forked from luck/tmp_suning_uos_patched
8c0a853770
There's no reason to call rcu_barrier() on every deactivate_locked_super(). We only need to make sure that all delayed rcu free inodes are flushed before we destroy related cache. Removing rcu_barrier() from deactivate_locked_super() affects some fast paths. E.g. on my machine exit_group() of a last process in IPC namespace takes 0.07538s. rcu_barrier() takes 0.05188s of that time. Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> |
||
---|---|---|
.. | ||
afs_cm.h | ||
afs_fs.h | ||
afs_vl.h | ||
afs.h | ||
cache.c | ||
callback.c | ||
cell.c | ||
cmservice.c | ||
dir.c | ||
file.c | ||
flock.c | ||
fsclient.c | ||
inode.c | ||
internal.h | ||
Kconfig | ||
main.c | ||
Makefile | ||
misc.c | ||
mntpt.c | ||
netdevices.c | ||
proc.c | ||
rxrpc.c | ||
security.c | ||
server.c | ||
super.c | ||
vlclient.c | ||
vlocation.c | ||
vnode.c | ||
volume.c | ||
write.c |