void prom_cif_interface(void);
+extern void prom_cif_direct(unsigned long *args);
 extern void prom_cif_callback(void);
 
 /*
@@ -36,114 +34,20 @@ extern void prom_cif_callback(void);
  */
 DEFINE_RAW_SPINLOCK(prom_entry_lock);
 
-long p1275_cmd(const char *service, long fmt, ...)
+void p1275_cmd_direct(unsigned long *args)
 {
-	char *p, *q;
 	unsigned long flags;
-	int nargs, nrets, i;
-	va_list list;
-	long attrs, x;
-	
-	p = p1275buf.prom_buffer;
 
 	raw_local_save_flags(flags);
 	raw_local_irq_restore(PIL_NMI);
 	raw_spin_lock(&prom_entry_lock);
 
-	p1275buf.prom_args[0] = (unsigned long)p;		/* service */
-	strcpy (p, service);
-	p = (char *)(((long)(strchr (p, 0) + 8)) & ~7);
-	p1275buf.prom_args[1] = nargs = (fmt & 0x0f);		/* nargs */
-	p1275buf.prom_args[2] = nrets = ((fmt & 0xf0) >> 4); 	/* nrets */
-	attrs = fmt >> 8;
-	va_start(list, fmt);
-	for (i = 0; i < nargs; i++, attrs >>= 3) {
-		switch (attrs & 0x7) {
-		case P1275_ARG_NUMBER:
-			p1275buf.prom_args[i + 3] =
-						(unsigned)va_arg(list, long);
-			break;
-		case P1275_ARG_IN_64B:
-			p1275buf.prom_args[i + 3] =
-				va_arg(list, unsigned long);
-			break;
-		case P1275_ARG_IN_STRING:
-			strcpy (p, va_arg(list, char *));
-			p1275buf.prom_args[i + 3] = (unsigned long)p;
-			p = (char *)(((long)(strchr (p, 0) + 8)) & ~7);
-			break;
-		case P1275_ARG_OUT_BUF:
-			(void) va_arg(list, char *);
-			p1275buf.prom_args[i + 3] = (unsigned long)p;
-			x = va_arg(list, long);
-			i++; attrs >>= 3;
-			p = (char *)(((long)(p + (int)x + 7)) & ~7);
-			p1275buf.prom_args[i + 3] = x;
-			break;
-		case P1275_ARG_IN_BUF:
-			q = va_arg(list, char *);
-			p1275buf.prom_args[i + 3] = (unsigned long)p;
-			x = va_arg(list, long);
-			i++; attrs >>= 3;
-			memcpy (p, q, (int)x);
-			p = (char *)(((long)(p + (int)x + 7)) & ~7);
-			p1275buf.prom_args[i + 3] = x;
-			break;
-		case P1275_ARG_OUT_32B:
-			(void) va_arg(list, char *);
-			p1275buf.prom_args[i + 3] = (unsigned long)p;
-			p += 32;
-			break;
-		case P1275_ARG_IN_FUNCTION:
-			p1275buf.prom_args[i + 3] =
-					(unsigned long)prom_cif_callback;
-			p1275buf.prom_callback = va_arg(list, long);
-			break;
-		}
-	}
-	va_end(list);
-
 	prom_world(1);
-	prom_cif_interface();
+	prom_cif_direct(args);
 	prom_world(0);
 
-	attrs = fmt >> 8;
-	va_start(list, fmt);
-	for (i = 0; i < nargs; i++, attrs >>= 3) {
-		switch (attrs & 0x7) {
-		case P1275_ARG_NUMBER:
-			(void) va_arg(list, long);
-			break;
-		case P1275_ARG_IN_STRING:
-			(void) va_arg(list, char *);
-			break;
-		case P1275_ARG_IN_FUNCTION:
-			(void) va_arg(list, long);
-			break;
-		case P1275_ARG_IN_BUF:
-			(void) va_arg(list, char *);
-			(void) va_arg(list, long);
-			i++; attrs >>= 3;
-			break;
-		case P1275_ARG_OUT_BUF:
-			p = va_arg(list, char *);
-			x = va_arg(list, long);
-			memcpy (p, (char *)(p1275buf.prom_args[i + 3]), (int)x);
-			i++; attrs >>= 3;
-			break;
-		case P1275_ARG_OUT_32B:
-			p = va_arg(list, char *);
-			memcpy (p, (char *)(p1275buf.prom_args[i + 3]), 32);
-			break;
-		}
-	}
-	va_end(list);
-	x = p1275buf.prom_args [nargs + 3];
-
 	raw_spin_unlock(&prom_entry_lock);
 	raw_local_irq_restore(flags);
-
-	return x;
 }
 
 void prom_cif_init(void *cif_handler, void *cif_stack)
diff --git a/arch/sparc/prom/tree_64.c b/arch/sparc/prom/tree_64.c
index 3c0d2dd..9d3f913 100644
--- a/arch/sparc/prom/tree_64.c
+++ b/arch/sparc/prom/tree_64.c
@@ -16,22 +16,39 @@
 #include <asm/oplib.h>
 #include <asm/ldc.h>
 
