From: Rik van Riel <riel@redhat.com>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mel@csn.ul.ie>,
Andrew Morton <akpm@linux-foundation.org>,
Linux-MM <linux-mm@kvack.org>,
LKML <linux-kernel@vger.kernel.org>,
Minchan Kim <minchan.kim@gmail.com>,
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>,
Christoph Lameter <cl@linux.com>,
Andrea Arcangeli <aarcange@redhat.com>
Subject: Re: [PATCH 1/2] mm,migration: Prevent rmap_walk_[anon|ksm] seeing the wrong VMA information
Date: Thu, 06 May 2010 09:40:56 -0400 [thread overview]
Message-ID: <4BE2C6E8.2030609@redhat.com> (raw)
In-Reply-To: <alpine.LFD.2.00.1005051007140.27218@i5.linux-foundation.org>
On 05/05/2010 01:34 PM, Linus Torvalds wrote:
> - you always lock the _deepest_ anon_vma you can find.
The emphasis should be on "always" :)
> That means just a single lock. And the "deepest" anon_vma is well-defined
> for all anon_vma's, because each same_anon_vma chain is always rooted in
> the original anon_vma that caused it.
It should work, but only if we always take the deepest
anon_vma lock.
Not just in the migration code, but also in mmap, munmap,
mprotect (for split_vma), expand_stack, etc...
Otherwise we will still not provide exclusion of migrate
vs. those events.
I'm guessing that means changing both anon_vma_lock and
page_lock_anon_vma to always take the deepest anon_vma
lock - not introducing a new function that is only called
by the migration code.
--
All rights reversed
WARNING: multiple messages have this Message-ID (diff)
From: Rik van Riel <riel@redhat.com>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mel@csn.ul.ie>,
Andrew Morton <akpm@linux-foundation.org>,
Linux-MM <linux-mm@kvack.org>,
LKML <linux-kernel@vger.kernel.org>,
Minchan Kim <minchan.kim@gmail.com>,
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>,
Christoph Lameter <cl@linux.com>,
Andrea Arcangeli <aarcange@redhat.com>
Subject: Re: [PATCH 1/2] mm,migration: Prevent rmap_walk_[anon|ksm] seeing the wrong VMA information
Date: Thu, 06 May 2010 09:40:56 -0400 [thread overview]
Message-ID: <4BE2C6E8.2030609@redhat.com> (raw)
In-Reply-To: <alpine.LFD.2.00.1005051007140.27218@i5.linux-foundation.org>
On 05/05/2010 01:34 PM, Linus Torvalds wrote:
> - you always lock the _deepest_ anon_vma you can find.
The emphasis should be on "always" :)
> That means just a single lock. And the "deepest" anon_vma is well-defined
> for all anon_vma's, because each same_anon_vma chain is always rooted in
> the original anon_vma that caused it.
It should work, but only if we always take the deepest
anon_vma lock.
Not just in the migration code, but also in mmap, munmap,
mprotect (for split_vma), expand_stack, etc...
Otherwise we will still not provide exclusion of migrate
vs. those events.
I'm guessing that means changing both anon_vma_lock and
page_lock_anon_vma to always take the deepest anon_vma
lock - not introducing a new function that is only called
by the migration code.
--
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>
next prev parent reply other threads:[~2010-05-06 13:42 UTC|newest]
Thread overview: 100+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-05-05 13:14 [PATCH 0/2] Fix migration races in rmap_walk() V5 Mel Gorman
2010-05-05 13:14 ` Mel Gorman
2010-05-05 13:14 ` [PATCH 1/2] mm,migration: Prevent rmap_walk_[anon|ksm] seeing the wrong VMA information Mel Gorman
2010-05-05 13:14 ` Mel Gorman
2010-05-05 14:34 ` Linus Torvalds
2010-05-05 14:34 ` Linus Torvalds
2010-05-05 14:56 ` Mel Gorman
2010-05-05 14:56 ` Mel Gorman
2010-05-05 15:31 ` Linus Torvalds
2010-05-05 15:31 ` Linus Torvalds
2010-05-05 15:54 ` Mel Gorman
2010-05-05 15:54 ` Mel Gorman
2010-05-05 16:13 ` Andrea Arcangeli
2010-05-05 16:13 ` Andrea Arcangeli
