From: Andrea Arcangeli <aarcange@redhat.com>
To: Mel Gorman <mel@csn.ul.ie>
Cc: 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>, Rik van Riel <riel@redhat.com>,
Andrew Morton <akpm@linux-foundation.org>
Subject: Re: [PATCH 1/3] mm,migration: During fork(), wait for migration to end if migration PTE is encountered
Date: Wed, 28 Apr 2010 00:22:45 +0200 [thread overview]
Message-ID: <20100427222245.GE8860@random.random> (raw)
In-Reply-To: <1272403852-10479-2-git-send-email-mel@csn.ul.ie>
Ok I had a first look:
On Tue, Apr 27, 2010 at 10:30:50PM +0100, Mel Gorman wrote:
> CPUA CPU B
> do_fork()
> copy_mm() (from process 1 to process2)
> insert new vma to mmap_list (if inode/anon_vma)
Insert to the tail of the anon_vma list...
> pte_lock(process1)
> unmap a page
> insert migration_entry
> pte_unlock(process1)
>
> migrate page copy
> copy_page_range
> remap new page by rmap_walk()
rmap_walk will walk process1 first! It's at the head, the vmas with
unmapped ptes are at the tail so process1 is walked before process2.
> pte_lock(process2)
> found no pte.
> pte_unlock(process2)
> pte lock(process2)
> pte lock(process1)
> copy migration entry to process2
> pte unlock(process1)
> pte unlokc(process2)
> pte_lock(process1)
> replace migration entry
> to new page's pte.
> pte_unlock(process1)
rmap_walk has to lock down process1 before process2, this is the
ordering issue I already mentioned in earlier email. So it cannot
happen and this patch is unnecessary.
The ordering is fundamental and as said anon_vma_link already adds new
vmas to the _tail_ of the anon-vma. And this is why it has to add to
the tail. If anon_vma_link would add new vmas to the head of the list,
the above bug could materialize, but it doesn't so it cannot happen.
In mainline anon_vma_link is called anon_vma_chain_link, see the
list_add_tail there to provide this guarantee.
Because process1 is walked first by CPU A, the migration entry is
replaced by the final pte before copy-migration-entry
runs. Alternatively if copy-migration-entry runs before before
process1 is walked, the migration entry will be copied and found in
process 2.
Comments welcome.
Andrea
--
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:[~2010-04-27 22:23 UTC|newest]
Thread overview: 66+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-04-27 21:30 [PATCH 0/3] Fix migration races in rmap_walk() V2 Mel Gorman
2010-04-27 21:30 ` [PATCH 1/3] mm,migration: During fork(), wait for migration to end if migration PTE is encountered Mel Gorman
2010-04-27 22:22 ` Andrea Arcangeli [this message]
2010-04-27 23:52 ` KAMEZAWA Hiroyuki
2010-04-28 0:18 ` Andrea Arcangeli
2010-04-28 0:19 ` Andrea Arcangeli
2010-04-28 0:28 ` KAMEZAWA Hiroyuki
2010-04-28 0:59 ` Andrea Arcangeli
2010-04-28 8:24 ` Mel Gorman
2010-04-27 21:30 ` [PATCH 2/3] mm,migration: Prevent rmap_walk_[anon|ksm] seeing the wrong VMA information Mel Gorman
2010-04-27 23:10 ` Andrea Arcangeli
2010-04-28 9:15 ` Mel Gorman
2010-04-28 15:35 ` Andrea Arcangeli
2010-04-28 15:39 ` Andrea Arcangeli
2010-04-28 15:55 ` Mel Gorman
2010-04-28 16:23 ` Andrea Arcangeli
2010-04-28 17:34 ` Mel Gorman
2010-04-28 17:58 ` Andrea Arcangeli
2010-04-28 17:47 ` [RFC PATCH] take all anon_vma locks in anon_vma_lock Rik van Riel
2010-04-28 18:03 ` Andrea Arcangeli
2010-04-28 18:09 ` Rik van Riel
2010-04-28 18:25 ` [RFC PATCH -v2] " Rik van Riel
2010-04-28 19:07 ` Mel Gorman
2010-04-28 20:17 ` [RFC PATCH -v3] " Rik van Riel
2010-04-28 20:57 ` Rik van Riel
2010-04-29 0:28 ` Minchan Kim
2010-04-29 2:10 ` Rik van Riel
2010-04-29 2:55 ` Minchan Kim
2010-04-29 6:42 ` Minchan Kim
2010-04-29 15:39 ` Rik van Riel
2010-04-29 7:37 ` Mel Gorman
2010-04-29 8:15 ` Mel Gorman
2010-04-29 8:32 ` Minchan Kim
2010-04-29 8:44 ` Mel Gorman
2010-04-27 21:30 ` [PATCH 3/3] mm,migration: Remove straggling migration PTEs when page tables are being moved after the VMA has already moved Mel Gorman
2010-04-27 22:30 ` Andrea Arcangeli
2010-04-27 22:58 ` Andrea Arcangeli
2010-04-28 0:39 ` KAMEZAWA Hiroyuki
2010-04-28 1:05 ` Andrea Arcangeli
2010-04-28 1:09 ` Andrea Arcangeli
2010-04-28 1:18 ` KAMEZAWA Hiroyuki
2010-04-28 1:36 ` Andrea Arcangeli
2010-04-28 1:29 ` KAMEZAWA Hiroyuki
2010-04-28 1:44 ` Andrea Arcangeli
2010-04-28 2:12 ` KAMEZAWA Hiroyuki
2010-04-28 2:42 ` Andrea Arcangeli
2010-04-28 2:49 ` KAMEZAWA Hiroyuki
2010-04-28 7:28 ` KAMEZAWA Hiroyuki
2010-04-28 10:48 ` Mel Gorman
2010-04-28 0:03 ` KAMEZAWA Hiroyuki
2010-04-28 0:08 ` Andrea Arcangeli
2010-04-28 0:36 ` KAMEZAWA Hiroyuki
2010-04-28 8:30 ` KAMEZAWA Hiroyuki
2010-04-28 14:46 ` Andrea Arcangeli
2010-04-27 22:27 ` [PATCH 0/3] Fix migration races in rmap_walk() V2 Christoph Lameter
2010-04-27 22:32 ` Andrea Arcangeli
2010-04-28 0:13 ` KAMEZAWA Hiroyuki
2010-04-28 0:20 ` Andrea Arcangeli
2010-04-28 14:23 ` Mel Gorman
2010-04-28 14:57 ` Mel Gorman
2010-04-28 15:16 ` Andrea Arcangeli
2010-04-28 15:23 ` Mel Gorman
2010-04-28 15:45 ` Andrea Arcangeli
2010-04-28 20:40 ` Andrea Arcangeli
2010-04-28 21:05 ` Andrea Arcangeli
2010-04-28 9:17 ` 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=20100427222245.GE8860@random.random \
--to=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=riel@redhat.com \
/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).