diff options
author | Matt Turner <mattst88@gmail.com> | 2015-08-04 14:35:05 -0700 |
---|---|---|
committer | Matt Turner <mattst88@gmail.com> | 2015-08-30 18:01:16 -0700 |
commit | 8f8dcb3f7fe4febbfa96e64d4ad47de958c5cc34 (patch) | |
tree | 58f9dca31e0fe7d0149ea90b9942594db6c38fae /arch | |
parent | 64291f7db5bd8150a74ad2036f1037e6a0428df2 (diff) | |
download | op-kernel-dev-8f8dcb3f7fe4febbfa96e64d4ad47de958c5cc34.zip op-kernel-dev-8f8dcb3f7fe4febbfa96e64d4ad47de958c5cc34.tar.gz |
alpha: select CONFIG_ARCH_USE_CMPXCHG_LOCKREF.
On Alpha we have spinlocks that are 32b in size and an efficient
cmpxchg64 implementation, so we qualify to make use of cmpxchg backed
lockrefs. Select the ARCH_USE_CMPXCHG_LOCKREF Kconfig symbol and provide
a trivial implementation of arch_spin_value_unlocked to satisfy the
lockref code.
Using Linus' simple testcase from
http://article.gmane.org/gmane.linux.file-systems/77466 on a dual CPU
ES47 system I see around an 8% gain:
N Min Max Median Avg Stddev
x 30 6194580 6295654 6272504 6272514 17694.232
+ 30 6731164 6786334 6767982 6764274 13738.863
Difference at 95.0% confidence
491760 +/- 8188.17
7.83992% +/- 0.130541%
(Student's t, pooled s = 15840.5)
Signed-off-by: Matt Turner <mattst88@gmail.com>
Diffstat (limited to 'arch')
-rw-r--r-- | arch/alpha/Kconfig | 1 | ||||
-rw-r--r-- | arch/alpha/include/asm/spinlock.h | 5 |
2 files changed, 6 insertions, 0 deletions
diff --git a/arch/alpha/Kconfig b/arch/alpha/Kconfig index bf9e9d3..f515a4d 100644 --- a/arch/alpha/Kconfig +++ b/arch/alpha/Kconfig @@ -3,6 +3,7 @@ config ALPHA default y select ARCH_MIGHT_HAVE_PC_PARPORT select ARCH_MIGHT_HAVE_PC_SERIO + select ARCH_USE_CMPXCHG_LOCKREF select HAVE_AOUT select HAVE_IDE select HAVE_OPROFILE diff --git a/arch/alpha/include/asm/spinlock.h b/arch/alpha/include/asm/spinlock.h index 37b570d..fed9c6f 100644 --- a/arch/alpha/include/asm/spinlock.h +++ b/arch/alpha/include/asm/spinlock.h @@ -16,6 +16,11 @@ #define arch_spin_unlock_wait(x) \ do { cpu_relax(); } while ((x)->lock) +static inline int arch_spin_value_unlocked(arch_spinlock_t lock) +{ + return lock.lock == 0; +} + static inline void arch_spin_unlock(arch_spinlock_t * lock) { mb(); |