Revert "fs/exec.c: use BUILD_BUG_ON for VM_STACK_FLAGS & VM_STACK_INCOMPLETE_SETUP"
authorLinus Torvalds <torvalds@linux-foundation.org>
Thu, 16 Jun 2011 04:53:52 +0000 (21:53 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Thu, 16 Jun 2011 04:53:52 +0000 (21:53 -0700)
commit13fca640bb8ab611a50e0ba120b186faa2994d6c
tree0a40a28e498b2e9abdef8c86aa0aef7f82096f00
parent67de0162fbb78713fcb23cb2502b380faa8bde73
Revert "fs/exec.c: use BUILD_BUG_ON for VM_STACK_FLAGS & VM_STACK_INCOMPLETE_SETUP"

This reverts commit 7f81c8890c15a10f5220bebae3b6dfae4961962a.

It turns out that it's not actually a build-time check on x86-64 UML,
which does some seriously crazy stuff with VM_STACK_FLAGS.

The VM_STACK_FLAGS define depends on the arch-supplied
VM_STACK_DEFAULT_FLAGS value, and on x86-64 UML we have

  arch/um/sys-x86_64/shared/sysdep/vm-flags.h:

#define VM_STACK_DEFAULT_FLAGS \
(test_thread_flag(TIF_IA32) ? vm_stack_flags32 : vm_stack_flags)

#define VM_STACK_DEFAULT_FLAGS vm_stack_flags

(yes, seriously: two different #define's for that thing, with the first
one being inside an "#ifdef TIF_IA32")

It's possible that it is UML that should just be fixed in this area, but
for now let's just undo the (very small) optimization.

Reported-by: Randy Dunlap <randy.dunlap@oracle.com>
Acked-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Richard Weinberger <richard@nod.at>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
fs/exec.c