linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Wu Fengguang <fengguang.wu@intel.com>
To: Andi Kleen <andi@firstfloor.org>
Cc: "akpm@linux-foundation.org" <akpm@linux-foundation.org>,
	"npiggin@suse.de" <npiggin@suse.de>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"linux-mm@kvack.org" <linux-mm@kvack.org>
Subject: Re: [PATCH] [6/16] HWPOISON: Add various poison checks in mm/memory.c
Date: Thu, 4 Jun 2009 19:55:33 +0800	[thread overview]
Message-ID: <20090604115533.GB22118@localhost> (raw)
In-Reply-To: <20090604051915.GN1065@one.firstfloor.org>

On Thu, Jun 04, 2009 at 01:19:15PM +0800, Andi Kleen wrote:
> On Thu, Jun 04, 2009 at 12:26:03PM +0800, Wu Fengguang wrote:
> > On Thu, Jun 04, 2009 at 02:46:38AM +0800, Andi Kleen wrote:
> > >
> > > Bail out early when hardware poisoned pages are found in page fault handling.
> >
> > I suspect this patch is also not absolutely necessary: the poisoned
> > page will normally have been isolated already.
>
> It's needed to prevent new pages comming in when there is a parallel
> fault while the memory failure handling is in process.
> Otherwise the pages could get remapped in that small window.

This patch makes no difference at least for file pages, including tmpfs.

In filemap_fault(), it will first do find_lock_page(), which will lock
the page and then double check if the page->mapping is NULL.  If so,
it drops that page and re-find/re-create one in the radix tree.

That logic automatically avoids the poisoned page that is being
processed, because the poisoned page is now being locked, and when
find_lock_page() eventually is able to lock it, its page->mapping
will be NULL. So the PageHWPoison(vmf.page) test will never be true
for filemap_fault.

shmem_fault() calls shmem_getpage() to get the page, which will return
EIO on !uptodate page. It then returns VM_FAULT_SIGBUS on its own.

> > > --- linux.orig/mm/memory.c	2009-06-03 19:36:23.000000000 +0200
> > > +++ linux/mm/memory.c	2009-06-03 19:36:23.000000000 +0200
> > > @@ -2797,6 +2797,9 @@
> > >  	if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE)))
> > >  		return ret;
> > >
> > > +	if (unlikely(PageHWPoison(vmf.page)))
> > > +		return VM_FAULT_HWPOISON;
> > > +
> >
> > Direct return with locked page could lockup someone later.
> > Either drop this patch or fix it with this check?
>
> Fair point. Fixed.

OK, thanks.

Fengguang

--
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>

  reply	other threads:[~2009-06-04 11:55 UTC|newest]

