From a8ddac7e53e89cb877965097d05adfeb1c91def3 Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Tue, 21 Oct 2008 10:59:15 +0200 Subject: mutex: speed up generic mutex implementations - atomic operations which both modify the variable and return something imply full smp memory barriers before and after the memory operations involved (failing atomic_cmpxchg, atomic_add_unless, etc don't imply a barrier because they don't modify the target). See Documentation/atomic_ops.txt. So remove extra barriers and branches. - All architectures support atomic_cmpxchg. This has no relation to __HAVE_ARCH_CMPXCHG. We can just take the atomic_cmpxchg path unconditionally This reduces a simple single threaded fastpath lock+unlock test from 590 cycles to 203 cycles on a ppc970 system. Signed-off-by: Nick Piggin Signed-off-by: Linus Torvalds --- include/asm-generic/mutex-xchg.h | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-) (limited to 'include/asm-generic/mutex-xchg.h') diff --git a/include/asm-generic/mutex-xchg.h b/include/asm-generic/mutex-xchg.h index 7b9cd2c..580a6d3 100644 --- a/include/asm-generic/mutex-xchg.h +++ b/include/asm-generic/mutex-xchg.h @@ -27,8 +27,6 @@ __mutex_fastpath_lock(atomic_t *count, void (*fail_fn)(atomic_t *)) { if (unlikely(atomic_xchg(count, 0) != 1)) fail_fn(count); - else - smp_mb(); } /** @@ -46,10 +44,7 @@ __mutex_fastpath_lock_retval(atomic_t *count, int (*fail_fn)(atomic_t *)) { if (unlikely(atomic_xchg(count, 0) != 1)) return fail_fn(count); - else { - smp_mb(); - return 0; - } + return 0; } /** @@ -67,7 +62,6 @@ __mutex_fastpath_lock_retval(atomic_t *count, int (*fail_fn)(atomic_t *)) static inline void __mutex_fastpath_unlock(atomic_t *count, void (*fail_fn)(atomic_t *)) { - smp_mb(); if (unlikely(atomic_xchg(count, 1) != 0)) fail_fn(count); } @@ -110,7 +104,6 @@ __mutex_fastpath_trylock(atomic_t *count, int (*fail_fn)(atomic_t *)) if (prev < 0) prev = 0; } - smp_mb(); return prev; } -- cgit v1.1