From: Wu Fengguang <fengguang.wu@intel.com>
To: Jan Kara <jack@suse.cz>
Cc: Masayoshi MIZUMA <m.mizuma@jp.fujitsu.com>,
"linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>,
"viro@zeniv.linux.org.uk" <viro@zeniv.linux.org.uk>,
Nick Piggin <npiggin@suse.de>
Subject: Re: [PATCH][BUG] Lack of mutex_lock in drop_pagecache_sb()
Date: Tue, 24 Mar 2009 20:11:29 +0800 [thread overview]
Message-ID: <20090324121129.GA25410@localhost> (raw)
In-Reply-To: <20090324120502.GC23439@duck.suse.cz>
On Tue, Mar 24, 2009 at 08:05:02PM +0800, Jan Kara wrote:
> On Tue 24-03-09 15:44:57, Wu Fengguang wrote:
> > Hi Masayoshi,
> >
> > On Tue, Mar 24, 2009 at 03:06:45PM +0800, Masayoshi MIZUMA wrote:
> > > Hi, Fengguang
> > >
> > > On Mon, 23 Mar 2009 18:38:46 +0800
> > > Wu Fengguang <fengguang.wu@intel.com> wrote:
> > >
> > > > Masasyoshi,
> > > >
> > > > On Wed, Mar 18, 2009 at 05:13:35PM +0900, Masasyoshi MIZUMA wrote:
> > > > > I create the patch which fixes lack of mutex_lock in drop_pagecache_sb().
> > > > > Please check the bug and the patch (below).
> > > >
> > > Thank you for your comment, and I apologize to you for my lack
> > > of explanation.
> > >
> > > > Is this a real producible bug or a theory one?
> > > This is a real bug.
> > >
> > > > IMHO the I_FREEING flag should avoid the race.
> > > I supplement the explanation for this problem.
> > >
> > > clear_inode() is called by dispose_list(), and sets the inode's
> > > i_state to I_CLEAR. Therefore, the following conditional expression
> > > doesn't match for the inode:
> > > "if (inode->i_state & (I_FREEING|I_WILL_FREE)) continue;"
> > > As the result, this problem can happen.
> > >
> > > >
> > > > > ----------------------------------------------------------------------
> > > > >
> > > > > When drop_pagecache_sb() frees inodes, it doesn't get mutex_lock of
> > > > > iprune_mutex. Therefore, if it races the process which frees inodes
> > > > > (ex. prune_icache()), OS panic may happen.
> > > > >
> > > > > An example of the panic flow is the following:
> > > > > ----------------------------------------------------------------------
> > > > > [process A] | [process B]
> > > > > | |
> > > > > | shrink_icache_memory() |
> > > > > | | |
> > > > > | V |
> > > > > | prune_icache() | drop_pagecache()
> > > > > | mutex_lock(&iprune_mutex) | |
> > > > > | spin_lock(&inode_lock) | |
> > > > > | | | V
> > > > > | | | drop_pagecache_sb()
> > > > > | | | |
> > > >
> > > > inode->i_state |= I_FREEING;
> > > >
> > > > > | V | V
> > > > > | spin_unlock(&inode_lock) | spin_lock(&inode_lock)
> > > > > | | | |
> > > >
> > > > if (inode->i_state & (I_FREEING|I_WILL_FREE))
> > > > continue;
> > > >
> > > > > | | | |
> > > > > | V | V
> > > > > | dispose_list() | __iget()
> > > > > | list_del() | |
> > > > > | | | |
> > > > > | V | V
> > > > > | spin_lock(&inode_lock) | list_move() <----- PANIC !!
> > > > > | |
> > > > > V |
> > > > > (time)
> > > > > ----------------------------------------------------------------------
> > > > > If the inode which Process B do list_move() with is the same as the one which
> > > > > Process A did list_del() with, OS may panic.
> > >
> > > I applied your comment and then modified the panic flow figure.