+static int prom_node_to_node(const char *type, int node)
+{
+	unsigned long args[5];
+
+	args[0] = (unsigned long) type;
+	args[1] = 1;
+	args[2] = 1;
+	args[3] = (unsigned int) node;
+	args[4] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	return (int) args[4];
+}
+
 /* Return the child of node 'node' or zero if no this node has no
  * direct descendent.
  */
 inline int __prom_getchild(int node)
 {
-	return p1275_cmd ("child", P1275_INOUT(1, 1), node);
+	return prom_node_to_node("child", node);
 }
 
 inline int prom_getchild(int node)
 {
 	int cnode;
 
-	if(node == -1) return 0;
+	if (node == -1)
+		return 0;
 	cnode = __prom_getchild(node);
-	if(cnode == -1) return 0;
-	return (int)cnode;
+	if (cnode == -1)
+		return 0;
+	return cnode;
 }
 EXPORT_SYMBOL(prom_getchild);
 
@@ -39,10 +56,12 @@ inline int prom_getparent(int node)
 {
 	int cnode;
 
-	if(node == -1) return 0;
-	cnode = p1275_cmd ("parent", P1275_INOUT(1, 1), node);
-	if(cnode == -1) return 0;
-	return (int)cnode;
+	if (node == -1)
+		return 0;
+	cnode = prom_node_to_node("parent", node);
+	if (cnode == -1)
+		return 0;
+	return cnode;
 }
 
 /* Return the next sibling of node 'node' or zero if no more siblings
@@ -50,7 +69,7 @@ inline int prom_getparent(int node)
  */
 inline int __prom_getsibling(int node)
 {
-	return p1275_cmd(prom_peer_name, P1275_INOUT(1, 1), node);
+	return prom_node_to_node(prom_peer_name, node);
 }
 
 inline int prom_getsibling(int node)
@@ -72,11 +91,21 @@ EXPORT_SYMBOL(prom_getsibling);
  */
 inline int prom_getproplen(int node, const char *prop)
 {
-	if((!node) || (!prop)) return -1;
-	return p1275_cmd ("getproplen", 
-			  P1275_ARG(1,P1275_ARG_IN_STRING)|
-			  P1275_INOUT(2, 1), 
-			  node, prop);
+	unsigned long args[6];
+
+	if (!node || !prop)
+		return -1;
+
+	args[0] = (unsigned long) "getproplen";
+	args[1] = 2;
+	args[2] = 1;
+	args[3] = (unsigned int) node;
+	args[4] = (unsigned long) prop;
+	args[5] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	return (int) args[5];
 }
 EXPORT_SYMBOL(prom_getproplen);
 
@@ -87,19 +116,25 @@ EXPORT_SYMBOL(prom_getproplen);
 inline int prom_getproperty(int node, const char *prop,
 			    char *buffer, int bufsize)
 {
+	unsigned long args[8];
 	int plen;
 
 	plen = prom_getproplen(node, prop);
-	if ((plen > bufsize) || (plen == 0) || (plen == -1)) {
+	if ((plen > bufsize) || (plen == 0) || (plen == -1))
 		return -1;
-	} else {
-		/* Ok, things seem all right. */
-		return p1275_cmd(prom_getprop_name, 
-				 P1275_ARG(1,P1275_ARG_IN_STRING)|
-				 P1275_ARG(2,P1275_ARG_OUT_BUF)|
-				 P1275_INOUT(4, 1), 
-				 node, prop, buffer, P1275_SIZE(plen));
-	}
+
+	args[0] = (unsigned long) prom_getprop_name;
+	args[1] = 4;
+	args[2] = 1;
+	args[3] = (unsigned int) node;
+	args[4] = (unsigned long) prop;
+	args[5] = (unsigned long) buffer;
+	args[6] = bufsize;
+	args[7] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	return (int) args[7];
 }
 EXPORT_SYMBOL(prom_getproperty);
 
@@ -110,7 +145,7 @@ inline int prom_getint(int node, const char *prop)
 {
 	int intprop;
 
-	if(prom_getproperty(node, prop, (char *) &intprop, sizeof(int)) != -1)
+	if (prom_getproperty(node, prop, (char *) &intprop, sizeof(int)) != -1)
 		return intprop;
 
 	return -1;
@@ -126,7 +161,8 @@ int prom_getintdefault(int node, const char *property, int deflt)
 	int retval;
 
 	retval = prom_getint(node, property);
-	if(retval == -1) return deflt;
+	if (retval == -1)
+		return deflt;
 
 	return retval;
 }
@@ -138,7 +174,8 @@ int prom_getbool(int node, const char *prop)
 	int retval;
 
 	retval = prom_getproplen(node, prop);
-	if(retval == -1) return 0;
+	if (retval == -1)
+		return 0;
 	return 1;
 }
 EXPORT_SYMBOL(prom_getbool);
@@ -152,7 +189,8 @@ void prom_getstring(int node, const char *prop, char *user_buf, int ubuf_size)
 	int len;
 
 	len = prom_getproperty(node, prop, user_buf, ubuf_size);
-	if(len != -1) return;
+	if (len != -1)
+		return;
 	user_buf[0] = 0;
 }
 EXPORT_SYMBOL(prom_getstring);
@@ -164,7 +202,8 @@ int prom_nodematch(int node, const char *name)
 {
 	char namebuf[128];
 	prom_getproperty(node, "name", namebuf, sizeof(namebuf));
-	if(strcmp(namebuf, name) == 0) return 1;
+	if (strcmp(namebuf, name) == 0)
+		return 1;
 	return 0;
 }
 
