Merge branch 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
  [MIPS] Fix sigset_t endianess swapping issues in 32-bit compat code.
  [MIPS] Fix uniprocessor Sibyte builds.
  [MIPS] Make entry.S a little more readable.
  [MIPS] Remove stray instruction from __get_user_asm_ll32.
  [MIPS] 32-bit: Fix warning about cast for fetching pointer from userspace.
  [MIPS] DECstation: Fix irq handling
  [MIPS] signals: make common _BLOCKABLE macro
  [MIPS] signal: Move sigframe definition for native O32/N64 into signal.c
  [MIPS] signal: Move {restore,setup}_sigcontext prototypes to their user
  [MIPS] signal: Fix warnings in o32 compat code.
  [MIPS] IP27: Enable N32 support in defconfig.
  Revert "[MIPS] Fix warning in get_user when fetching pointer object from userspace."
  [MIPS] Don't claim we support dma_declare_coherent_memory - we don't.
  [MIPS] Unify dma-{coherent,noncoherent.ip27,ip32}
  [MIPS] Improve branch prediction in ll/sc atomic operations.

12 files changed, 396 insertions, 34 deletions

diff --git a/include/asm-mips/atomic.h b/include/asm-mips/atomic.h
index c1a2409bb52a..8578869a8bcf 100644
--- a/include/asm-mips/atomic.h
+++ b/include/asm-mips/atomic.h
@@ -69,7 +69,10 @@ static __inline__ void atomic_add(int i, atomic_t * v)
 		"1:	ll	%0, %1		# atomic_add		\n"
 		"	addu	%0, %2					\n"
 		"	sc	%0, %1					\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (temp), "=m" (v->counter)
 		: "Ir" (i), "m" (v->counter));
@@ -111,7 +114,10 @@ static __inline__ void atomic_sub(int i, atomic_t * v)
 		"1:	ll	%0, %1		# atomic_sub		\n"
 		"	subu	%0, %2					\n"
 		"	sc	%0, %1					\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (temp), "=m" (v->counter)
 		: "Ir" (i), "m" (v->counter));
@@ -155,8 +161,11 @@ static __inline__ int atomic_add_return(int i, atomic_t * v)
 		"1:	ll	%1, %2		# atomic_add_return	\n"
 		"	addu	%0, %1, %3				\n"
 		"	sc	%0, %2					\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
 		"	addu	%0, %1, %3				\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (result), "=&r" (temp), "=m" (v->counter)
 		: "Ir" (i), "m" (v->counter)
@@ -204,8 +213,11 @@ static __inline__ int atomic_sub_return(int i, atomic_t * v)
 		"1:	ll	%1, %2		# atomic_sub_return	\n"
 		"	subu	%0, %1, %3				\n"
 		"	sc	%0, %2					\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
 		"	subu	%0, %1, %3				\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (result), "=&r" (temp), "=m" (v->counter)
 		: "Ir" (i), "m" (v->counter)
@@ -267,10 +279,13 @@ static __inline__ int atomic_sub_if_positive(int i, atomic_t * v)
 		"	bltz	%0, 1f					\n"
 		"	sc	%0, %2					\n"
 		"	.set	noreorder				\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
 		"	 subu	%0, %1, %3				\n"
 		"	.set	reorder					\n"
 		"1:							\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (result), "=&r" (temp), "=m" (v->counter)
 		: "Ir" (i), "m" (v->counter)
@@ -429,7 +444,10 @@ static __inline__ void atomic64_add(long i, atomic64_t * v)
 		"1:	lld	%0, %1		# atomic64_add		\n"
 		"	addu	%0, %2					\n"
 		"	scd	%0, %1					\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (temp), "=m" (v->counter)
 		: "Ir" (i), "m" (v->counter));
@@ -471,7 +489,10 @@ static __inline__ void atomic64_sub(long i, atomic64_t * v)
