[PATCH] spinlock consolidation

This patch (written by me and also containing many suggestions of Arjan van
de Ven) does a major cleanup of the spinlock code.  It does the following
things:

 - consolidates and enhances the spinlock/rwlock debugging code

 - simplifies the asm/spinlock.h files

 - encapsulates the raw spinlock type and moves generic spinlock
   features (such as ->break_lock) into the generic code.

 - cleans up the spinlock code hierarchy to get rid of the spaghetti.

Most notably there's now only a single variant of the debugging code,
located in lib/spinlock_debug.c.  (previously we had one SMP debugging
variant per architecture, plus a separate generic one for UP builds)

Also, i've enhanced the rwlock debugging facility, it will now track
write-owners.  There is new spinlock-owner/CPU-tracking on SMP builds too.
All locks have lockup detection now, which will work for both soft and hard
spin/rwlock lockups.

The arch-level include files now only contain the minimally necessary
subset of the spinlock code - all the rest that can be generalized now
lives in the generic headers:

 include/asm-i386/spinlock_types.h       |   16
 include/asm-x86_64/spinlock_types.h     |   16

I have also split up the various spinlock variants into separate files,
making it easier to see which does what. The new layout is:

   SMP                         |  UP
   ----------------------------|-----------------------------------
   asm/spinlock_types_smp.h    |  linux/spinlock_types_up.h
   linux/spinlock_types.h      |  linux/spinlock_types.h
   asm/spinlock_smp.h          |  linux/spinlock_up.h
   linux/spinlock_api_smp.h    |  linux/spinlock_api_up.h
   linux/spinlock.h            |  linux/spinlock.h

/*
 * here's the role of the various spinlock/rwlock related include files:
 *
 * on SMP builds:
 *
 *  asm/spinlock_types.h: contains the raw_spinlock_t/raw_rwlock_t and the
 *                        initializers
 *
 *  linux/spinlock_types.h:
 *                        defines the generic type and initializers
 *
 *  asm/spinlock.h:       contains the __raw_spin_*()/etc. lowlevel
 *                        implementations, mostly inline assembly code
 *
 *   (also included on UP-debug builds:)
 *
 *  linux/spinlock_api_smp.h:
 *                        contains the prototypes for the _spin_*() APIs.
 *
 *  linux/spinlock.h:     builds the final spin_*() APIs.
 *
 * on UP builds:
 *
 *  linux/spinlock_type_up.h:
 *                        contains the generic, simplified UP spinlock type.
 *                        (which is an empty structure on non-debug builds)
 *
 *  linux/spinlock_types.h:
 *                        defines the generic type and initializers
 *
 *  linux/spinlock_up.h:
 *                        contains the __raw_spin_*()/etc. version of UP
 *                        builds. (which are NOPs on non-debug, non-preempt
 *                        builds)
 *
 *   (included on UP-non-debug builds:)
 *
 *  linux/spinlock_api_up.h:
 *                        builds the _spin_*() APIs.
 *
 *  linux/spinlock.h:     builds the final spin_*() APIs.
 */

All SMP and UP architectures are converted by this patch.

arm, i386, ia64, ppc, ppc64, s390/s390x, x64 was build-tested via
crosscompilers.  m32r, mips, sh, sparc, have not been tested yet, but should
be mostly fine.

From: Grant Grundler <grundler@parisc-linux.org>

  Booted and lightly tested on a500-44 (64-bit, SMP kernel, dual CPU).
  Builds 32-bit SMP kernel (not booted or tested).  I did not try to build
  non-SMP kernels.  That should be trivial to fix up later if necessary.

  I converted bit ops atomic_hash lock to raw_spinlock_t.  Doing so avoids
  some ugly nesting of linux/*.h and asm/*.h files.  Those particular locks
  are well tested and contained entirely inside arch specific code.  I do NOT
  expect any new issues to arise with them.

 If someone does ever need to use debug/metrics with them, then they will
  need to unravel this hairball between spinlocks, atomic ops, and bit ops
  that exist only because parisc has exactly one atomic instruction: LDCW
  (load and clear word).

From: "Luck, Tony" <tony.luck@intel.com>

   ia64 fix

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Arjan van de Ven <arjanv@infradead.org>
Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
Cc: Matthew Wilcox <willy@debian.org>
Signed-off-by: Hirokazu Takata <takata@linux-m32r.org>
Signed-off-by: Mikael Pettersson <mikpe@csd.uu.se>
Signed-off-by: Benoit Boissinot <benoit.boissinot@ens-lyon.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

41 files changed, 1289 insertions, 1723 deletions

diff --git a/include/asm-alpha/spinlock.h b/include/asm-alpha/spinlock.h
index 80780dba998..8197c69eff4 100644
--- a/include/asm-alpha/spinlock.h
+++ b/include/asm-alpha/spinlock.h
@@ -6,7 +6,6 @@
 #include <linux/kernel.h>
 #include <asm/current.h>
 
-
 /*
  * Simple spin lock operations.  There are two variants, one clears IRQ's
  * on the local processor, one does not.
@@ -14,43 +13,18 @@
  * We make no fairness assumptions. They have a cost.
  */
 
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_DEBUG_SPINLOCK
-	int on_cpu;
-	int line_no;
-	void *previous;
-	struct task_struct * task;
-	const char *base_file;
-#endif
-} spinlock_t;
-
-#ifdef CONFIG_DEBUG_SPINLOCK
-#define SPIN_LOCK_UNLOCKED	(spinlock_t){ 0, -1, 0, NULL, NULL, NULL }
-#else
-#define SPIN_LOCK_UNLOCKED	(spinlock_t){ 0 }
-#endif
-
-#define spin_lock_init(x)	do { *(x) = SPIN_LOCK_UNLOCKED; } while(0)
-#define spin_is_locked(x)	((x)->lock != 0)
-#define spin_unlock_wait(x)	do { barrier(); } while ((x)->lock)
-
-#ifdef CONFIG_DEBUG_SPINLOCK
-extern void _raw_spin_unlock(spinlock_t * lock);
-extern void debug_spin_lock(spinlock_t * lock, const char *, int);
-extern int debug_spin_trylock(spinlock_t * lock, const char *, int);
-#define _raw_spin_lock(LOCK) \
-	debug_spin_lock(LOCK, __BASE_FILE__, __LINE__)
-#define _raw_spin_trylock(LOCK) \
-	debug_spin_trylock(LOCK, __BASE_FILE__, __LINE__)
-#else
-static inline void _raw_spin_unlock(spinlock_t * lock)
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
+#define __raw_spin_is_locked(x)	((x)->lock != 0)
+#define __raw_spin_unlock_wait(x) \
+		do { cpu_relax(); } while ((x)->lock)
+
+static inline void __raw_spin_unlock(raw_spinlock_t * lock)
 {
 	mb();
 	lock->lock = 0;
 }
 
-static inline void _raw_spin_lock(spinlock_t * lock)
+static inline void __raw_spin_lock(raw_spinlock_t * lock)
 {
 	long tmp;
 
@@ -70,80 +44,64 @@ static inline void _raw_spin_lock(spinlock_t * lock)
 	: "m"(lock->lock) : "memory");
 }
 
-static inline int _raw_spin_trylock(spinlock_t *lock)
+static inline int __raw_spin_trylock(raw_spinlock_t *lock)
 {
 	return !test_and_set_bit(0, &lock->lock);
 }
-#endif /* CONFIG_DEBUG_SPINLOCK */
-
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
 
 /***********************************************************/
 
-typedef struct {
-	volatile unsigned int lock;
-} rwlock_t;
-
-#define RW_LOCK_UNLOCKED	(rwlock_t){ 0 }
-
-#define rwlock_init(x)		do { *(x) = RW_LOCK_UNLOCKED; } while(0)
-
-static inline int read_can_lock(rwlock_t *lock)
+static inline int __raw_read_can_lock(raw_rwlock_t *lock)
 {
 	return (lock->lock & 1) == 0;
 }
 
-static inline int write_can_lock(rwlock_t *lock)
+static inline int __raw_write_can_lock(raw_rwlock_t *lock)
 {
 	return lock->lock == 0;
 }
 
-#ifdef CONFIG_DEBUG_RWLOCK
-extern void _raw_write_lock(rwlock_t * lock);
-extern void _raw_read_lock(rwlock_t * lock);
-#else
-static inline void _raw_write_lock(rwlock_t * lock)
+static inline void __raw_read_lock(raw_rwlock_t *lock)
 {
 	long regx;
 
 	__asm__ __volatile__(
 	"1:	ldl_l	%1,%0\n"
-	"	bne	%1,6f\n"
-	"	lda	%1,1\n"
+	"	blbs	%1,6f\n"
+	"	subl	%1,2,%1\n"
 	"	stl_c	%1,%0\n"
 	"	beq	%1,6f\n"
 	"	mb\n"
 	".subsection 2\n"
 	"6:	ldl	%1,%0\n"
-	"	bne	%1,6b\n"
+	"	blbs	%1,6b\n"
 	"	br	1b\n"
 	".previous"
 	: "=m" (*lock), "=&r" (regx)
 	: "m" (*lock) : "memory");
 }
 
-static inline void _raw_read_lock(rwlock_t * lock)
+static inline void __raw_write_lock(raw_rwlock_t *lock)
 {
 	long regx;
 
 	__asm__ __volatile__(
 	"1:	ldl_l	%1,%0\n"
-	"	blbs	%1,6f\n"
-	"	subl	%1,2,%1\n"
+	"	bne	%1,6f\n"
+	"	lda	%1,1\n"
 	"	stl_c	%1,%0\n"
 	"	beq	%1,6f\n"
 	"	mb\n"
 	".subsection 2\n"
 	"6:	ldl	%1,%0\n"
-	"	blbs	%1,6b\n"
+	"	bne	%1,6b\n"
 	"	br	1b\n"
 	".previous"
 	: "=m" (*lock), "=&r" (regx)
 	: "m" (*lock) : "memory");
 }
-#endif /* CONFIG_DEBUG_RWLOCK */
 
-static inline int _raw_read_trylock(rwlock_t * lock)
+static inline int __raw_read_trylock(raw_rwlock_t * lock)
 {
 	long regx;
 	int success;
@@ -165,7 +123,7 @@ static inline int _raw_read_trylock(rwlock_t * lock)
 	return success;
 }
 
-static inline int _raw_write_trylock(rwlock_t * lock)
+static inline int __raw_write_trylock(raw_rwlock_t * lock)
 {
 	long regx;
 	int success;
@@ -187,13 +145,7 @@ static inline int _raw_write_trylock(rwlock_t * lock)
 	return success;
 }
 
-static inline void _raw_write_unlock(rwlock_t * lock)
-{
-	mb();
-	lock->lock = 0;
-}
-
-static inline void _raw_read_unlock(rwlock_t * lock)
+static inline void __raw_read_unlock(raw_rwlock_t * lock)
 {
 	long regx;
 	__asm__ __volatile__(
@@ -209,4 +161,10 @@ static inline void _raw_read_unlock(rwlock_t * lock)
 	: "m" (*lock) : "memory");
 }
 
+static inline void __raw_write_unlock(raw_rwlock_t * lock)
+{
+	mb();
+	lock->lock = 0;
+}
+
 #endif /* _ALPHA_SPINLOCK_H */
diff --git a/include/asm-alpha/spinlock_types.h b/include/asm-alpha/spinlock_types.h
new file mode 100644
index 00000000000..8141eb5ebf0
--- /dev/null
+++ b/include/asm-alpha/spinlock_types.h
@@ -0,0 +1,20 @@
+#ifndef _ALPHA_SPINLOCK_TYPES_H
+#define _ALPHA_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned int lock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 0 }
+
+typedef struct {
+	volatile unsigned int lock;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ 0 }
+
+#endif
diff --git a/include/asm-arm/spinlock.h b/include/asm-arm/spinlock.h
index 1f906d09b68..cb4906b4555 100644
--- a/include/asm-arm/spinlock.h
+++ b/include/asm-arm/spinlock.h
@@ -16,21 +16,14 @@
  * Unlocked value: 0
  * Locked value: 1
  */
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
 
-#define SPIN_LOCK_UNLOCKED	(spinlock_t) { 0 }
+#define __raw_spin_is_locked(x)		((x)->lock != 0)
+#define __raw_spin_unlock_wait(lock) \
+	do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
 
-#define spin_lock_init(x)	do { *(x) = SPIN_LOCK_UNLOCKED; } while (0)
-#define spin_is_locked(x)	((x)->lock != 0)
-#define spin_unlock_wait(x)	do { barrier(); } while (spin_is_locked(x))
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
 
-static inline void _raw_spin_lock(spinlock_t *lock)
+static inline void __raw_spin_lock(raw_spinlock_t *lock)
 {
 	unsigned long tmp;
 
@@ -47,7 +40,7 @@ static inline void _raw_spin_lock(spinlock_t *lock)
 	smp_mb();
 }
 
-static inline int _raw_spin_trylock(spinlock_t *lock)
+static inline int __raw_spin_trylock(raw_spinlock_t *lock)
 {
 	unsigned long tmp;
 
@@ -67,7 +60,7 @@ static inline int _raw_spin_trylock(spinlock_t *lock)
 	}
 }
 
-static inline void _raw_spin_unlock(spinlock_t *lock)
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
 {
 	smp_mb();
 
@@ -80,23 +73,14 @@ static inline void _raw_spin_unlock(spinlock_t *lock)
 
 /*
  * RWLOCKS
- */
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-
-#define RW_LOCK_UNLOCKED	(rwlock_t) { 0 }
-#define rwlock_init(x)		do { *(x) = RW_LOCK_UNLOCKED; } while (0)
-#define rwlock_is_locked(x)	(*((volatile unsigned int *)(x)) != 0)
-
-/*
+ *
+ *
  * Write locks are easy - we just set bit 31.  When unlocking, we can
  * just write zero since the lock is exclusively held.
  */
-static inline void _raw_write_lock(rwlock_t *rw)
+#define rwlock_is_locked(x)	(*((volatile unsigned int *)(x)) != 0)
+
+static inline void __raw_write_lock(rwlock_t *rw)
 {
 	unsigned long tmp;
 
@@ -113,7 +97,7 @@ static inline void _raw_write_lock(rwlock_t *rw)
 	smp_mb();
 }
 
-static inline int _raw_write_trylock(rwlock_t *rw)
+static inline int __raw_write_trylock(rwlock_t *rw)
 {
 	unsigned long tmp;
 
@@ -133,7 +117,7 @@ static inline int _raw_write_trylock(rwlock_t *rw)
 	}
 }
 
-static inline void _raw_write_unlock(rwlock_t *rw)
+static inline void __raw_write_unlock(raw_rwlock_t *rw)
 {
 	smp_mb();
 
@@ -156,7 +140,7 @@ static inline void _raw_write_unlock(rwlock_t *rw)
  * currently active.  However, we know we won't have any write
  * locks.
  */
-static inline void _raw_read_lock(rwlock_t *rw)
+static inline void __raw_read_lock(raw_rwlock_t *rw)
 {
 	unsigned long tmp, tmp2;
 
@@ -173,7 +157,7 @@ static inline void _raw_read_lock(rwlock_t *rw)
 	smp_mb();
 }
 
-static inline void _raw_read_unlock(rwlock_t *rw)
+static inline void __raw_read_unlock(rwlock_t *rw)
 {
 	unsigned long tmp, tmp2;
 
@@ -190,6 +174,6 @@ static inline void _raw_read_unlock(rwlock_t *rw)
 	: "cc");
 }
 
-#define _raw_read_trylock(lock) generic_raw_read_trylock(lock)
+#define __raw_read_trylock(lock) generic__raw_read_trylock(lock)
 
 #endif /* __ASM_SPINLOCK_H */
diff --git a/include/asm-arm/spinlock_types.h b/include/asm-arm/spinlock_types.h
new file mode 100644
index 00000000000..43e83f6d2ee
--- /dev/null
+++ b/include/asm-arm/spinlock_types.h
@@ -0,0 +1,20 @@
+#ifndef __ASM_SPINLOCK_TYPES_H
+#define __ASM_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned int lock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 0 }
+
+typedef struct {
+	volatile unsigned int lock;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ 0 }
+
+#endif
diff --git a/include/asm-i386/spinlock.h b/include/asm-i386/spinlock.h
index f9ff31f4003..23604350cdf 100644
--- a/include/asm-i386/spinlock.h
+++ b/include/asm-i386/spinlock.h
@@ -7,46 +7,21 @@
 #include <linux/config.h>
 #include <linux/compiler.h>
 
-asmlinkage int printk(const char * fmt, ...)
-	__attribute__ ((format (printf, 1, 2)));
-
 /*
  * Your basic SMP spinlocks, allowing only a single CPU anywhere
- */
-
-typedef struct {
-	volatile unsigned int slock;
-#ifdef CONFIG_DEBUG_SPINLOCK
-	unsigned magic;
-#endif
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
-
-#define SPINLOCK_MAGIC	0xdead4ead
-
-#ifdef CONFIG_DEBUG_SPINLOCK
-#define SPINLOCK_MAGIC_INIT	, SPINLOCK_MAGIC
-#else
-#define SPINLOCK_MAGIC_INIT	/* */
-#endif
-
-#define SPIN_LOCK_UNLOCKED (spinlock_t) { 1 SPINLOCK_MAGIC_INIT }
-
-#define spin_lock_init(x)	do { *(x) = SPIN_LOCK_UNLOCKED; } while(0)
-
-/*
+ *
  * Simple spin lock operations.  There are two variants, one clears IRQ's
  * on the local processor, one does not.
  *
  * We make no fairness assumptions. They have a cost.
+ *
+ * (the type definitions are in asm/spinlock_types.h)
  */
 
-#define spin_is_locked(x)	(*(volatile signed char *)(&(x)->slock) <= 0)
-#define spin_unlock_wait(x)	do { barrier(); } while(spin_is_locked(x))
+#define __raw_spin_is_locked(x) \
+		(*(volatile signed char *)(&(x)->slock) <= 0)
 
-#define spin_lock_string \
+#define __raw_spin_lock_string \
 	"\n1:\t" \
 	"lock ; decb %0\n\t" \
 	"jns 3f\n" \
@@ -57,7 +32,7 @@ typedef struct {
 	"jmp 1b\n" \
 	"3:\n\t"
 
-#define spin_lock_string_flags \
+#define __raw_spin_lock_string_flags \
 	"\n1:\t" \
 	"lock ; decb %0\n\t" \
 	"jns 4f\n\t" \
@@ -73,86 +48,71 @@ typedef struct {
 	"jmp 1b\n" \
 	"4:\n\t"
 
+static inline void __raw_spin_lock(raw_spinlock_t *lock)
+{
+	__asm__ __volatile__(
+		__raw_spin_lock_string
+		:"=m" (lock->slock) : : "memory");
+}
+
+static inline void __raw_spin_lock_flags(raw_spinlock_t *lock, unsigned long flags)
+{
+	__asm__ __volatile__(
+		__raw_spin_lock_string_flags
+		:"=m" (lock->slock) : "r" (flags) : "memory");
+}
+
+static inline int __raw_spin_trylock(raw_spinlock_t *lock)
+{
+	char oldval;
+	__asm__ __volatile__(
+		"xchgb %b0,%1"
+		:"=q" (oldval), "=m" (lock->slock)
+		:"0" (0) : "memory");
+	return oldval > 0;
+}
+
 /*
- * This works. Despite all the confusion.
- * (except on PPro SMP or if we are using OOSTORE)
+ * __raw_spin_unlock based on writing $1 to the low byte.
+ * This method works. Despite all the confusion.
+ * (except on PPro SMP or if we are using OOSTORE, so we use xchgb there)
  * (PPro errata 66, 92)
  */
 
 #if !defined(CONFIG_X86_OOSTORE) && !defined(CONFIG_X86_PPRO_FENCE)
 
-#define spin_unlock_string \
+#define __raw_spin_unlock_string \
 	"movb $1,%0" \
 		:"=m" (lock->slock) : : "memory"
 
 
-static inline void _raw_spin_unlock(spinlock_t *lock)
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
 {
-#ifdef CONFIG_DEBUG_SPINLOCK
-	BUG_ON(lock->magic != SPINLOCK_MAGIC);
-	BUG_ON(!spin_is_locked(lock));
-#endif
 	__asm__ __volatile__(
-		spin_unlock_string
+		__raw_spin_unlock_string
 	);
 }
 
 #else
 
-#define spin_unlock_string \
+#define __raw_spin_unlock_string \
 	"xchgb %b0, %1" \
 		:"=q" (oldval), "=m" (lock->slock) \
 		:"0" (oldval) : "memory"
 
-static inline void _raw_spin_unlock(spinlock_t *lock)
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
 {
 	char oldval = 1;
-#ifdef CONFIG_DEBUG_SPINLOCK
-	BUG_ON(lock->magic != SPINLOCK_MAGIC);
-	BUG_ON(!spin_is_locked(lock));
-#endif
-	__asm__ __volatile__(
-		spin_unlock_string
-	);
-}
 
-#endif
-
-static inline int _raw_spin_trylock(spinlock_t *lock)
-{
-	char oldval;
 	__asm__ __volatile__(
-		"xchgb %b0,%1"
-		:"=q" (oldval), "=m" (lock->slock)
-		:"0" (0) : "memory");
-	return oldval > 0;
+		__raw_spin_unlock_string
+	);
 }
 
-static inline void _raw_spin_lock(spinlock_t *lock)
-{
-#ifdef CONFIG_DEBUG_SPINLOCK
-	if (unlikely(lock->magic != SPINLOCK_MAGIC)) {
-		printk("eip: %p\n", __builtin_return_address(0));
-		BUG();
-	}
 #endif
-	__asm__ __volatile__(
-		spin_lock_string
-		:"=m" (lock->slock) : : "memory");
-}
 
