linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
* [BUGFIX][PATCH] Fix false positive BUG_ON in __page_set_anon_rmap
@ 2010-07-22  7:41 KAMEZAWA Hiroyuki
  2010-07-22 13:16 ` Rik van Riel
  0 siblings, 1 reply; 3+ messages in thread
From: KAMEZAWA Hiroyuki @ 2010-07-22  7:41 UTC (permalink / raw)
  To: akpm@linux-foundation.org
  Cc: linux-mm@kvack.org, Rik van Riel, kosaki.motohiro,
	linux-kernel@vger.kernel.org

Rik, how do you think ?

==
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

Problem: wrong BUG_ON() in  __page_set_anon_rmap().
Kernel version: mmotm-0719

How to reproduce:
  create a small cgroup
  # mount -t cgroup none /cgroup -o memory
  # mkdir /cgroup/A
  # echo 30M > /cgroup/A/memory.limit_in_bytes

and run a malloc() program to cause swap-in v.s. swap-out ping-pong.

Description:
  Even if SwapCache is fully unmapped and mapcount goes down to 0,
  page->mapping is not cleared and will remain on memory until kswapd or some
  finds it. If a thread cause a page fault onto such "unmapped-but-not-discarded"
  swapcache, it will see a swap cache whose mapcount is 0 but page->mapping has a
  valid value.

  When it's reused at do_swap_page(), __page_set_anon_rmap() is called with
  "exclusive==1" and hits BUG_ON(). But this BUG_ON() is wrong. Nothing bad
  with rmapping a page which has page->mapping isn't 0.


Log:
Jul 22 16:06:02 ubuntu kernel: [  892.542485] 
Jul 22 16:06:02 ubuntu kernel: [  892.542488] Pid: 2951, comm: malloc Not tainted 2.6.35-rc5-mm1 #3 440BX Desktop Reference Platform/VMware Virtual Platform
Jul 22 16:06:02 ubuntu kernel: [  892.542491] RIP: 0010:[<ffffffff810e676c>]  [<ffffffff810e676c>] __page_set_anon_rmap+0x39/0x54
Jul 22 16:06:02 ubuntu kernel: [  892.542503] RSP: 0000:ffff880056a31dd8  EFLAGS: 00010202
Jul 22 16:06:02 ubuntu kernel: [  892.542505] RAX: ffff880059be7810 RBX: ffffea0001228038 RCX: ffff880059be7801
Jul 22 16:06:02 ubuntu kernel: [  892.542507] RDX: 0000000002823000 RSI: ffff8800569c5580 RDI: ffff8800569c5580
Jul 22 16:06:02 ubuntu kernel: [  892.542509] RBP: ffff880056a31de8 R08: ffff880056a31db8 R09: 00000000ffffffec
Jul 22 16:06:02 ubuntu kernel: [  892.542511] R10: ffff880056a31ec0 R11: ffffffff00000000 R12: ffff8800569c5580
Jul 22 16:06:02 ubuntu kernel: [  892.542514] R13: 0000000000000001 R14: 0000000000000008 R15: ffffea0001228038
Jul 22 16:06:02 ubuntu kernel: [  892.542517] FS:  00007f8c8d618700(0000) GS:ffff880001e00000(0000) knlGS:0000000000000000
Jul 22 16:06:02 ubuntu kernel: [  892.542520] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Jul 22 16:06:02 ubuntu kernel: [  892.542522] CR2: 0000000002823000 CR3: 00000000607e2000 CR4: 00000000000006f0
Jul 22 16:06:02 ubuntu kernel: [  892.542527] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
Jul 22 16:06:02 ubuntu kernel: [  892.542532] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Jul 22 16:06:02 ubuntu kernel: [  892.542535] Process malloc (pid: 2951, threadinfo ffff880056a30000, task ffff880037b796d0)
Jul 22 16:06:02 ubuntu kernel: [  892.542538]  ffff880056a31de8 ffffea0001228038 ffff880056a31e28 ffffffff810e6b3f
Jul 22 16:06:02 ubuntu kernel: [  892.542544] <0> ffff880000000001 0000000002823000 ffff8800569c5580 ffff8800569c5580
Jul 22 16:06:02 ubuntu kernel: [  892.542548] <0> ffff88005ed2f0a0 ffff88006359e300 ffff880056a31ef8 ffffffff810e07b8
Jul 22 16:06:02 ubuntu kernel: [  892.542556]  [<ffffffff810e6b3f>] do_page_add_anon_rmap+0x62/0x6d
Jul 22 16:06:02 ubuntu kernel: [  892.542560]  [<ffffffff810e07b8>] handle_mm_fault+0x716/0x8d7
Jul 22 16:06:02 ubuntu kernel: [  892.542567]  [<ffffffff8100870a>] ? __switch_to+0x215/0x227
Jul 22 16:06:02 ubuntu kernel: [  892.542571]  [<ffffffff81040d03>] ? pick_next_task_fair+0xdb/0xec
Jul 22 16:06:02 ubuntu kernel: [  892.542576]  [<ffffffff8144e42a>] ? schedule+0x589/0x5db
Jul 22 16:06:02 ubuntu kernel: [  892.542579]  [<ffffffff81453192>] do_page_fault+0x2c4/0x2dc
Jul 22 16:06:02 ubuntu kernel: [  892.542582]  [<ffffffff814502b5>] page_fault+0x25/0x30
Jul 22 16:06:02 ubuntu kernel: [  892.542619]  RSP <ffff880056a31dd8>
Jul 22 16:06:02 ubuntu kernel: [  892.542622] ---[ end trace 3e21bbaadd2d0799 ]---


Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
--
---
 mm/rmap.c |   12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

Index: mmotm-2.6.35-0719/mm/rmap.c
===================================================================
--- mmotm-2.6.35-0719.orig/mm/rmap.c
+++ mmotm-2.6.35-0719/mm/rmap.c
@@ -783,8 +783,16 @@ static void __page_set_anon_rmap(struct 
 		if (PageAnon(page))
 			return;
 		anon_vma = anon_vma->root;
-	} else
-		BUG_ON(PageAnon(page));
+	} else {
+		/*
+ 		 * In this case, swapped-out-but-not-discarded swap-cache
+ 		 * is remapped. So, no need to update page->mapping here.
+ 		 * We convice anon_vma poitned by page->mapping is not obsolete
+ 		 * because vma->anon_vma is necessary to be a family of it.
+ 		 */
+		if (PageAnon(page))
+			return;
+	}
 
 	anon_vma = (void *) anon_vma + PAGE_MAPPING_ANON;
 	page->mapping = (struct address_space *) anon_vma;

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [BUGFIX][PATCH] Fix false positive BUG_ON in __page_set_anon_rmap
  2010-07-22  7:41 [BUGFIX][PATCH] Fix false positive BUG_ON in __page_set_anon_rmap KAMEZAWA Hiroyuki
@ 2010-07-22 13:16 ` Rik van Riel
  2010-07-22 17:39   ` Andrea Arcangeli
  0 siblings, 1 reply; 3+ messages in thread
From: Rik van Riel @ 2010-07-22 13:16 UTC (permalink / raw)
  To: KAMEZAWA Hiroyuki
  Cc: akpm@linux-foundation.org, linux-mm@kvack.org, kosaki.motohiro,
	linux-kernel@vger.kernel.org, Andrea Arcangeli

On 07/22/2010 03:41 AM, KAMEZAWA Hiroyuki wrote:
> Rik, how do you think ?
>
> ==
> From: KAMEZAWA Hiroyuki<kamezawa.hiroyu@jp.fujitsu.com>
>
> Problem: wrong BUG_ON() in  __page_set_anon_rmap().
> Kernel version: mmotm-0719

> Description:
>    Even if SwapCache is fully unmapped and mapcount goes down to 0,
>    page->mapping is not cleared and will remain on memory until kswapd or some
>    finds it. If a thread cause a page fault onto such "unmapped-but-not-discarded"
>    swapcache, it will see a swap cache whose mapcount is 0 but page->mapping has a
>    valid value.
>
>    When it's reused at do_swap_page(), __page_set_anon_rmap() is called with
>    "exclusive==1" and hits BUG_ON(). But this BUG_ON() is wrong. Nothing bad
>    with rmapping a page which has page->mapping isn't 0.