@@ -190,16 +229,29 @@ int prom_searchsiblings(int node_start, const char *nodename)
 }
 EXPORT_SYMBOL(prom_searchsiblings);
 
+static const char *prom_nextprop_name = "nextprop";
+
 /* Return the first property type for node 'node'.
  * buffer should be at least 32B in length
  */
 inline char *prom_firstprop(int node, char *buffer)
 {
+	unsigned long args[7];
+
 	*buffer = 0;
-	if(node == -1) return buffer;
-	p1275_cmd ("nextprop", P1275_ARG(2,P1275_ARG_OUT_32B)|
-			       P1275_INOUT(3, 0), 
-			       node, (char *) 0x0, buffer);
+	if (node == -1)
+		return buffer;
+
+	args[0] = (unsigned long) prom_nextprop_name;
+	args[1] = 3;
+	args[2] = 1;
+	args[3] = (unsigned int) node;
+	args[4] = 0;
+	args[5] = (unsigned long) buffer;
+	args[6] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
 	return buffer;
 }
 EXPORT_SYMBOL(prom_firstprop);
@@ -210,9 +262,10 @@ EXPORT_SYMBOL(prom_firstprop);
  */
 inline char *prom_nextprop(int node, const char *oprop, char *buffer)
 {
+	unsigned long args[7];
 	char buf[32];
 
-	if(node == -1) {
+	if (node == -1) {
 		*buffer = 0;
 		return buffer;
 	}
@@ -220,10 +273,17 @@ inline char *prom_nextprop(int node, const char *oprop, char *buffer)
 		strcpy (buf, oprop);
 		oprop = buf;
 	}
-	p1275_cmd ("nextprop", P1275_ARG(1,P1275_ARG_IN_STRING)|
-				    P1275_ARG(2,P1275_ARG_OUT_32B)|
-				    P1275_INOUT(3, 0), 
-				    node, oprop, buffer); 
+
+	args[0] = (unsigned long) prom_nextprop_name;
+	args[1] = 3;
+	args[2] = 1;
+	args[3] = (unsigned int) node;
+	args[4] = (unsigned long) oprop;
+	args[5] = (unsigned long) buffer;
+	args[6] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
 	return buffer;
 }
 EXPORT_SYMBOL(prom_nextprop);
@@ -231,12 +291,19 @@ EXPORT_SYMBOL(prom_nextprop);
 int
 prom_finddevice(const char *name)
 {
+	unsigned long args[5];
+
 	if (!name)
 		return 0;
-	return p1275_cmd(prom_finddev_name,
-			 P1275_ARG(0,P1275_ARG_IN_STRING)|
-			 P1275_INOUT(1, 1), 
-			 name);
+	args[0] = (unsigned long) "finddevice";
+	args[1] = 1;
+	args[2] = 1;
+	args[3] = (unsigned long) name;
+	args[4] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	return (int) args[4];
 }
 EXPORT_SYMBOL(prom_finddevice);
 
@@ -247,7 +314,7 @@ int prom_node_has_property(int node, const char *prop)
 	*buf = 0;
 	do {
 		prom_nextprop(node, buf, buf);
-		if(!strcmp(buf, prop))
+		if (!strcmp(buf, prop))
 			return 1;
 	} while (*buf);
 	return 0;
@@ -260,6 +327,8 @@ EXPORT_SYMBOL(prom_node_has_property);
 int
 prom_setprop(int node, const char *pname, char *value, int size)
 {
+	unsigned long args[8];
+
 	if (size == 0)
 		return 0;
 	if ((pname == 0) || (value == 0))
@@ -271,19 +340,37 @@ prom_setprop(int node, const char *pname, char *value, int size)
 		return 0;
 	}
 #endif
-	return p1275_cmd ("setprop", P1275_ARG(1,P1275_ARG_IN_STRING)|
-					  P1275_ARG(2,P1275_ARG_IN_BUF)|
-					  P1275_INOUT(4, 1), 
-					  node, pname, value, P1275_SIZE(size));
+	args[0] = (unsigned long) "setprop";
+	args[1] = 4;
+	args[2] = 1;
+	args[3] = (unsigned int) node;
+	args[4] = (unsigned long) pname;
+	args[5] = (unsigned long) value;
+	args[6] = size;
+	args[7] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	return (int) args[7];
 }
 EXPORT_SYMBOL(prom_setprop);
 
 inline int prom_inst2pkg(int inst)
 {
+	unsigned long args[5];
 	int node;
 	
-	node = p1275_cmd ("instance-to-package", P1275_INOUT(1, 1), inst);
-	if (node == -1) return 0;
+	args[0] = (unsigned long) "instance-to-package";
+	args[1] = 1;
+	args[2] = 1;
+	args[3] = (unsigned int) inst;
+	args[4] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	node = (int) args[4];
+	if (node == -1)
+		return 0;
 	return node;
 }
 
