X-Spam-Checker-Version: SpamAssassin 3.4.0-r929098 (2010-03-30) on oss.sgi.com
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00,FREEMAIL_FROM,
	T_DKIM_INVALID autolearn=unavailable version=3.4.0-r929098
Received: from cuda.sgi.com (cuda1.sgi.com [192.48.157.11])
	by oss.sgi.com (8.14.3/8.14.3/SuSE Linux 0.8) with ESMTP id q1H5cps9055447
	for <xfs-masters@oss.sgi.com>; Thu, 16 Feb 2012 23:38:52 -0600
X-ASG-Debug-ID: 1329457128-04bdf002ce1f790001-7TOuyN
Received: from mail-bk0-f53.google.com (mail-bk0-f53.google.com [209.85.214.53]) by cuda.sgi.com with ESMTP id dLHRVRUuCdcFmuwG (version=TLSv1 cipher=RC4-SHA bits=128 verify=NO); Thu, 16 Feb 2012 21:38:49 -0800 (PST)
X-Barracuda-Envelope-From: linkinjeon@gmail.com
X-Barracuda-Apparent-Source-IP: 209.85.214.53
Received: by bkcit16 with SMTP id it16so3040877bkc.26
        for <multiple recipients>; Thu, 16 Feb 2012 21:38:48 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:date:message-id:subject:from:to
         :cc:content-type:content-transfer-encoding;
        bh=qdGLMu/h6UdTWfq14kTLvs8zPQ+2CS7QoFgBY4EuB5g=;
        b=jGH4OYhYkW0f2WLw5au88HKXlfC1NFZCK6S5pOmTkZFD+OtBUeraetq2K3yO/S/him
         Pv6JPEHQtLTlPCnZQTim1sSY9DpLH1vhgAbtFZx4QXyms+gkJZC0MpORiJO3ou0TBshP
         idlrWJP3qXaBxNyUVOj9qbwg43PpnlEDRp0ug=
MIME-Version: 1.0
Received: by 10.204.128.143 with SMTP id k15mr3593403bks.101.1329457128171;
 Thu, 16 Feb 2012 21:38:48 -0800 (PST)
Received: by 10.204.58.80 with HTTP; Thu, 16 Feb 2012 21:38:48 -0800 (PST)
In-Reply-To: <CADDb1s2_rQJsvLxCFgc1jreMx1vNvQV2YA50Z6k8BcMinQbsiw@mail.gmail.com>
References: <1329306980-17997-1-git-send-email-amit.sahrawat83@gmail.com>
	<CADDb1s2_rQJsvLxCFgc1jreMx1vNvQV2YA50Z6k8BcMinQbsiw@mail.gmail.com>
Date: Fri, 17 Feb 2012 14:38:48 +0900
Message-ID: <CAKYAXd83T3fxEF+_Q7BOfaOg0ngvh41-q=qNOfXFNROB3gCU1w@mail.gmail.com>
Subject: Re: [PATCH 1/1] xfs: fix buffer flushing during log unmount
From: Namjae Jeon <linkinjeon@gmail.com>
X-ASG-Orig-Subj: Re: [PATCH 1/1] xfs: fix buffer flushing during log unmount
To: Dave Chinner <david@fromorbit.com>, Christoph Hellwig <hch@infradead.org>
Cc: Ben Myers <bpm@sgi.com>, Alex Elder <elder@kernel.org>,
        xfs-masters@oss.sgi.com, xfs@oss.sgi.com, linux-kernel@vger.kernel.org,
        Amit Sahrawat <amit.sahrawat83@gmail.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
X-Barracuda-Connect: mail-bk0-f53.google.com[209.85.214.53]
X-Barracuda-Start-Time: 1329457129
X-Barracuda-Encrypted: RC4-SHA
X-Barracuda-URL: http://192.48.157.11:80/cgi-mod/mark.cgi
X-Virus-Scanned: by bsmtpd at sgi.com
X-Barracuda-Spam-Score: 0.00
X-Barracuda-Spam-Status: No, SCORE=0.00 using per-user scores of TAG_LEVEL=1000.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=1.3 tests=DKIM_SIGNED, DKIM_VERIFIED
X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.2.88752
	Rule breakdown below
	 pts rule name              description
	---- ---------------------- --------------------------------------------------
	-0.00 DKIM_VERIFIED          Domain Keys Identified Mail: signature passes
	                           verification
	0.00 DKIM_SIGNED            Domain Keys Identified Mail: message has a signature

Hi. Dave, Christoph.

Would you please review this patch ?

Thanks.

2012/2/15 Amit Sahrawat <amit.sahrawat83@gmail.com>:
> Errror logs on 3.0.18
> Architecture: ARM
> Just tried to mount a corrupted USB HDD - which resulted in the below
> mentioned behaviour.