-static inline void _raw_spin_lock_flags (spinlock_t *lock, unsigned long flags)
-{
-#ifdef CONFIG_DEBUG_SPINLOCK
-	if (unlikely(lock->magic != SPINLOCK_MAGIC)) {
-		printk("eip: %p\n", __builtin_return_address(0));
-		BUG();
-	}
-#endif
-	__asm__ __volatile__(
-		spin_lock_string_flags
-		:"=m" (lock->slock) : "r" (flags) : "memory");
-}
+#define __raw_spin_unlock_wait(lock) \
+	do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
 
 /*
  * Read-write spinlocks, allowing multiple readers
@@ -163,72 +123,41 @@ static inline void _raw_spin_lock_flags (spinlock_t *lock, unsigned long flags)
  * can "mix" irq-safe locks - any writer needs to get a
  * irq-safe write-lock, but readers can get non-irqsafe
  * read-locks.
+ *
+ * On x86, we implement read-write locks as a 32-bit counter
+ * with the high bit (sign) being the "contended" bit.
+ *
+ * The inline assembly is non-obvious. Think about it.
+ *
+ * Changed to use the same technique as rw semaphores.  See
+ * semaphore.h for details.  -ben
+ *
+ * the helpers are in arch/i386/kernel/semaphore.c
  */
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_DEBUG_SPINLOCK
-	unsigned magic;
-#endif
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-
-#define RWLOCK_MAGIC	0xdeaf1eed
-
-#ifdef CONFIG_DEBUG_SPINLOCK
-#define RWLOCK_MAGIC_INIT	, RWLOCK_MAGIC
-#else
-#define RWLOCK_MAGIC_INIT	/* */
-#endif
-
-#define RW_LOCK_UNLOCKED (rwlock_t) { RW_LOCK_BIAS RWLOCK_MAGIC_INIT }
-
-#define rwlock_init(x)	do { *(x) = RW_LOCK_UNLOCKED; } while(0)
 
 /**
  * read_can_lock - would read_trylock() succeed?
  * @lock: the rwlock in question.
  */
-#define read_can_lock(x) ((int)(x)->lock > 0)
+#define __raw_read_can_lock(x)		((int)(x)->lock > 0)
 
 /**
  * write_can_lock - would write_trylock() succeed?
  * @lock: the rwlock in question.
  */
-#define write_can_lock(x) ((x)->lock == RW_LOCK_BIAS)
+#define __raw_write_can_lock(x)		((x)->lock == RW_LOCK_BIAS)
 
-/*
- * On x86, we implement read-write locks as a 32-bit counter
- * with the high bit (sign) being the "contended" bit.
- *
- * The inline assembly is non-obvious. Think about it.
- *
- * Changed to use the same technique as rw semaphores.  See
- * semaphore.h for details.  -ben
- */
-/* the spinlock helpers are in arch/i386/kernel/semaphore.c */
-
-static inline void _raw_read_lock(rwlock_t *rw)
+static inline void __raw_read_lock(raw_rwlock_t *rw)
 {
-#ifdef CONFIG_DEBUG_SPINLOCK
-	BUG_ON(rw->magic != RWLOCK_MAGIC);
-#endif
 	__build_read_lock(rw, "__read_lock_failed");
 }
 
-static inline void _raw_write_lock(rwlock_t *rw)
+static inline void __raw_write_lock(raw_rwlock_t *rw)
 {
-#ifdef CONFIG_DEBUG_SPINLOCK
-	BUG_ON(rw->magic != RWLOCK_MAGIC);
-#endif
 	__build_write_lock(rw, "__write_lock_failed");
 }
 
-#define _raw_read_unlock(rw)		asm volatile("lock ; incl %0" :"=m" ((rw)->lock) : : "memory")
-#define _raw_write_unlock(rw)	asm volatile("lock ; addl $" RW_LOCK_BIAS_STR ",%0":"=m" ((rw)->lock) : : "memory")
-
-static inline int _raw_read_trylock(rwlock_t *lock)
+static inline int __raw_read_trylock(raw_rwlock_t *lock)
 {
 	atomic_t *count = (atomic_t *)lock;
 	atomic_dec(count);
@@ -238,7 +167,7 @@ static inline int _raw_read_trylock(rwlock_t *lock)
 	return 0;
 }
 
-static inline int _raw_write_trylock(rwlock_t *lock)
+static inline int __raw_write_trylock(raw_rwlock_t *lock)
 {
 	atomic_t *count = (atomic_t *)lock;
 	if (atomic_sub_and_test(RW_LOCK_BIAS, count))
@@ -247,4 +176,15 @@ static inline int _raw_write_trylock(rwlock_t *lock)
 	return 0;
 }
 
+static inline void __raw_read_unlock(raw_rwlock_t *rw)
+{
+	asm volatile("lock ; incl %0" :"=m" (rw->lock) : : "memory");
+}
+
+static inline void __raw_write_unlock(raw_rwlock_t *rw)
+{
+	asm volatile("lock ; addl $" RW_LOCK_BIAS_STR ", %0"
+				 : "=m" (rw->lock) : : "memory");
+}
+
 #endif /* __ASM_SPINLOCK_H */
diff --git a/include/asm-i386/spinlock_types.h b/include/asm-i386/spinlock_types.h
new file mode 100644
index 00000000000..59efe849f35
--- /dev/null
+++ b/include/asm-i386/spinlock_types.h
@@ -0,0 +1,20 @@
+#ifndef __ASM_SPINLOCK_TYPES_H
+#define __ASM_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned int slock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 1 }
+
+typedef struct {
+	volatile unsigned int lock;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ RW_LOCK_BIAS }
+
+#endif
diff --git a/include/asm-ia64/spinlock.h b/include/asm-ia64/spinlock.h
index d2430aa0d49..5b78611411c 100644
--- a/include/asm-ia64/spinlock.h
+++ b/include/asm-ia64/spinlock.h
@@ -17,28 +17,20 @@
 #include <asm/intrinsics.h>
 #include <asm/system.h>
 
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
-
-#define SPIN_LOCK_UNLOCKED			(spinlock_t) { 0 }
-#define spin_lock_init(x)			((x)->lock = 0)
+#define __raw_spin_lock_init(x)			((x)->lock = 0)
 
 #ifdef ASM_SUPPORTED
 /*
  * Try to get the lock.  If we fail to get the lock, make a non-standard call to
  * ia64_spinlock_contention().  We do not use a normal call because that would force all
- * callers of spin_lock() to be non-leaf routines.  Instead, ia64_spinlock_contention() is
- * carefully coded to touch only those registers that spin_lock() marks "clobbered".
+ * callers of __raw_spin_lock() to be non-leaf routines.  Instead, ia64_spinlock_contention() is
+ * carefully coded to touch only those registers that __raw_spin_lock() marks "clobbered".
  */
 
 #define IA64_SPINLOCK_CLOBBERS "ar.ccv", "ar.pfs", "p14", "p15", "r27", "r28", "r29", "r30", "b6", "memory"
 
 static inline void
-_raw_spin_lock_flags (spinlock_t *lock, unsigned long flags)
+__raw_spin_lock_flags (raw_spinlock_t *lock, unsigned long flags)
 {
 	register volatile unsigned int *ptr asm ("r31") = &lock->lock;
 
@@ -94,17 +86,17 @@ _raw_spin_lock_flags (spinlock_t *lock, unsigned long flags)
 #endif
 }
 
-#define _raw_spin_lock(lock) _raw_spin_lock_flags(lock, 0)
+#define __raw_spin_lock(lock) __raw_spin_lock_flags(lock, 0)
 
 /* Unlock by doing an ordered store and releasing the cacheline with nta */
-static inline void _raw_spin_unlock(spinlock_t *x) {
+static inline void __raw_spin_unlock(raw_spinlock_t *x) {
 	barrier();
 	asm volatile ("st4.rel.nta [%0] = r0\n\t" :: "r"(x));
 }
 
 #else /* !ASM_SUPPORTED */
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
-# define _raw_spin_lock(x)								\
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
+# define __raw_spin_lock(x)								\
 do {											\
 	__u32 *ia64_spinlock_ptr = (__u32 *) (x);					\
 	__u64 ia64_spinlock_val;							\
@@ -117,29 +109,20 @@ do {											\
 		} while (ia64_spinlock_val);						\
 	}										\
 } while (0)
-#define _raw_spin_unlock(x)	do { barrier(); ((spinlock_t *) x)->lock = 0; } while (0)
+#define __raw_spin_unlock(x)	do { barrier(); ((raw_spinlock_t *) x)->lock = 0; } while (0)
 #endif /* !ASM_SUPPORTED */
 
-#define spin_is_locked(x)	((x)->lock != 0)
-#define _raw_spin_trylock(x)	(cmpxchg_acq(&(x)->lock, 0, 1) == 0)
-#define spin_unlock_wait(x)	do { barrier(); } while ((x)->lock)
-
-typedef struct {
-	volatile unsigned int read_counter	: 24;
-	volatile unsigned int write_lock	:  8;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-#define RW_LOCK_UNLOCKED (rwlock_t) { 0, 0 }
+#define __raw_spin_is_locked(x)		((x)->lock != 0)
+#define __raw_spin_trylock(x)		(cmpxchg_acq(&(x)->lock, 0, 1) == 0)
+#define __raw_spin_unlock_wait(lock) \
+	do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
 
-#define rwlock_init(x)		do { *(x) = RW_LOCK_UNLOCKED; } while(0)
-#define read_can_lock(rw)	(*(volatile int *)(rw) >= 0)
-#define write_can_lock(rw)	(*(volatile int *)(rw) == 0)
+#define __raw_read_can_lock(rw)		(*(volatile int *)(rw) >= 0)
+#define __raw_write_can_lock(rw)	(*(volatile int *)(rw) == 0)
 
-#define _raw_read_lock(rw)								\
+#define __raw_read_lock(rw)								\
 do {											\
-	rwlock_t *__read_lock_ptr = (rw);						\
+	raw_rwlock_t *__read_lock_ptr = (rw);						\
 											\
 	while (unlikely(ia64_fetchadd(1, (int *) __read_lock_ptr, acq) < 0)) {		\
 		ia64_fetchadd(-1, (int *) __read_lock_ptr, rel);			\
@@ -148,14 +131,14 @@ do {											\
 	}										\
 } while (0)
 
-#define _raw_read_unlock(rw)					\
+#define __raw_read_unlock(rw)					\
 do {								\
-	rwlock_t *__read_lock_ptr = (rw);			\
+	raw_rwlock_t *__read_lock_ptr = (rw);			\
 	ia64_fetchadd(-1, (int *) __read_lock_ptr, rel);	\
 } while (0)
 
 #ifdef ASM_SUPPORTED
-#define _raw_write_lock(rw)							\
+#define __raw_write_lock(rw)							\
 do {										\
  	__asm__ __volatile__ (							\
 		"mov ar.ccv = r0\n"						\
@@ -170,7 +153,7 @@ do {										\
 		:: "r"(rw) : "ar.ccv", "p7", "r2", "r29", "memory");		\
 } while(0)
 
-#define _raw_write_trylock(rw)							\
+#define __raw_write_trylock(rw)							\
 ({										\
 	register long result;							\
 										\
@@ -182,7 +165,7 @@ do {										\
 	(result == 0);								\
 })
 
-static inline void _raw_write_unlock(rwlock_t *x)
+static inline void __raw_write_unlock(raw_rwlock_t *x)
 {
 	u8 *y = (u8 *)x;
 	barrier();
@@ -191,7 +174,7 @@ static inline void _raw_write_unlock(rwlock_t *x)
 
 #else /* !ASM_SUPPORTED */
 
-#define _raw_write_lock(l)								\
+#define __raw_write_lock(l)								\
 ({											\
 	__u64 ia64_val, ia64_set_val = ia64_dep_mi(-1, 0, 31, 1);			\
 	__u32 *ia64_write_lock_ptr = (__u32 *) (l);					\
@@ -202,7 +185,7 @@ static inline void _raw_write_unlock(rwlock_t *x)
 	} while (ia64_val);								\
 })
 
-#define _raw_write_trylock(rw)						\
+#define __raw_write_trylock(rw)						\
 ({									\
 	__u64 ia64_val;							\
 	__u64 ia64_set_val = ia64_dep_mi(-1, 0, 31,1);			\
@@ -210,7 +193,7 @@ static inline void _raw_write_unlock(rwlock_t *x)
 	(ia64_val == 0);						\
 })
 
-static inline void _raw_write_unlock(rwlock_t *x)
+static inline void __raw_write_unlock(raw_rwlock_t *x)
 {
 	barrier();
 	x->write_lock = 0;
@@ -218,6 +201,6 @@ static inline void _raw_write_unlock(rwlock_t *x)
 
 #endif /* !ASM_SUPPORTED */
 
-#define _raw_read_trylock(lock) generic_raw_read_trylock(lock)
+#define __raw_read_trylock(lock) generic__raw_read_trylock(lock)
 
 #endif /*  _ASM_IA64_SPINLOCK_H */
diff --git a/include/asm-ia64/spinlock_types.h b/include/asm-ia64/spinlock_types.h
new file mode 100644
index 00000000000..474e46f1ab4
--- /dev/null
+++ b/include/asm-ia64/spinlock_types.h
@@ -0,0 +1,21 @@
+#ifndef _ASM_IA64_SPINLOCK_TYPES_H
+#define _ASM_IA64_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned int lock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 0 }
+
+typedef struct {
+	volatile unsigned int read_counter	: 31;
+	volatile unsigned int write_lock	:  1;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ 0, 0 }
+
+#endif
diff --git a/include/asm-m32r/spinlock.h b/include/asm-m32r/spinlock.h
index 6608d8371c5..7de7def28da 100644
--- a/include/asm-m32r/spinlock.h
+++ b/include/asm-m32r/spinlock.h
@@ -14,57 +14,30 @@
 #include <asm/atomic.h>
 #include <asm/page.h>
 
-extern int printk(const char * fmt, ...)
-	__attribute__ ((format (printf, 1, 2)));
-
-#define RW_LOCK_BIAS		 0x01000000
-#define RW_LOCK_BIAS_STR	"0x01000000"
-
 /*
  * Your basic SMP spinlocks, allowing only a single CPU anywhere
- */
-
-typedef struct {
-	volatile int slock;
-#ifdef CONFIG_DEBUG_SPINLOCK
-	unsigned magic;
-#endif
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
-
-#define SPINLOCK_MAGIC	0xdead4ead
-
-#ifdef CONFIG_DEBUG_SPINLOCK
-#define SPINLOCK_MAGIC_INIT	, SPINLOCK_MAGIC
-#else
-#define SPINLOCK_MAGIC_INIT	/* */
-#endif
-
-#define SPIN_LOCK_UNLOCKED (spinlock_t) { 1 SPINLOCK_MAGIC_INIT }
-
-#define spin_lock_init(x)	do { *(x) = SPIN_LOCK_UNLOCKED; } while(0)
-
-/*
+ *
+ * (the type definitions are in asm/spinlock_types.h)
+ *
  * Simple spin lock operations.  There are two variants, one clears IRQ's
  * on the local processor, one does not.
  *
  * We make no fairness assumptions. They have a cost.
  */
 
-#define spin_is_locked(x)	(*(volatile int *)(&(x)->slock) <= 0)
-#define spin_unlock_wait(x)	do { barrier(); } while(spin_is_locked(x))
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
+#define __raw_spin_is_locked(x)		(*(volatile int *)(&(x)->slock) <= 0)
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
+#define __raw_spin_unlock_wait(x) \
+		do { cpu_relax(); } while (__raw_spin_is_locked(x))
 
 /**
- * _raw_spin_trylock - Try spin lock and return a result
+ * __raw_spin_trylock - Try spin lock and return a result
  * @lock: Pointer to the lock variable
  *
- * _raw_spin_trylock() tries to get the lock and returns a result.
+ * __raw_spin_trylock() tries to get the lock and returns a result.
  * On the m32r, the result value is 1 (= Success) or 0 (= Failure).
  */
-static inline int _raw_spin_trylock(spinlock_t *lock)
+static inline int __raw_spin_trylock(raw_spinlock_t *lock)
 {
 	int oldval;
 	unsigned long tmp1, tmp2;
@@ -78,7 +51,7 @@ static inline int _raw_spin_trylock(spinlock_t *lock)
 	 * }
 	 */
 	__asm__ __volatile__ (
-		"# spin_trylock			\n\t"
+		"# __raw_spin_trylock		\n\t"
 		"ldi	%1, #0;			\n\t"
 		"mvfc	%2, psw;		\n\t"
 		"clrpsw	#0x40 -> nop;		\n\t"
@@ -97,16 +70,10 @@ static inline int _raw_spin_trylock(spinlock_t *lock)
 	return (oldval > 0);
 }
 
-static inline void _raw_spin_lock(spinlock_t *lock)
+static inline void __raw_spin_lock(raw_spinlock_t *lock)
 {
 	unsigned long tmp0, tmp1;
 
-#ifdef CONFIG_DEBUG_SPINLOCK
-	if (unlikely(lock->magic != SPINLOCK_MAGIC)) {
-		printk("pc: %p\n", __builtin_return_address(0));
-		BUG();
-	}
-#endif
 	/*
 	 * lock->slock :  =1 : unlock
 	 *             : <=0 : lock
@@ -118,7 +85,7 @@ static inline void _raw_spin_lock(spinlock_t *lock)
 	 * }
 	 */
 	__asm__ __volatile__ (
-		"# spin_lock			\n\t"
+		"# __raw_spin_lock		\n\t"
 		".fillinsn			\n"
 		"1:				\n\t"
 		"mvfc	%1, psw;		\n\t"
@@ -145,12 +112,8 @@ static inline void _raw_spin_lock(spinlock_t *lock)
 	);
 }
 
-static inline void _raw_spin_unlock(spinlock_t *lock)
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
 {
-#ifdef CONFIG_DEBUG_SPINLOCK
-	BUG_ON(lock->magic != SPINLOCK_MAGIC);
-	BUG_ON(!spin_is_locked(lock));
-#endif
 	mb();
 	lock->slock = 1;
 }
@@ -164,59 +127,32 @@ static inline void _raw_spin_unlock(spinlock_t *lock)
  * can "mix" irq-safe locks - any writer needs to get a
  * irq-safe write-lock, but readers can get non-irqsafe
  * read-locks.
+ *
+ * On x86, we implement read-write locks as a 32-bit counter
+ * with the high bit (sign) being the "contended" bit.
+ *
+ * The inline assembly is non-obvious. Think about it.
+ *
+ * Changed to use the same technique as rw semaphores.  See
+ * semaphore.h for details.  -ben
  */
-typedef struct {
-	volatile int lock;
-#ifdef CONFIG_DEBUG_SPINLOCK
-	unsigned magic;
-#endif
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-
-#define RWLOCK_MAGIC	0xdeaf1eed
-
-#ifdef CONFIG_DEBUG_SPINLOCK
-#define RWLOCK_MAGIC_INIT	, RWLOCK_MAGIC
-#else
-#define RWLOCK_MAGIC_INIT	/* */
-#endif
-
-#define RW_LOCK_UNLOCKED (rwlock_t) { RW_LOCK_BIAS RWLOCK_MAGIC_INIT }
-
-#define rwlock_init(x)	do { *(x) = RW_LOCK_UNLOCKED; } while(0)
 
 /**
  * read_can_lock - would read_trylock() succeed?
  * @lock: the rwlock in question.
  */
-#define read_can_lock(x) ((int)(x)->lock > 0)
+#define __raw_read_can_lock(x) ((int)(x)->lock > 0)
 
 /**
  * write_can_lock - would write_trylock() succeed?
  * @lock: the rwlock in question.
  */
-#define write_can_lock(x) ((x)->lock == RW_LOCK_BIAS)
-
-/*
- * On x86, we implement read-write locks as a 32-bit counter
- * with the high bit (sign) being the "contended" bit.
- *
- * The inline assembly is non-obvious. Think about it.
- *
- * Changed to use the same technique as rw semaphores.  See
- * semaphore.h for details.  -ben
- */
-/* the spinlock helpers are in arch/i386/kernel/semaphore.c */
+#define __raw_write_can_lock(x) ((x)->lock == RW_LOCK_BIAS)
 
-static inline void _raw_read_lock(rwlock_t *rw)
+static inline void __raw_read_lock(raw_rwlock_t *rw)
 {
 	unsigned long tmp0, tmp1;
 
-#ifdef CONFIG_DEBUG_SPINLOCK
-	BUG_ON(rw->magic != RWLOCK_MAGIC);
-#endif
 	/*
 	 * rw->lock :  >0 : unlock
 	 *          : <=0 : lock
@@ -264,13 +200,10 @@ static inline void _raw_read_lock(rwlock_t *rw)
 	);
 }
 
-static inline void _raw_write_lock(rwlock_t *rw)
+static inline void __raw_write_lock(raw_rwlock_t *rw)
 {
 	unsigned long tmp0, tmp1, tmp2;
 
-#ifdef CONFIG_DEBUG_SPINLOCK
-	BUG_ON(rw->magic != RWLOCK_MAGIC);
-#endif
 	/*
 	 * rw->lock :  =RW_LOCK_BIAS_STR : unlock
 	 *          : !=RW_LOCK_BIAS_STR : lock
@@ -320,7 +253,7 @@ static inline void _raw_write_lock(rwlock_t *rw)
 	);
 }
 
-static inline void _raw_read_unlock(rwlock_t *rw)
+static inline void __raw_read_unlock(raw_rwlock_t *rw)
 {
 	unsigned long tmp0, tmp1;
 
@@ -342,7 +275,7 @@ static inline void _raw_read_unlock(rwlock_t *rw)
 	);
 }
 
-static inline void _raw_write_unlock(rwlock_t *rw)
+static inline void __raw_write_unlock(raw_rwlock_t *rw)
 {
 	unsigned long tmp0, tmp1, tmp2;
 