@@ -296,17 +383,28 @@ prom_pathtoinode(const char *path)
 	int node, inst;
 
 	inst = prom_devopen (path);
-	if (inst == 0) return 0;
-	node = prom_inst2pkg (inst);
-	prom_devclose (inst);
-	if (node == -1) return 0;
+	if (inst == 0)
+		return 0;
+	node = prom_inst2pkg(inst);
+	prom_devclose(inst);
+	if (node == -1)
+		return 0;
 	return node;
 }
 
 int prom_ihandle2path(int handle, char *buffer, int bufsize)
 {
-	return p1275_cmd("instance-to-path",
-			 P1275_ARG(1,P1275_ARG_OUT_BUF)|
-			 P1275_INOUT(3, 1),
-			 handle, buffer, P1275_SIZE(bufsize));
+	unsigned long args[7];
+
+	args[0] = (unsigned long) "instance-to-path";
+	args[1] = 3;
+	args[2] = 1;
+	args[3] = (unsigned int) handle;
+	args[4] = (unsigned long) buffer;
+	args[5] = bufsize;
+	args[6] = (unsigned long) -1;
+
+	p1275_cmd_direct(args);
+
+	return (int) args[6];
 }
diff --git a/arch/tile/kernel/process.c b/arch/tile/kernel/process.c
index ed590ad..985cc28 100644
--- a/arch/tile/kernel/process.c
+++ b/arch/tile/kernel/process.c
@@ -543,8 +543,9 @@ long _sys_vfork(struct pt_regs *regs)
 /*
  * sys_execve() executes a new program.
  */
-long _sys_execve(char __user *path, char __user *__user *argv,
-		 char __user *__user *envp, struct pt_regs *regs)
+long _sys_execve(const char __user *path,
+		 const char __user *const __user *argv,
+		 const char __user *const __user *envp, struct pt_regs *regs)
 {
 	long error;
 	char *filename;
diff --git a/arch/um/include/asm/dma-mapping.h b/arch/um/include/asm/dma-mapping.h
index 17a2cb5..1f469e80 100644
--- a/arch/um/include/asm/dma-mapping.h
+++ b/arch/um/include/asm/dma-mapping.h
@@ -95,13 +95,6 @@ dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, int nelems,
 #define dma_alloc_noncoherent(d, s, h, f) dma_alloc_coherent(d, s, h, f)
 #define dma_free_noncoherent(d, s, v, h) dma_free_coherent(d, s, v, h)
 
-static inline int
-dma_get_cache_alignment(void)
-{
-	BUG();
-	return(0);
-}
-
 static inline void
 dma_cache_sync(struct device *dev, void *vaddr, size_t size,
 	       enum dma_data_direction direction)
diff --git a/arch/um/kernel/exec.c b/arch/um/kernel/exec.c
index 59b20d9..cd145ed 100644
--- a/arch/um/kernel/exec.c
+++ b/arch/um/kernel/exec.c
@@ -44,8 +44,9 @@ void start_thread(struct pt_regs *regs, unsigned long eip, unsigned long esp)
 	PT_REGS_SP(regs) = esp;
 }
 
-static long execve1(const char *file, char __user * __user *argv,
-		    char __user *__user *env)
+static long execve1(const char *file,
+		    const char __user *const __user *argv,
+		    const char __user *const __user *env)
 {
 	long error;
 
diff --git a/arch/um/kernel/syscall.c b/arch/um/kernel/syscall.c
index 7427c0b..5ddb246 100644
--- a/arch/um/kernel/syscall.c
+++ b/arch/um/kernel/syscall.c
@@ -51,7 +51,9 @@ long old_mmap(unsigned long addr, unsigned long len,
 	return err;
 }
 
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	mm_segment_t fs;
 	int ret;
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index a84fc34..cea0cd9 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -245,6 +245,11 @@ config ARCH_HWEIGHT_CFLAGS
 
 config KTIME_SCALAR
 	def_bool X86_32
+
+config ARCH_CPU_PROBE_RELEASE
+	def_bool y
+	depends on HOTPLUG_CPU
+
 source "init/Kconfig"
 source "kernel/Kconfig.freezer"
 
@@ -749,11 +754,11 @@ config IOMMU_API
 	def_bool (AMD_IOMMU || DMAR)
 
 config MAXSMP
-	bool "Configure Maximum number of SMP Processors and NUMA Nodes"
+	bool "Enable Maximum number of SMP Processors and NUMA Nodes"
 	depends on X86_64 && SMP && DEBUG_KERNEL && EXPERIMENTAL
 	select CPUMASK_OFFSTACK
 	---help---
-	  Configure maximum number of CPUS and NUMA Nodes for this architecture.
+	  Enable maximum number of CPUS and NUMA Nodes for this architecture.
 	  If unsure, say N.
 