>
> #> mount -t xfs /dev/sdb3 /mnt/
> XFS (sdb3): Mounting Filesystem
> XFS (sdb3): Starting recovery (logdev: internal)
> e420d000: 3f b5 ce 5d 15 3b 64 e2 bb b4 f2 9b a0 97 f5 f4 =C2=A0?..].;d..=
.......
> XFS (sdb3): Internal error xfs_btree_check_sblock at line 120 of file
> fs/xfs/xfs_btree.c. =C2=A0Caller 0xc012c444
>
> [<c003a008>] (unwind_backtrace+0x0/0xe4) from [<c013c990>]
> (xfs_corruption_error+0x54/0x70)
> [<c013c990>] (xfs_corruption_error+0x54/0x70) from [<c012c310>]
> (xfs_btree_check_sblock+0xe4/0xf8)
> [<c012c310>] (xfs_btree_check_sblock+0xe4/0xf8) from [<c012c444>]
> (xfs_btree_read_buf_block+0x78/0x98)
> [<c012c444>] (xfs_btree_read_buf_block+0x78/0x98) from [<c012e0b0>]
> (xfs_btree_rshift+0xb0/0x508)
> [<c012e0b0>] (xfs_btree_rshift+0xb0/0x508) from [<c012e5c4>]
> (xfs_btree_make_block_unfull+0xbc/0x168)
> [<c012e5c4>] (xfs_btree_make_block_unfull+0xbc/0x168) from
> [<c012e854>] (xfs_btree_insrec+0x1e4/0x504)
> [<c012e854>] (xfs_btree_insrec+0x1e4/0x504) from [<c012ebd8>]
> (xfs_btree_insert+0x64/0x15c)
> [<c012ebd8>] (xfs_btree_insert+0x64/0x15c) from [<c011a51c>]
> (xfs_free_ag_extent+0x478/0x5a8)
> [<c011a51c>] (xfs_free_ag_extent+0x478/0x5a8) from [<c011af1c>]
> (xfs_free_extent+0xcc/0x108)
> [<c011af1c>] (xfs_free_extent+0xcc/0x108) from [<c014d2b4>]
> (xlog_recover_process_efi+0x168/0x1d4)
> [<c014d2b4>] (xlog_recover_process_efi+0x168/0x1d4) from [<c014d380>]
> (xlog_recover_process_efis+0x60/0xac)
> [<c014d380>] (xlog_recover_process_efis+0x60/0xac) from [<c014d8b4>]
> (xlog_recover_finish+0x18/0x90)
> [<c014d8b4>] (xlog_recover_finish+0x18/0x90) from [<c0154390>]
> (xfs_mountfs+0x4c8/0x5c4)
> [<c0154390>] (xfs_mountfs+0x4c8/0x5c4) from [<c0167c5c>]
> (xfs_fs_fill_super+0x150/0x244)
> [<c0167c5c>] (xfs_fs_fill_super+0x150/0x244) from [<c00c05f4>]
> (mount_bdev+0x120/0x19c)
> [<c00c05f4>] (mount_bdev+0x120/0x19c) from [<c0166198>] (xfs_fs_mount+0x1=
0/0x18)
> [<c0166198>] (xfs_fs_mount+0x10/0x18) from [<c00bf39c>] (mount_fs+0x10/0x=
b8)
> [<c00bf39c>] (mount_fs+0x10/0xb8) from [<c00d6d50>] (vfs_kern_mount+0x50/=
0x88)
> [<c00d6d50>] (vfs_kern_mount+0x50/0x88) from [<c00d700c>]
> (do_kern_mount+0x34/0xc8)
> [<c00d700c>] (do_kern_mount+0x34/0xc8) from [<c00d8420>] (do_mount+0x620/=
0x688)
> [<c00d8420>] (do_mount+0x620/0x688) from [<c00d850c>] (sys_mount+0x84/0xc=
4)
> [<c00d850c>] (sys_mount+0x84/0xc4) from [<c0034260>] (ret_fast_syscall+0x=
0/0x30)
> XFS (sdb3): Corruption detected. Unmount and run xfs_repair
> XFS (sdb3): Internal error xfs_trans_cancel at line 1928 of file
> fs/xfs/xfs_trans.c. =C2=A0Caller 0xc014d314
>
> [<c003a008>] (unwind_backtrace+0x0/0xe4) from [<c0156e34>]
> (xfs_trans_cancel+0x70/0xfc)
> [<c0156e34>] (xfs_trans_cancel+0x70/0xfc) from [<c014d314>]
> (xlog_recover_process_efi+0x1c8/0x1d4)
> [<c014d314>] (xlog_recover_process_efi+0x1c8/0x1d4) from [<c014d380>]
> (xlog_recover_process_efis+0x60/0xac)
> [<c014d380>] (xlog_recover_process_efis+0x60/0xac) from [<c014d8b4>]
> (xlog_recover_finish+0x18/0x90)