 		"1:	lld	%0, %1		# atomic64_sub		\n"
 		"	subu	%0, %2					\n"
 		"	scd	%0, %1					\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (temp), "=m" (v->counter)
 		: "Ir" (i), "m" (v->counter));
@@ -515,8 +536,11 @@ static __inline__ long atomic64_add_return(long i, atomic64_t * v)
 		"1:	lld	%1, %2		# atomic64_add_return	\n"
 		"	addu	%0, %1, %3				\n"
 		"	scd	%0, %2					\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
 		"	addu	%0, %1, %3				\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (result), "=&r" (temp), "=m" (v->counter)
 		: "Ir" (i), "m" (v->counter)
@@ -564,8 +588,11 @@ static __inline__ long atomic64_sub_return(long i, atomic64_t * v)
 		"1:	lld	%1, %2		# atomic64_sub_return	\n"
 		"	subu	%0, %1, %3				\n"
 		"	scd	%0, %2					\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
 		"	subu	%0, %1, %3				\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (result), "=&r" (temp), "=m" (v->counter)
 		: "Ir" (i), "m" (v->counter)
@@ -627,10 +654,13 @@ static __inline__ long atomic64_sub_if_positive(long i, atomic64_t * v)
 		"	bltz	%0, 1f					\n"
 		"	scd	%0, %2					\n"
 		"	.set	noreorder				\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
 		"	 dsubu	%0, %1, %3				\n"
 		"	.set	reorder					\n"
 		"1:							\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (result), "=&r" (temp), "=m" (v->counter)
 		: "Ir" (i), "m" (v->counter)
diff --git a/include/asm-mips/bitops.h b/include/asm-mips/bitops.h
index 06445de1324b..06c08228a525 100644
--- a/include/asm-mips/bitops.h
+++ b/include/asm-mips/bitops.h
@@ -68,7 +68,10 @@ static inline void set_bit(unsigned long nr, volatile unsigned long *addr)
 		"1:	" __LL "%0, %1			# set_bit	\n"
 		"	or	%0, %2					\n"
 		"	" __SC	"%0, %1					\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (temp), "=m" (*m)
 		: "ir" (1UL << (nr & SZLONG_MASK)), "m" (*m));
@@ -116,7 +119,10 @@ static inline void clear_bit(unsigned long nr, volatile unsigned long *addr)
 		"1:	" __LL "%0, %1			# clear_bit	\n"
 		"	and	%0, %2					\n"
 		"	" __SC "%0, %1					\n"
-		"	beqz	%0, 1b					\n"
+		"	beqz	%0, 2f					\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (temp), "=m" (*m)
 		: "ir" (~(1UL << (nr & SZLONG_MASK))), "m" (*m));
@@ -166,7 +172,10 @@ static inline void change_bit(unsigned long nr, volatile unsigned long *addr)
 		"1:	" __LL "%0, %1		# change_bit	\n"
 		"	xor	%0, %2				\n"
 		"	" __SC	"%0, %1				\n"
-		"	beqz	%0, 1b				\n"
+		"	beqz	%0, 2f				\n"
+		"	.subsection 2				\n"
+		"2:	b	1b				\n"
+		"	.previous				\n"
 		"	.set	mips0				\n"
 		: "=&r" (temp), "=m" (*m)
 		: "ir" (1UL << (nr & SZLONG_MASK)), "m" (*m));
@@ -222,8 +231,12 @@ static inline int test_and_set_bit(unsigned long nr,
 		"1:	" __LL "%0, %1		# test_and_set_bit	\n"
 		"	or	%2, %0, %3				\n"
 		"	" __SC	"%2, %1					\n"
-		"	beqz	%2, 1b					\n"
+		"	beqz	%2, 2f					\n"
 		"	 and	%2, %0, %3				\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	 nop						\n"
+		"	.previous					\n"
 		"	.set	pop					\n"
 		: "=&r" (temp), "=m" (*m), "=&r" (res)
 		: "r" (1UL << (nr & SZLONG_MASK)), "m" (*m)
@@ -290,8 +303,12 @@ static inline int test_and_clear_bit(unsigned long nr,
 		"	or	%2, %0, %3				\n"
 		"	xor	%2, %3					\n"
 		"	" __SC 	"%2, %1					\n"
