mm, sched: Allow uaccess in atomic with pagefault_disable()
authorMichael S. Tsirkin <mst@redhat.com>
Sun, 26 May 2013 14:32:23 +0000 (17:32 +0300)
committerIngo Molnar <mingo@kernel.org>
Tue, 28 May 2013 07:41:11 +0000 (09:41 +0200)
commit662bbcb2747c2422cf98d3d97619509379eee466
treed733c7257e59bde654d963cce0e26454a134d787
parent114276ac0a3beb9c391a410349bd770653e185ce
mm, sched: Allow uaccess in atomic with pagefault_disable()

This changes might_fault() so that it does not
trigger a false positive diagnostic for e.g. the following
sequence:

spin_lock_irqsave()
pagefault_disable()
copy_to_user()
pagefault_enable()
spin_unlock_irqrestore()

In particular vhost wants to do this, to call
socket ops from under a lock.

There are 3 cases to consider:

 - CONFIG_PROVE_LOCKING - might_fault is non-inline
   so it's easy to move the in_atomic test to fix
   up the false positive warning.

 - CONFIG_DEBUG_ATOMIC_SLEEP - might_fault
   is currently inline, but we are calling a
   non-inline __might_sleep anyway,
   so let's use the non-line version of might_fault
   that does the right thing.

 - !CONFIG_DEBUG_ATOMIC_SLEEP && !CONFIG_PROVE_LOCKING
   __might_sleep is a nop so might_fault is a nop.

Make this explicit.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1369577426-26721-11-git-send-email-mst@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
include/linux/kernel.h
mm/memory.c