]> git.karo-electronics.de Git - karo-tx-linux.git/commit
futex: Fix spurious wakeup for requeue_pi really
authorThomas Gleixner <tglx@linutronix.de>
Wed, 28 Oct 2009 19:26:48 +0000 (20:26 +0100)
committerGreg Kroah-Hartman <gregkh@suse.de>
Tue, 10 Nov 2009 00:21:54 +0000 (16:21 -0800)
commitfc7df048bf93e52c04f4c39104a5a388b73725b8
tree09a841f881558f710677b962240160ac19065442
parent81e6fd571d574da95cf977ff3266d95a802b1820
futex: Fix spurious wakeup for requeue_pi really

commit 11df6dddcbc38affb7473aad3d962baf8414a947 upstream.

The requeue_pi path doesn't use unqueue_me() (and the racy lock_ptr ==
NULL test) nor does it use the wake_list of futex_wake() which where
the reason for commit 41890f2 (futex: Handle spurious wake up)

See debugging discussing on LKML Message-ID: <4AD4080C.20703@us.ibm.com>

The changes in this fix to the wait_requeue_pi path were considered to
be a likely unecessary, but harmless safety net. But it turns out that
due to the fact that for unknown $@#!*( reasons EWOULDBLOCK is defined
as EAGAIN we built an endless loop in the code path which returns
correctly EWOULDBLOCK.

Spurious wakeups in wait_requeue_pi code path are unlikely so we do
the easy solution and return EWOULDBLOCK^WEAGAIN to user space and let
it deal with the spurious wakeup.

Cc: Darren Hart <dvhltc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: John Stultz <johnstul@linux.vnet.ibm.com>
Cc: Dinakar Guniguntala <dino@in.ibm.com>
LKML-Reference: <4AE23C74.1090502@us.ibm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
kernel/futex.c