-		"	beqz	%2, 1b					\n"
+		"	beqz	%2, 2f					\n"
 		"	 and	%2, %0, %3				\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	 nop						\n"
+		"	.previous					\n"
 		"	.set	pop					\n"
 		: "=&r" (temp), "=m" (*m), "=&r" (res)
 		: "r" (1UL << (nr & SZLONG_MASK)), "m" (*m)
@@ -356,8 +373,12 @@ static inline int test_and_change_bit(unsigned long nr,
 		"1:	" __LL	"%0, %1		# test_and_change_bit	\n"
 		"	xor	%2, %0, %3				\n"
 		"	" __SC	"\t%2, %1				\n"
-		"	beqz	%2, 1b					\n"
+		"	beqz	%2, 2f					\n"
 		"	 and	%2, %0, %3				\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	 nop						\n"
+		"	.previous					\n"
 		"	.set	pop					\n"
 		: "=&r" (temp), "=m" (*m), "=&r" (res)
 		: "r" (1UL << (nr & SZLONG_MASK)), "m" (*m)
diff --git a/include/asm-mips/compat-signal.h b/include/asm-mips/compat-signal.h
new file mode 100644
index 000000000000..672077084aa1
--- /dev/null
+++ b/include/asm-mips/compat-signal.h
@@ -0,0 +1,55 @@
+#ifndef __ASM_COMPAT_SIGNAL_H
+#define __ASM_COMPAT_SIGNAL_H
+
+#include <linux/bug.h>
+#include <linux/compat.h>
+#include <linux/compiler.h>
+
+static inline int __copy_conv_sigset_to_user(compat_sigset_t __user *d,
+	const sigset_t *s)
+{
+	int err;
+
+	BUG_ON(sizeof(*d) != sizeof(*s));
+	BUG_ON(_NSIG_WORDS != 2);
+
+	err  = __put_user(s->sig[0],       &d->sig[0]);
+	err |= __put_user(s->sig[0] >> 32, &d->sig[1]);
+	err |= __put_user(s->sig[1],       &d->sig[2]);
+	err |= __put_user(s->sig[1] >> 32, &d->sig[3]);
+
+	return err;
+}
+
+static inline int __copy_conv_sigset_from_user(sigset_t *d,
+	const compat_sigset_t __user *s)
+{
+	int err;
+	union sigset_u {
+		sigset_t	s;
+		compat_sigset_t c;
+	} *u = (union sigset_u *) d;
+
+	BUG_ON(sizeof(*d) != sizeof(*s));
+	BUG_ON(_NSIG_WORDS != 2);
+
+	if (unlikely(!access_ok(VERIFY_READ, d, sizeof(*d))))
+		return -EFAULT;
+
+#ifdef CONFIG_CPU_BIG_ENDIAN
+	err  = __get_user(u->c.sig[1], &s->sig[0]);
+	err |= __get_user(u->c.sig[0], &s->sig[1]);
+	err |= __get_user(u->c.sig[3], &s->sig[2]);
+	err |= __get_user(u->c.sig[2], &s->sig[3]);
+#endif
+#ifdef CONFIG_CPU_LITTLE_ENDIAN
+	err  = __get_user(u->c.sig[0], &s->sig[0]);
+	err |= __get_user(u->c.sig[1], &s->sig[1]);
+	err |= __get_user(u->c.sig[2], &s->sig[2]);
+	err |= __get_user(u->c.sig[3], &s->sig[3]);
+#endif
+
+	return err;
+}
+
+#endif /* __ASM_COMPAT_SIGNAL_H */
diff --git a/include/asm-mips/dma-mapping.h b/include/asm-mips/dma-mapping.h
index 236d1a467cc7..230b3f1b69b1 100644
--- a/include/asm-mips/dma-mapping.h
+++ b/include/asm-mips/dma-mapping.h
@@ -68,6 +68,7 @@ extern int dma_is_consistent(struct device *dev, dma_addr_t dma_addr);
 extern void dma_cache_sync(struct device *dev, void *vaddr, size_t size,
 	       enum dma_data_direction direction);
 
+#if 0
 #define ARCH_HAS_DMA_DECLARE_COHERENT_MEMORY
 
 extern int dma_declare_coherent_memory(struct device *dev, dma_addr_t bus_addr,
@@ -75,5 +76,6 @@ extern int dma_declare_coherent_memory(struct device *dev, dma_addr_t bus_addr,
 extern void dma_release_declared_memory(struct device *dev);
 extern void * dma_mark_declared_memory_occupied(struct device *dev,
 	dma_addr_t device_addr, size_t size);
+#endif
 
 #endif /* _ASM_DMA_MAPPING_H */
diff --git a/include/asm-mips/mach-generic/dma-coherence.h b/include/asm-mips/mach-generic/dma-coherence.h