@@ -366,9 +299,9 @@ static inline void _raw_write_unlock(rwlock_t *rw)
 	);
 }
 
-#define _raw_read_trylock(lock) generic_raw_read_trylock(lock)
+#define __raw_read_trylock(lock) generic__raw_read_trylock(lock)
 
-static inline int _raw_write_trylock(rwlock_t *lock)
+static inline int __raw_write_trylock(raw_rwlock_t *lock)
 {
 	atomic_t *count = (atomic_t *)lock;
 	if (atomic_sub_and_test(RW_LOCK_BIAS, count))
diff --git a/include/asm-m32r/spinlock_types.h b/include/asm-m32r/spinlock_types.h
new file mode 100644
index 00000000000..7e9941c45f4
--- /dev/null
+++ b/include/asm-m32r/spinlock_types.h
@@ -0,0 +1,23 @@
+#ifndef _ASM_M32R_SPINLOCK_TYPES_H
+#define _ASM_M32R_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile int slock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 1 }
+
+typedef struct {
+	volatile int lock;
+} raw_rwlock_t;
+
+#define RW_LOCK_BIAS			0x01000000
+#define RW_LOCK_BIAS_STR		"0x01000000"
+
+#define __RAW_RW_LOCK_UNLOCKED		{ RW_LOCK_BIAS }
+
+#endif
diff --git a/include/asm-mips/spinlock.h b/include/asm-mips/spinlock.h
index 114d3eb98a6..4d0135b1115 100644
--- a/include/asm-mips/spinlock.h
+++ b/include/asm-mips/spinlock.h
@@ -16,20 +16,10 @@
  * Your basic SMP spinlocks, allowing only a single CPU anywhere
  */
 
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
-
-#define SPIN_LOCK_UNLOCKED (spinlock_t) { 0 }
-
-#define spin_lock_init(x)	do { (x)->lock = 0; } while(0)
-
-#define spin_is_locked(x)	((x)->lock != 0)
-#define spin_unlock_wait(x)	do { barrier(); } while ((x)->lock)
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
+#define __raw_spin_is_locked(x)	((x)->lock != 0)
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
+#define __raw_spin_unlock_wait(x) \
+		do { cpu_relax(); } while ((x)->lock)
 
 /*
  * Simple spin lock operations.  There are two variants, one clears IRQ's
@@ -38,13 +28,13 @@ typedef struct {
  * We make no fairness assumptions.  They have a cost.
  */
 
-static inline void _raw_spin_lock(spinlock_t *lock)
+static inline void __raw_spin_lock(raw_spinlock_t *lock)
 {
 	unsigned int tmp;
 
 	if (R10000_LLSC_WAR) {
 		__asm__ __volatile__(
-		"	.set	noreorder	# _raw_spin_lock	\n"
+		"	.set	noreorder	# __raw_spin_lock	\n"
 		"1:	ll	%1, %2					\n"
 		"	bnez	%1, 1b					\n"
 		"	 li	%1, 1					\n"
@@ -58,7 +48,7 @@ static inline void _raw_spin_lock(spinlock_t *lock)
 		: "memory");
 	} else {
 		__asm__ __volatile__(
-		"	.set	noreorder	# _raw_spin_lock	\n"
+		"	.set	noreorder	# __raw_spin_lock	\n"
 		"1:	ll	%1, %2					\n"
 		"	bnez	%1, 1b					\n"
 		"	 li	%1, 1					\n"
@@ -72,10 +62,10 @@ static inline void _raw_spin_lock(spinlock_t *lock)
 	}
 }
 
-static inline void _raw_spin_unlock(spinlock_t *lock)
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
 {
 	__asm__ __volatile__(
-	"	.set	noreorder	# _raw_spin_unlock	\n"
+	"	.set	noreorder	# __raw_spin_unlock	\n"
 	"	sync						\n"
 	"	sw	$0, %0					\n"
 	"	.set\treorder					\n"
@@ -84,13 +74,13 @@ static inline void _raw_spin_unlock(spinlock_t *lock)
 	: "memory");
 }
 
-static inline unsigned int _raw_spin_trylock(spinlock_t *lock)
+static inline unsigned int __raw_spin_trylock(raw_spinlock_t *lock)
 {
 	unsigned int temp, res;
 
 	if (R10000_LLSC_WAR) {
 		__asm__ __volatile__(
-		"	.set	noreorder	# _raw_spin_trylock	\n"
+		"	.set	noreorder	# __raw_spin_trylock	\n"
 		"1:	ll	%0, %3					\n"
 		"	ori	%2, %0, 1				\n"
 		"	sc	%2, %1					\n"
@@ -104,7 +94,7 @@ static inline unsigned int _raw_spin_trylock(spinlock_t *lock)
 		: "memory");
 	} else {
 		__asm__ __volatile__(
-		"	.set	noreorder	# _raw_spin_trylock	\n"
+		"	.set	noreorder	# __raw_spin_trylock	\n"
 		"1:	ll	%0, %3					\n"
 		"	ori	%2, %0, 1				\n"
 		"	sc	%2, %1					\n"
@@ -129,24 +119,13 @@ static inline unsigned int _raw_spin_trylock(spinlock_t *lock)
  * read-locks.
  */
 
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-
-#define RW_LOCK_UNLOCKED (rwlock_t) { 0 }
-
-#define rwlock_init(x)  do { *(x) = RW_LOCK_UNLOCKED; } while(0)
-
-static inline void _raw_read_lock(rwlock_t *rw)
+static inline void __raw_read_lock(raw_rwlock_t *rw)
 {
 	unsigned int tmp;
 
 	if (R10000_LLSC_WAR) {
 		__asm__ __volatile__(
-		"	.set	noreorder	# _raw_read_lock	\n"
+		"	.set	noreorder	# __raw_read_lock	\n"
 		"1:	ll	%1, %2					\n"
 		"	bltz	%1, 1b					\n"
 		"	 addu	%1, 1					\n"
@@ -160,7 +139,7 @@ static inline void _raw_read_lock(rwlock_t *rw)
 		: "memory");
 	} else {
 		__asm__ __volatile__(
-		"	.set	noreorder	# _raw_read_lock	\n"
+		"	.set	noreorder	# __raw_read_lock	\n"
 		"1:	ll	%1, %2					\n"
 		"	bltz	%1, 1b					\n"
 		"	 addu	%1, 1					\n"
@@ -177,13 +156,13 @@ static inline void _raw_read_lock(rwlock_t *rw)
 /* Note the use of sub, not subu which will make the kernel die with an
    overflow exception if we ever try to unlock an rwlock that is already
    unlocked or is being held by a writer.  */
-static inline void _raw_read_unlock(rwlock_t *rw)
+static inline void __raw_read_unlock(raw_rwlock_t *rw)
 {
 	unsigned int tmp;
 
 	if (R10000_LLSC_WAR) {
 		__asm__ __volatile__(
-		"1:	ll	%1, %2		# _raw_read_unlock	\n"
+		"1:	ll	%1, %2		# __raw_read_unlock	\n"
 		"	sub	%1, 1					\n"
 		"	sc	%1, %0					\n"
 		"	beqzl	%1, 1b					\n"
@@ -193,7 +172,7 @@ static inline void _raw_read_unlock(rwlock_t *rw)
 		: "memory");
 	} else {
 		__asm__ __volatile__(
-		"	.set	noreorder	# _raw_read_unlock	\n"
+		"	.set	noreorder	# __raw_read_unlock	\n"
 		"1:	ll	%1, %2					\n"
 		"	sub	%1, 1					\n"
 		"	sc	%1, %0					\n"
@@ -206,13 +185,13 @@ static inline void _raw_read_unlock(rwlock_t *rw)
 	}
 }
 
-static inline void _raw_write_lock(rwlock_t *rw)
+static inline void __raw_write_lock(raw_rwlock_t *rw)
 {
 	unsigned int tmp;
 
 	if (R10000_LLSC_WAR) {
 		__asm__ __volatile__(
-		"	.set	noreorder	# _raw_write_lock	\n"
+		"	.set	noreorder	# __raw_write_lock	\n"
 		"1:	ll	%1, %2					\n"
 		"	bnez	%1, 1b					\n"
 		"	 lui	%1, 0x8000				\n"
@@ -226,7 +205,7 @@ static inline void _raw_write_lock(rwlock_t *rw)
 		: "memory");
 	} else {
 		__asm__ __volatile__(
-		"	.set	noreorder	# _raw_write_lock	\n"
+		"	.set	noreorder	# __raw_write_lock	\n"
 		"1:	ll	%1, %2					\n"
 		"	bnez	%1, 1b					\n"
 		"	 lui	%1, 0x8000				\n"
@@ -241,26 +220,26 @@ static inline void _raw_write_lock(rwlock_t *rw)
 	}
 }
 
-static inline void _raw_write_unlock(rwlock_t *rw)
+static inline void __raw_write_unlock(raw_rwlock_t *rw)
 {
 	__asm__ __volatile__(
-	"	sync			# _raw_write_unlock	\n"
+	"	sync			# __raw_write_unlock	\n"
 	"	sw	$0, %0					\n"
 	: "=m" (rw->lock)
 	: "m" (rw->lock)
 	: "memory");
 }
 
-#define _raw_read_trylock(lock) generic_raw_read_trylock(lock)
+#define __raw_read_trylock(lock) generic__raw_read_trylock(lock)
 
-static inline int _raw_write_trylock(rwlock_t *rw)
+static inline int __raw_write_trylock(raw_rwlock_t *rw)
 {
 	unsigned int tmp;
 	int ret;
 
 	if (R10000_LLSC_WAR) {
 		__asm__ __volatile__(
-		"	.set	noreorder	# _raw_write_trylock	\n"
+		"	.set	noreorder	# __raw_write_trylock	\n"
 		"	li	%2, 0					\n"
 		"1:	ll	%1, %3					\n"
 		"	bnez	%1, 2f					\n"
@@ -277,7 +256,7 @@ static inline int _raw_write_trylock(rwlock_t *rw)
 		: "memory");
 	} else {
 		__asm__ __volatile__(
-		"	.set	noreorder	# _raw_write_trylock	\n"
+		"	.set	noreorder	# __raw_write_trylock	\n"
 		"	li	%2, 0					\n"
 		"1:	ll	%1, %3					\n"
 		"	bnez	%1, 2f					\n"
diff --git a/include/asm-mips/spinlock_types.h b/include/asm-mips/spinlock_types.h
new file mode 100644
index 00000000000..ce26c5048b1
--- /dev/null
+++ b/include/asm-mips/spinlock_types.h
@@ -0,0 +1,20 @@
+#ifndef _ASM_SPINLOCK_TYPES_H
+#define _ASM_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned int lock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 0 }
+
+typedef struct {
+	volatile unsigned int lock;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ 0 }
+
+#endif
diff --git a/include/asm-parisc/atomic.h b/include/asm-parisc/atomic.h
index e24f7579adb..048a2c7fd0c 100644
--- a/include/asm-parisc/atomic.h
+++ b/include/asm-parisc/atomic.h
@@ -24,19 +24,19 @@
 #  define ATOMIC_HASH_SIZE 4
 #  define ATOMIC_HASH(a) (&(__atomic_hash[ (((unsigned long) a)/L1_CACHE_BYTES) & (ATOMIC_HASH_SIZE-1) ]))
 
-extern spinlock_t __atomic_hash[ATOMIC_HASH_SIZE] __lock_aligned;
+extern raw_spinlock_t __atomic_hash[ATOMIC_HASH_SIZE] __lock_aligned;
 
-/* Can't use _raw_spin_lock_irq because of #include problems, so
+/* Can't use raw_spin_lock_irq because of #include problems, so
  * this is the substitute */
 #define _atomic_spin_lock_irqsave(l,f) do {	\
-	spinlock_t *s = ATOMIC_HASH(l);		\
+	raw_spinlock_t *s = ATOMIC_HASH(l);		\
 	local_irq_save(f);			\
-	_raw_spin_lock(s);			\
+	__raw_spin_lock(s);			\
 } while(0)
 
 #define _atomic_spin_unlock_irqrestore(l,f) do {	\
-	spinlock_t *s = ATOMIC_HASH(l);			\
-	_raw_spin_unlock(s);				\
+	raw_spinlock_t *s = ATOMIC_HASH(l);			\
+	__raw_spin_unlock(s);				\
 	local_irq_restore(f);				\
 } while(0)
 
diff --git a/include/asm-parisc/bitops.h b/include/asm-parisc/bitops.h
index 928e5ef850b..af7db694b22 100644
--- a/include/asm-parisc/bitops.h
+++ b/include/asm-parisc/bitops.h
@@ -2,7 +2,7 @@
 #define _PARISC_BITOPS_H
 
 #include <linux/compiler.h>
-#include <asm/system.h>
+#include <asm/spinlock.h>
 #include <asm/byteorder.h>
 #include <asm/atomic.h>
 
diff --git a/include/asm-parisc/cacheflush.h b/include/asm-parisc/cacheflush.h
index 06732719d92..aa592d8c0e3 100644
--- a/include/asm-parisc/cacheflush.h
+++ b/include/asm-parisc/cacheflush.h
@@ -3,6 +3,7 @@
 
 #include <linux/config.h>
 #include <linux/mm.h>
+#include <asm/cache.h>	/* for flush_user_dcache_range_asm() proto */
 
 /* The usual comment is "Caches aren't brain-dead on the <architecture>".
  * Unfortunately, that doesn't apply to PA-RISC. */
diff --git a/include/asm-parisc/processor.h b/include/asm-parisc/processor.h
index 0b61f51d846..a9dfadd0565 100644
--- a/include/asm-parisc/processor.h
+++ b/include/asm-parisc/processor.h
@@ -11,6 +11,7 @@
 #ifndef __ASSEMBLY__
 #include <linux/config.h>
 #include <linux/threads.h>
+#include <linux/spinlock_types.h>
 
 #include <asm/hardware.h>
 #include <asm/page.h>
diff --git a/include/asm-parisc/spinlock.h b/include/asm-parisc/spinlock.h
index 679ea1c651e..43eaa6e742e 100644
--- a/include/asm-parisc/spinlock.h
+++ b/include/asm-parisc/spinlock.h
@@ -2,30 +2,25 @@
 #define __ASM_SPINLOCK_H
 
 #include <asm/system.h>
+#include <asm/processor.h>
+#include <asm/spinlock_types.h>
 
 /* Note that PA-RISC has to use `1' to mean unlocked and `0' to mean locked
  * since it only has load-and-zero. Moreover, at least on some PA processors,
  * the semaphore address has to be 16-byte aligned.
  */
 
-#ifndef CONFIG_DEBUG_SPINLOCK
-
-#define __SPIN_LOCK_UNLOCKED	{ { 1, 1, 1, 1 } }
-#undef SPIN_LOCK_UNLOCKED
-#define SPIN_LOCK_UNLOCKED (spinlock_t) __SPIN_LOCK_UNLOCKED
-
-#define spin_lock_init(x)	do { *(x) = SPIN_LOCK_UNLOCKED; } while(0)
-
-static inline int spin_is_locked(spinlock_t *x)
+static inline int __raw_spin_is_locked(raw_spinlock_t *x)
 {
 	volatile unsigned int *a = __ldcw_align(x);
 	return *a == 0;
 }
 
-#define spin_unlock_wait(x)	do { barrier(); } while(spin_is_locked(x))
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
+#define __raw_spin_unlock_wait(x) \
+		do { cpu_relax(); } while (__raw_spin_is_locked(x))
 
-static inline void _raw_spin_lock(spinlock_t *x)
+static inline void __raw_spin_lock(raw_spinlock_t *x)
 {
 	volatile unsigned int *a;
 
@@ -36,7 +31,7 @@ static inline void _raw_spin_lock(spinlock_t *x)
 	mb();
 }
 
-static inline void _raw_spin_unlock(spinlock_t *x)
+static inline void __raw_spin_unlock(raw_spinlock_t *x)
 {
 	volatile unsigned int *a;
 	mb();
@@ -45,7 +40,7 @@ static inline void _raw_spin_unlock(spinlock_t *x)
 	mb();
 }
 
-static inline int _raw_spin_trylock(spinlock_t *x)
+static inline int __raw_spin_trylock(raw_spinlock_t *x)
 {
 	volatile unsigned int *a;
 	int ret;
@@ -57,131 +52,38 @@ static inline int _raw_spin_trylock(spinlock_t *x)
 
 	return ret;
 }
-	
-#define spin_lock_own(LOCK, LOCATION)	((void)0)
-
-#else /* !(CONFIG_DEBUG_SPINLOCK) */
-
-#define SPINLOCK_MAGIC	0x1D244B3C
-
-#define __SPIN_LOCK_UNLOCKED	{ { 1, 1, 1, 1 }, SPINLOCK_MAGIC, 10, __FILE__ , NULL, 0, -1, NULL, NULL }
-#undef SPIN_LOCK_UNLOCKED
-#define SPIN_LOCK_UNLOCKED (spinlock_t) __SPIN_LOCK_UNLOCKED
-
-#define spin_lock_init(x)	do { *(x) = SPIN_LOCK_UNLOCKED; } while(0)
-
-#define CHECK_LOCK(x)							\
-	do {								\
-	 	if (unlikely((x)->magic != SPINLOCK_MAGIC)) {			\
-			printk(KERN_ERR "%s:%d: spin_is_locked"		\
-			" on uninitialized spinlock %p.\n",		\
-				__FILE__, __LINE__, (x)); 		\
-		} 							\
-	} while(0)
-
-#define spin_is_locked(x)						\
-	({								\
-	 	CHECK_LOCK(x);						\
-		volatile unsigned int *a = __ldcw_align(x);		\
-		if (unlikely((*a == 0) && (x)->babble)) {				\
-			(x)->babble--;					\
-			printk("KERN_WARNING				\
-				%s:%d: spin_is_locked(%s/%p) already"	\
-				" locked by %s:%d in %s at %p(%d)\n",	\
-				__FILE__,__LINE__, (x)->module,	(x),	\
-				(x)->bfile, (x)->bline, (x)->task->comm,\
-				(x)->previous, (x)->oncpu);		\
-		}							\
-		*a == 0;						\
-	})
-
-#define spin_unlock_wait(x)						\
-	do {								\
-	 	CHECK_LOCK(x);						\
-		volatile unsigned int *a = __ldcw_align(x);		\
-		if (unlikely((*a == 0) && (x)->babble)) {				\
-			(x)->babble--;					\
-			printk("KERN_WARNING				\
-				%s:%d: spin_unlock_wait(%s/%p)"		\
-				" owned by %s:%d in %s at %p(%d)\n",	\
-				__FILE__,__LINE__, (x)->module, (x),	\
-				(x)->bfile, (x)->bline, (x)->task->comm,\
-				(x)->previous, (x)->oncpu);		\
-		}							\
-		barrier();						\
-	} while (*((volatile unsigned char *)(__ldcw_align(x))) == 0)
-
-extern void _dbg_spin_lock(spinlock_t *lock, const char *base_file, int line_no);
-extern void _dbg_spin_unlock(spinlock_t *lock, const char *, int);
-extern int _dbg_spin_trylock(spinlock_t * lock, const char *, int);
-
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
-
-#define _raw_spin_unlock(lock)	_dbg_spin_unlock(lock, __FILE__, __LINE__)
-#define _raw_spin_lock(lock) _dbg_spin_lock(lock, __FILE__, __LINE__)
-#define _raw_spin_trylock(lock) _dbg_spin_trylock(lock, __FILE__, __LINE__)
-
-/* just in case we need it */
-#define spin_lock_own(LOCK, LOCATION)					\
-do {									\
-	volatile unsigned int *a = __ldcw_align(LOCK);			\
-	if (!((*a == 0) && ((LOCK)->oncpu == smp_processor_id())))	\
-		printk("KERN_WARNING					\
-			%s: called on %d from %p but lock %s on %d\n",	\
-			LOCATION, smp_processor_id(),			\
-			__builtin_return_address(0),			\
-			(*a == 0) ? "taken" : "freed", (LOCK)->on_cpu);	\
-} while (0)
-
-#endif /* !(CONFIG_DEBUG_SPINLOCK) */
 
 /*
  * Read-write spinlocks, allowing multiple readers
  * but only one writer.
  */
-typedef struct {
-	spinlock_t lock;
-	volatile int counter;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-
-#define RW_LOCK_UNLOCKED (rwlock_t) { __SPIN_LOCK_UNLOCKED, 0 }
-
-#define rwlock_init(lp)	do { *(lp) = RW_LOCK_UNLOCKED; } while (0)
 
-#define _raw_read_trylock(lock) generic_raw_read_trylock(lock)
+#define __raw_read_trylock(lock) generic__raw_read_trylock(lock)
 
 /* read_lock, read_unlock are pretty straightforward.  Of course it somehow
  * sucks we end up saving/restoring flags twice for read_lock_irqsave aso. */
 
-#ifdef CONFIG_DEBUG_RWLOCK
-extern void _dbg_read_lock(rwlock_t * rw, const char *bfile, int bline);
-#define _raw_read_lock(rw) _dbg_read_lock(rw, __FILE__, __LINE__)
-#else
-static  __inline__ void _raw_read_lock(rwlock_t *rw)
+static  __inline__ void __raw_read_lock(raw_rwlock_t *rw)
 {
 	unsigned long flags;
 	local_irq_save(flags);
-	_raw_spin_lock(&rw->lock); 
+	__raw_spin_lock(&rw->lock);
 
 	rw->counter++;
 
-	_raw_spin_unlock(&rw->lock);
+	__raw_spin_unlock(&rw->lock);
 	local_irq_restore(flags);
 }
-#endif	/* CONFIG_DEBUG_RWLOCK */
 