 config NR_CPUS
diff --git a/arch/x86/include/asm/pgtable_32.h b/arch/x86/include/asm/pgtable_32.h
index 2984a25..f686f49 100644
--- a/arch/x86/include/asm/pgtable_32.h
+++ b/arch/x86/include/asm/pgtable_32.h
@@ -26,6 +26,7 @@ struct mm_struct;
 struct vm_area_struct;
 
 extern pgd_t swapper_pg_dir[1024];
+extern pgd_t trampoline_pg_dir[1024];
 
 static inline void pgtable_cache_init(void) { }
 static inline void check_pgt_cache(void) { }
diff --git a/arch/x86/include/asm/syscalls.h b/arch/x86/include/asm/syscalls.h
index feb2ff9..f1d8b44 100644
--- a/arch/x86/include/asm/syscalls.h
+++ b/arch/x86/include/asm/syscalls.h
@@ -23,8 +23,9 @@ long sys_iopl(unsigned int, struct pt_regs *);
 /* kernel/process.c */
 int sys_fork(struct pt_regs *);
 int sys_vfork(struct pt_regs *);
-long sys_execve(const char __user *, char __user * __user *,
-		char __user * __user *, struct pt_regs *);
+long sys_execve(const char __user *,
+		const char __user *const __user *,
+		const char __user *const __user *, struct pt_regs *);
 long sys_clone(unsigned long, unsigned long, void __user *,
 	       void __user *, struct pt_regs *);
 
diff --git a/arch/x86/include/asm/trampoline.h b/arch/x86/include/asm/trampoline.h
index cb507bb..4dde797 100644
--- a/arch/x86/include/asm/trampoline.h
+++ b/arch/x86/include/asm/trampoline.h
@@ -13,14 +13,17 @@ extern unsigned char *trampoline_base;
 
 extern unsigned long init_rsp;
 extern unsigned long initial_code;
+extern unsigned long initial_page_table;
 extern unsigned long initial_gs;
 
 #define TRAMPOLINE_SIZE roundup(trampoline_end - trampoline_data, PAGE_SIZE)
 
 extern unsigned long setup_trampoline(void);
+extern void __init setup_trampoline_page_table(void);
 extern void __init reserve_trampoline_memory(void);
 #else
-static inline void reserve_trampoline_memory(void) {};
+static inline void setup_trampoline_page_table(void) {}
+static inline void reserve_trampoline_memory(void) {}
 #endif /* CONFIG_X86_TRAMPOLINE */
 
 #endif /* __ASSEMBLY__ */
diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
index 4dc0084..f1efeba 100644
--- a/arch/x86/kernel/apic/io_apic.c
+++ b/arch/x86/kernel/apic/io_apic.c
@@ -1728,6 +1728,8 @@ __apicdebuginit(void) print_IO_APIC(void)
 		struct irq_pin_list *entry;
 
 		cfg = desc->chip_data;
+		if (!cfg)
+			continue;
 		entry = cfg->irq_2_pin;
 		if (!entry)
 			continue;
diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c
index 60a57b1..ba5f62f 100644
--- a/arch/x86/kernel/cpu/amd.c
+++ b/arch/x86/kernel/cpu/amd.c
@@ -669,7 +669,7 @@ bool cpu_has_amd_erratum(const int *erratum)
 	}
 
 	/* OSVW unavailable or ID unknown, match family-model-stepping range */
-	ms = (cpu->x86_model << 8) | cpu->x86_mask;
+	ms = (cpu->x86_model << 4) | cpu->x86_mask;
 	while ((range = *erratum++))
 		if ((cpu->x86 == AMD_MODEL_RANGE_FAMILY(range)) &&
 		    (ms >= AMD_MODEL_RANGE_START(range)) &&
diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S
index ff4c453..fa8c1b8 100644
--- a/arch/x86/kernel/head_32.S
+++ b/arch/x86/kernel/head_32.S
@@ -334,7 +334,7 @@ ENTRY(startup_32_smp)
 /*
  * Enable paging
  */
-	movl $pa(swapper_pg_dir),%eax
+	movl pa(initial_page_table), %eax
 	movl %eax,%cr3		/* set the page table pointer.. */
 	movl %cr0,%eax
 	orl  $X86_CR0_PG,%eax
@@ -614,6 +614,8 @@ ignore_int:
 .align 4
 ENTRY(initial_code)
 	.long i386_start_kernel
+ENTRY(initial_page_table)
+	.long pa(swapper_pg_dir)
 
 /*
  * BSS section
@@ -629,6 +631,10 @@ ENTRY(swapper_pg_dir)
 #endif
 swapper_pg_fixmap:
 	.fill 1024,4,0
+#ifdef CONFIG_X86_TRAMPOLINE
+ENTRY(trampoline_pg_dir)
+	.fill 1024,4,0
+#endif
 ENTRY(empty_zero_page)
 	.fill 4096,1,0
 
diff --git a/arch/x86/kernel/i387.c b/arch/x86/kernel/i387.c
index 1f11f5c..a46cb35 100644
--- a/arch/x86/kernel/i387.c
+++ b/arch/x86/kernel/i387.c
@@ -40,6 +40,7 @@
 
 static unsigned int		mxcsr_feature_mask __read_mostly = 0xffffffffu;
 unsigned int xstate_size;
+EXPORT_SYMBOL_GPL(xstate_size);
 unsigned int sig_xstate_ia32_size = sizeof(struct _fpstate_ia32);
 static struct i387_fxsave_struct fx_scratch __cpuinitdata;
 