new file mode 100644
index 000000000000..df71822fd27b
--- /dev/null
+++ b/include/asm-mips/mach-generic/dma-coherence.h
@@ -0,0 +1,43 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2006  Ralf Baechle <ralf@linux-mips.org>
+ *
+ */
+#ifndef __ASM_MACH_GENERIC_DMA_COHERENCE_H
+#define __ASM_MACH_GENERIC_DMA_COHERENCE_H
+
+struct device;
+
+static dma_addr_t plat_map_dma_mem(struct device *dev, void *addr, size_t size)
+{
+	return virt_to_phys(addr);
+}
+
+static dma_addr_t plat_map_dma_mem_page(struct device *dev, struct page *page)
+{
+	return page_to_phys(page);
+}
+
+static unsigned long plat_dma_addr_to_phys(dma_addr_t dma_addr)
+{
+	return dma_addr;
+}
+
+static void plat_unmap_dma_mem(dma_addr_t dma_addr)
+{
+}
+
+static inline int plat_device_is_coherent(struct device *dev)
+{
+#ifdef CONFIG_DMA_COHERENT
+	return 1;
+#endif
+#ifdef CONFIG_DMA_NONCOHERENT
+	return 0;
+#endif
+}
+
+#endif /* __ASM_MACH_GENERIC_DMA_COHERENCE_H */
diff --git a/include/asm-mips/mach-generic/kmalloc.h b/include/asm-mips/mach-generic/kmalloc.h
index 410ab5f6c563..b8e6deba352f 100644
--- a/include/asm-mips/mach-generic/kmalloc.h
+++ b/include/asm-mips/mach-generic/kmalloc.h
@@ -5,6 +5,7 @@
 #ifndef CONFIG_DMA_COHERENT
 /*
  * Total overkill for most systems but need as a safe default.
+ * Set this one if any device in the system might do non-coherent DMA.
  */
 #define ARCH_KMALLOC_MINALIGN	128
 #endif
diff --git a/include/asm-mips/mach-ip27/dma-coherence.h b/include/asm-mips/mach-ip27/dma-coherence.h
new file mode 100644
index 000000000000..659816e200d4
--- /dev/null
+++ b/include/asm-mips/mach-ip27/dma-coherence.h
@@ -0,0 +1,49 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2006  Ralf Baechle <ralf@linux-mips.org>
+ *
+ */
+#ifndef __ASM_MACH_IP27_DMA_COHERENCE_H
+#define __ASM_MACH_IP27_DMA_COHERENCE_H
+
+#include <asm/pci/bridge.h>
+
+#define pdev_to_baddr(pdev, addr) \
+	(BRIDGE_CONTROLLER(pdev->bus)->baddr + (addr))
+#define dev_to_baddr(dev, addr) \
+	pdev_to_baddr(to_pci_dev(dev), (addr))
+
+struct device;
+
+static dma_addr_t plat_map_dma_mem(struct device *dev, void *addr, size_t size)
+{
+	dma_addr_t pa = dev_to_baddr(dev, virt_to_phys(addr));
+
+	return pa;
+}
+
+static dma_addr_t plat_map_dma_mem_page(struct device *dev, struct page *page)
+{
+	dma_addr_t pa = dev_to_baddr(dev, page_to_phys(page));
+
+	return pa;
+}
+
+static unsigned long plat_dma_addr_to_phys(dma_addr_t dma_addr)
+{
+	return dma_addr & (0xffUL << 56);
+}
+
+static void plat_unmap_dma_mem(dma_addr_t dma_addr)
+{
+}
+
+static inline int plat_device_is_coherent(struct device *dev)
+{
+	return 1;		/* IP27 non-cohernet mode is unsupported */
+}
+
+#endif /* __ASM_MACH_IP27_DMA_COHERENCE_H */
diff --git a/include/asm-mips/mach-ip32/dma-coherence.h b/include/asm-mips/mach-ip32/dma-coherence.h
new file mode 100644
index 000000000000..950be17bbb86
--- /dev/null
+++ b/include/asm-mips/mach-ip32/dma-coherence.h
@@ -0,0 +1,71 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2006  Ralf Baechle <ralf@linux-mips.org>
+ *
+ */
+#ifndef __ASM_MACH_IP35_DMA_COHERENCE_H
+#define __ASM_MACH_IP35_DMA_COHERENCE_H
+
+#include <asm/ip32/crime.h>
+
+struct device;
+
+/*
+ * Few notes.