-static  __inline__ void _raw_read_unlock(rwlock_t *rw)
+static  __inline__ void __raw_read_unlock(raw_rwlock_t *rw)
 {
 	unsigned long flags;
 	local_irq_save(flags);
-	_raw_spin_lock(&rw->lock); 
+	__raw_spin_lock(&rw->lock);
 
 	rw->counter--;
 
-	_raw_spin_unlock(&rw->lock);
+	__raw_spin_unlock(&rw->lock);
 	local_irq_restore(flags);
 }
 
@@ -194,20 +96,17 @@ static  __inline__ void _raw_read_unlock(rwlock_t *rw)
  * writers) in interrupt handlers someone fucked up and we'd dead-lock
  * sooner or later anyway.   prumpf */
 
-#ifdef CONFIG_DEBUG_RWLOCK
-extern void _dbg_write_lock(rwlock_t * rw, const char *bfile, int bline);
-#define _raw_write_lock(rw) _dbg_write_lock(rw, __FILE__, __LINE__)
-#else
-static  __inline__ void _raw_write_lock(rwlock_t *rw)
+static  __inline__ void __raw_write_lock(raw_rwlock_t *rw)
 {
 retry:
-	_raw_spin_lock(&rw->lock);
+	__raw_spin_lock(&rw->lock);
 
 	if(rw->counter != 0) {
 		/* this basically never happens */
-		_raw_spin_unlock(&rw->lock);
+		__raw_spin_unlock(&rw->lock);
 
-		while(rw->counter != 0);
+		while (rw->counter != 0)
+			cpu_relax();
 
 		goto retry;
 	}
@@ -215,26 +114,21 @@ retry:
 	/* got it.  now leave without unlocking */
 	rw->counter = -1; /* remember we are locked */
 }
-#endif /* CONFIG_DEBUG_RWLOCK */
 
 /* write_unlock is absolutely trivial - we don't have to wait for anything */
 
-static  __inline__ void _raw_write_unlock(rwlock_t *rw)
+static  __inline__ void __raw_write_unlock(raw_rwlock_t *rw)
 {
 	rw->counter = 0;
-	_raw_spin_unlock(&rw->lock);
+	__raw_spin_unlock(&rw->lock);
 }
 
-#ifdef CONFIG_DEBUG_RWLOCK
-extern int _dbg_write_trylock(rwlock_t * rw, const char *bfile, int bline);
-#define _raw_write_trylock(rw) _dbg_write_trylock(rw, __FILE__, __LINE__)
-#else
-static  __inline__ int _raw_write_trylock(rwlock_t *rw)
+static  __inline__ int __raw_write_trylock(raw_rwlock_t *rw)
 {
-	_raw_spin_lock(&rw->lock);
+	__raw_spin_lock(&rw->lock);
 	if (rw->counter != 0) {
 		/* this basically never happens */
-		_raw_spin_unlock(&rw->lock);
+		__raw_spin_unlock(&rw->lock);
 
 		return 0;
 	}
@@ -243,14 +137,13 @@ static  __inline__ int _raw_write_trylock(rwlock_t *rw)
 	rw->counter = -1; /* remember we are locked */
 	return 1;
 }
-#endif /* CONFIG_DEBUG_RWLOCK */
 
-static __inline__ int is_read_locked(rwlock_t *rw)
+static __inline__ int __raw_is_read_locked(raw_rwlock_t *rw)
 {
 	return rw->counter > 0;
 }
 
-static __inline__ int is_write_locked(rwlock_t *rw)
+static __inline__ int __raw_is_write_locked(raw_rwlock_t *rw)
 {
 	return rw->counter < 0;
 }
diff --git a/include/asm-parisc/spinlock_types.h b/include/asm-parisc/spinlock_types.h
new file mode 100644
index 00000000000..785bba822fb
--- /dev/null
+++ b/include/asm-parisc/spinlock_types.h
@@ -0,0 +1,21 @@
+#ifndef __ASM_SPINLOCK_TYPES_H
+#define __ASM_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned int lock[4];
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ { 1, 1, 1, 1 } }
+
+typedef struct {
+	raw_spinlock_t lock;
+	volatile int counter;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ __RAW_SPIN_LOCK_UNLOCKED, 0 }
+
+#endif
diff --git a/include/asm-parisc/system.h b/include/asm-parisc/system.h
index 81c54333903..26ff844a21c 100644
--- a/include/asm-parisc/system.h
+++ b/include/asm-parisc/system.h
@@ -160,29 +160,7 @@ static inline void set_eiem(unsigned long val)
 })
 
 #ifdef CONFIG_SMP
-/*
- * Your basic SMP spinlocks, allowing only a single CPU anywhere
- */
-
-typedef struct {
-	volatile unsigned int lock[4];
-#ifdef CONFIG_DEBUG_SPINLOCK
-	unsigned long magic;
-	volatile unsigned int babble;
-	const char *module;
-	char *bfile;
-	int bline;
-	int oncpu;
-	void *previous;
-	struct task_struct * task;
-#endif
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
-
-#define __lock_aligned __attribute__((__section__(".data.lock_aligned")))
-
+# define __lock_aligned __attribute__((__section__(".data.lock_aligned")))
 #endif
 
 #define KERNEL_START (0x10100000 - 0x1000)
diff --git a/include/asm-ppc/spinlock.h b/include/asm-ppc/spinlock.h
index 909199aae10..20edcf2a6e0 100644
--- a/include/asm-ppc/spinlock.h
+++ b/include/asm-ppc/spinlock.h
@@ -5,41 +5,21 @@
 
 /*
  * Simple spin lock operations.
+ *
+ * (the type definitions are in asm/raw_spinlock_types.h)
  */
 
-typedef struct {
-	volatile unsigned long lock;
-#ifdef CONFIG_DEBUG_SPINLOCK
-	volatile unsigned long owner_pc;
-	volatile unsigned long owner_cpu;
-#endif
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
-
-#ifdef __KERNEL__
-#ifdef CONFIG_DEBUG_SPINLOCK
-#define SPINLOCK_DEBUG_INIT     , 0, 0
-#else
-#define SPINLOCK_DEBUG_INIT     /* */
-#endif
-
-#define SPIN_LOCK_UNLOCKED	(spinlock_t) { 0 SPINLOCK_DEBUG_INIT }
-
-#define spin_lock_init(x) 	do { *(x) = SPIN_LOCK_UNLOCKED; } while(0)
-#define spin_is_locked(x)	((x)->lock != 0)
-#define spin_unlock_wait(x)	do { barrier(); } while(spin_is_locked(x))
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
-
-#ifndef CONFIG_DEBUG_SPINLOCK
-
-static inline void _raw_spin_lock(spinlock_t *lock)
+#define __raw_spin_is_locked(x)		((x)->lock != 0)
+#define __raw_spin_unlock_wait(lock) \
+	do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
+
+static inline void __raw_spin_lock(raw_spinlock_t *lock)
 {
 	unsigned long tmp;
 
 	__asm__ __volatile__(
-	"b	1f		# spin_lock\n\
+	"b	1f		# __raw_spin_lock\n\
 2:	lwzx	%0,0,%1\n\
 	cmpwi	0,%0,0\n\
 	bne+	2b\n\
@@ -55,21 +35,13 @@ static inline void _raw_spin_lock(spinlock_t *lock)
 	: "cr0", "memory");
 }
 
-static inline void _raw_spin_unlock(spinlock_t *lock)
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
 {
-	__asm__ __volatile__("eieio		# spin_unlock": : :"memory");
+	__asm__ __volatile__("eieio	# __raw_spin_unlock": : :"memory");
 	lock->lock = 0;
 }
 
-#define _raw_spin_trylock(l) (!test_and_set_bit(0,&(l)->lock))
-
-#else
-
-extern void _raw_spin_lock(spinlock_t *lock);
-extern void _raw_spin_unlock(spinlock_t *lock);
-extern int _raw_spin_trylock(spinlock_t *lock);
-
-#endif
+#define __raw_spin_trylock(l) (!test_and_set_bit(0,&(l)->lock))
 
 /*
  * Read-write spinlocks, allowing multiple readers
@@ -81,22 +53,11 @@ extern int _raw_spin_trylock(spinlock_t *lock);
  * irq-safe write-lock, but readers can get non-irqsafe
  * read-locks.
  */
-typedef struct {
-	volatile signed int lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
 
-#define RW_LOCK_UNLOCKED (rwlock_t) { 0 }
-#define rwlock_init(lp) do { *(lp) = RW_LOCK_UNLOCKED; } while(0)
+#define __raw_read_can_lock(rw)	((rw)->lock >= 0)
+#define __raw_write_can_lock(rw)	(!(rw)->lock)
 
-#define read_can_lock(rw)	((rw)->lock >= 0)
-#define write_can_lock(rw)	(!(rw)->lock)
-
-#ifndef CONFIG_DEBUG_SPINLOCK
-
-static __inline__ int _raw_read_trylock(rwlock_t *rw)
+static __inline__ int __raw_read_trylock(raw_rwlock_t *rw)
 {
 	signed int tmp;
 
@@ -116,7 +77,7 @@ static __inline__ int _raw_read_trylock(rwlock_t *rw)
 	return tmp > 0;
 }
 
-static __inline__ void _raw_read_lock(rwlock_t *rw)
+static __inline__ void __raw_read_lock(raw_rwlock_t *rw)
 {
 	signed int tmp;
 
@@ -137,7 +98,7 @@ static __inline__ void _raw_read_lock(rwlock_t *rw)
 	: "cr0", "memory");
 }
 
-static __inline__ void _raw_read_unlock(rwlock_t *rw)
+static __inline__ void __raw_read_unlock(raw_rwlock_t *rw)
 {
 	signed int tmp;
 
@@ -153,7 +114,7 @@ static __inline__ void _raw_read_unlock(rwlock_t *rw)
 	: "cr0", "memory");
 }
 
-static __inline__ int _raw_write_trylock(rwlock_t *rw)
+static __inline__ int __raw_write_trylock(raw_rwlock_t *rw)
 {
 	signed int tmp;
 
@@ -173,7 +134,7 @@ static __inline__ int _raw_write_trylock(rwlock_t *rw)
 	return tmp == 0;
 }
 
-static __inline__ void _raw_write_lock(rwlock_t *rw)
+static __inline__ void __raw_write_lock(raw_rwlock_t *rw)
 {
 	signed int tmp;
 
@@ -194,22 +155,10 @@ static __inline__ void _raw_write_lock(rwlock_t *rw)
 	: "cr0", "memory");
 }
 
-static __inline__ void _raw_write_unlock(rwlock_t *rw)
+static __inline__ void __raw_write_unlock(raw_rwlock_t *rw)
 {
 	__asm__ __volatile__("eieio		# write_unlock": : :"memory");
 	rw->lock = 0;
 }
 
-#else
-
-extern void _raw_read_lock(rwlock_t *rw);
-extern void _raw_read_unlock(rwlock_t *rw);
-extern void _raw_write_lock(rwlock_t *rw);
-extern void _raw_write_unlock(rwlock_t *rw);
-extern int _raw_read_trylock(rwlock_t *rw);
-extern int _raw_write_trylock(rwlock_t *rw);
-
-#endif
-
 #endif /* __ASM_SPINLOCK_H */
-#endif /* __KERNEL__ */
diff --git a/include/asm-ppc/spinlock_types.h b/include/asm-ppc/spinlock_types.h
new file mode 100644
index 00000000000..7919ccc75b8
--- /dev/null
+++ b/include/asm-ppc/spinlock_types.h
@@ -0,0 +1,20 @@
+#ifndef __ASM_SPINLOCK_TYPES_H
+#define __ASM_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned long lock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 0 }
+
+typedef struct {
+	volatile signed int lock;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ 0 }
+
+#endif
diff --git a/include/asm-ppc64/spinlock.h b/include/asm-ppc64/spinlock.h
index acd11564dd7..14cb895bb60 100644
--- a/include/asm-ppc64/spinlock.h
+++ b/include/asm-ppc64/spinlock.h
@@ -15,36 +15,42 @@
  * modify it under the terms of the GNU General Public License
  * as published by the Free Software Foundation; either version
  * 2 of the License, or (at your option) any later version.
+ *
+ * (the type definitions are in asm/spinlock_types.h)
  */
 #include <linux/config.h>
 #include <asm/paca.h>
 #include <asm/hvcall.h>
 #include <asm/iSeries/HvCall.h>
 
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
+#define __raw_spin_is_locked(x)		((x)->slock != 0)
 
-typedef struct {
-	volatile signed int lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
+/*
+ * This returns the old value in the lock, so we succeeded
+ * in getting the lock if the return value is 0.
+ */
+static __inline__ unsigned long __spin_trylock(raw_spinlock_t *lock)
+{
+	unsigned long tmp, tmp2;
 
-#ifdef __KERNEL__
-#define SPIN_LOCK_UNLOCKED	(spinlock_t) { 0 }
+	__asm__ __volatile__(
+"	lwz		%1,%3(13)		# __spin_trylock\n\
+1:	lwarx		%0,0,%2\n\
+	cmpwi		0,%0,0\n\
+	bne-		2f\n\
+	stwcx.		%1,0,%2\n\
+	bne-		1b\n\
+	isync\n\
+2:"	: "=&r" (tmp), "=&r" (tmp2)
+	: "r" (&lock->slock), "i" (offsetof(struct paca_struct, lock_token))
+	: "cr0", "memory");
 
-#define spin_is_locked(x)	((x)->lock != 0)
-#define spin_lock_init(x)	do { *(x) = SPIN_LOCK_UNLOCKED; } while(0)
+	return tmp;
+}
 
-static __inline__ void _raw_spin_unlock(spinlock_t *lock)
+static int __inline__ __raw_spin_trylock(raw_spinlock_t *lock)
 {
-	__asm__ __volatile__("lwsync	# spin_unlock": : :"memory");
-	lock->lock = 0;
+	return __spin_trylock(lock) == 0;
 }
 
 /*
@@ -64,44 +70,15 @@ static __inline__ void _raw_spin_unlock(spinlock_t *lock)
 #if defined(CONFIG_PPC_SPLPAR) || defined(CONFIG_PPC_ISERIES)
 /* We only yield to the hypervisor if we are in shared processor mode */
 #define SHARED_PROCESSOR (get_paca()->lppaca.shared_proc)
-extern void __spin_yield(spinlock_t *lock);
-extern void __rw_yield(rwlock_t *lock);
+extern void __spin_yield(raw_spinlock_t *lock);
+extern void __rw_yield(raw_rwlock_t *lock);
 #else /* SPLPAR || ISERIES */
 #define __spin_yield(x)	barrier()
 #define __rw_yield(x)	barrier()
 #define SHARED_PROCESSOR	0
 #endif
-extern void spin_unlock_wait(spinlock_t *lock);
-
-/*
- * This returns the old value in the lock, so we succeeded
- * in getting the lock if the return value is 0.
- */
-static __inline__ unsigned long __spin_trylock(spinlock_t *lock)
-{
-	unsigned long tmp, tmp2;
-
-	__asm__ __volatile__(
-"	lwz		%1,%3(13)		# __spin_trylock\n\
-1:	lwarx		%0,0,%2\n\
-	cmpwi		0,%0,0\n\
-	bne-		2f\n\
-	stwcx.		%1,0,%2\n\
-	bne-		1b\n\
-	isync\n\
-2:"	: "=&r" (tmp), "=&r" (tmp2)
-	: "r" (&lock->lock), "i" (offsetof(struct paca_struct, lock_token))
-	: "cr0", "memory");
-
-	return tmp;
-}
-
-static int __inline__ _raw_spin_trylock(spinlock_t *lock)
-{
-	return __spin_trylock(lock) == 0;
-}
 
-static void __inline__ _raw_spin_lock(spinlock_t *lock)
+static void __inline__ __raw_spin_lock(raw_spinlock_t *lock)
 {
 	while (1) {
 		if (likely(__spin_trylock(lock) == 0))
@@ -110,12 +87,12 @@ static void __inline__ _raw_spin_lock(spinlock_t *lock)
 			HMT_low();
 			if (SHARED_PROCESSOR)
 				__spin_yield(lock);
-		} while (unlikely(lock->lock != 0));
+		} while (unlikely(lock->slock != 0));
 		HMT_medium();
 	}
 }
 
-static void __inline__ _raw_spin_lock_flags(spinlock_t *lock, unsigned long flags)
+static void __inline__ __raw_spin_lock_flags(raw_spinlock_t *lock, unsigned long flags)
 {
 	unsigned long flags_dis;
 
@@ -128,12 +105,20 @@ static void __inline__ _raw_spin_lock_flags(spinlock_t *lock, unsigned long flag
 			HMT_low();
 			if (SHARED_PROCESSOR)
 				__spin_yield(lock);
-		} while (unlikely(lock->lock != 0));
+		} while (unlikely(lock->slock != 0));
 		HMT_medium();
 		local_irq_restore(flags_dis);
 	}
 }
 
+static __inline__ void __raw_spin_unlock(raw_spinlock_t *lock)
+{
+	__asm__ __volatile__("lwsync	# __raw_spin_unlock": : :"memory");
+	lock->slock = 0;
+}
+
+extern void __raw_spin_unlock_wait(raw_spinlock_t *lock);
+
 /*
  * Read-write spinlocks, allowing multiple readers
  * but only one writer.
@@ -144,24 +129,15 @@ static void __inline__ _raw_spin_lock_flags(spinlock_t *lock, unsigned long flag
  * irq-safe write-lock, but readers can get non-irqsafe
  * read-locks.
  */
-#define RW_LOCK_UNLOCKED (rwlock_t) { 0 }
 
-#define rwlock_init(x)		do { *(x) = RW_LOCK_UNLOCKED; } while(0)
-
-#define read_can_lock(rw)	((rw)->lock >= 0)
-#define write_can_lock(rw)	(!(rw)->lock)
-
-static __inline__ void _raw_write_unlock(rwlock_t *rw)
-{
-	__asm__ __volatile__("lwsync		# write_unlock": : :"memory");
-	rw->lock = 0;
-}
+#define __raw_read_can_lock(rw)		((rw)->lock >= 0)
+#define __raw_write_can_lock(rw)	(!(rw)->lock)
 
 /*
  * This returns the old value in the lock + 1,
  * so we got a read lock if the return value is > 0.
  */
-static long __inline__ __read_trylock(rwlock_t *rw)
+static long __inline__ __read_trylock(raw_rwlock_t *rw)
 {
 	long tmp;
 
@@ -180,45 +156,11 @@ static long __inline__ __read_trylock(rwlock_t *rw)
 	return tmp;
 }
 
-static int __inline__ _raw_read_trylock(rwlock_t *rw)
-{
-	return __read_trylock(rw) > 0;
-}
-
-static void __inline__ _raw_read_lock(rwlock_t *rw)
-{
-	while (1) {
-		if (likely(__read_trylock(rw) > 0))
-			break;
-		do {
-			HMT_low();
-			if (SHARED_PROCESSOR)
-				__rw_yield(rw);
-		} while (unlikely(rw->lock < 0));
-		HMT_medium();
-	}
-}
-
-static void __inline__ _raw_read_unlock(rwlock_t *rw)
-{
-	long tmp;
-
-	__asm__ __volatile__(
-	"eieio				# read_unlock\n\
-1:	lwarx		%0,0,%1\n\
-	addic		%0,%0,-1\n\
-	stwcx.		%0,0,%1\n\
-	bne-		1b"
-	: "=&r"(tmp)
-	: "r"(&rw->lock)
-	: "cr0", "memory");
-}
-
 /*
  * This returns the old value in the lock,
  * so we got the write lock if the return value is 0.
  */
-static __inline__ long __write_trylock(rwlock_t *rw)
+static __inline__ long __write_trylock(raw_rwlock_t *rw)
 {
 	long tmp, tmp2;
 
@@ -237,12 +179,21 @@ static __inline__ long __write_trylock(rwlock_t *rw)
 	return tmp;
 }
 
-static int __inline__ _raw_write_trylock(rwlock_t *rw)
+static void __inline__ __raw_read_lock(raw_rwlock_t *rw)
 {
-	return __write_trylock(rw) == 0;
+	while (1) {
+		if (likely(__read_trylock(rw) > 0))
+			break;
+		do {
+			HMT_low();
+			if (SHARED_PROCESSOR)
+				__rw_yield(rw);
+		} while (unlikely(rw->lock < 0));
+		HMT_medium();
+	}
 }
 
-static void __inline__ _raw_write_lock(rwlock_t *rw)
+static void __inline__ __raw_write_lock(raw_rwlock_t *rw)
 {
 	while (1) {
 		if (likely(__write_trylock(rw) == 0))
@@ -256,5 +207,35 @@ static void __inline__ _raw_write_lock(rwlock_t *rw)
 	}
 }
 
-#endif /* __KERNEL__ */
+static int __inline__ __raw_read_trylock(raw_rwlock_t *rw)
+{
+	return __read_trylock(rw) > 0;
+}
+
+static int __inline__ __raw_write_trylock(raw_rwlock_t *rw)
+{
+	return __write_trylock(rw) == 0;
+}
+
+static void __inline__ __raw_read_unlock(raw_rwlock_t *rw)
+{
+	long tmp;
+
+	__asm__ __volatile__(
+	"eieio				# read_unlock\n\
+1:	lwarx		%0,0,%1\n\
+	addic		%0,%0,-1\n\
+	stwcx.		%0,0,%1\n\
+	bne-		1b"
+	: "=&r"(tmp)
+	: "r"(&rw->lock)
+	: "cr0", "memory");
+}
+
+static __inline__ void __raw_write_unlock(raw_rwlock_t *rw)
+{
+	__asm__ __volatile__("lwsync	# write_unlock": : :"memory");
+	rw->lock = 0;
+}
+
 #endif /* __ASM_SPINLOCK_H */
diff --git a/include/asm-ppc64/spinlock_types.h b/include/asm-ppc64/spinlock_types.h
new file mode 100644
index 00000000000..a37c8eabb9f
--- /dev/null
+++ b/include/asm-ppc64/spinlock_types.h
@@ -0,0 +1,20 @@
+#ifndef __ASM_SPINLOCK_TYPES_H
+#define __ASM_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned int slock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 0 }
+
+typedef struct {
+	volatile signed int lock;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ 0 }
+
+#endif
diff --git a/include/asm-s390/spinlock.h b/include/asm-s390/spinlock.h
index 321b23bba1e..273dbecf8ac 100644
--- a/include/asm-s390/spinlock.h
+++ b/include/asm-s390/spinlock.h
@@ -27,25 +27,19 @@ _raw_compare_and_swap(volatile unsigned int *lock,
  * on the local processor, one does not.
  *
  * We make no fairness assumptions. They have a cost.
+ *
+ * (the type definitions are in asm/spinlock_types.h)
  */
 
