Commit-ID: bf7b4c472db44413251bcef79ca1f6bf1ec81475 Gitweb: http://git.kernel.org/tip/bf7b4c472db44413251bcef79ca1f6bf1ec81475 Author: Waiman Long <waiman.l...@hpe.com> AuthorDate: Tue, 17 May 2016 21:26:22 -0400 Committer: Ingo Molnar <mi...@kernel.org> CommitDate: Wed, 8 Jun 2016 15:17:00 +0200
locking/rwsem: Improve reader wakeup code In __rwsem_do_wake(), the reader wakeup code will assume a writer has stolen the lock if the active reader/writer count is not 0. However, this is not as reliable an indicator as the original "< RWSEM_WAITING_BIAS" check. If another reader is present, the code will still break out and exit even if the writer is gone. This patch changes it to check the same "< RWSEM_WAITING_BIAS" condition to reduce the chance of false positive. Signed-off-by: Waiman Long <waiman.l...@hpe.com> Signed-off-by: Peter Zijlstra (Intel) <pet...@infradead.org> Reviewed-by: Peter Hurley <pe...@hurleysoftware.com> Cc: Andrew Morton <a...@linux-foundation.org> Cc: Dave Chinner <da...@fromorbit.com> Cc: Davidlohr Bueso <d...@stgolabs.net> Cc: Douglas Hatch <doug.ha...@hpe.com> Cc: Jason Low <jason.l...@hp.com> Cc: Linus Torvalds <torva...@linux-foundation.org> Cc: Paul E. McKenney <paul...@linux.vnet.ibm.com> Cc: Peter Zijlstra <pet...@infradead.org> Cc: Scott J Norton <scott.nor...@hpe.com> Cc: Thomas Gleixner <t...@linutronix.de> Link: http://lkml.kernel.org/r/1463534783-38814-5-git-send-email-waiman.l...@hpe.com Signed-off-by: Ingo Molnar <mi...@kernel.org> --- kernel/locking/rwsem-xadd.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/kernel/locking/rwsem-xadd.c b/kernel/locking/rwsem-xadd.c index 6b0d060..4f1daf5 100644 --- a/kernel/locking/rwsem-xadd.c +++ b/kernel/locking/rwsem-xadd.c @@ -156,9 +156,14 @@ __rwsem_mark_wake(struct rw_semaphore *sem, oldcount = atomic_long_add_return(adjustment, &sem->count) - adjustment; if (unlikely(oldcount < RWSEM_WAITING_BIAS)) { - /* A writer stole the lock. Undo our reader grant. */ - if (atomic_long_sub_return(adjustment, &sem->count) & - RWSEM_ACTIVE_MASK) + /* + * If the count is still less than RWSEM_WAITING_BIAS + * after removing the adjustment, it is assumed that + * a writer has stolen the lock. We have to undo our + * reader grant. + */ + if (atomic_long_add_return(-adjustment, &sem->count) < + RWSEM_WAITING_BIAS) goto out; /* Last active locker left. Retry waking readers. */ goto try_reader_grant;
