oom: fix constraint deadlock
authorDavid Rientjes <[email protected]>
Sun, 6 May 2007 21:50:00 +0000 (14:50 -0700)
committerLinus Torvalds <[email protected]>
Mon, 7 May 2007 19:12:55 +0000 (12:12 -0700)
commit2b45ab3398a0ba119b1f672c7c56fd5a431b7f0a
tree302238b786d2137111244702345acf6d4a17d4b7
parent2b744c01a54fe0c9974ff1b29522f25f07084053
oom: fix constraint deadlock

Fixes a deadlock in the OOM killer for allocations that are not
__GFP_HARDWALL.

Before the OOM killer checks for the allocation constraint, it takes
callback_mutex.

constrained_alloc() iterates through each zone in the allocation zonelist
and calls cpuset_zone_allowed_softwall() to determine whether an allocation
for gfp_mask is possible.  If a zone's node is not in the OOM-triggering
task's mems_allowed, it is not exiting, and we did not fail on a
__GFP_HARDWALL allocation, cpuset_zone_allowed_softwall() attempts to take
callback_mutex to check the nearest exclusive ancestor of current's cpuset.
 This results in deadlock.

We now take callback_mutex after iterating through the zonelist since we
don't need it yet.

Cc: Andi Kleen <[email protected]>
Cc: Nick Piggin <[email protected]>
Cc: Christoph Lameter <[email protected]>
Cc: Martin J. Bligh <[email protected]>
Signed-off-by: David Rientjes <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
mm/oom_kill.c