diff --git a/arch/x86/kernel/kgdb.c b/arch/x86/kernel/kgdb.c
index ef10940..852b819 100644
--- a/arch/x86/kernel/kgdb.c
+++ b/arch/x86/kernel/kgdb.c
@@ -194,7 +194,7 @@ static struct hw_breakpoint {
 	unsigned long		addr;
 	int			len;
 	int			type;
-	struct perf_event	**pev;
+	struct perf_event	* __percpu *pev;
 } breakinfo[HBP_NUM];
 
 static unsigned long early_dr7;
diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes.c
index 1bfb6cf..770ebfb 100644
--- a/arch/x86/kernel/kprobes.c
+++ b/arch/x86/kernel/kprobes.c
@@ -709,6 +709,7 @@ static __used __kprobes void *trampoline_handler(struct pt_regs *regs)
 	struct hlist_node *node, *tmp;
 	unsigned long flags, orig_ret_address = 0;
 	unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline;
+	kprobe_opcode_t *correct_ret_addr = NULL;
 
 	INIT_HLIST_HEAD(&empty_rp);
 	kretprobe_hash_lock(current, &head, &flags);
@@ -740,14 +741,34 @@ static __used __kprobes void *trampoline_handler(struct pt_regs *regs)
 			/* another task is sharing our hash bucket */
 			continue;
 
+		orig_ret_address = (unsigned long)ri->ret_addr;
+
+		if (orig_ret_address != trampoline_address)
+			/*
+			 * This is the real return address. Any other
+			 * instances associated with this task are for
+			 * other calls deeper on the call stack
+			 */
+			break;
+	}
+
+	kretprobe_assert(ri, orig_ret_address, trampoline_address);
+
+	correct_ret_addr = ri->ret_addr;
+	hlist_for_each_entry_safe(ri, node, tmp, head, hlist) {
+		if (ri->task != current)
+			/* another task is sharing our hash bucket */
+			continue;
+
+		orig_ret_address = (unsigned long)ri->ret_addr;
 		if (ri->rp && ri->rp->handler) {
 			__get_cpu_var(current_kprobe) = &ri->rp->kp;
 			get_kprobe_ctlblk()->kprobe_status = KPROBE_HIT_ACTIVE;
+			ri->ret_addr = correct_ret_addr;
 			ri->rp->handler(ri, regs);
 			__get_cpu_var(current_kprobe) = NULL;
 		}
 
-		orig_ret_address = (unsigned long)ri->ret_addr;
 		recycle_rp_inst(ri, &empty_rp);
 
 		if (orig_ret_address != trampoline_address)
@@ -759,8 +780,6 @@ static __used __kprobes void *trampoline_handler(struct pt_regs *regs)
 			break;
 	}
 
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
-
 	kretprobe_hash_unlock(current, &flags);
 
 	hlist_for_each_entry_safe(ri, node, tmp, &empty_rp, hlist) {
diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
index 64ecaf0..57d1868 100644
--- a/arch/x86/kernel/process.c
+++ b/arch/x86/kernel/process.c
@@ -301,8 +301,9 @@ EXPORT_SYMBOL(kernel_thread);
 /*
  * sys_execve() executes a new program.
  */
-long sys_execve(const char __user *name, char __user * __user *argv,
-		char __user * __user *envp, struct pt_regs *regs)
+long sys_execve(const char __user *name,
+		const char __user *const __user *argv,
+		const char __user *const __user *envp, struct pt_regs *regs)
 {
 	long error;
 	char *filename;
diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c
index b008e78..c3a4fbb 100644
--- a/arch/x86/kernel/setup.c
+++ b/arch/x86/kernel/setup.c
@@ -1014,6 +1014,8 @@ void __init setup_arch(char **cmdline_p)
 	paging_init();
 	x86_init.paging.pagetable_setup_done(swapper_pg_dir);
 
+	setup_trampoline_page_table();
+
 	tboot_probe();
 
 #ifdef CONFIG_X86_64
diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index a5e928b..8b3bfc4 100644
--- a/arch/x86/kernel/smpboot.c
+++ b/arch/x86/kernel/smpboot.c
@@ -73,7 +73,6 @@
 
 #ifdef CONFIG_X86_32
 u8 apicid_2_node[MAX_APICID];
-static int low_mappings;
 #endif
 