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} __attribute__ ((aligned (4))) spinlock_t;
-
-#define SPIN_LOCK_UNLOCKED	(spinlock_t) { 0 }
-#define spin_lock_init(lp)	do { (lp)->lock = 0; } while(0)
-#define spin_unlock_wait(lp)	do { barrier(); } while(((volatile spinlock_t *)(lp))->lock)
-#define spin_is_locked(x)	((x)->lock != 0)
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
+#define __raw_spin_is_locked(x) ((x)->lock != 0)
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
+#define __raw_spin_unlock_wait(lock) \
+	do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
 
-extern void _raw_spin_lock_wait(spinlock_t *lp, unsigned int pc);
-extern int _raw_spin_trylock_retry(spinlock_t *lp, unsigned int pc);
+extern void _raw_spin_lock_wait(raw_spinlock_t *lp, unsigned int pc);
+extern int _raw_spin_trylock_retry(raw_spinlock_t *lp, unsigned int pc);
 
-static inline void _raw_spin_lock(spinlock_t *lp)
+static inline void __raw_spin_lock(raw_spinlock_t *lp)
 {
 	unsigned long pc = 1 | (unsigned long) __builtin_return_address(0);
 
@@ -53,7 +47,7 @@ static inline void _raw_spin_lock(spinlock_t *lp)
 		_raw_spin_lock_wait(lp, pc);
 }
 
-static inline int _raw_spin_trylock(spinlock_t *lp)
+static inline int __raw_spin_trylock(raw_spinlock_t *lp)
 {
 	unsigned long pc = 1 | (unsigned long) __builtin_return_address(0);
 
@@ -62,7 +56,7 @@ static inline int _raw_spin_trylock(spinlock_t *lp)
 	return _raw_spin_trylock_retry(lp, pc);
 }
 
-static inline void _raw_spin_unlock(spinlock_t *lp)
+static inline void __raw_spin_unlock(raw_spinlock_t *lp)
 {
 	_raw_compare_and_swap(&lp->lock, lp->lock, 0);
 }
@@ -77,36 +71,25 @@ static inline void _raw_spin_unlock(spinlock_t *lp)
  * irq-safe write-lock, but readers can get non-irqsafe
  * read-locks.
  */
-typedef struct {
-	volatile unsigned int lock;
-	volatile unsigned long owner_pc;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-
-#define RW_LOCK_UNLOCKED (rwlock_t) { 0, 0 }
-
-#define rwlock_init(x)	do { *(x) = RW_LOCK_UNLOCKED; } while(0)
 
 /**
  * read_can_lock - would read_trylock() succeed?
  * @lock: the rwlock in question.
  */
-#define read_can_lock(x) ((int)(x)->lock >= 0)
+#define __raw_read_can_lock(x) ((int)(x)->lock >= 0)
 
 /**
  * write_can_lock - would write_trylock() succeed?
  * @lock: the rwlock in question.
  */
-#define write_can_lock(x) ((x)->lock == 0)
+#define __raw_write_can_lock(x) ((x)->lock == 0)
 
-extern void _raw_read_lock_wait(rwlock_t *lp);
-extern int _raw_read_trylock_retry(rwlock_t *lp);
-extern void _raw_write_lock_wait(rwlock_t *lp);
-extern int _raw_write_trylock_retry(rwlock_t *lp);
+extern void _raw_read_lock_wait(raw_rwlock_t *lp);
+extern int _raw_read_trylock_retry(raw_rwlock_t *lp);
+extern void _raw_write_lock_wait(raw_rwlock_t *lp);
+extern int _raw_write_trylock_retry(raw_rwlock_t *lp);
 
-static inline void _raw_read_lock(rwlock_t *rw)
+static inline void __raw_read_lock(raw_rwlock_t *rw)
 {
 	unsigned int old;
 	old = rw->lock & 0x7fffffffU;
@@ -114,7 +97,7 @@ static inline void _raw_read_lock(rwlock_t *rw)
 		_raw_read_lock_wait(rw);
 }
 
-static inline void _raw_read_unlock(rwlock_t *rw)
+static inline void __raw_read_unlock(raw_rwlock_t *rw)
 {
 	unsigned int old, cmp;
 
@@ -125,18 +108,18 @@ static inline void _raw_read_unlock(rwlock_t *rw)
 	} while (cmp != old);
 }
 
-static inline void _raw_write_lock(rwlock_t *rw)
+static inline void __raw_write_lock(raw_rwlock_t *rw)
 {
 	if (unlikely(_raw_compare_and_swap(&rw->lock, 0, 0x80000000) != 0))
 		_raw_write_lock_wait(rw);
 }
 
-static inline void _raw_write_unlock(rwlock_t *rw)
+static inline void __raw_write_unlock(raw_rwlock_t *rw)
 {
 	_raw_compare_and_swap(&rw->lock, 0x80000000, 0);
 }
 
-static inline int _raw_read_trylock(rwlock_t *rw)
+static inline int __raw_read_trylock(raw_rwlock_t *rw)
 {
 	unsigned int old;
 	old = rw->lock & 0x7fffffffU;
@@ -145,7 +128,7 @@ static inline int _raw_read_trylock(rwlock_t *rw)
 	return _raw_read_trylock_retry(rw);
 }
 
-static inline int _raw_write_trylock(rwlock_t *rw)
+static inline int __raw_write_trylock(raw_rwlock_t *rw)
 {
 	if (likely(_raw_compare_and_swap(&rw->lock, 0, 0x80000000) == 0))
 		return 1;
diff --git a/include/asm-s390/spinlock_types.h b/include/asm-s390/spinlock_types.h
new file mode 100644
index 00000000000..f79a2216204
--- /dev/null
+++ b/include/asm-s390/spinlock_types.h
@@ -0,0 +1,21 @@
+#ifndef __ASM_SPINLOCK_TYPES_H
+#define __ASM_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned int lock;
+} __attribute__ ((aligned (4))) raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 0 }
+
+typedef struct {
+	volatile unsigned int lock;
+	volatile unsigned int owner_pc;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ 0, 0 }
+
+#endif
diff --git a/include/asm-sh/spinlock.h b/include/asm-sh/spinlock.h
index e770b55649e..846322d4c35 100644
--- a/include/asm-sh/spinlock.h
+++ b/include/asm-sh/spinlock.h
@@ -15,20 +15,11 @@
 /*
  * Your basic SMP spinlocks, allowing only a single CPU anywhere
  */
-typedef struct {
-	volatile unsigned long lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
 
-#define SPIN_LOCK_UNLOCKED	(spinlock_t) { 0 }
-
-#define spin_lock_init(x)	do { *(x) = SPIN_LOCK_UNLOCKED; } while(0)
-
-#define spin_is_locked(x)	((x)->lock != 0)
-#define spin_unlock_wait(x)	do { barrier(); } while (spin_is_locked(x))
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
+#define __raw_spin_is_locked(x)	((x)->lock != 0)
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
+#define __raw_spin_unlock_wait(x) \
+	do { cpu_relax(); } while (__raw_spin_is_locked(x))
 
 /*
  * Simple spin lock operations.  There are two variants, one clears IRQ's
@@ -36,7 +27,7 @@ typedef struct {
  *
  * We make no fairness assumptions.  They have a cost.
  */
-static inline void _raw_spin_lock(spinlock_t *lock)
+static inline void __raw_spin_lock(raw_spinlock_t *lock)
 {
 	__asm__ __volatile__ (
 		"1:\n\t"
@@ -49,14 +40,14 @@ static inline void _raw_spin_lock(spinlock_t *lock)
 	);
 }
 
-static inline void _raw_spin_unlock(spinlock_t *lock)
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
 {
 	assert_spin_locked(lock);
 
 	lock->lock = 0;
 }
 
-#define _raw_spin_trylock(x) (!test_and_set_bit(0, &(x)->lock))
+#define __raw_spin_trylock(x) (!test_and_set_bit(0, &(x)->lock))
 
 /*
  * Read-write spinlocks, allowing multiple readers but only one writer.
@@ -66,51 +57,40 @@ static inline void _raw_spin_unlock(spinlock_t *lock)
  * needs to get a irq-safe write-lock, but readers can get non-irqsafe
  * read-locks.
  */
-typedef struct {
-	spinlock_t lock;
-	atomic_t counter;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-
-#define RW_LOCK_BIAS		0x01000000
-#define RW_LOCK_UNLOCKED	(rwlock_t) { { 0 }, { RW_LOCK_BIAS } }
-#define rwlock_init(x)		do { *(x) = RW_LOCK_UNLOCKED; } while (0)
-
-static inline void _raw_read_lock(rwlock_t *rw)
+
+static inline void __raw_read_lock(raw_rwlock_t *rw)
 {
-	_raw_spin_lock(&rw->lock);
+	__raw_spin_lock(&rw->lock);
 
 	atomic_inc(&rw->counter);
 
-	_raw_spin_unlock(&rw->lock);
+	__raw_spin_unlock(&rw->lock);
 }
 
-static inline void _raw_read_unlock(rwlock_t *rw)
+static inline void __raw_read_unlock(raw_rwlock_t *rw)
 {
-	_raw_spin_lock(&rw->lock);
+	__raw_spin_lock(&rw->lock);
 
 	atomic_dec(&rw->counter);
 
-	_raw_spin_unlock(&rw->lock);
+	__raw_spin_unlock(&rw->lock);
 }
 
-static inline void _raw_write_lock(rwlock_t *rw)
+static inline void __raw_write_lock(raw_rwlock_t *rw)
 {
-	_raw_spin_lock(&rw->lock);
+	__raw_spin_lock(&rw->lock);
 	atomic_set(&rw->counter, -1);
 }
 
-static inline void _raw_write_unlock(rwlock_t *rw)
+static inline void __raw_write_unlock(raw_rwlock_t *rw)
 {
 	atomic_set(&rw->counter, 0);
-	_raw_spin_unlock(&rw->lock);
+	__raw_spin_unlock(&rw->lock);
 }
 
-#define _raw_read_trylock(lock) generic_raw_read_trylock(lock)
+#define __raw_read_trylock(lock) generic__raw_read_trylock(lock)
 
-static inline int _raw_write_trylock(rwlock_t *rw)
+static inline int __raw_write_trylock(raw_rwlock_t *rw)
 {
 	if (atomic_sub_and_test(RW_LOCK_BIAS, &rw->counter))
 		return 1;
@@ -121,4 +101,3 @@ static inline int _raw_write_trylock(rwlock_t *rw)
 }
 
 #endif /* __ASM_SH_SPINLOCK_H */
-
diff --git a/include/asm-sh/spinlock_types.h b/include/asm-sh/spinlock_types.h
new file mode 100644
index 00000000000..8c41b6c3aac
--- /dev/null
+++ b/include/asm-sh/spinlock_types.h
@@ -0,0 +1,22 @@
+#ifndef __ASM_SH_SPINLOCK_TYPES_H
+#define __ASM_SH_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned long lock;
+} raw_spinlock_t;
+
+#define __SPIN_LOCK_UNLOCKED		{ 0 }
+
+typedef struct {
+	raw_spinlock_t lock;
+	atomic_t counter;
+} raw_rwlock_t;
+
+#define RW_LOCK_BIAS			0x01000000
+#define __RAW_RW_LOCK_UNLOCKED		{ { 0 }, { RW_LOCK_BIAS } }
+
+#endif
diff --git a/include/asm-sparc/spinlock.h b/include/asm-sparc/spinlock.h
index 0cbd87ad491..111727a2bb4 100644
--- a/include/asm-sparc/spinlock.h
+++ b/include/asm-sparc/spinlock.h
@@ -12,96 +12,12 @@
 
 #include <asm/psr.h>
 
-#ifdef CONFIG_DEBUG_SPINLOCK
-struct _spinlock_debug {
-	unsigned char lock;
-	unsigned long owner_pc;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-};
-typedef struct _spinlock_debug spinlock_t;
-
-#define SPIN_LOCK_UNLOCKED	(spinlock_t) { 0, 0 }
-#define spin_lock_init(lp)	do { *(lp)= SPIN_LOCK_UNLOCKED; } while(0)
-#define spin_is_locked(lp)  (*((volatile unsigned char *)(&((lp)->lock))) != 0)
-#define spin_unlock_wait(lp)	do { barrier(); } while(*(volatile unsigned char *)(&(lp)->lock))
-
-extern void _do_spin_lock(spinlock_t *lock, char *str);
-extern int _spin_trylock(spinlock_t *lock);
-extern void _do_spin_unlock(spinlock_t *lock);
-
-#define _raw_spin_trylock(lp)	_spin_trylock(lp)
-#define _raw_spin_lock(lock)	_do_spin_lock(lock, "spin_lock")
-#define _raw_spin_unlock(lock)	_do_spin_unlock(lock)
-
-struct _rwlock_debug {
-	volatile unsigned int lock;
-	unsigned long owner_pc;
-	unsigned long reader_pc[NR_CPUS];
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-};
-typedef struct _rwlock_debug rwlock_t;
-
-#define RW_LOCK_UNLOCKED (rwlock_t) { 0, 0, {0} }
-
-#define rwlock_init(lp)	do { *(lp)= RW_LOCK_UNLOCKED; } while(0)
-
-extern void _do_read_lock(rwlock_t *rw, char *str);
-extern void _do_read_unlock(rwlock_t *rw, char *str);
-extern void _do_write_lock(rwlock_t *rw, char *str);
-extern void _do_write_unlock(rwlock_t *rw);
-
-#define _raw_read_lock(lock)	\
-do {	unsigned long flags; \
-	local_irq_save(flags); \
-	_do_read_lock(lock, "read_lock"); \
-	local_irq_restore(flags); \
-} while(0)
-
-#define _raw_read_unlock(lock) \
-do {	unsigned long flags; \
-	local_irq_save(flags); \
-	_do_read_unlock(lock, "read_unlock"); \
-	local_irq_restore(flags); \
-} while(0)
-
-#define _raw_write_lock(lock) \
-do {	unsigned long flags; \
-	local_irq_save(flags); \
-	_do_write_lock(lock, "write_lock"); \
-	local_irq_restore(flags); \
-} while(0)
-
-#define _raw_write_unlock(lock) \
-do {	unsigned long flags; \
-	local_irq_save(flags); \
-	_do_write_unlock(lock); \
-	local_irq_restore(flags); \
-} while(0)
-
-#else /* !CONFIG_DEBUG_SPINLOCK */
-
-typedef struct {
-	unsigned char lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
-
-#define SPIN_LOCK_UNLOCKED	(spinlock_t) { 0 }
-
-#define spin_lock_init(lock)   (*((unsigned char *)(lock)) = 0)
-#define spin_is_locked(lock)    (*((volatile unsigned char *)(lock)) != 0)
+#define __raw_spin_is_locked(lock) (*((volatile unsigned char *)(lock)) != 0)
 
-#define spin_unlock_wait(lock) \
-do { \
-	barrier(); \
-} while(*((volatile unsigned char *)lock))
+#define __raw_spin_unlock_wait(lock) \
+	do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
 
-extern __inline__ void _raw_spin_lock(spinlock_t *lock)
+extern __inline__ void __raw_spin_lock(raw_spinlock_t *lock)
 {
 	__asm__ __volatile__(
 	"\n1:\n\t"
@@ -121,7 +37,7 @@ extern __inline__ void _raw_spin_lock(spinlock_t *lock)
 	: "g2", "memory", "cc");
 }
 
-extern __inline__ int _raw_spin_trylock(spinlock_t *lock)
+extern __inline__ int __raw_spin_trylock(raw_spinlock_t *lock)
 {
 	unsigned int result;
 	__asm__ __volatile__("ldstub [%1], %0"
@@ -131,7 +47,7 @@ extern __inline__ int _raw_spin_trylock(spinlock_t *lock)
 	return (result == 0);
 }
 
-extern __inline__ void _raw_spin_unlock(spinlock_t *lock)
+extern __inline__ void __raw_spin_unlock(raw_spinlock_t *lock)
 {
 	__asm__ __volatile__("stb %%g0, [%0]" : : "r" (lock) : "memory");
 }
@@ -147,23 +63,11 @@ extern __inline__ void _raw_spin_unlock(spinlock_t *lock)
  *
  * XXX This might create some problems with my dual spinlock
  * XXX scheme, deadlocks etc. -DaveM
- */
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-
-#define RW_LOCK_UNLOCKED (rwlock_t) { 0 }
-
-#define rwlock_init(lp)	do { *(lp)= RW_LOCK_UNLOCKED; } while(0)
-
-
-/* Sort of like atomic_t's on Sparc, but even more clever.
+ *
+ * Sort of like atomic_t's on Sparc, but even more clever.
  *
  *	------------------------------------
- *	| 24-bit counter           | wlock |  rwlock_t
+ *	| 24-bit counter           | wlock |  raw_rwlock_t
  *	------------------------------------
  *	 31                       8 7     0
  *
@@ -174,9 +78,9 @@ typedef struct {
  *
  * Unfortunately this scheme limits us to ~16,000,000 cpus.
  */
-extern __inline__ void _read_lock(rwlock_t *rw)
+extern __inline__ void __read_lock(raw_rwlock_t *rw)
 {
-	register rwlock_t *lp asm("g1");
+	register raw_rwlock_t *lp asm("g1");
 	lp = rw;
 	__asm__ __volatile__(
 	"mov	%%o7, %%g4\n\t"
@@ -187,16 +91,16 @@ extern __inline__ void _read_lock(rwlock_t *rw)
 	: "g2", "g4", "memory", "cc");
 }
 
-#define _raw_read_lock(lock) \
+#define __raw_read_lock(lock) \
 do {	unsigned long flags; \
 	local_irq_save(flags); \
-	_read_lock(lock); \
+	__raw_read_lock(lock); \
 	local_irq_restore(flags); \
 } while(0)
 
-extern __inline__ void _read_unlock(rwlock_t *rw)
+extern __inline__ void __read_unlock(raw_rwlock_t *rw)
 {
-	register rwlock_t *lp asm("g1");
+	register raw_rwlock_t *lp asm("g1");
 	lp = rw;
 	__asm__ __volatile__(
 	"mov	%%o7, %%g4\n\t"
@@ -207,16 +111,16 @@ extern __inline__ void _read_unlock(rwlock_t *rw)
 	: "g2", "g4", "memory", "cc");
 }
 
-#define _raw_read_unlock(lock) \
+#define __raw_read_unlock(lock) \
 do {	unsigned long flags; \
 	local_irq_save(flags); \
-	_read_unlock(lock); \
+	__raw_read_unlock(lock); \
 	local_irq_restore(flags); \
 } while(0)
 
-extern __inline__ void _raw_write_lock(rwlock_t *rw)
+extern __inline__ void __raw_write_lock(raw_rwlock_t *rw)
 {
-	register rwlock_t *lp asm("g1");
+	register raw_rwlock_t *lp asm("g1");
 	lp = rw;
 	__asm__ __volatile__(
 	"mov	%%o7, %%g4\n\t"
@@ -227,11 +131,9 @@ extern __inline__ void _raw_write_lock(rwlock_t *rw)
 	: "g2", "g4", "memory", "cc");
 }
 
-#define _raw_write_unlock(rw)	do { (rw)->lock = 0; } while(0)
-
-#endif /* CONFIG_DEBUG_SPINLOCK */
+#define __raw_write_unlock(rw)	do { (rw)->lock = 0; } while(0)
 
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
 
 #endif /* !(__ASSEMBLY__) */
 
diff --git a/include/asm-sparc/spinlock_types.h b/include/asm-sparc/spinlock_types.h
new file mode 100644
index 00000000000..0a0fb116c4e
--- /dev/null
+++ b/include/asm-sparc/spinlock_types.h
@@ -0,0 +1,20 @@
+#ifndef __SPARC_SPINLOCK_TYPES_H
+#define __SPARC_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	unsigned char lock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 0 }
+
+typedef struct {
+	volatile unsigned int lock;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ 0 }
+
+#endif
diff --git a/include/asm-sparc64/spinlock.h b/include/asm-sparc64/spinlock.h
index a02c4370eb4..ec85d12d73b 100644
--- a/include/asm-sparc64/spinlock.h
+++ b/include/asm-sparc64/spinlock.h
@@ -29,24 +29,13 @@
  * must be pre-V9 branches.
  */
 
-#ifndef CONFIG_DEBUG_SPINLOCK
+#define __raw_spin_is_locked(lp)	((lp)->lock != 0)
 
-typedef struct {
-	volatile unsigned char lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
-#define SPIN_LOCK_UNLOCKED	(spinlock_t) {0,}
+#define __raw_spin_unlock_wait(lp)	\
+	do {	rmb();			\
+	} while((lp)->lock)
 