Yes, you are absolutely right.

Acked-by: Rik van Riel <riel@redhat.com>

> Index: mmotm-2.6.35-0719/mm/rmap.c
> ===================================================================
> --- mmotm-2.6.35-0719.orig/mm/rmap.c
> +++ mmotm-2.6.35-0719/mm/rmap.c
> @@ -783,8 +783,16 @@ static void __page_set_anon_rmap(struct
>   		if (PageAnon(page))
>   			return;
>   		anon_vma = anon_vma->root;
> -	} else
> -		BUG_ON(PageAnon(page));
> +	} else {
> +		/*
> + 		 * In this case, swapped-out-but-not-discarded swap-cache
> + 		 * is remapped. So, no need to update page->mapping here.
> + 		 * We convice anon_vma poitned by page->mapping is not obsolete
> + 		 * because vma->anon_vma is necessary to be a family of it.
> + 		 */
> +		if (PageAnon(page))
> +			return;
> +	}
>
>   	anon_vma = (void *) anon_vma + PAGE_MAPPING_ANON;
>   	page->mapping = (struct address_space *) anon_vma;
>


-- 
All rights reversed

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [BUGFIX][PATCH] Fix false positive BUG_ON in __page_set_anon_rmap
  2010-07-22 13:16 ` Rik van Riel
@ 2010-07-22 17:39   ` Andrea Arcangeli
  0 siblings, 0 replies; 3+ messages in thread
From: Andrea Arcangeli @ 2010-07-22 17:39 UTC (permalink / raw)
  To: Rik van Riel
  Cc: KAMEZAWA Hiroyuki, akpm@linux-foundation.org, linux-mm@kvack.org,
	kosaki.motohiro, linux-kernel@vger.kernel.org

On Thu, Jul 22, 2010 at 09:16:44AM -0400, Rik van Riel wrote:
> On 07/22/2010 03:41 AM, KAMEZAWA Hiroyuki wrote:
> > Rik, how do you think ?
> >
> > ==
> > From: KAMEZAWA Hiroyuki<kamezawa.hiroyu@jp.fujitsu.com>
> >
> > Problem: wrong BUG_ON() in  __page_set_anon_rmap().
> > Kernel version: mmotm-0719
> 
> > Description:
> >    Even if SwapCache is fully unmapped and mapcount goes down to 0,
> >    page->mapping is not cleared and will remain on memory until kswapd or some
> >    finds it. If a thread cause a page fault onto such "unmapped-but-not-discarded"
> >    swapcache, it will see a swap cache whose mapcount is 0 but page->mapping has a
> >    valid value.
> >
> >    When it's reused at do_swap_page(), __page_set_anon_rmap() is called with
> >    "exclusive==1" and hits BUG_ON(). But this BUG_ON() is wrong. Nothing bad
> >    with rmapping a page which has page->mapping isn't 0.
> 
> Yes, you are absolutely right.
> 

I already noticed the problem when I merged your patch in aa.git
(before it would only be exclusive=0 in do_swap_page so it wasn't a
false positive), and I fixed it this way:

http://git.kernel.org/?p=linux/kernel/git/andrea/aa.git;a=commitdiff;h=2fe4f42f0f17498984b3f86b2339d583004b45de;hp=ffd146080305632406d97c7f6f984a648854d755

So I retained the BUG_ON for the real page_add_anon_rmap. Maybe not
worth it but you can have a look at my solution if you're interested
to retain it too.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2010-07-22 17:39 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-07-22  7:41 [BUGFIX][PATCH] Fix false positive BUG_ON in __page_set_anon_rmap KAMEZAWA Hiroyuki
2010-07-22 13:16 ` Rik van Riel
2010-07-22 17:39   ` Andrea Arcangeli

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).