 /* State of each CPU */
@@ -91,6 +90,25 @@ DEFINE_PER_CPU(int, cpu_state) = { 0 };
 static DEFINE_PER_CPU(struct task_struct *, idle_thread_array);
 #define get_idle_for_cpu(x)      (per_cpu(idle_thread_array, x))
 #define set_idle_for_cpu(x, p)   (per_cpu(idle_thread_array, x) = (p))
+
+/*
+ * We need this for trampoline_base protection from concurrent accesses when
+ * off- and onlining cores wildly.
+ */
+static DEFINE_MUTEX(x86_cpu_hotplug_driver_mutex);
+
+void cpu_hotplug_driver_lock()
+{
+        mutex_lock(&x86_cpu_hotplug_driver_mutex);
+}
+
+void cpu_hotplug_driver_unlock()
+{
+        mutex_unlock(&x86_cpu_hotplug_driver_mutex);
+}
+
+ssize_t arch_cpu_probe(const char *buf, size_t count) { return -1; }
+ssize_t arch_cpu_release(const char *buf, size_t count) { return -1; }
 #else
 static struct task_struct *idle_thread_array[NR_CPUS] __cpuinitdata ;
 #define get_idle_for_cpu(x)      (idle_thread_array[(x)])
@@ -281,6 +299,18 @@ notrace static void __cpuinit start_secondary(void *unused)
 	 * fragile that we want to limit the things done here to the
 	 * most necessary things.
 	 */
+
+#ifdef CONFIG_X86_32
+	/*
+	 * Switch away from the trampoline page-table
+	 *
+	 * Do this before cpu_init() because it needs to access per-cpu
+	 * data which may not be mapped in the trampoline page-table.
+	 */
+	load_cr3(swapper_pg_dir);
+	__flush_tlb_all();
+#endif
+
 	vmi_bringup();
 	cpu_init();
 	preempt_disable();
@@ -299,12 +329,6 @@ notrace static void __cpuinit start_secondary(void *unused)
 		legacy_pic->chip->unmask(0);
 	}
 
-#ifdef CONFIG_X86_32
-	while (low_mappings)
-		cpu_relax();
-	__flush_tlb_all();
-#endif
-
 	/* This must be done before setting cpu_online_mask */
 	set_cpu_sibling_map(raw_smp_processor_id());
 	wmb();
@@ -750,6 +774,7 @@ do_rest:
 #ifdef CONFIG_X86_32
 	/* Stack for startup_32 can be just as for start_secondary onwards */
 	irq_ctx_init(cpu);
+	initial_page_table = __pa(&trampoline_pg_dir);
 #else
 	clear_tsk_thread_flag(c_idle.idle, TIF_FORK);
 	initial_gs = per_cpu_offset(cpu);
@@ -897,20 +922,8 @@ int __cpuinit native_cpu_up(unsigned int cpu)
 
 	per_cpu(cpu_state, cpu) = CPU_UP_PREPARE;
 
-#ifdef CONFIG_X86_32
-	/* init low mem mapping */
-	clone_pgd_range(swapper_pg_dir, swapper_pg_dir + KERNEL_PGD_BOUNDARY,
-		min_t(unsigned long, KERNEL_PGD_PTRS, KERNEL_PGD_BOUNDARY));
-	flush_tlb_all();
-	low_mappings = 1;
-
 	err = do_boot_cpu(apicid, cpu);
 
-	zap_low_mappings(false);
-	low_mappings = 0;
-#else
-	err = do_boot_cpu(apicid, cpu);
-#endif
 	if (err) {
 		pr_debug("do_boot_cpu failed %d\n", err);
 		return -EIO;
diff --git a/arch/x86/kernel/sys_i386_32.c b/arch/x86/kernel/sys_i386_32.c
index 196552b..d5e0662 100644
--- a/arch/x86/kernel/sys_i386_32.c
+++ b/arch/x86/kernel/sys_i386_32.c
@@ -28,7 +28,9 @@
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
-int kernel_execve(const char *filename, char *const argv[], char *const envp[])
+int kernel_execve(const char *filename,
+		  const char *const argv[],
+		  const char *const envp[])
 {
 	long __res;
 	asm volatile ("push %%ebx ; movl %2,%%ebx ; int $0x80 ; pop %%ebx"
diff --git a/arch/x86/kernel/trampoline.c b/arch/x86/kernel/trampoline.c
index c652ef6..a874495 100644
--- a/arch/x86/kernel/trampoline.c
+++ b/arch/x86/kernel/trampoline.c
@@ -1,6 +1,7 @@
 #include <linux/io.h>
 
 #include <asm/trampoline.h>
+#include <asm/pgtable.h>
 #include <asm/e820.h>
 
 #if defined(CONFIG_X86_64) && defined(CONFIG_ACPI_SLEEP)
@@ -37,3 +38,20 @@ unsigned long __trampinit setup_trampoline(void)
 	memcpy(trampoline_base, trampoline_data, TRAMPOLINE_SIZE);
 	return virt_to_phys(trampoline_base);
 }
+
+void __init setup_trampoline_page_table(void)
+{
+#ifdef CONFIG_X86_32
+	/* Copy kernel address range */
+	clone_pgd_range(trampoline_pg_dir + KERNEL_PGD_BOUNDARY,
+			swapper_pg_dir + KERNEL_PGD_BOUNDARY,
+			min_t(unsigned long, KERNEL_PGD_PTRS,
+			      KERNEL_PGD_BOUNDARY));
+
+	/* Initialize low mappings */
+	clone_pgd_range(trampoline_pg_dir,
+			swapper_pg_dir + KERNEL_PGD_BOUNDARY,
+			min_t(unsigned long, KERNEL_PGD_PTRS,
+			      KERNEL_PGD_BOUNDARY));
+#endif
+}
diff --git a/arch/x86/kvm/i8254.c b/arch/x86/kvm/i8254.c
index 0fd6378..ddeb231 100644
--- a/arch/x86/kvm/i8254.c
+++ b/arch/x86/kvm/i8254.c
@@ -697,6 +697,7 @@ struct kvm_pit *kvm_create_pit(struct kvm *kvm, u32 flags)
 	pit->wq = create_singlethread_workqueue("kvm-pit-wq");
 	if (!pit->wq) {
 		mutex_unlock(&pit->pit_state.lock);
+		kvm_free_irq_source_id(kvm, pit->irq_source_id);
 		kfree(pit);
 		return NULL;
 	}
@@ -742,7 +743,7 @@ fail:
 	kvm_unregister_irq_mask_notifier(kvm, 0, &pit->mask_notifier);
 	kvm_unregister_irq_ack_notifier(kvm, &pit_state->irq_ack_notifier);
 	kvm_free_irq_source_id(kvm, pit->irq_source_id);
-
+	destroy_workqueue(pit->wq);
 	kfree(pit);
 	return NULL;
 }
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 25f1907..3a09c62 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -2387,7 +2387,7 @@ static void kvm_vcpu_ioctl_x86_get_xsave(struct kvm_vcpu *vcpu,
 	if (cpu_has_xsave)
 		memcpy(guest_xsave->region,
 			&vcpu->arch.guest_fpu.state->xsave,
-			sizeof(struct xsave_struct));
+			xstate_size);
 	else {
 		memcpy(guest_xsave->region,
 			&vcpu->arch.guest_fpu.state->fxsave,
@@ -2405,7 +2405,7 @@ static int kvm_vcpu_ioctl_x86_set_xsave(struct kvm_vcpu *vcpu,
 