+ * 1. CPU sees memory as two chunks: 0-256M@0x0, and the rest @0x40000000+256M
+ * 2. PCI sees memory as one big chunk @0x0 (or we could use 0x40000000 for
+ *    native-endian)
+ * 3. All other devices see memory as one big chunk at 0x40000000
+ * 4. Non-PCI devices will pass NULL as struct device*
+ *
+ * Thus we translate differently, depending on device.
+ */
+
+#define RAM_OFFSET_MASK 0x3fffffffUL
+
+static dma_addr_t plat_map_dma_mem(struct device *dev, void *addr, size_t size)
+{
+	dma_addr_t pa = virt_to_phys(addr) & RAM_OFFSET_MASK;
+
+	if (dev == NULL)
+		pa += CRIME_HI_MEM_BASE;
+
+	return pa;
+}
+
+static dma_addr_t plat_map_dma_mem_page(struct device *dev, struct page *page)
+{
+	dma_addr_t pa;
+
+	pa = page_to_phys(page) & RAM_OFFSET_MASK;
+
+	if (dev == NULL)
+		pa += CRIME_HI_MEM_BASE;
+
+	return pa;
+}
+
+/* This is almost certainly wrong but it's what dma-ip32.c used to use  */
+static unsigned long plat_dma_addr_to_phys(dma_addr_t dma_addr)
+{
+	unsigned long addr = dma_addr & RAM_OFFSET_MASK;
+
+	if (dma_addr >= 256*1024*1024)
+		addr += CRIME_HI_MEM_BASE;
+
+	return addr;
+}
+
+static void plat_unmap_dma_mem(dma_addr_t dma_addr)
+{
+}
+
+static inline int plat_device_is_coherent(struct device *dev)
+{
+	return 0;		/* IP32 is non-cohernet */
+}
+
+#endif /* __ASM_MACH_IP35_DMA_COHERENCE_H */
diff --git a/include/asm-mips/mach-jazz/dma-coherence.h b/include/asm-mips/mach-jazz/dma-coherence.h
new file mode 100644
index 000000000000..d66979a124a8
--- /dev/null
+++ b/include/asm-mips/mach-jazz/dma-coherence.h
@@ -0,0 +1,40 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2006  Ralf Baechle <ralf@linux-mips.org>
+ */
+#ifndef __ASM_MACH_JAZZ_DMA_COHERENCE_H
+#define __ASM_MACH_JAZZ_DMA_COHERENCE_H
+
+#include <asm/jazzdma.h>
+
+struct device;
+
+static dma_addr_t plat_map_dma_mem(struct device *dev, void *addr, size_t size)
+{
+	return vdma_alloc(virt_to_phys(addr), size);
+}
+
+static dma_addr_t plat_map_dma_mem_page(struct device *dev, struct page *page)
+{
+	return vdma_alloc(page_to_phys(page), PAGE_SIZE);
+}
+
+static unsigned long plat_dma_addr_to_phys(dma_addr_t dma_addr)
+{
+	return vdma_log2phys(dma_addr);
+}
+
+static void plat_unmap_dma_mem(dma_addr_t dma_addr)
+{
+	vdma_free(dma_addr);
+}
+
+static inline int plat_device_is_coherent(struct device *dev)
+{
+	return 0;
+}
+
+#endif /* __ASM_MACH_JAZZ_DMA_COHERENCE_H */
diff --git a/include/asm-mips/spinlock.h b/include/asm-mips/spinlock.h
index fc3217fc1118..f1755d28a36a 100644
--- a/include/asm-mips/spinlock.h
+++ b/include/asm-mips/spinlock.h
@@ -3,7 +3,7 @@
  * License.  See the file "COPYING" in the main directory of this archive
  * for more details.
  *
- * Copyright (C) 1999, 2000, 06 by Ralf Baechle
+ * Copyright (C) 1999, 2000, 06 Ralf Baechle (ralf@linux-mips.org)
  * Copyright (C) 1999, 2000 Silicon Graphics, Inc.