2010-05-05 19:11 ` Peter Zijlstra
2010-05-05 19:11 ` Peter Zijlstra
2010-05-05 19:57 ` Andrea Arcangeli
2010-05-05 19:57 ` Andrea Arcangeli
2010-05-21 0:27 ` Andrea Arcangeli
2010-05-21 0:27 ` Andrea Arcangeli
2010-05-06 10:37 ` Mel Gorman
2010-05-06 10:37 ` Mel Gorman
2010-05-05 17:34 ` Linus Torvalds
2010-05-05 17:34 ` Linus Torvalds
2010-05-05 17:57 ` Linus Torvalds
2010-05-05 17:57 ` Linus Torvalds
2010-05-05 18:14 ` Mel Gorman
2010-05-05 18:14 ` Mel Gorman
2010-05-05 18:34 ` Linus Torvalds
2010-05-05 18:34 ` Linus Torvalds
2010-05-06 11:03 ` Mel Gorman
2010-05-06 11:03 ` Mel Gorman
2010-05-06 13:40 ` Rik van Riel [this message]
2010-05-06 13:40 ` Rik van Riel
2010-05-06 13:45 ` Mel Gorman
2010-05-06 13:45 ` Mel Gorman
2010-05-05 17:53 ` Mel Gorman
2010-05-05 17:53 ` Mel Gorman
2010-05-05 18:02 ` Linus Torvalds
2010-05-05 18:02 ` Linus Torvalds
2010-05-05 18:17 ` Mel Gorman
2010-05-05 18:17 ` Mel Gorman
2010-05-06 0:22 ` Mel Gorman
2010-05-06 0:22 ` Mel Gorman
2010-05-06 0:42 ` Linus Torvalds
2010-05-06 0:42 ` Linus Torvalds
2010-05-06 10:02 ` Mel Gorman
2010-05-06 10:02 ` Mel Gorman
2010-05-06 14:15 ` Linus Torvalds
2010-05-06 14:15 ` Linus Torvalds
2010-05-06 14:25 ` Mel Gorman
2010-05-06 14:25 ` Mel Gorman
2010-05-06 9:47 ` Minchan Kim
2010-05-06 9:47 ` Minchan Kim
2010-05-06 9:54 ` Mel Gorman
2010-05-06 9:54 ` Mel Gorman
2010-05-06 10:01 ` Minchan Kim
2010-05-06 10:01 ` Minchan Kim
2010-05-06 10:10 ` Mel Gorman
2010-05-06 10:10 ` Mel Gorman
2010-05-06 14:06 ` Linus Torvalds
2010-05-06 14:06 ` Linus Torvalds
2010-05-06 15:59 ` Minchan Kim
2010-05-06 15:59 ` Minchan Kim
2010-05-06 7:38 ` KAMEZAWA Hiroyuki
2010-05-06 7:38 ` KAMEZAWA Hiroyuki
2010-05-06 9:46 ` Mel Gorman
2010-05-06 9:46 ` Mel Gorman
2010-05-06 23:52 ` KAMEZAWA Hiroyuki
2010-05-06 23:52 ` KAMEZAWA Hiroyuki
2010-05-07 5:49 ` KAMEZAWA Hiroyuki
2010-05-07 5:49 ` KAMEZAWA Hiroyuki
2010-05-05 13:14 ` [PATCH 2/2] mm,migration: Fix race between shift_arg_pages and rmap_walk by guaranteeing rmap_walk finds PTEs created within the temporary stack Mel Gorman
2010-05-05 13:14 ` Mel Gorman
-- strict thread matches above, loose matches on Subject: below --
2010-05-06 15:33 [PATCH 0/2] Fix migration races in rmap_walk() V6 Mel Gorman
2010-05-06 15:33 ` [PATCH 1/2] mm,migration: Prevent rmap_walk_[anon|ksm] seeing the wrong VMA information Mel Gorman
2010-05-06 15:33 ` Mel Gorman
2010-05-06 15:44 ` Rik van Riel
2010-05-06 15:44 ` Rik van Riel
2010-05-06 15:51 ` Mel Gorman
2010-05-06 15:51 ` Mel Gorman
2010-05-06 15:59 ` Linus Torvalds
2010-05-06 15:59 ` Linus Torvalds
2010-05-06 17:07 ` Mel Gorman
2010-05-06 17:07 ` Mel Gorman
2010-05-06 23:20 [PATCH 0/2] Fix migration races in rmap_walk() V7 Mel Gorman
2010-05-06 23:20 ` [PATCH 1/2] mm,migration: Prevent rmap_walk_[anon|ksm] seeing the wrong VMA information Mel Gorman
2010-05-06 23:20 ` Mel Gorman
2010-05-07 0:56 ` KAMEZAWA Hiroyuki
2010-05-07 0:56 ` KAMEZAWA Hiroyuki
2010-05-07 16:26 ` Mel Gorman
2010-05-07 16:26 ` Mel Gorman
2010-05-08 15:39 ` Andrea Arcangeli
2010-05-08 15:39 ` Andrea Arcangeli
2010-05-08 17:02 ` Linus Torvalds
2010-05-08 17:02 ` Linus Torvalds
2010-05-08 18:04 ` Andrea Arcangeli
2010-05-08 18:04 ` Andrea Arcangeli
2010-05-08 19:51 ` Linus Torvalds
2010-05-08 19:51 ` Linus Torvalds
2010-05-09 19:23 ` Mel Gorman
2010-05-09 19:23 ` Mel Gorman
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=4BE2C6E8.2030609@redhat.com \
--to=riel@redhat.com \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=cl@linux.com \
--cc=kamezawa.hiroyu@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mel@csn.ul.ie \
--cc=minchan.kim@gmail.com \
--cc=torvalds@linux-foundation.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.