 	if (cpu_has_xsave)
 		memcpy(&vcpu->arch.guest_fpu.state->xsave,
-			guest_xsave->region, sizeof(struct xsave_struct));
+			guest_xsave->region, xstate_size);
 	else {
 		if (xstate_bv & ~XSTATE_FPSSE)
 			return -EINVAL;
diff --git a/arch/x86/xen/platform-pci-unplug.c b/arch/x86/xen/platform-pci-unplug.c
index 554c002..0f45638 100644
--- a/arch/x86/xen/platform-pci-unplug.c
+++ b/arch/x86/xen/platform-pci-unplug.c
@@ -72,13 +72,17 @@ void __init xen_unplug_emulated_devices(void)
 {
 	int r;
 
+	/* user explicitly requested no unplug */
+	if (xen_emul_unplug & XEN_UNPLUG_NEVER)
+		return;
 	/* check the version of the xen platform PCI device */
 	r = check_platform_magic();
 	/* If the version matches enable the Xen platform PCI driver.
-	 * Also enable the Xen platform PCI driver if the version is really old
-	 * and the user told us to ignore it. */
+	 * Also enable the Xen platform PCI driver if the host does
+	 * not support the unplug protocol (XEN_PLATFORM_ERR_MAGIC)
+	 * but the user told us that unplugging is unnecessary. */
 	if (r && !(r == XEN_PLATFORM_ERR_MAGIC &&
-			(xen_emul_unplug & XEN_UNPLUG_IGNORE)))
+			(xen_emul_unplug & XEN_UNPLUG_UNNECESSARY)))
 		return;
 	/* Set the default value of xen_emul_unplug depending on whether or
 	 * not the Xen PV frontends and the Xen platform PCI driver have
@@ -99,7 +103,7 @@ void __init xen_unplug_emulated_devices(void)
 		}
 	}
 	/* Now unplug the emulated devices */
-	if (!(xen_emul_unplug & XEN_UNPLUG_IGNORE))
+	if (!(xen_emul_unplug & XEN_UNPLUG_UNNECESSARY))
 		outw(xen_emul_unplug, XEN_IOPORT_UNPLUG);
 	xen_platform_pci_unplug = xen_emul_unplug;
 }
@@ -125,8 +129,10 @@ static int __init parse_xen_emul_unplug(char *arg)
 			xen_emul_unplug |= XEN_UNPLUG_AUX_IDE_DISKS;
 		else if (!strncmp(p, "nics", l))
 			xen_emul_unplug |= XEN_UNPLUG_ALL_NICS;
-		else if (!strncmp(p, "ignore", l))
-			xen_emul_unplug |= XEN_UNPLUG_IGNORE;
+		else if (!strncmp(p, "unnecessary", l))
+			xen_emul_unplug |= XEN_UNPLUG_UNNECESSARY;
+		else if (!strncmp(p, "never", l))
+			xen_emul_unplug |= XEN_UNPLUG_NEVER;
 		else
 			printk(KERN_WARNING "unrecognised option '%s' "
 				 "in parameter 'xen_emul_unplug'\n", p);
diff --git a/arch/xtensa/kernel/process.c b/arch/xtensa/kernel/process.c
index 7c2f38f..e3558b9 100644
--- a/arch/xtensa/kernel/process.c
+++ b/arch/xtensa/kernel/process.c
@@ -318,8 +318,9 @@ long xtensa_clone(unsigned long clone_flags, unsigned long newsp,
  */
 
 asmlinkage
-long xtensa_execve(const char __user *name, char __user * __user *argv,
-                   char __user * __user *envp,
+long xtensa_execve(const char __user *name,
+		   const char __user *const __user *argv,
+                   const char __user *const __user *envp,
                    long a3, long a4, long a5,
                    struct pt_regs *regs)
 {