This is a note to let you know that I've just added the patch titled
mm/memory.c: use entry = ACCESS_ONCE(*pte) in handle_pte_fault()
to the 3.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary
The filename of the patch is:
mm-memory.c-use-entry-access_once-pte-in-handle_pte_fault.patch
and it can be found in the queue-3.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@xxxxxxxxxxxxxxx> know about it.
>From c0d73261f5c1355a35b8b40e871d31578ce0c044 Mon Sep 17 00:00:00 2001
From: Hugh Dickins <hughd@xxxxxxxxxx>
Date: Wed, 6 Aug 2014 16:05:08 -0700
Subject: mm/memory.c: use entry = ACCESS_ONCE(*pte) in handle_pte_fault()
From: Hugh Dickins <hughd@xxxxxxxxxx>
commit c0d73261f5c1355a35b8b40e871d31578ce0c044 upstream.
Use ACCESS_ONCE() in handle_pte_fault() when getting the entry or
orig_pte upon which all subsequent decisions and pte_same() tests will
be made.
I have no evidence that its lack is responsible for the mm/filemap.c:202
BUG_ON(page_mapped(page)) in __delete_from_page_cache() found by
trinity, and I am not optimistic that it will fix it. But I have found
no other explanation, and ACCESS_ONCE() here will surely not hurt.
If gcc does re-access the pte before passing it down, then that would be
disastrous for correct page fault handling, and certainly could explain
the page_mapped() BUGs seen (concurrent fault causing page to be mapped
in a second time on top of itself: mapcount 2 for a single pte).
Signed-off-by: Hugh Dickins <hughd@xxxxxxxxxx>
Cc: Sasha Levin <sasha.levin@xxxxxxxxxx>
Cc: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Cc: "Kirill A. Shutemov" <kirill.shutemov@xxxxxxxxxxxxxxx>
Cc: Konstantin Khlebnikov <koct9i@xxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Signed-off-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Signed-off-by: Mel Gorman <mgorman@xxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
mm/memory.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -3646,7 +3646,7 @@ static int handle_pte_fault(struct mm_st
pte_t entry;
spinlock_t *ptl;
- entry = *pte;
+ entry = ACCESS_ONCE(*pte);
if (!pte_present(entry)) {
if (pte_none(entry)) {
if (vma->vm_ops) {
Patches currently in stable-queue which might be from hughd@xxxxxxxxxx are
queue-3.14/mm-page_alloc-use-jump-labels-to-avoid-checking-number_of_cpusets.patch
queue-3.14/mm-non-atomically-mark-page-accessed-during-page-cache-allocation-where-possible.patch
queue-3.14/mm-page_alloc-convert-hot-cold-parameter-and-immediate-callers-to-bool.patch
queue-3.14/mm-page_alloc-only-check-the-zone-id-check-if-pages-are-buddies.patch
queue-3.14/mm-page_alloc-only-check-the-alloc-flags-and-gfp_mask-for-dirty-once.patch
queue-3.14/mm-page_alloc-take-the-alloc_no_watermark-check-out-of-the-fast-path.patch
queue-3.14/fs-buffer-do-not-use-unnecessary-atomic-operations-when-discarding-buffers.patch
queue-3.14/mm-do-not-use-atomic-operations-when-releasing-pages.patch
queue-3.14/mm-page_alloc-reduce-number-of-times-page_to_pfn-is-called.patch
queue-3.14/mm-page_alloc-use-unsigned-int-for-order-in-more-places.patch
queue-3.14/mm-shmem-avoid-atomic-operation-during-shmem_getpage_gfp.patch
queue-3.14/shmem-fix-init_page_accessed-use-to-stop-pagelru-bug.patch
queue-3.14/mm-memory.c-use-entry-access_once-pte-in-handle_pte_fault.patch
queue-3.14/mm-do-not-use-unnecessary-atomic-operations-when-adding-pages-to-the-lru.patch
queue-3.14/include-linux-jump_label.h-expose-the-reference-count.patch
queue-3.14/mm-page_alloc-calculate-classzone_idx-once-from-the.patch
queue-3.14/mm-page_alloc-lookup-pageblock-migratetype-with-irqs-enabled-during-free.patch
--
To unsubscribe from this list: send the line "unsubscribe stable" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html