From: Ingo Molnar <mingo@elte.hu>
To: Trond Myklebust <trond.myklebust@fys.uio.no>
Cc: Yinghai Lu <yhlu.kernel@gmail.com>,
Pekka Enberg <penberg@cs.helsinki.fi>,
Arjan van de Ven <arjan@infradead.org>,
David Miller <davem@davemloft.net>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: [re-send PATCH] nfs: Fix nfs_parse_mount_options() double kfree()
Date: Mon, 19 Oct 2009 09:04:21 +0200 [thread overview]
Message-ID: <20091019070421.GA7256@elte.hu> (raw)
In-Reply-To: <20091019065433.GA29550@elte.hu>
* Ingo Molnar <mingo@elte.hu> wrote:
> * Trond Myklebust <trond.myklebust@fys.uio.no> wrote:
>
> > > yes. something miss merged again...
> > >
> > > need change some lines.
> >
> > This doesn't match mainline either. To do so, the above kfree() has to
> > be at the end of the "Opt_xprt_rdma:" case...
>
> it's from a test patch in tip:out-of-tree:
>
> d40bc6b: <not-for-merge> nfs: fix nfs_parse_mount_options() double kfree()
>
> (attached below)
>
> that fix is wrong apparently - is there a correct fix upstream perhaps?
Looks like this might be related to a long-standing NFS bug not fixed
upstream yet. You were Cc:-ed to the original thread and to the original
patch, see this thread on lkml:
nfs mount fail on linus 20090402 git
This newer "nfs mount fail" thread got started because the surrounding
code changed and the pending fix got mismerged. The merged up (and
fixed) version is attached below - what remains of it by today is a
memory leak fix.
The original NFS bug was apparently fixed upstream without crediting
Yinghai and Pekka for finding it, and without a reply to the "nfs mount
fail on linus 20090402 git" thread, so the patch stayed pending.
Ingo
-------------->
>From 6914a677dc32b99b5be4e76b84b465fa3e417b94 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Fri, 3 Apr 2009 09:06:15 +0200
Subject: [PATCH] <not-for-merge> nfs: fix nfs_parse_mount_options() double kfree()
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
Impact: fix crash
Yinghai Lu reported the following crash:
> mpk12-3214-189-158:~ # sh x
> [ 63.198629] ------------[ cut here ]------------
> [ 63.202589] kernel BUG at mm/slub.c:2753!
> [ 63.202589] invalid opcode: 0000 [#1] SMP
> [ 63.202589] last sysfs file: /sys/devices/virtual/net/sit0/type
> [ 63.202589] CPU 0
> [ 63.202589] Modules linked in:
> [ 63.202589] Pid: 10027, comm: mount.nfs Not tainted 2.6.29-07100-g833bb30 #21 Sun Fire X4440
> [ 63.202589] RIP: 0010:[<ffffffff802e0015>] [<ffffffff802e0015>] kfree+0x5a/0xcd
> [ 63.202589] RSP: 0018:ffff882042ceb9f8 EFLAGS: 00010246
> [ 63.202589] RAX: 0200000000000000 RBX: 0000000000000005 RCX: ffffffff80a7dc1f
> [ 63.202589] RDX: ffffe20000000000 RSI: ffffc2000000f470 RDI: ffffe2001c018950
> [ 63.202589] RBP: ffff882042ceba18 R08: 0000000000000000 R09: ffffffff811019c0
> [ 63.202589] R10: 000000004262ce02 R11: ffff882042ceba18 R12: ffff880800706475
> [ 63.202589] R13: ffff882042886000 R14: ffff882042cebbd8 R15: ffff882042cebbf0
> [ 63.202589] FS: 00007fac729ed6f0(0000) GS:ffffc20000000000(0000) knlGS:0000000000000000
> [ 63.202589] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
> [ 63.202589] CR2: 00007fac72c12000 CR3: 0000001841cbb000 CR4: 00000000000006e0
> [ 63.202589] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> [ 63.202589] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
> [ 63.202589] Process mount.nfs (pid: 10027, threadinfo ffff882042cea000, task ffff8820434dc290)
> [ 63.202589] Stack:
> [ 63.202589] ffff882042ceba18 000000004262ce02 0000000000000005 ffff882042886028
> [ 63.202589] ffff882042ceba58 ffffffff80a7dc1f 000000004262ce02 ffff882042886000
> [ 63.202589] 000000004262ce02 ffff882042886000 ffffffff80a7b4a6 ffff882042c9ee18
> [ 63.202589] Call Trace:
> [ 63.202589] [<ffffffff80a7dc1f>] xs_destroy+0x67/0xac
