mm: fix kernel BUG at mm/rmap.c:1017!
authorHugh Dickins <hughd@google.com>
Sat, 28 May 2011 20:17:04 +0000 (13:17 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Sat, 28 May 2011 23:09:26 +0000 (16:09 -0700)
commit5dbe0af47f8a8f968bac2991c3ec974c6e3eaabc
treee936955d61fc2a69e36b2e9276f5ad7585134255
parent826267cf1e6c6899eda1325a19f1b1d15c558b20
mm: fix kernel BUG at mm/rmap.c:1017!

I've hit the "address >= vma->vm_end" check in do_page_add_anon_rmap()
just once.  The stack showed khugepaged allocation trying to compact
pages: the call to page_add_anon_rmap() coming from remove_migration_pte().

That path holds anon_vma lock, but does not hold mmap_sem: it can
therefore race with a split_vma(), and in commit 5f70b962ccc2 "mmap:
avoid unnecessary anon_vma lock" we just took away the anon_vma lock
protection when adjusting vma->vm_end.

I don't think that particular BUG_ON ever caught anything interesting,
so better replace it by a comment, than reinstate the anon_vma locking.

Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/rmap.c