> > > Please check below:
> > > ----------------------------------------------------------------------
> > > [process A] | [process B]
> > > | |
> > > | shrink_icache_memory() |
> > > | | |
> > > | V |
> > > | prune_icache() | drop_pagecache()
> > > | mutex_lock(&iprune_mutex) | |
> > > | spin_lock(&inode_lock) | |
> > > | | | V
> > > | | | drop_pagecache_sb()
> > > | | | |
> > > | V | |
> > > | inode->i_state |= I_FREEING; | |
> > > | | | |
> > > | V | V
> > > | spin_unlock(&inode_lock) | spin_lock(&inode_lock)
> > > | | | |
> > > | | | |
> > > | V | |
> > > | dispose_list() | |
> > > | list_del() | |
> > > | | | |
> > > | V | |
> > > | clear_inode() | |
> > > | inode->i_state = I_CLEAR | |
> > > | | | |
> > > | | | V
> > > | | | if (inode->i_state & (I_FREEING|I_WILL_FREE))
> > > | | | continue; <---- NOT MATCH
> > > | | | |
> > > | | | V
> > > | | | __iget()
> > > | | | |
> > > | V | V
> > > | spin_lock(&inode_lock) | list_move() <----- PANIC !!
> > > | |
> > > V |
> > > (time)
> > > ----------------------------------------------------------------------
> >
> > Ah thanks for the explanation!
> >
> > How about this lightweight fix? Since s_umount is already taken in
> > drop_pagecache(), it's not necessary to take iprune_mutex again.
> Yes, the patch looks fine like this. But I believe there are more places
> which might miss similar check. Probably dquot.c: add_dquot_ref() needs
> the same check and fs-writeback.c: generic_sync_sb_inodes() also needs
> adding I_CLEAR to the test, doesn't it?
Exactly! Basically any test on I_FREEING shall be coupled with I_CLEAR,
because of the I_FREEING => I_CLEAR transition _outside_ of inode_lock.
Thanks,
Fengguang
> > --- mm.orig/fs/drop_caches.c
> > +++ mm/fs/drop_caches.c
> > @@ -18,7 +18,7 @@ static void drop_pagecache_sb(struct sup
> >
> > spin_lock(&inode_lock);
> > list_for_each_entry(inode, &sb->s_inodes, i_sb_list) {
> > - if (inode->i_state & (I_FREEING|I_WILL_FREE))
> > + if (inode->i_state & (I_FREEING|I_CLEAR|I_WILL_FREE))
> > continue;
> > if (inode->i_mapping->nrpages == 0)
> > continue;
> --
> Jan Kara <jack@suse.cz>
> SUSE Labs, CR
next prev parent reply other threads:[~2009-03-24 12:11 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-03-18 8:13 [PATCH][BUG] Lack of mutex_lock in drop_pagecache_sb() Masasyoshi MIZUMA
2009-03-23 10:38 ` Wu Fengguang
2009-03-24 7:06 ` Masayoshi MIZUMA
2009-03-24 7:44 ` Wu Fengguang
2009-03-24 12:05 ` Jan Kara
2009-03-24 12:11 ` Wu Fengguang [this message]
2009-03-24 12:40 ` [PATCH] skip I_CLEAR state inodes Wu Fengguang
2009-03-30 7:18 ` [PATCH][RESEND for 2.6.29-rc8-mm1] " Wu Fengguang
2009-03-31 23:43 ` Andrew Morton
2009-04-01 0:53 ` Wu Fengguang
2009-06-01 21:38 ` [PATCH] " Eric Sandeen
2009-06-02 8:55 ` Wu Fengguang
2009-06-02 10:27 ` Jeff Layton
2009-06-02 11:37 ` Jan Kara
2009-06-02 21:48 ` Eric Sandeen
2009-06-03 10:45 ` Jeff Layton
2009-06-03 13:32 ` Wu Fengguang
2009-06-03 14:00 ` Jan Kara
2009-06-03 14:10 ` Wu Fengguang
2009-06-03 14:16 ` Jan Kara
2009-06-03 14:47 ` Wu Fengguang
2009-06-06 3:07 ` [PATCH] writeback: skip new or to-be-freed inodes Wu Fengguang
2009-06-08 7:03 ` Artem Bityutskiy
2009-06-08 9:29 ` Wu Fengguang
2009-06-08 10:45 ` Christoph Hellwig
2009-06-09 7:24 ` Artem Bityutskiy
2009-06-09 7:03 ` Artem Bityutskiy
2009-06-08 17:07 ` Jan Kara
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=20090324121129.GA25410@localhost \
--to=fengguang.wu@intel.com \
--cc=jack@suse.cz \
--cc=linux-fsdevel@vger.kernel.org \
--cc=m.mizuma@jp.fujitsu.com \
--cc=npiggin@suse.de \
--cc=viro@zeniv.linux.org.uk \
/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).