-#define spin_lock_init(lp)	do { *(lp)= SPIN_LOCK_UNLOCKED; } while(0)
-#define spin_is_locked(lp)  ((lp)->lock != 0)
-
-#define spin_unlock_wait(lp)	\
-do {	rmb();			\
-} while((lp)->lock)
-
-static inline void _raw_spin_lock(spinlock_t *lock)
+static inline void __raw_spin_lock(raw_spinlock_t *lock)
 {
 	unsigned long tmp;
 
@@ -67,7 +56,7 @@ static inline void _raw_spin_lock(spinlock_t *lock)
 	: "memory");
 }
 
-static inline int _raw_spin_trylock(spinlock_t *lock)
+static inline int __raw_spin_trylock(raw_spinlock_t *lock)
 {
 	unsigned long result;
 
@@ -81,7 +70,7 @@ static inline int _raw_spin_trylock(spinlock_t *lock)
 	return (result == 0UL);
 }
 
-static inline void _raw_spin_unlock(spinlock_t *lock)
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
 {
 	__asm__ __volatile__(
 "	membar		#StoreStore | #LoadStore\n"
@@ -91,7 +80,7 @@ static inline void _raw_spin_unlock(spinlock_t *lock)
 	: "memory");
 }
 
-static inline void _raw_spin_lock_flags(spinlock_t *lock, unsigned long flags)
+static inline void __raw_spin_lock_flags(raw_spinlock_t *lock, unsigned long flags)
 {
 	unsigned long tmp1, tmp2;
 
@@ -115,51 +104,9 @@ static inline void _raw_spin_lock_flags(spinlock_t *lock, unsigned long flags)
 	: "memory");
 }
 
-#else /* !(CONFIG_DEBUG_SPINLOCK) */
-
-typedef struct {
-	volatile unsigned char lock;
-	unsigned int owner_pc, owner_cpu;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
-#define SPIN_LOCK_UNLOCKED (spinlock_t) { 0, 0, 0xff }
-#define spin_lock_init(lp)	do { *(lp)= SPIN_LOCK_UNLOCKED; } while(0)
-#define spin_is_locked(__lock)	((__lock)->lock != 0)
-#define spin_unlock_wait(__lock)	\
-do { \
-	rmb(); \
-} while((__lock)->lock)
-
-extern void _do_spin_lock(spinlock_t *lock, char *str, unsigned long caller);
-extern void _do_spin_unlock(spinlock_t *lock);
-extern int _do_spin_trylock(spinlock_t *lock, unsigned long caller);
-
-#define _raw_spin_trylock(lp)	\
-	_do_spin_trylock(lp, (unsigned long) __builtin_return_address(0))
-#define _raw_spin_lock(lock)	\
-	_do_spin_lock(lock, "spin_lock", \
-		      (unsigned long) __builtin_return_address(0))
-#define _raw_spin_unlock(lock)	_do_spin_unlock(lock)
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
-
-#endif /* CONFIG_DEBUG_SPINLOCK */
-
 /* Multi-reader locks, these are much saner than the 32-bit Sparc ones... */
 
-#ifndef CONFIG_DEBUG_SPINLOCK
-
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-#define RW_LOCK_UNLOCKED	(rwlock_t) {0,}
-#define rwlock_init(lp) do { *(lp) = RW_LOCK_UNLOCKED; } while(0)
-
-static void inline __read_lock(rwlock_t *lock)
+static void inline __read_lock(raw_rwlock_t *lock)
 {
 	unsigned long tmp1, tmp2;
 
@@ -184,7 +131,7 @@ static void inline __read_lock(rwlock_t *lock)
 	: "memory");
 }
 
-static void inline __read_unlock(rwlock_t *lock)
+static void inline __read_unlock(raw_rwlock_t *lock)
 {
 	unsigned long tmp1, tmp2;
 
@@ -201,7 +148,7 @@ static void inline __read_unlock(rwlock_t *lock)
 	: "memory");
 }
 
-static void inline __write_lock(rwlock_t *lock)
+static void inline __write_lock(raw_rwlock_t *lock)
 {
 	unsigned long mask, tmp1, tmp2;
 
@@ -228,7 +175,7 @@ static void inline __write_lock(rwlock_t *lock)
 	: "memory");
 }
 
-static void inline __write_unlock(rwlock_t *lock)
+static void inline __write_unlock(raw_rwlock_t *lock)
 {
 	__asm__ __volatile__(
 "	membar		#LoadStore | #StoreStore\n"
@@ -238,7 +185,7 @@ static void inline __write_unlock(rwlock_t *lock)
 	: "memory");
 }
 
-static int inline __write_trylock(rwlock_t *lock)
+static int inline __write_trylock(raw_rwlock_t *lock)
 {
 	unsigned long mask, tmp1, tmp2, result;
 
@@ -263,78 +210,15 @@ static int inline __write_trylock(rwlock_t *lock)
 	return result;
 }
 
-#define _raw_read_lock(p)	__read_lock(p)
-#define _raw_read_unlock(p)	__read_unlock(p)
-#define _raw_write_lock(p)	__write_lock(p)
-#define _raw_write_unlock(p)	__write_unlock(p)
-#define _raw_write_trylock(p)	__write_trylock(p)
-
-#else /* !(CONFIG_DEBUG_SPINLOCK) */
-
-typedef struct {
-	volatile unsigned long lock;
-	unsigned int writer_pc, writer_cpu;
-	unsigned int reader_pc[NR_CPUS];
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-#define RW_LOCK_UNLOCKED	(rwlock_t) { 0, 0, 0xff, { } }
-#define rwlock_init(lp) do { *(lp) = RW_LOCK_UNLOCKED; } while(0)
-
-extern void _do_read_lock(rwlock_t *rw, char *str, unsigned long caller);
-extern void _do_read_unlock(rwlock_t *rw, char *str, unsigned long caller);
-extern void _do_write_lock(rwlock_t *rw, char *str, unsigned long caller);
-extern void _do_write_unlock(rwlock_t *rw, unsigned long caller);
-extern int _do_write_trylock(rwlock_t *rw, char *str, unsigned long caller);
-
-#define _raw_read_lock(lock) \
-do {	unsigned long flags; \
-	local_irq_save(flags); \
-	_do_read_lock(lock, "read_lock", \
-		      (unsigned long) __builtin_return_address(0)); \
-	local_irq_restore(flags); \
-} while(0)
-
-#define _raw_read_unlock(lock) \
-do {	unsigned long flags; \
-	local_irq_save(flags); \
-	_do_read_unlock(lock, "read_unlock", \
-		      (unsigned long) __builtin_return_address(0)); \
-	local_irq_restore(flags); \
-} while(0)
-
-#define _raw_write_lock(lock) \
-do {	unsigned long flags; \
-	local_irq_save(flags); \
-	_do_write_lock(lock, "write_lock", \
-		      (unsigned long) __builtin_return_address(0)); \
-	local_irq_restore(flags); \
-} while(0)
-
-#define _raw_write_unlock(lock) \
-do {	unsigned long flags; \
-	local_irq_save(flags); \
-	_do_write_unlock(lock, \
-		      (unsigned long) __builtin_return_address(0)); \
-	local_irq_restore(flags); \
-} while(0)
-
-#define _raw_write_trylock(lock) \
-({	unsigned long flags; \
-	int val; \
-	local_irq_save(flags); \
-	val = _do_write_trylock(lock, "write_trylock", \
-				(unsigned long) __builtin_return_address(0)); \
-	local_irq_restore(flags); \
-	val; \
-})
-
-#endif /* CONFIG_DEBUG_SPINLOCK */
-
-#define _raw_read_trylock(lock) generic_raw_read_trylock(lock)
-#define read_can_lock(rw)	(!((rw)->lock & 0x80000000UL))
-#define write_can_lock(rw)	(!(rw)->lock)
+#define __raw_read_lock(p)	__read_lock(p)
+#define __raw_read_unlock(p)	__read_unlock(p)
+#define __raw_write_lock(p)	__write_lock(p)
+#define __raw_write_unlock(p)	__write_unlock(p)
+#define __raw_write_trylock(p)	__write_trylock(p)
+
+#define __raw_read_trylock(lock)	generic__raw_read_trylock(lock)
+#define __raw_read_can_lock(rw)		(!((rw)->lock & 0x80000000UL))
+#define __raw_write_can_lock(rw)	(!(rw)->lock)
 
 #endif /* !(__ASSEMBLY__) */
 
diff --git a/include/asm-sparc64/spinlock_types.h b/include/asm-sparc64/spinlock_types.h
new file mode 100644
index 00000000000..e128112a0d7
--- /dev/null
+++ b/include/asm-sparc64/spinlock_types.h
@@ -0,0 +1,20 @@
+#ifndef __SPARC64_SPINLOCK_TYPES_H
+#define __SPARC64_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned char lock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 0 }
+
+typedef struct {
+	volatile unsigned int lock;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ 0 }
+
+#endif
diff --git a/include/asm-x86_64/spinlock.h b/include/asm-x86_64/spinlock.h
index 5aeb57a3baa..69636831ad2 100644
--- a/include/asm-x86_64/spinlock.h
+++ b/include/asm-x86_64/spinlock.h
@@ -6,47 +6,21 @@
 #include <asm/page.h>
 #include <linux/config.h>
 
-extern int printk(const char * fmt, ...)
-	__attribute__ ((format (printf, 1, 2)));
-
 /*
  * Your basic SMP spinlocks, allowing only a single CPU anywhere
- */
-
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_DEBUG_SPINLOCK
-	unsigned magic;
-#endif
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} spinlock_t;
-
-#define SPINLOCK_MAGIC	0xdead4ead
-
-#ifdef CONFIG_DEBUG_SPINLOCK
-#define SPINLOCK_MAGIC_INIT	, SPINLOCK_MAGIC
-#else
-#define SPINLOCK_MAGIC_INIT	/* */
-#endif
-
-#define SPIN_LOCK_UNLOCKED (spinlock_t) { 1 SPINLOCK_MAGIC_INIT }
-
-#define spin_lock_init(x)	do { *(x) = SPIN_LOCK_UNLOCKED; } while(0)
-
-/*
+ *
  * Simple spin lock operations.  There are two variants, one clears IRQ's
  * on the local processor, one does not.
  *
  * We make no fairness assumptions. They have a cost.
+ *
+ * (the type definitions are in asm/spinlock_types.h)
  */
 
-#define spin_is_locked(x)	(*(volatile signed char *)(&(x)->lock) <= 0)
-#define spin_unlock_wait(x)	do { barrier(); } while(spin_is_locked(x))
-#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
+#define __raw_spin_is_locked(x) \
+		(*(volatile signed char *)(&(x)->slock) <= 0)
 
-#define spin_lock_string \
+#define __raw_spin_lock_string \
 	"\n1:\t" \
 	"lock ; decb %0\n\t" \
 	"js 2f\n" \
@@ -58,74 +32,40 @@ typedef struct {
 	"jmp 1b\n" \
 	LOCK_SECTION_END
 
-/*
- * This works. Despite all the confusion.
- * (except on PPro SMP or if we are using OOSTORE)
- * (PPro errata 66, 92)
- */
- 
-#if !defined(CONFIG_X86_OOSTORE) && !defined(CONFIG_X86_PPRO_FENCE)
-
-#define spin_unlock_string \
+#define __raw_spin_unlock_string \
 	"movb $1,%0" \
-		:"=m" (lock->lock) : : "memory"
-
-
-static inline void _raw_spin_unlock(spinlock_t *lock)
-{
-#ifdef CONFIG_DEBUG_SPINLOCK
-	BUG_ON(lock->magic != SPINLOCK_MAGIC);
-	assert_spin_locked(lock);
-#endif
-	__asm__ __volatile__(
-		spin_unlock_string
-	);
-}
-
-#else
-
-#define spin_unlock_string \
-	"xchgb %b0, %1" \
-		:"=q" (oldval), "=m" (lock->lock) \
-		:"0" (oldval) : "memory"
+		:"=m" (lock->slock) : : "memory"
 
-static inline void _raw_spin_unlock(spinlock_t *lock)
+static inline void __raw_spin_lock(raw_spinlock_t *lock)
 {
-	char oldval = 1;
-#ifdef CONFIG_DEBUG_SPINLOCK
-	BUG_ON(lock->magic != SPINLOCK_MAGIC);
-	assert_spin_locked(lock);
-#endif
 	__asm__ __volatile__(
-		spin_unlock_string
-	);
+		__raw_spin_lock_string
+		:"=m" (lock->slock) : : "memory");
 }
 
-#endif
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
 
-static inline int _raw_spin_trylock(spinlock_t *lock)
+static inline int __raw_spin_trylock(raw_spinlock_t *lock)
 {
 	char oldval;
+
 	__asm__ __volatile__(
 		"xchgb %b0,%1"
-		:"=q" (oldval), "=m" (lock->lock)
+		:"=q" (oldval), "=m" (lock->slock)
 		:"0" (0) : "memory");
+
 	return oldval > 0;
 }
 
-static inline void _raw_spin_lock(spinlock_t *lock)
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
 {
-#ifdef CONFIG_DEBUG_SPINLOCK
-	if (lock->magic != SPINLOCK_MAGIC) {
-		printk("eip: %p\n", __builtin_return_address(0));
-		BUG();
-	}
-#endif
 	__asm__ __volatile__(
-		spin_lock_string
-		:"=m" (lock->lock) : : "memory");
+		__raw_spin_unlock_string
+	);
 }
 
+#define __raw_spin_unlock_wait(lock) \
+	do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
 
 /*
  * Read-write spinlocks, allowing multiple readers
@@ -136,33 +76,7 @@ static inline void _raw_spin_lock(spinlock_t *lock)
  * can "mix" irq-safe locks - any writer needs to get a
  * irq-safe write-lock, but readers can get non-irqsafe
  * read-locks.
- */
-typedef struct {
-	volatile unsigned int lock;
-#ifdef CONFIG_DEBUG_SPINLOCK
-	unsigned magic;
-#endif
-#ifdef CONFIG_PREEMPT
-	unsigned int break_lock;
-#endif
-} rwlock_t;
-
-#define RWLOCK_MAGIC	0xdeaf1eed
-
-#ifdef CONFIG_DEBUG_SPINLOCK
-#define RWLOCK_MAGIC_INIT	, RWLOCK_MAGIC
-#else
-#define RWLOCK_MAGIC_INIT	/* */
-#endif
-
-#define RW_LOCK_UNLOCKED (rwlock_t) { RW_LOCK_BIAS RWLOCK_MAGIC_INIT }
-
-#define rwlock_init(x)	do { *(x) = RW_LOCK_UNLOCKED; } while(0)
-
-#define read_can_lock(x)	((int)(x)->lock > 0)
-#define write_can_lock(x)	((x)->lock == RW_LOCK_BIAS)
-
-/*
+ *
  * On x86, we implement read-write locks as a 32-bit counter
  * with the high bit (sign) being the "contended" bit.
  *
@@ -170,29 +84,24 @@ typedef struct {
  *
  * Changed to use the same technique as rw semaphores.  See
  * semaphore.h for details.  -ben
+ *
+ * the helpers are in arch/i386/kernel/semaphore.c
  */
-/* the spinlock helpers are in arch/i386/kernel/semaphore.c */
 
-static inline void _raw_read_lock(rwlock_t *rw)
+#define __raw_read_can_lock(x)		((int)(x)->lock > 0)
+#define __raw_write_can_lock(x)		((x)->lock == RW_LOCK_BIAS)
+
+static inline void __raw_read_lock(raw_rwlock_t *rw)
 {
-#ifdef CONFIG_DEBUG_SPINLOCK
-	BUG_ON(rw->magic != RWLOCK_MAGIC);
-#endif
 	__build_read_lock(rw, "__read_lock_failed");
 }
 
-static inline void _raw_write_lock(rwlock_t *rw)
+static inline void __raw_write_lock(raw_rwlock_t *rw)
 {
-#ifdef CONFIG_DEBUG_SPINLOCK
-	BUG_ON(rw->magic != RWLOCK_MAGIC);
-#endif
 	__build_write_lock(rw, "__write_lock_failed");
 }
 
-#define _raw_read_unlock(rw)		asm volatile("lock ; incl %0" :"=m" ((rw)->lock) : : "memory")
-#define _raw_write_unlock(rw)	asm volatile("lock ; addl $" RW_LOCK_BIAS_STR ",%0":"=m" ((rw)->lock) : : "memory")
-
-static inline int _raw_read_trylock(rwlock_t *lock)
+static inline int __raw_read_trylock(raw_rwlock_t *lock)
 {
 	atomic_t *count = (atomic_t *)lock;
 	atomic_dec(count);
@@ -202,7 +111,7 @@ static inline int _raw_read_trylock(rwlock_t *lock)
 	return 0;
 }
 
-static inline int _raw_write_trylock(rwlock_t *lock)
+static inline int __raw_write_trylock(raw_rwlock_t *lock)
 {
 	atomic_t *count = (atomic_t *)lock;
 	if (atomic_sub_and_test(RW_LOCK_BIAS, count))
@@ -211,4 +120,15 @@ static inline int _raw_write_trylock(rwlock_t *lock)
 	return 0;
 }
 
+static inline void __raw_read_unlock(raw_rwlock_t *rw)
+{
+	asm volatile("lock ; incl %0" :"=m" (rw->lock) : : "memory");
+}
+
+static inline void __raw_write_unlock(raw_rwlock_t *rw)
+{
+	asm volatile("lock ; addl $" RW_LOCK_BIAS_STR ",%0"
+				: "=m" (rw->lock) : : "memory");
+}
+
 #endif /* __ASM_SPINLOCK_H */
diff --git a/include/asm-x86_64/spinlock_types.h b/include/asm-x86_64/spinlock_types.h
new file mode 100644
index 00000000000..59efe849f35
--- /dev/null
+++ b/include/asm-x86_64/spinlock_types.h
@@ -0,0 +1,20 @@
+#ifndef __ASM_SPINLOCK_TYPES_H
+#define __ASM_SPINLOCK_TYPES_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+typedef struct {
+	volatile unsigned int slock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED	{ 1 }
+
+typedef struct {
+	volatile unsigned int lock;
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED		{ RW_LOCK_BIAS }
+
+#endif
diff --git a/include/linux/bit_spinlock.h b/include/linux/bit_spinlock.h
new file mode 100644
index 00000000000..6b20af0bbb7
--- /dev/null
+++ b/include/linux/bit_spinlock.h
@@ -0,0 +1,77 @@
+#ifndef __LINUX_BIT_SPINLOCK_H
+#define __LINUX_BIT_SPINLOCK_H
+
+/*
+ *  bit-based spin_lock()
+ *
+ * Don't use this unless you really need to: spin_lock() and spin_unlock()
+ * are significantly faster.
+ */
+static inline void bit_spin_lock(int bitnum, unsigned long *addr)
+{
+	/*
+	 * Assuming the lock is uncontended, this never enters
+	 * the body of the outer loop. If it is contended, then
+	 * within the inner loop a non-atomic test is used to
+	 * busywait with less bus contention for a good time to
+	 * attempt to acquire the lock bit.
+	 */
+	preempt_disable();
+#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
+	while (test_and_set_bit(bitnum, addr)) {
+		while (test_bit(bitnum, addr)) {
+			preempt_enable();
+			cpu_relax();
+			preempt_disable();
+		}
+	}
+#endif
+	__acquire(bitlock);
+}
+
+/*
+ * Return true if it was acquired
+ */
+static inline int bit_spin_trylock(int bitnum, unsigned long *addr)
+{
+	preempt_disable();
+#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
+	if (test_and_set_bit(bitnum, addr)) {
+		preempt_enable();
+		return 0;
+	}
+#endif
+	__acquire(bitlock);
+	return 1;
+}
+
+/*
+ *  bit-based spin_unlock()
+ */
+static inline void bit_spin_unlock(int bitnum, unsigned long *addr)
+{
+#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
+	BUG_ON(!test_bit(bitnum, addr));
+	smp_mb__before_clear_bit();
+	clear_bit(bitnum, addr);
+#endif
+	preempt_enable();
+	__release(bitlock);
+}
+
+/*
+ * Return true if the lock is held.
+ */
+static inline int bit_spin_is_locked(int bitnum, unsigned long *addr)
+{
+#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
+	return test_bit(bitnum, addr);
+#elif defined CONFIG_PREEMPT
+	return preempt_count();
+#else
+	return 1;
+#endif
+}
+
+#endif /* __LINUX_BIT_SPINLOCK_H */
+
diff --git a/include/linux/jbd.h b/include/linux/jbd.h
index 84321a4cac9..de097269bd7 100644
--- a/include/linux/jbd.h
+++ b/include/linux/jbd.h
@@ -28,6 +28,7 @@
 #include <linux/buffer_head.h>
 #include <linux/journal-head.h>
 #include <linux/stddef.h>
+#include <linux/bit_spinlock.h>
 #include <asm/semaphore.h>
 #endif
 
diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h
index d6ba068719b..cdc99a27840 100644
--- a/include/linux/spinlock.h
+++ b/include/linux/spinlock.h
@@ -2,7 +2,48 @@
 #define __LINUX_SPINLOCK_H
 
