projects / oweals / musl.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: c13f2af) | patch
mitigate performance regression in libc-internal locks on x86_64
authorRich Felker <dalias@aerifal.cx>
Sun, 16 Aug 2015 18:15:18 +0000 (18:15 +0000)
committerRich Felker <dalias@aerifal.cx>
Sun, 16 Aug 2015 18:15:18 +0000 (18:15 +0000)
commit5a9c8c05a5a0cdced4122589184fd795b761bb4a
tree34b9f909c151de0d24ec29ef194815a5fa4daedetree | snapshot
parentc13f2af1fe1856e36dd1b2773cac05d5d72641dccommit | diff
mitigate performance regression in libc-internal locks on x86_64

commit 3c43c0761e1725fd5f89a9c028cbf43250abb913 fixed missing
synchronization in the atomic store operation for i386 and x86_64, but
opted to use mfence for the barrier on x86_64 where it's always
available. however, in practice mfence is significantly slower than
the barrier approach used on i386 (a nop-like lock orl operation).
this commit changes x86_64 (and x32) to use the faster barrier.
arch/x32/atomic.h diff | blob | history
arch/x86_64/atomic.h diff | blob | history
Unnamed repository; edit this file 'description' to name the repository.