  */
 #ifndef _ASM_SPINLOCK_H
@@ -49,11 +49,18 @@ static inline void __raw_spin_lock(raw_spinlock_t *lock)
 		__asm__ __volatile__(
 		"	.set	noreorder	# __raw_spin_lock	\n"
 		"1:	ll	%1, %2					\n"
-		"	bnez	%1, 1b					\n"
+		"	bnez	%1, 2f					\n"
 		"	 li	%1, 1					\n"
 		"	sc	%1, %0					\n"
-		"	beqz	%1, 1b					\n"
+		"	beqz	%1, 2f					\n"
 		"	 nop						\n"
+		"	.subsection 2					\n"
+		"2:	ll	%1, %2					\n"
+		"	bnez	%1, 2b					\n"
+		"	 li	%1, 1					\n"
+		"	b	1b					\n"
+		"	 nop						\n"
+		"	.previous					\n"
 		"	.set	reorder					\n"
 		: "=m" (lock->lock), "=&r" (tmp)
 		: "m" (lock->lock)
@@ -99,8 +106,12 @@ static inline unsigned int __raw_spin_trylock(raw_spinlock_t *lock)
 		"1:	ll	%0, %3					\n"
 		"	ori	%2, %0, 1				\n"
 		"	sc	%2, %1					\n"
-		"	beqz	%2, 1b					\n"
+		"	beqz	%2, 2f					\n"
 		"	 andi	%2, %0, 1				\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	 nop						\n"
+		"	.previous					\n"
 		"	.set	reorder"
 		: "=&r" (temp), "=m" (lock->lock), "=&r" (res)
 		: "m" (lock->lock)
@@ -154,11 +165,18 @@ static inline void __raw_read_lock(raw_rwlock_t *rw)
 		__asm__ __volatile__(
 		"	.set	noreorder	# __raw_read_lock	\n"
 		"1:	ll	%1, %2					\n"
-		"	bltz	%1, 1b					\n"
+		"	bltz	%1, 2f					\n"
 		"	 addu	%1, 1					\n"
 		"	sc	%1, %0					\n"
 		"	beqz	%1, 1b					\n"
 		"	 nop						\n"
+		"	.subsection 2					\n"
+		"2:	ll	%1, %2					\n"
+		"	bltz	%1, 2b					\n"
+		"	 addu	%1, 1					\n"
+		"	b	1b					\n"
+		"	 nop						\n"
+		"	.previous					\n"
 		"	.set	reorder					\n"
 		: "=m" (rw->lock), "=&r" (tmp)
 		: "m" (rw->lock)
@@ -192,8 +210,12 @@ static inline void __raw_read_unlock(raw_rwlock_t *rw)
 		"1:	ll	%1, %2					\n"
 		"	sub	%1, 1					\n"
 		"	sc	%1, %0					\n"
-		"	beqz	%1, 1b					\n"
+		"	beqz	%1, 2f					\n"
+		"	 nop						\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
 		"	 nop						\n"
+		"	.previous					\n"
 		"	.set	reorder					\n"
 		: "=m" (rw->lock), "=&r" (tmp)
 		: "m" (rw->lock)
@@ -222,11 +244,18 @@ static inline void __raw_write_lock(raw_rwlock_t *rw)
 		__asm__ __volatile__(
 		"	.set	noreorder	# __raw_write_lock	\n"
 		"1:	ll	%1, %2					\n"
-		"	bnez	%1, 1b					\n"
+		"	bnez	%1, 2f					\n"
 		"	 lui	%1, 0x8000				\n"
 		"	sc	%1, %0					\n"
-		"	beqz	%1, 1b					\n"
+		"	beqz	%1, 2f					\n"
+		"	 nop						\n"
+		"	.subsection 2					\n"
+		"2:	ll	%1, %2					\n"
+		"	bnez	%1, 2b					\n"
+		"	 lui	%1, 0x8000				\n"
+		"	b	1b					\n"
 		"	 nop						\n"
+		"	.previous					\n"
 		"	.set	reorder					\n"
 		: "=m" (rw->lock), "=&r" (tmp)
 		: "m" (rw->lock)
@@ -322,12 +351,15 @@ static inline int __raw_write_trylock(raw_rwlock_t *rw)
 		"	bnez	%1, 2f					\n"
 		"	lui	%1, 0x8000				\n"
 		"	sc	%1, %0					\n"
-		"	beqz	%1, 1b					\n"
-		"	 nop						\n"
+		"	beqz	%1, 3f					\n"
+		"	 li	%2, 1					\n"
+		"2:							\n"
 		__WEAK_ORDERING_MB