 /*
- * include/linux/spinlock.h - generic locking declarations
+ * include/linux/spinlock.h - generic spinlock/rwlock declarations
+ *
+ * here's the role of the various spinlock/rwlock related include files:
+ *
+ * on SMP builds:
+ *
+ *  asm/spinlock_types.h: contains the raw_spinlock_t/raw_rwlock_t and the
+ *                        initializers
+ *
+ *  linux/spinlock_types.h:
+ *                        defines the generic type and initializers
+ *
+ *  asm/spinlock.h:       contains the __raw_spin_*()/etc. lowlevel
+ *                        implementations, mostly inline assembly code
+ *
+ *   (also included on UP-debug builds:)
+ *
+ *  linux/spinlock_api_smp.h:
+ *                        contains the prototypes for the _spin_*() APIs.
+ *
+ *  linux/spinlock.h:     builds the final spin_*() APIs.
+ *
+ * on UP builds:
+ *
+ *  linux/spinlock_type_up.h:
+ *                        contains the generic, simplified UP spinlock type.
+ *                        (which is an empty structure on non-debug builds)
+ *
+ *  linux/spinlock_types.h:
+ *                        defines the generic type and initializers
+ *
+ *  linux/spinlock_up.h:
+ *                        contains the __raw_spin_*()/etc. version of UP
+ *                        builds. (which are NOPs on non-debug, non-preempt
+ *                        builds)
+ *
+ *   (included on UP-non-debug builds:)
+ *
+ *  linux/spinlock_api_up.h:
+ *                        builds the _spin_*() APIs.
+ *
+ *  linux/spinlock.h:     builds the final spin_*() APIs.
  */
 
 #include <linux/config.h>
@@ -13,7 +54,6 @@
 #include <linux/kernel.h>
 #include <linux/stringify.h>
 
-#include <asm/processor.h>	/* for cpu relax */
 #include <asm/system.h>
 
 /*
@@ -35,423 +75,84 @@
 #define __lockfunc fastcall __attribute__((section(".spinlock.text")))
 
 /*
- * If CONFIG_SMP is set, pull in the _raw_* definitions
+ * Pull the raw_spinlock_t and raw_rwlock_t definitions:
  */
-#ifdef CONFIG_SMP
-
-#define assert_spin_locked(x)	BUG_ON(!spin_is_locked(x))
-#include <asm/spinlock.h>
-
-int __lockfunc _spin_trylock(spinlock_t *lock);
-int __lockfunc _read_trylock(rwlock_t *lock);
-int __lockfunc _write_trylock(rwlock_t *lock);
-
-void __lockfunc _spin_lock(spinlock_t *lock)	__acquires(spinlock_t);
-void __lockfunc _read_lock(rwlock_t *lock)	__acquires(rwlock_t);
-void __lockfunc _write_lock(rwlock_t *lock)	__acquires(rwlock_t);
-
-void __lockfunc _spin_unlock(spinlock_t *lock)	__releases(spinlock_t);
-void __lockfunc _read_unlock(rwlock_t *lock)	__releases(rwlock_t);
-void __lockfunc _write_unlock(rwlock_t *lock)	__releases(rwlock_t);
-
-unsigned long __lockfunc _spin_lock_irqsave(spinlock_t *lock)	__acquires(spinlock_t);
-unsigned long __lockfunc _read_lock_irqsave(rwlock_t *lock)	__acquires(rwlock_t);
-unsigned long __lockfunc _write_lock_irqsave(rwlock_t *lock)	__acquires(rwlock_t);
-
-void __lockfunc _spin_lock_irq(spinlock_t *lock)	__acquires(spinlock_t);
-void __lockfunc _spin_lock_bh(spinlock_t *lock)		__acquires(spinlock_t);
-void __lockfunc _read_lock_irq(rwlock_t *lock)		__acquires(rwlock_t);
-void __lockfunc _read_lock_bh(rwlock_t *lock)		__acquires(rwlock_t);
-void __lockfunc _write_lock_irq(rwlock_t *lock)		__acquires(rwlock_t);
-void __lockfunc _write_lock_bh(rwlock_t *lock)		__acquires(rwlock_t);
-
-void __lockfunc _spin_unlock_irqrestore(spinlock_t *lock, unsigned long flags)	__releases(spinlock_t);
-void __lockfunc _spin_unlock_irq(spinlock_t *lock)				__releases(spinlock_t);
-void __lockfunc _spin_unlock_bh(spinlock_t *lock)				__releases(spinlock_t);
-void __lockfunc _read_unlock_irqrestore(rwlock_t *lock, unsigned long flags)	__releases(rwlock_t);
-void __lockfunc _read_unlock_irq(rwlock_t *lock)				__releases(rwlock_t);
-void __lockfunc _read_unlock_bh(rwlock_t *lock)					__releases(rwlock_t);
-void __lockfunc _write_unlock_irqrestore(rwlock_t *lock, unsigned long flags)	__releases(rwlock_t);
-void __lockfunc _write_unlock_irq(rwlock_t *lock)				__releases(rwlock_t);
-void __lockfunc _write_unlock_bh(rwlock_t *lock)				__releases(rwlock_t);
-
-int __lockfunc _spin_trylock_bh(spinlock_t *lock);
-int __lockfunc generic_raw_read_trylock(rwlock_t *lock);
-int in_lock_functions(unsigned long addr);
-
-#else
+#include <linux/spinlock_types.h>
 
-#define in_lock_functions(ADDR) 0
+extern int __lockfunc generic__raw_read_trylock(raw_rwlock_t *lock);
 
-#if !defined(CONFIG_PREEMPT) && !defined(CONFIG_DEBUG_SPINLOCK)
-# define _atomic_dec_and_lock(atomic,lock) atomic_dec_and_test(atomic)
-# define ATOMIC_DEC_AND_LOCK
-#endif
-
-#ifdef CONFIG_DEBUG_SPINLOCK
- 
-#define SPINLOCK_MAGIC	0x1D244B3C
-typedef struct {
-	unsigned long magic;
-	volatile unsigned long lock;
-	volatile unsigned int babble;
-	const char *module;
-	char *owner;
-	int oline;
-} spinlock_t;
-#define SPIN_LOCK_UNLOCKED (spinlock_t) { SPINLOCK_MAGIC, 0, 10, __FILE__ , NULL, 0}
-
-#define spin_lock_init(x) \
-	do { \
-		(x)->magic = SPINLOCK_MAGIC; \
-		(x)->lock = 0; \
-		(x)->babble = 5; \
-		(x)->module = __FILE__; \
-		(x)->owner = NULL; \
-		(x)->oline = 0; \
-	} while (0)
-
-#define CHECK_LOCK(x) \
-	do { \
-	 	if ((x)->magic != SPINLOCK_MAGIC) { \
-			printk(KERN_ERR "%s:%d: spin_is_locked on uninitialized spinlock %p.\n", \
-					__FILE__, __LINE__, (x)); \
-		} \
-	} while(0)
-
-#define _raw_spin_lock(x)		\
-	do { \
-	 	CHECK_LOCK(x); \
-		if ((x)->lock&&(x)->babble) { \
-			(x)->babble--; \
-			printk("%s:%d: spin_lock(%s:%p) already locked by %s/%d\n", \
-					__FILE__,__LINE__, (x)->module, \
-					(x), (x)->owner, (x)->oline); \
-		} \
-		(x)->lock = 1; \
-		(x)->owner = __FILE__; \
-		(x)->oline = __LINE__; \
-	} while (0)
-
-/* without debugging, spin_is_locked on UP always says
- * FALSE. --> printk if already locked. */
-#define spin_is_locked(x) \
-	({ \
-	 	CHECK_LOCK(x); \
-		if ((x)->lock&&(x)->babble) { \
-			(x)->babble--; \
-			printk("%s:%d: spin_is_locked(%s:%p) already locked by %s/%d\n", \
-					__FILE__,__LINE__, (x)->module, \
-					(x), (x)->owner, (x)->oline); \
-		} \
-		0; \
-	})
-
-/* with debugging, assert_spin_locked() on UP does check
- * the lock value properly */
-#define assert_spin_locked(x) \
-	({ \
-		CHECK_LOCK(x); \
-		BUG_ON(!(x)->lock); \
-	})
-
-/* without debugging, spin_trylock on UP always says
- * TRUE. --> printk if already locked. */
-#define _raw_spin_trylock(x) \
-	({ \
-	 	CHECK_LOCK(x); \
-		if ((x)->lock&&(x)->babble) { \
-			(x)->babble--; \
-			printk("%s:%d: spin_trylock(%s:%p) already locked by %s/%d\n", \
-					__FILE__,__LINE__, (x)->module, \
-					(x), (x)->owner, (x)->oline); \
-		} \
-		(x)->lock = 1; \
-		(x)->owner = __FILE__; \
-		(x)->oline = __LINE__; \
-		1; \
-	})
-
-#define spin_unlock_wait(x)	\
-	do { \
-	 	CHECK_LOCK(x); \
-		if ((x)->lock&&(x)->babble) { \
-			(x)->babble--; \
-			printk("%s:%d: spin_unlock_wait(%s:%p) owned by %s/%d\n", \
-					__FILE__,__LINE__, (x)->module, (x), \
-					(x)->owner, (x)->oline); \
-		}\
-	} while (0)
-
-#define _raw_spin_unlock(x) \
-	do { \
-	 	CHECK_LOCK(x); \
-		if (!(x)->lock&&(x)->babble) { \
-			(x)->babble--; \
-			printk("%s:%d: spin_unlock(%s:%p) not locked\n", \
-					__FILE__,__LINE__, (x)->module, (x));\
-		} \
-		(x)->lock = 0; \
-	} while (0)
-#else
 /*
- * gcc versions before ~2.95 have a nasty bug with empty initializers.
+ * Pull the __raw*() functions/declarations (UP-nondebug doesnt need them):
  */
-#if (__GNUC__ > 2)
-  typedef struct { } spinlock_t;
-  #define SPIN_LOCK_UNLOCKED (spinlock_t) { }
+#if defined(CONFIG_SMP)
+# include <asm/spinlock.h>
 #else
-  typedef struct { int gcc_is_buggy; } spinlock_t;
-  #define SPIN_LOCK_UNLOCKED (spinlock_t) { 0 }
+# include <linux/spinlock_up.h>
 #endif
 
+#define spin_lock_init(lock)	do { *(lock) = SPIN_LOCK_UNLOCKED; } while (0)
+#define rwlock_init(lock)	do { *(lock) = RW_LOCK_UNLOCKED; } while (0)
+
+#define spin_is_locked(lock)	__raw_spin_is_locked(&(lock)->raw_lock)
+
+/**
+ * spin_unlock_wait - wait until the spinlock gets unlocked
+ * @lock: the spinlock in question.
+ */
+#define spin_unlock_wait(lock)	__raw_spin_unlock_wait(&(lock)->raw_lock)
+
 /*
- * If CONFIG_SMP is unset, declare the _raw_* definitions as nops
+ * Pull the _spin_*()/_read_*()/_write_*() functions/declarations:
  */
-#define spin_lock_init(lock)	do { (void)(lock); } while(0)
-#define _raw_spin_lock(lock)	do { (void)(lock); } while(0)
-#define spin_is_locked(lock)	((void)(lock), 0)
-#define assert_spin_locked(lock)	do { (void)(lock); } while(0)
-#define _raw_spin_trylock(lock)	(((void)(lock), 1))
-#define spin_unlock_wait(lock)	(void)(lock)
-#define _raw_spin_unlock(lock) do { (void)(lock); } while(0)
-#endif /* CONFIG_DEBUG_SPINLOCK */
-
-/* RW spinlocks: No debug version */
-
-#if (__GNUC__ > 2)
-  typedef struct { } rwlock_t;
-  #define RW_LOCK_UNLOCKED (rwlock_t) { }
+#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
+# include <linux/spinlock_api_smp.h>
 #else
-  typedef struct { int gcc_is_buggy; } rwlock_t;
-  #define RW_LOCK_UNLOCKED (rwlock_t) { 0 }
+# include <linux/spinlock_api_up.h>
 #endif
 
-#define rwlock_init(lock)	do { (void)(lock); } while(0)
-#define _raw_read_lock(lock)	do { (void)(lock); } while(0)
-#define _raw_read_unlock(lock)	do { (void)(lock); } while(0)
-#define _raw_write_lock(lock)	do { (void)(lock); } while(0)
-#define _raw_write_unlock(lock)	do { (void)(lock); } while(0)
-#define read_can_lock(lock)	(((void)(lock), 1))
-#define write_can_lock(lock)	(((void)(lock), 1))
-#define _raw_read_trylock(lock) ({ (void)(lock); (1); })
-#define _raw_write_trylock(lock) ({ (void)(lock); (1); })
-
-#define _spin_trylock(lock)	({preempt_disable(); _raw_spin_trylock(lock) ? \
-				1 : ({preempt_enable(); 0;});})
-
-#define _read_trylock(lock)	({preempt_disable();_raw_read_trylock(lock) ? \
-				1 : ({preempt_enable(); 0;});})
-
-#define _write_trylock(lock)	({preempt_disable(); _raw_write_trylock(lock) ? \
-				1 : ({preempt_enable(); 0;});})
-
-#define _spin_trylock_bh(lock)	({preempt_disable(); local_bh_disable(); \
-				_raw_spin_trylock(lock) ? \
-				1 : ({preempt_enable_no_resched(); local_bh_enable(); 0;});})
-
-#define _spin_lock(lock)	\
-do { \
-	preempt_disable(); \
-	_raw_spin_lock(lock); \
-	__acquire(lock); \
-} while(0)
-
-#define _write_lock(lock) \
-do { \
-	preempt_disable(); \
-	_raw_write_lock(lock); \
-	__acquire(lock); \
-} while(0)
- 
-#define _read_lock(lock)	\
-do { \
-	preempt_disable(); \
-	_raw_read_lock(lock); \
-	__acquire(lock); \
-} while(0)
-
-#define _spin_unlock(lock) \
-do { \
-	_raw_spin_unlock(lock); \
-	preempt_enable(); \
-	__release(lock); \
-} while (0)
-
-#define _write_unlock(lock) \
-do { \
-	_raw_write_unlock(lock); \
-	preempt_enable(); \
-	__release(lock); \
-} while(0)
-
-#define _read_unlock(lock) \
-do { \
-	_raw_read_unlock(lock); \
-	preempt_enable(); \
-	__release(lock); \
-} while(0)
-
-#define _spin_lock_irqsave(lock, flags) \
-do {	\
-	local_irq_save(flags); \
-	preempt_disable(); \
-	_raw_spin_lock(lock); \
-	__acquire(lock); \
-} while (0)
-
-#define _spin_lock_irq(lock) \
-do { \
-	local_irq_disable(); \
-	preempt_disable(); \
-	_raw_spin_lock(lock); \
-	__acquire(lock); \
-} while (0)
-
-#define _spin_lock_bh(lock) \
-do { \
-	local_bh_disable(); \
-	preempt_disable(); \
-	_raw_spin_lock(lock); \
-	__acquire(lock); \
-} while (0)
-
-#define _read_lock_irqsave(lock, flags) \
-do {	\
-	local_irq_save(flags); \
-	preempt_disable(); \
-	_raw_read_lock(lock); \
-	__acquire(lock); \
-} while (0)
-
-#define _read_lock_irq(lock) \
-do { \
-	local_irq_disable(); \
-	preempt_disable(); \
-	_raw_read_lock(lock); \
-	__acquire(lock); \
-} while (0)
-
-#define _read_lock_bh(lock) \
-do { \
-	local_bh_disable(); \
-	preempt_disable(); \
-	_raw_read_lock(lock); \
-	__acquire(lock); \
-} while (0)
-
-#define _write_lock_irqsave(lock, flags) \
-do {	\
-	local_irq_save(flags); \
-	preempt_disable(); \
-	_raw_write_lock(lock); \
-	__acquire(lock); \
-} while (0)
+#ifdef CONFIG_DEBUG_SPINLOCK
+ extern void _raw_spin_lock(spinlock_t *lock);
+#define _raw_spin_lock_flags(lock, flags) _raw_spin_lock(lock)
+ extern int _raw_spin_trylock(spinlock_t *lock);
+ extern void _raw_spin_unlock(spinlock_t *lock);
+
+ extern void _raw_read_lock(rwlock_t *lock);
+ extern int _raw_read_trylock(rwlock_t *lock);
+ extern void _raw_read_unlock(rwlock_t *lock);
+ extern void _raw_write_lock(rwlock_t *lock);
+ extern int _raw_write_trylock(rwlock_t *lock);
+ extern void _raw_write_unlock(rwlock_t *lock);
+#else
+# define _raw_spin_unlock(lock)		__raw_spin_unlock(&(lock)->raw_lock)
+# define _raw_spin_trylock(lock)	__raw_spin_trylock(&(lock)->raw_lock)
+# define _raw_spin_lock(lock)		__raw_spin_lock(&(lock)->raw_lock)
+# define _raw_spin_lock_flags(lock, flags) \
+		__raw_spin_lock_flags(&(lock)->raw_lock, *(flags))
+# define _raw_read_lock(rwlock)		__raw_read_lock(&(rwlock)->raw_lock)
+# define _raw_write_lock(rwlock)	__raw_write_lock(&(rwlock)->raw_lock)
+# define _raw_read_unlock(rwlock)	__raw_read_unlock(&(rwlock)->raw_lock)
+# define _raw_write_unlock(rwlock)	__raw_write_unlock(&(rwlock)->raw_lock)
+# define _raw_read_trylock(rwlock)	__raw_read_trylock(&(rwlock)->raw_lock)
+# define _raw_write_trylock(rwlock)	__raw_write_trylock(&(rwlock)->raw_lock)
+#endif
 
-#define _write_lock_irq(lock) \
-do { \
-	local_irq_disable(); \
-	preempt_disable(); \
-	_raw_write_lock(lock); \
-	__acquire(lock); \
-} while (0)
-
-#define _write_lock_bh(lock) \
-do { \
-	local_bh_disable(); \
-	preempt_disable(); \
-	_raw_write_lock(lock); \
-	__acquire(lock); \
-} while (0)
-
-#define _spin_unlock_irqrestore(lock, flags) \
-do { \
-	_raw_spin_unlock(lock); \
-	local_irq_restore(flags); \
-	preempt_enable(); \
-	__release(lock); \
-} while (0)
-
-#define _spin_unlock_irq(lock) \
-do { \
-	_raw_spin_unlock(lock); \
-	local_irq_enable(); \
-	preempt_enable(); \
-	__release(lock); \
-} while (0)
-
-#define _spin_unlock_bh(lock) \
-do { \
-	_raw_spin_unlock(lock); \
-	preempt_enable_no_resched(); \
-	local_bh_enable(); \
-	__release(lock); \
-} while (0)
-
-#define _write_unlock_bh(lock) \
-do { \
-	_raw_write_unlock(lock); \
-	preempt_enable_no_resched(); \
-	local_bh_enable(); \
-	__release(lock); \
-} while (0)
-
-#define _read_unlock_irqrestore(lock, flags) \
-do { \
-	_raw_read_unlock(lock); \
-	local_irq_restore(flags); \
-	preempt_enable(); \
-	__release(lock); \
-} while (0)
-
-#define _write_unlock_irqrestore(lock, flags) \
-do { \
-	_raw_write_unlock(lock); \
-	local_irq_restore(flags); \
-	preempt_enable(); \
-	__release(lock); \
-} while (0)
-
-#define _read_unlock_irq(lock)	\
-do { \
-	_raw_read_unlock(lock);	\
-	local_irq_enable();	\
-	preempt_enable();	\
-	__release(lock); \
-} while (0)
-
-#define _read_unlock_bh(lock)	\
-do { \
-	_raw_read_unlock(lock);	\
-	preempt_enable_no_resched();	\
-	local_bh_enable();	\
-	__release(lock); \
-} while (0)
-
-#define _write_unlock_irq(lock)	\
-do { \
-	_raw_write_unlock(lock);	\
-	local_irq_enable();	\
-	preempt_enable();	\
-	__release(lock); \
-} while (0)
-
-#endif /* !SMP */
+#define read_can_lock(rwlock)		__raw_read_can_lock(&(rwlock)->raw_lock)
+#define write_can_lock(rwlock)		__raw_write_can_lock(&(rwlock)->raw_lock)
 
 /*
  * Define the various spin_lock and rw_lock methods.  Note we define these
  * regardless of whether CONFIG_SMP or CONFIG_PREEMPT are set. The various
  * methods are defined as nops in the case they are not required.
  */
-#define spin_trylock(lock)	__cond_lock(_spin_trylock(lock))
-#define read_trylock(lock)	__cond_lock(_read_trylock(lock))
-#define write_trylock(lock)	__cond_lock(_write_trylock(lock))
+#define spin_trylock(lock)		__cond_lock(_spin_trylock(lock))
+#define read_trylock(lock)		__cond_lock(_read_trylock(lock))
+#define write_trylock(lock)		__cond_lock(_write_trylock(lock))
 
-#define spin_lock(lock)		_spin_lock(lock)
-#define write_lock(lock)	_write_lock(lock)
-#define read_lock(lock)		_read_lock(lock)
+#define spin_lock(lock)			_spin_lock(lock)
+#define write_lock(lock)		_write_lock(lock)
+#define read_lock(lock)			_read_lock(lock)
 
-#ifdef CONFIG_SMP
+#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
 #define spin_lock_irqsave(lock, flags)	flags = _spin_lock_irqsave(lock)
 #define read_lock_irqsave(lock, flags)	flags = _read_lock_irqsave(lock)
 #define write_lock_irqsave(lock, flags)	flags = _write_lock_irqsave(lock)
@@ -470,137 +171,59 @@ do { \
 #define write_lock_irq(lock)		_write_lock_irq(lock)
 #define write_lock_bh(lock)		_write_lock_bh(lock)
 
-#define spin_unlock(lock)	_spin_unlock(lock)
-#define write_unlock(lock)	_write_unlock(lock)
-#define read_unlock(lock)	_read_unlock(lock)
+#define spin_unlock(lock)		_spin_unlock(lock)
+#define write_unlock(lock)		_write_unlock(lock)
+#define read_unlock(lock)		_read_unlock(lock)
 
-#define spin_unlock_irqrestore(lock, flags)	_spin_unlock_irqrestore(lock, flags)
+#define spin_unlock_irqrestore(lock, flags) \
+					_spin_unlock_irqrestore(lock, flags)
 #define spin_unlock_irq(lock)		_spin_unlock_irq(lock)
 #define spin_unlock_bh(lock)		_spin_unlock_bh(lock)
 