Thread overview: 71+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-06-03 18:46 [PATCH] [0/16] HWPOISON: Intro Andi Kleen
2009-06-03 18:46 ` [PATCH] [1/16] HWPOISON: Add page flag for poisoned pages Andi Kleen
2009-06-03 18:46 ` [PATCH] [2/16] HWPOISON: Export some rmap vma locking to outside world Andi Kleen
2009-06-03 18:46 ` [PATCH] [3/16] HWPOISON: Add support for poison swap entries v2 Andi Kleen
2009-06-03 18:46 ` [PATCH] [4/16] HWPOISON: Add new SIGBUS error codes for hardware poison signals Andi Kleen
2009-06-03 18:46 ` [PATCH] [5/16] HWPOISON: Add basic support for poisoned pages in fault handler v3 Andi Kleen
2009-06-03 18:46 ` [PATCH] [6/16] HWPOISON: Add various poison checks in mm/memory.c Andi Kleen
2009-06-04  4:26   ` Wu Fengguang
2009-06-04  5:19     ` Andi Kleen
2009-06-04 11:55       ` Wu Fengguang [this message]
2009-06-04 12:52         ` Andi Kleen
2009-06-04 12:50           ` Wu Fengguang
2009-06-04 13:02             ` Andi Kleen
2009-06-04 13:16               ` Wu Fengguang
2009-06-09 10:25   ` Nick Piggin
2009-06-09 12:21     ` Wu Fengguang
2009-06-09 12:35       ` Nick Piggin
2009-06-03 18:46 ` [PATCH] [7/16] HWPOISON: x86: Add VM_FAULT_HWPOISON handling to x86 page fault handler v2 Andi Kleen
2009-06-09  9:54   ` Nick Piggin
2009-06-09 12:34     ` [PATCH] HWPOISON: define VM_FAULT_HWPOISON to 0 when feature is disabled Wu Fengguang
2009-06-03 18:46 ` [PATCH] [8/16] HWPOISON: Use bitmask/action code for try_to_unmap behaviour Andi Kleen
2009-06-09  9:57   ` Nick Piggin
2009-06-10  2:27     ` Wu Fengguang
2009-06-10  6:07       ` Nick Piggin
2009-06-03 18:46 ` [PATCH] [9/16] HWPOISON: Handle hardware poisoned pages in try_to_unmap Andi Kleen
2009-06-04  4:35   ` Wu Fengguang
2009-06-04  5:21     ` Andi Kleen
2009-06-03 18:46 ` [PATCH] [10/16] HWPOISON: Handle poisoned pages in set_page_dirty() Andi Kleen
2009-06-04  0:36   ` Wu Fengguang
2009-06-04  5:27     ` Andi Kleen
2009-06-09  9:59   ` Nick Piggin
2009-06-09 12:51     ` Wu Fengguang
2009-06-03 18:46 ` [PATCH] [11/16] HWPOISON: check and isolate corrupted free pages v2 Andi Kleen
2009-06-09 10:02   ` Nick Piggin
2009-06-09 13:03     ` Wu Fengguang
2009-06-09 13:28       ` Nick Piggin
2009-06-09 13:49         ` Wu Fengguang
2009-06-09 13:55           ` Nick Piggin
2009-06-09 14:56             ` Wu Fengguang
2009-06-09 15:31               ` Nick Piggin
2009-06-03 18:46 ` [PATCH] [12/16] Refactor truncate to allow direct truncating of page Andi Kleen
2009-06-04  4:32   ` Wu Fengguang
2009-06-04  5:20     ` Andi Kleen
2009-06-03 18:46 ` [PATCH] [13/16] HWPOISON: The high level memory error handler in the VM v5 Andi Kleen
2009-06-04  3:24   ` Wu Fengguang
2009-06-04  5:13     ` Andi Kleen
2009-06-04  9:07       ` Wu Fengguang
2009-06-04  9:26         ` Andi Kleen
2009-06-09  9:51   ` Nick Piggin
2009-06-09 11:14     ` Nick Piggin
2009-06-09 10:09   ` Nick Piggin
2009-06-09 16:05     ` Hugh Dickins
2009-06-09 16:35       ` Nick Piggin
2009-06-10  8:38       ` Wu Fengguang
2009-06-10  8:59         ` Nick Piggin
2009-06-10  9:20           ` Wu Fengguang
2009-06-10 11:03             ` Nick Piggin
2009-06-10 12:16               ` Wu Fengguang
2009-06-10 12:36                 ` Nick Piggin
2009-06-12  9:58       ` Andi Kleen
2009-06-10  3:10     ` [PATCH] HWPOISON: fix tasklist_lock/anon_vma locking order Wu Fengguang
2009-06-03 18:46 ` [PATCH] [14/16] HWPOISON: FOR TESTING: Enable memory failure code unconditionally Andi Kleen
2009-06-03 18:46 ` [PATCH] [15/16] HWPOISON: Add madvise() based injector for hardware poisoned pages v3 Andi Kleen
2009-06-03 18:46 ` [PATCH] [16/16] HWPOISON: Add simple debugfs interface to inject hwpoison on arbitary PFNs Andi Kleen
2009-06-09 10:20 ` [PATCH] [0/16] HWPOISON: Intro Nick Piggin
2009-06-10  9:07   ` Wu Fengguang
2009-06-10  9:18     ` Nick Piggin
2009-06-10  9:45       ` Wu Fengguang
2009-06-10 11:15         ` Nick Piggin
2009-06-10 12:36           ` Wu Fengguang
2009-06-10 12:47             ` Nick Piggin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20090604115533.GB22118@localhost \
    --to=fengguang.wu@intel.com \
    --cc=akpm@linux-foundation.org \
    --cc=andi@firstfloor.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=npiggin@suse.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).