Skip to content

Commit

Permalink
tmpfs: fix shmem_getpage_gfp() VM_BUG_ON
Browse files Browse the repository at this point in the history
Fuzzing with trinity hit the "impossible" VM_BUG_ON(error) (which Fedora
has converted to WARNING) in shmem_getpage_gfp():

  WARNING: at mm/shmem.c:1151 shmem_getpage_gfp+0xa5c/0xa70()
  Pid: 29795, comm: trinity-child4 Not tainted 3.7.0-rc2+ #49
  Call Trace:
    warn_slowpath_common+0x7f/0xc0
    warn_slowpath_null+0x1a/0x20
    shmem_getpage_gfp+0xa5c/0xa70
    shmem_fault+0x4f/0xa0
    __do_fault+0x71/0x5c0
    handle_pte_fault+0x97/0xae0
    handle_mm_fault+0x289/0x350
    __do_page_fault+0x18e/0x530
    do_page_fault+0x2b/0x50
    page_fault+0x28/0x30
    tracesys+0xe1/0xe6

Thanks to Johannes for pointing to truncation: free_swap_and_cache()
only does a trylock on the page, so the page lock we've held since
before confirming swap is not enough to protect against truncation.

What cleanup is needed in this case? Just delete_from_swap_cache(),
which takes care of the memcg uncharge.

Signed-off-by: Hugh Dickins <hughd@google.com>
Reported-by: Dave Jones <davej@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
  • Loading branch information
Hugh Dickins authored and torvalds committed Nov 16, 2012
1 parent 498c228 commit 215c02b
Showing 1 changed file with 14 additions and 2 deletions.
16 changes: 14 additions & 2 deletions mm/shmem.c
Original file line number Diff line number Diff line change
Expand Up @@ -1145,8 +1145,20 @@ static int shmem_getpage_gfp(struct inode *inode, pgoff_t index,
if (!error) {
error = shmem_add_to_page_cache(page, mapping, index,
gfp, swp_to_radix_entry(swap));
/* We already confirmed swap, and make no allocation */
VM_BUG_ON(error);
/*
* We already confirmed swap under page lock, and make
* no memory allocation here, so usually no possibility
* of error; but free_swap_and_cache() only trylocks a
* page, so it is just possible that the entry has been
* truncated or holepunched since swap was confirmed.
* shmem_undo_range() will have done some of the
* unaccounting, now delete_from_swap_cache() will do
* the rest (including mem_cgroup_uncharge_swapcache).
* Reset swap.val? No, leave it so "failed" goes back to
* "repeat": reading a hole and writing should succeed.
*/
if (error)
delete_from_swap_cache(page);
}
if (error)
goto failed;
Expand Down

0 comments on commit 215c02b

Please sign in to comment.