> [ 63.202589] [<ffffffff80a7b4a6>] ? xprt_destroy+0x0/0xa7
> [ 63.202589] [<ffffffff80a7b532>] xprt_destroy+0x8c/0xa7
> [ 63.202589] [<ffffffff80a823b2>] ? put_rpccred+0x112/0x131
> [ 63.202589] [<ffffffff8051cdd5>] kref_put+0x65/0x87
> [ 63.202589] [<ffffffff80a7a9a9>] ? rpc_free_client+0x0/0xf9
> [ 63.202589] [<ffffffff80a7b490>] xprt_put+0x23/0x39
> [ 63.202589] [<ffffffff80a7aa7a>] rpc_free_client+0xd1/0xf9
> [ 63.202589] [<ffffffff80a83345>] ? unx_destroy+0x3c/0x57
> [ 63.202589] [<ffffffff8051cdd5>] kref_put+0x65/0x87
> [ 63.202589] [<ffffffff80a7aaa2>] ? rpc_free_auth+0x0/0x69
> [ 63.202589] [<ffffffff80a7aaf0>] rpc_free_auth+0x4e/0x69
> [ 63.202589] [<ffffffff8025b827>] ? __wake_up+0x52/0x75
> [ 63.202589] [<ffffffff8051cdd5>] kref_put+0x65/0x87
> [ 63.202589] [<ffffffff80a7a98e>] rpc_release_client+0x64/0x7f
> [ 63.202589] [<ffffffff80a8061c>] ? rpc_put_task+0xb0/0xcb
> [ 63.202589] [<ffffffff80a7abe0>] rpc_shutdown_client+0xd5/0xf8
> [ 63.202589] [<ffffffff80a7a893>] ? rpc_call_sync+0x63/0x80
> [ 63.202589] [<ffffffff803fc4ab>] nfs_mount+0x11f/0x1bf
> [ 63.202589] [<ffffffff803f3036>] nfs_get_sb+0x4ac/0x82a
> [ 63.202589] [<ffffffff802e8f24>] vfs_kern_mount+0x61/0xbf
> [ 63.202589] [<ffffffff802fea1d>] ? get_fs_type+0x58/0xc5
> [ 63.202589] [<ffffffff802e9015>] do_kern_mount+0x56/0x108
> [ 63.202589] [<ffffffff80302195>] do_mount+0x729/0x788
> [ 63.202589] [<ffffffff80300025>] ? copy_mount_options+0xdf/0x155
> [ 63.202589] [<ffffffff8030228c>] sys_mount+0x98/0xf8
> [ 63.202589] [<ffffffff80230d6b>] system_call_fastpath+0x16/0x1b
> [ 63.202589] Code: 0c 48 ba 00 00 00 00 00 e2 ff ff 48 6b c0 38 48 8d 3c 10 48 8b 07 f6 c4 40 74 04 48 8b 7f 10 48 8b 07 84 c0 78 10 f6 c4 60 75 04 <0f> 0b eb fe e8 90 75 fd ff eb 4c 48 8b 4d 08 4c 8b 4f 10 9c 5b
> [ 63.202589] RIP [<ffffffff802e0015>] kfree+0x5a/0xcd
> [ 63.202589] RSP <ffff882042ceb9f8>
> [ 63.524555] ---[ end trace cd0d38e02ad11d61 ]---
Pekka observed that a bogus pointer was passed to kfree().
This commit:
a67d18f: NFS: load the rpc/rdma transport module automatically
Moved a kfree() of the options strings in nfs_parse_mount_options()
inadvertently and introduced a double kfree(). Fix it.
Reported-by: Yinghai Lu <yinghai@kernel.org>
Analyzed-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
fs/nfs/super.c | 2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index a2c18ac..1fcb375 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -1253,6 +1253,7 @@ static int nfs_parse_mount_options(char *raw,
default:
dfprintk(MOUNT, "NFS: unrecognized "
"transport protocol\n");
+ kfree(string);
return 0;
}
break;
next prev parent reply other threads:[~2009-10-19 7:04 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-10-18 1:25 nfs mount fail Yinghai Lu
2009-10-18 4:22 ` Yinghai Lu
2009-10-19 2:35 ` Pekka Enberg
2009-10-19 2:43 ` Arjan van de Ven
2009-10-19 2:58 ` Pekka Enberg
2009-10-19 5:52 ` Yinghai Lu
2009-10-19 6:30 ` Trond Myklebust
2009-10-19 6:54 ` Ingo Molnar
2009-10-19 6:58 ` Pekka Enberg
2009-10-19 7:08 ` Ingo Molnar
2009-10-19 7:31 ` Yinghai Lu
2009-10-19 7:52 ` Trond Myklebust
2009-10-19 8:04 ` Ingo Molnar
2009-10-19 8:23 ` Ingo Molnar
2009-10-19 8:29 ` Trond Myklebust
2009-10-19 8:39 ` [PATCH] nfs: Fix nfs_parse_mount_options() kfree() leak Ingo Molnar
2009-10-19 7:04 ` Ingo Molnar [this message]
2009-10-19 7:10 ` [re-send PATCH] nfs: Fix nfs_parse_mount_options() double kfree() Ingo Molnar
2009-10-19 7:04 ` nfs mount fail Trond Myklebust
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=20091019070421.GA7256@elte.hu \
--to=mingo@elte.hu \
--cc=arjan@infradead.org \
--cc=davem@davemloft.net \
--cc=linux-kernel@vger.kernel.org \
--cc=penberg@cs.helsinki.fi \
--cc=trond.myklebust@fys.uio.no \
--cc=yhlu.kernel@gmail.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
all inboxes | Powered by JetHome®