-		"	li	%2, 1					\n"
+		"	.subsection 2					\n"
+		"3:	b	1b					\n"
+		"	 li	%2, 0					\n"
+		"	.previous					\n"
 		"	.set	reorder					\n"
-		"2:							\n"
 		: "=m" (rw->lock), "=&r" (tmp), "=&r" (ret)
 		: "m" (rw->lock)
 		: "memory");
diff --git a/include/asm-mips/system.h b/include/asm-mips/system.h
index 5e1289c85ed9..597a3743f6a1 100644
--- a/include/asm-mips/system.h
+++ b/include/asm-mips/system.h
@@ -110,7 +110,10 @@ static inline unsigned long __xchg_u32(volatile int * m, unsigned int val)
 		"	move	%2, %z4					\n"
 		"	.set	mips3					\n"
 		"	sc	%2, %1					\n"
-		"	beqz	%2, 1b					\n"
+		"	beqz	%2, 2f					\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (retval), "=m" (*m), "=&r" (dummy)
 		: "R" (*m), "Jr" (val)
@@ -155,7 +158,10 @@ static inline __u64 __xchg_u64(volatile __u64 * m, __u64 val)
 		"1:	lld	%0, %3			# xchg_u64	\n"
 		"	move	%2, %z4					\n"
 		"	scd	%2, %1					\n"
-		"	beqz	%2, 1b					\n"
+		"	beqz	%2, 2f					\n"
+		"	.subsection 2					\n"
+		"2:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	mips0					\n"
 		: "=&r" (retval), "=m" (*m), "=&r" (dummy)
 		: "R" (*m), "Jr" (val)
@@ -232,8 +238,11 @@ static inline unsigned long __cmpxchg_u32(volatile int * m, unsigned long old,
 		"	move	$1, %z4					\n"
 		"	.set	mips3					\n"
 		"	sc	$1, %1					\n"
-		"	beqz	$1, 1b					\n"
+		"	beqz	$1, 3f					\n"
 		"2:							\n"
+		"	.subsection 2					\n"
+		"3:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	pop					\n"
 		: "=&r" (retval), "=R" (*m)
 		: "R" (*m), "Jr" (old), "Jr" (new)
@@ -283,8 +292,11 @@ static inline unsigned long __cmpxchg_u64(volatile int * m, unsigned long old,
 		"	bne	%0, %z3, 2f				\n"
 		"	move	$1, %z4					\n"
 		"	scd	$1, %1					\n"
-		"	beqz	$1, 1b					\n"
+		"	beqz	$1, 3f					\n"
 		"2:							\n"
+		"	.subsection 2					\n"
+		"3:	b	1b					\n"
+		"	.previous					\n"
 		"	.set	pop					\n"
 		: "=&r" (retval), "=R" (*m)
 		: "R" (*m), "Jr" (old), "Jr" (new)
diff --git a/include/asm-mips/uaccess.h b/include/asm-mips/uaccess.h
index 825fcbd9eabd..3eff8d8fe28a 100644
--- a/include/asm-mips/uaccess.h
+++ b/include/asm-mips/uaccess.h
@@ -265,10 +265,14 @@ do {									\
  */
 #define __get_user_asm_ll32(val, addr)					\
 {									\
+	union {								\
+		unsigned long long	l;				\
+		__typeof__(*(addr))	t;				\
+	} __gu_tmp;							\
+									\
 	__asm__ __volatile__(						\
 	"1:	lw	%1, (%3)				\n"	\
 	"2:	lw	%D1, 4(%3)				\n"	\
-	"	move	%0, $0					\n"	\
 	"3:	.section	.fixup,\"ax\"			\n"	\
 	"4:	li	%0, %4					\n"	\
 	"	move	%1, $0					\n"	\
@@ -279,8 +283,10 @@ do {									\
 	"	" __UA_ADDR "	1b, 4b				\n"	\
 	"	" __UA_ADDR "	2b, 4b				\n"	\
 	"	.previous					\n"	\
-	: "=r" (__gu_err), "=&r" (val)					\
+	: "=r" (__gu_err), "=&r" (__gu_tmp.l)				\
 	: "0" (0), "r" (addr), "i" (-EFAULT));				\
+									\
+	(val) = __gu_tmp.t;						\
 }
 
 /*