-#define read_unlock_irqrestore(lock, flags)	_read_unlock_irqrestore(lock, flags)
-#define read_unlock_irq(lock)			_read_unlock_irq(lock)
-#define read_unlock_bh(lock)			_read_unlock_bh(lock)
+#define read_unlock_irqrestore(lock, flags) \
+					_read_unlock_irqrestore(lock, flags)
+#define read_unlock_irq(lock)		_read_unlock_irq(lock)
+#define read_unlock_bh(lock)		_read_unlock_bh(lock)
 
-#define write_unlock_irqrestore(lock, flags)	_write_unlock_irqrestore(lock, flags)
-#define write_unlock_irq(lock)			_write_unlock_irq(lock)
-#define write_unlock_bh(lock)			_write_unlock_bh(lock)
+#define write_unlock_irqrestore(lock, flags) \
+					_write_unlock_irqrestore(lock, flags)
+#define write_unlock_irq(lock)		_write_unlock_irq(lock)
+#define write_unlock_bh(lock)		_write_unlock_bh(lock)
 
-#define spin_trylock_bh(lock)			__cond_lock(_spin_trylock_bh(lock))
+#define spin_trylock_bh(lock)		__cond_lock(_spin_trylock_bh(lock))
 
 #define spin_trylock_irq(lock) \
 ({ \
 	local_irq_disable(); \
 	_spin_trylock(lock) ? \
-	1 : ({local_irq_enable(); 0; }); \
+	1 : ({ local_irq_enable(); 0;  }); \
 })
 
 #define spin_trylock_irqsave(lock, flags) \
 ({ \
 	local_irq_save(flags); \
 	_spin_trylock(lock) ? \
-	1 : ({local_irq_restore(flags); 0;}); \
+	1 : ({ local_irq_restore(flags); 0; }); \
 })
 
-#ifdef CONFIG_LOCKMETER
-extern void _metered_spin_lock   (spinlock_t *lock);
-extern void _metered_spin_unlock (spinlock_t *lock);
-extern int  _metered_spin_trylock(spinlock_t *lock);
-extern void _metered_read_lock    (rwlock_t *lock);
-extern void _metered_read_unlock  (rwlock_t *lock);
-extern void _metered_write_lock   (rwlock_t *lock);
-extern void _metered_write_unlock (rwlock_t *lock);
-extern int  _metered_read_trylock (rwlock_t *lock);
-extern int  _metered_write_trylock(rwlock_t *lock);
-#endif
-
-/* "lock on reference count zero" */
-#ifndef ATOMIC_DEC_AND_LOCK
-#include <asm/atomic.h>
-extern int _atomic_dec_and_lock(atomic_t *atomic, spinlock_t *lock);
-#endif
-
-#define atomic_dec_and_lock(atomic,lock) __cond_lock(_atomic_dec_and_lock(atomic,lock))
-
-/*
- *  bit-based spin_lock()
- *
- * Don't use this unless you really need to: spin_lock() and spin_unlock()
- * are significantly faster.
- */
-static inline void bit_spin_lock(int bitnum, unsigned long *addr)
-{
-	/*
-	 * Assuming the lock is uncontended, this never enters
-	 * the body of the outer loop. If it is contended, then
-	 * within the inner loop a non-atomic test is used to
-	 * busywait with less bus contention for a good time to
-	 * attempt to acquire the lock bit.
-	 */
-	preempt_disable();
-#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
-	while (test_and_set_bit(bitnum, addr)) {
-		while (test_bit(bitnum, addr)) {
-			preempt_enable();
-			cpu_relax();
-			preempt_disable();
-		}
-	}
-#endif
-	__acquire(bitlock);
-}
-
 /*
- * Return true if it was acquired
+ * Pull the atomic_t declaration:
+ * (asm-mips/atomic.h needs above definitions)
  */
-static inline int bit_spin_trylock(int bitnum, unsigned long *addr)
-{
-	preempt_disable();	
-#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
-	if (test_and_set_bit(bitnum, addr)) {
-		preempt_enable();
-		return 0;
-	}
-#endif
-	__acquire(bitlock);
-	return 1;
-}
-
-/*
- *  bit-based spin_unlock()
- */
-static inline void bit_spin_unlock(int bitnum, unsigned long *addr)
-{
-#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
-	BUG_ON(!test_bit(bitnum, addr));
-	smp_mb__before_clear_bit();
-	clear_bit(bitnum, addr);
-#endif
-	preempt_enable();
-	__release(bitlock);
-}
-
-/*
- * Return true if the lock is held.
+#include <asm/atomic.h>
+/**
+ * atomic_dec_and_lock - lock on reaching reference count zero
+ * @atomic: the atomic counter
+ * @lock: the spinlock in question
  */
-static inline int bit_spin_is_locked(int bitnum, unsigned long *addr)
-{
-#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
-	return test_bit(bitnum, addr);
-#elif defined CONFIG_PREEMPT
-	return preempt_count();
-#else
-	return 1;
-#endif
-}
-
-#define DEFINE_SPINLOCK(x) spinlock_t x = SPIN_LOCK_UNLOCKED
-#define DEFINE_RWLOCK(x) rwlock_t x = RW_LOCK_UNLOCKED
+extern int _atomic_dec_and_lock(atomic_t *atomic, spinlock_t *lock);
+#define atomic_dec_and_lock(atomic, lock) \
+		__cond_lock(_atomic_dec_and_lock(atomic, lock))
 
 /**
  * spin_can_lock - would spin_trylock() succeed?
  * @lock: the spinlock in question.
  */
-#define spin_can_lock(lock)		(!spin_is_locked(lock))
+#define spin_can_lock(lock)	(!spin_is_locked(lock))
 
 #endif /* __LINUX_SPINLOCK_H */
diff --git a/include/linux/spinlock_api_smp.h b/include/linux/spinlock_api_smp.h
new file mode 100644
index 00000000000..78e6989ffb5
--- /dev/null
+++ b/include/linux/spinlock_api_smp.h
@@ -0,0 +1,57 @@
+#ifndef __LINUX_SPINLOCK_API_SMP_H
+#define __LINUX_SPINLOCK_API_SMP_H
+
+#ifndef __LINUX_SPINLOCK_H
+# error "please don't include this file directly"
+#endif
+
+/*
+ * include/linux/spinlock_api_smp.h
+ *
+ * spinlock API declarations on SMP (and debug)
+ * (implemented in kernel/spinlock.c)
+ *
+ * portions Copyright 2005, Red Hat, Inc., Ingo Molnar
+ * Released under the General Public License (GPL).
+ */
+
+int in_lock_functions(unsigned long addr);
+
+#define assert_spin_locked(x)	BUG_ON(!spin_is_locked(x))
+
+void __lockfunc _spin_lock(spinlock_t *lock)		__acquires(spinlock_t);
+void __lockfunc _read_lock(rwlock_t *lock)		__acquires(rwlock_t);
+void __lockfunc _write_lock(rwlock_t *lock)		__acquires(rwlock_t);
+void __lockfunc _spin_lock_bh(spinlock_t *lock)		__acquires(spinlock_t);
+void __lockfunc _read_lock_bh(rwlock_t *lock)		__acquires(rwlock_t);
+void __lockfunc _write_lock_bh(rwlock_t *lock)		__acquires(rwlock_t);
+void __lockfunc _spin_lock_irq(spinlock_t *lock)	__acquires(spinlock_t);
+void __lockfunc _read_lock_irq(rwlock_t *lock)		__acquires(rwlock_t);
+void __lockfunc _write_lock_irq(rwlock_t *lock)		__acquires(rwlock_t);
+unsigned long __lockfunc _spin_lock_irqsave(spinlock_t *lock)
+							__acquires(spinlock_t);
+unsigned long __lockfunc _read_lock_irqsave(rwlock_t *lock)
+							__acquires(rwlock_t);
+unsigned long __lockfunc _write_lock_irqsave(rwlock_t *lock)
+							__acquires(rwlock_t);
+int __lockfunc _spin_trylock(spinlock_t *lock);
+int __lockfunc _read_trylock(rwlock_t *lock);
+int __lockfunc _write_trylock(rwlock_t *lock);
+int __lockfunc _spin_trylock_bh(spinlock_t *lock);
+void __lockfunc _spin_unlock(spinlock_t *lock)		__releases(spinlock_t);
+void __lockfunc _read_unlock(rwlock_t *lock)		__releases(rwlock_t);
+void __lockfunc _write_unlock(rwlock_t *lock)		__releases(rwlock_t);
+void __lockfunc _spin_unlock_bh(spinlock_t *lock)	__releases(spinlock_t);
+void __lockfunc _read_unlock_bh(rwlock_t *lock)		__releases(rwlock_t);
+void __lockfunc _write_unlock_bh(rwlock_t *lock)	__releases(rwlock_t);
+void __lockfunc _spin_unlock_irq(spinlock_t *lock)	__releases(spinlock_t);
+void __lockfunc _read_unlock_irq(rwlock_t *lock)	__releases(rwlock_t);
+void __lockfunc _write_unlock_irq(rwlock_t *lock)	__releases(rwlock_t);
+void __lockfunc _spin_unlock_irqrestore(spinlock_t *lock, unsigned long flags)
+							__releases(spinlock_t);
+void __lockfunc _read_unlock_irqrestore(rwlock_t *lock, unsigned long flags)
+							__releases(rwlock_t);
+void __lockfunc _write_unlock_irqrestore(rwlock_t *lock, unsigned long flags)
+							__releases(rwlock_t);
+
+#endif /* __LINUX_SPINLOCK_API_SMP_H */
diff --git a/include/linux/spinlock_api_up.h b/include/linux/spinlock_api_up.h
new file mode 100644
index 00000000000..cd81cee566f
--- /dev/null
+++ b/include/linux/spinlock_api_up.h
@@ -0,0 +1,80 @@
+#ifndef __LINUX_SPINLOCK_API_UP_H
+#define __LINUX_SPINLOCK_API_UP_H
+
+#ifndef __LINUX_SPINLOCK_H
+# error "please don't include this file directly"
+#endif
+
+/*
+ * include/linux/spinlock_api_up.h
+ *
+ * spinlock API implementation on UP-nondebug (inlined implementation)
+ *
+ * portions Copyright 2005, Red Hat, Inc., Ingo Molnar
+ * Released under the General Public License (GPL).
+ */
+
+#define in_lock_functions(ADDR)		0
+
+#define assert_spin_locked(lock)	do { (void)(lock); } while (0)
+
+/*
+ * In the UP-nondebug case there's no real locking going on, so the
+ * only thing we have to do is to keep the preempt counts and irq
+ * flags straight, to supress compiler warnings of unused lock
+ * variables, and to add the proper checker annotations:
+ */
+#define __LOCK(lock) \
+  do { preempt_disable(); __acquire(lock); (void)(lock); } while (0)
+
+#define __LOCK_BH(lock) \
+  do { local_bh_disable(); __LOCK(lock); } while (0)
+
+#define __LOCK_IRQ(lock) \
+  do { local_irq_disable(); __LOCK(lock); } while (0)
+
+#define __LOCK_IRQSAVE(lock, flags) \
+  do { local_irq_save(flags); __LOCK(lock); } while (0)
+
+#define __UNLOCK(lock) \
+  do { preempt_enable(); __release(lock); (void)(lock); } while (0)
+
+#define __UNLOCK_BH(lock) \
+  do { preempt_enable_no_resched(); local_bh_enable(); __release(lock); (void)(lock); } while (0)
+
+#define __UNLOCK_IRQ(lock) \
+  do { local_irq_enable(); __UNLOCK(lock); } while (0)
+
+#define __UNLOCK_IRQRESTORE(lock, flags) \
+  do { local_irq_restore(flags); __UNLOCK(lock); } while (0)
+
+#define _spin_lock(lock)			__LOCK(lock)
+#define _read_lock(lock)			__LOCK(lock)
+#define _write_lock(lock)			__LOCK(lock)
+#define _spin_lock_bh(lock)			__LOCK_BH(lock)
+#define _read_lock_bh(lock)			__LOCK_BH(lock)
+#define _write_lock_bh(lock)			__LOCK_BH(lock)
+#define _spin_lock_irq(lock)			__LOCK_IRQ(lock)
+#define _read_lock_irq(lock)			__LOCK_IRQ(lock)
+#define _write_lock_irq(lock)			__LOCK_IRQ(lock)
+#define _spin_lock_irqsave(lock, flags)		__LOCK_IRQSAVE(lock, flags)
+#define _read_lock_irqsave(lock, flags)		__LOCK_IRQSAVE(lock, flags)
+#define _write_lock_irqsave(lock, flags)	__LOCK_IRQSAVE(lock, flags)
+#define _spin_trylock(lock)			({ __LOCK(lock); 1; })
+#define _read_trylock(lock)			({ __LOCK(lock); 1; })
+#define _write_trylock(lock)			({ __LOCK(lock); 1; })
+#define _spin_trylock_bh(lock)			({ __LOCK_BH(lock); 1; })
+#define _spin_unlock(lock)			__UNLOCK(lock)
+#define _read_unlock(lock)			__UNLOCK(lock)
+#define _write_unlock(lock)			__UNLOCK(lock)
+#define _spin_unlock_bh(lock)			__UNLOCK_BH(lock)
+#define _write_unlock_bh(lock)			__UNLOCK_BH(lock)
+#define _read_unlock_bh(lock)			__UNLOCK_BH(lock)
+#define _spin_unlock_irq(lock)			__UNLOCK_IRQ(lock)
+#define _read_unlock_irq(lock)			__UNLOCK_IRQ(lock)
+#define _write_unlock_irq(lock)			__UNLOCK_IRQ(lock)
+#define _spin_unlock_irqrestore(lock, flags)	__UNLOCK_IRQRESTORE(lock, flags)
+#define _read_unlock_irqrestore(lock, flags)	__UNLOCK_IRQRESTORE(lock, flags)
+#define _write_unlock_irqrestore(lock, flags)	__UNLOCK_IRQRESTORE(lock, flags)
+
+#endif /* __LINUX_SPINLOCK_API_UP_H */
diff --git a/include/linux/spinlock_types.h b/include/linux/spinlock_types.h
new file mode 100644
index 00000000000..9cb51e07039
--- /dev/null
+++ b/include/linux/spinlock_types.h
@@ -0,0 +1,67 @@
+#ifndef __LINUX_SPINLOCK_TYPES_H
+#define __LINUX_SPINLOCK_TYPES_H
+
+/*
+ * include/linux/spinlock_types.h - generic spinlock type definitions
+ *                                  and initializers
+ *
+ * portions Copyright 2005, Red Hat, Inc., Ingo Molnar
+ * Released under the General Public License (GPL).
+ */
+
+#if defined(CONFIG_SMP)
+# include <asm/spinlock_types.h>
+#else
+# include <linux/spinlock_types_up.h>
+#endif
+
+typedef struct {
+	raw_spinlock_t raw_lock;
+#if defined(CONFIG_PREEMPT) && defined(CONFIG_SMP)
+	unsigned int break_lock;
+#endif
+#ifdef CONFIG_DEBUG_SPINLOCK
+	unsigned int magic, owner_cpu;
+	void *owner;
+#endif
+} spinlock_t;
+
+#define SPINLOCK_MAGIC		0xdead4ead
+
+typedef struct {
+	raw_rwlock_t raw_lock;
+#if defined(CONFIG_PREEMPT) && defined(CONFIG_SMP)
+	unsigned int break_lock;
+#endif
+#ifdef CONFIG_DEBUG_SPINLOCK
+	unsigned int magic, owner_cpu;
+	void *owner;
+#endif
+} rwlock_t;
+
+#define RWLOCK_MAGIC		0xdeaf1eed
+
+#define SPINLOCK_OWNER_INIT	((void *)-1L)
+
+#ifdef CONFIG_DEBUG_SPINLOCK
+# define SPIN_LOCK_UNLOCKED						\
+	(spinlock_t)	{	.raw_lock = __RAW_SPIN_LOCK_UNLOCKED,	\
+				.magic = SPINLOCK_MAGIC,		\
+				.owner = SPINLOCK_OWNER_INIT,		\
+				.owner_cpu = -1 }
+#define RW_LOCK_UNLOCKED						\
+	(rwlock_t)	{	.raw_lock = __RAW_RW_LOCK_UNLOCKED,	\
+				.magic = RWLOCK_MAGIC,			\
+				.owner = SPINLOCK_OWNER_INIT,		\
+				.owner_cpu = -1 }
+#else
+# define SPIN_LOCK_UNLOCKED \
+	(spinlock_t)	{	.raw_lock = __RAW_SPIN_LOCK_UNLOCKED }
+#define RW_LOCK_UNLOCKED \
+	(rwlock_t)	{	.raw_lock = __RAW_RW_LOCK_UNLOCKED }
+#endif
+
+#define DEFINE_SPINLOCK(x)	spinlock_t x = SPIN_LOCK_UNLOCKED
+#define DEFINE_RWLOCK(x)	rwlock_t x = RW_LOCK_UNLOCKED
+
+#endif /* __LINUX_SPINLOCK_TYPES_H */
diff --git a/include/linux/spinlock_types_up.h b/include/linux/spinlock_types_up.h
new file mode 100644
index 00000000000..def2d173a8d
--- /dev/null
+++ b/include/linux/spinlock_types_up.h
@@ -0,0 +1,51 @@
+#ifndef __LINUX_SPINLOCK_TYPES_UP_H
+#define __LINUX_SPINLOCK_TYPES_UP_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+# error "please don't include this file directly"
+#endif
+
+/*
+ * include/linux/spinlock_types_up.h - spinlock type definitions for UP
+ *
+ * portions Copyright 2005, Red Hat, Inc., Ingo Molnar
+ * Released under the General Public License (GPL).
+ */
+
+#ifdef CONFIG_DEBUG_SPINLOCK
+
+typedef struct {
+	volatile unsigned int slock;
+} raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED { 1 }
+
+#else
+
+/*
+ * All gcc 2.95 versions and early versions of 2.96 have a nasty bug
+ * with empty initializers.
+ */
+#if (__GNUC__ > 2)
+typedef struct { } raw_spinlock_t;
+
+#define __RAW_SPIN_LOCK_UNLOCKED { }
+#else
+typedef struct { int gcc_is_buggy; } raw_spinlock_t;
+#define __RAW_SPIN_LOCK_UNLOCKED (raw_spinlock_t) { 0 }
+#endif
+
+#endif
+
+#if (__GNUC__ > 2)
+typedef struct {
+	/* no debug version on UP */
+} raw_rwlock_t;
+
+#define __RAW_RW_LOCK_UNLOCKED { }
+#else
+typedef struct { int gcc_is_buggy; } raw_rwlock_t;
+#define __RAW_RW_LOCK_UNLOCKED (raw_rwlock_t) { 0 }
+#endif
+
+#endif /* __LINUX_SPINLOCK_TYPES_UP_H */
diff --git a/include/linux/spinlock_up.h b/include/linux/spinlock_up.h
new file mode 100644
index 00000000000..31accf2f0b1
--- /dev/null
+++ b/include/linux/spinlock_up.h
@@ -0,0 +1,74 @@
+#ifndef __LINUX_SPINLOCK_UP_H
+#define __LINUX_SPINLOCK_UP_H
+
+#ifndef __LINUX_SPINLOCK_H
+# error "please don't include this file directly"
+#endif
+
+/*
+ * include/linux/spinlock_up.h - UP-debug version of spinlocks.
+ *
+ * portions Copyright 2005, Red Hat, Inc., Ingo Molnar
+ * Released under the General Public License (GPL).
+ *
+ * In the debug case, 1 means unlocked, 0 means locked. (the values
+ * are inverted, to catch initialization bugs)
+ *
+ * No atomicity anywhere, we are on UP.
+ */
+
+#ifdef CONFIG_DEBUG_SPINLOCK
+
+#define __raw_spin_is_locked(x)		((x)->slock == 0)
+
+static inline void __raw_spin_lock(raw_spinlock_t *lock)
+{
+	lock->slock = 0;
+}
+
+static inline void
+__raw_spin_lock_flags(raw_spinlock_t *lock, unsigned long flags)
+{
+	local_irq_save(flags);
+	lock->slock = 0;
+}
+
+static inline int __raw_spin_trylock(raw_spinlock_t *lock)
+{
+	char oldval = lock->slock;
+
+	lock->slock = 0;
+
+	return oldval > 0;
+}
+
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
+{
+	lock->slock = 1;
+}
+
+/*
+ * Read-write spinlocks. No debug version.
+ */
+#define __raw_read_lock(lock)		do { (void)(lock); } while (0)
+#define __raw_write_lock(lock)		do { (void)(lock); } while (0)
+#define __raw_read_trylock(lock)	({ (void)(lock); 1; })
+#define __raw_write_trylock(lock)	({ (void)(lock); 1; })
+#define __raw_read_unlock(lock)		do { (void)(lock); } while (0)
+#define __raw_write_unlock(lock)	do { (void)(lock); } while (0)
+
+#else /* DEBUG_SPINLOCK */
+#define __raw_spin_is_locked(lock)	((void)(lock), 0)
+/* for sched.c and kernel_lock.c: */
+# define __raw_spin_lock(lock)		do { (void)(lock); } while (0)
+# define __raw_spin_unlock(lock)	do { (void)(lock); } while (0)
+# define __raw_spin_trylock(lock)	({ (void)(lock); 1; })
+#endif /* DEBUG_SPINLOCK */
+
+#define __raw_read_can_lock(lock)	(((void)(lock), 1))
+#define __raw_write_can_lock(lock)	(((void)(lock), 1))
+
+#define __raw_spin_unlock_wait(lock) \
+		do { cpu_relax(); } while (__raw_spin_is_locked(lock))
+
+#endif /* __LINUX_SPINLOCK_UP_H */