> [<c014d8b4>] (xlog_recover_finish+0x18/0x90) from [<c0154390>]
> (xfs_mountfs+0x4c8/0x5c4)
> [<c0154390>] (xfs_mountfs+0x4c8/0x5c4) from [<c0167c5c>]
> (xfs_fs_fill_super+0x150/0x244)
> [<c0167c5c>] (xfs_fs_fill_super+0x150/0x244) from [<c00c05f4>]
> (mount_bdev+0x120/0x19c)
> [<c00c05f4>] (mount_bdev+0x120/0x19c) from [<c0166198>] (xfs_fs_mount+0x1=
0/0x18)
> [<c0166198>] (xfs_fs_mount+0x10/0x18) from [<c00bf39c>] (mount_fs+0x10/0x=
b8)
> [<c00bf39c>] (mount_fs+0x10/0xb8) from [<c00d6d50>] (vfs_kern_mount+0x50/=
0x88)
> [<c00d6d50>] (vfs_kern_mount+0x50/0x88) from [<c00d700c>]
> (do_kern_mount+0x34/0xc8)
> [<c00d700c>] (do_kern_mount+0x34/0xc8) from [<c00d8420>] (do_mount+0x620/=
0x688)
> [<c00d8420>] (do_mount+0x620/0x688) from [<c00d850c>] (sys_mount+0x84/0xc=
4)
> [<c00d850c>] (sys_mount+0x84/0xc4) from [<c0034260>] (ret_fast_syscall+0x=
0/0x30)
> XFS (sdb3): xfs_do_force_shutdown(0x8) called from line 1929 of file
> fs/xfs/xfs_trans.c. =C2=A0Return address =3D 0xc0156e48
> XFS (sdb3): Corruption of in-memory data detected. =C2=A0Shutting down fi=
lesystem
> XFS (sdb3): Please umount the filesystem and rectify the problem(s)
> XFS (sdb3): Failed to recover EFIs
> XFS (sdb3): log mount finish failed
> Unable to handle kernel paging request at virtual address ffffffff
> pgd =3D e80bc000
> [ffffffff] *pgd=3D68ffc821, *pte=3D00000000, *ppte=3D00000000
> Internal error: Oops: 17 [#1] PREEMPT SMP
> Modules linked in:
> CPU: 1 =C2=A0 =C2=A0Not tainted =C2=A0(3.0.18 #17)
> PC is at strnlen+0x10/0x28
> LR is at string+0x34/0xcc
> pc : [<c01964e0>] =C2=A0 =C2=A0lr : [<c0197ad8>] =C2=A0 =C2=A0psr: a00000=
93
> sp : e424fca0 =C2=A0ip : 00000000 =C2=A0fp : 00000400
> r10: e424fd8c =C2=A0r9 : 00000002 =C2=A0r8 : ffffffff
> r7 : 00000000 =C2=A0r6 : 0000ffff =C2=A0r5 : c03abca8 =C2=A0r4 : c03ab8b0
> r3 : 00000000 =C2=A0r2 : ffffffff =C2=A0r1 : ffffffff =C2=A0r0 : ffffffff
> Flags: NzCv =C2=A0IRQs off =C2=A0FIQs on =C2=A0Mode SVC_32 =C2=A0ISA ARM =
=C2=A0Segment user
> Control: 10c53c7d =C2=A0Table: 680bc04a =C2=A0DAC: 00000015
> Process mount (pid: 656, stack limit =3D 0xe424e2f0)
> Stack: (0xe424fca0 to 0xe4250000)
> ...
> ...
> ffe0: 00000000 be88470c 000436dc 00009604 a0000010 be884b4f 45d65600 0400=
0000
> [<c01964e0>] (strnlen+0x10/0x28) from [<c0197ad8>] (string+0x34/0xcc)
> [<c0197ad8>] (string+0x34/0xcc) from [<c0198874>] (vsnprintf+0x1bc/0x344)
> [<c0198874>] (vsnprintf+0x1bc/0x344) from [<c0198a68>] (vscnprintf+0xc/0x=
24)
> [<c0198a68>] (vscnprintf+0xc/0x24) from [<c0057424>] (vprintk+0x14c/0x3fc=
)
> [<c0057424>] (vprintk+0x14c/0x3fc) from [<c0293820>] (printk+0x18/0x24)
> [<c0293820>] (printk+0x18/0x24) from [<c01660c8>] (xfs_alert_tag+0x64/0x9=
8)
> [<c01660c8>] (xfs_alert_tag+0x64/0x98) from [<c0158034>]
> (xfs_trans_ail_delete_bulk+0x74/0x118)
> [<c0158034>] (xfs_trans_ail_delete_bulk+0x74/0x118) from [<c012fe80>]
> (xfs_buf_iodone+0x2c/0x38)
> [<c012fe80>] (xfs_buf_iodone+0x2c/0x38) from [<c012fe30>]
> (xfs_buf_do_callbacks+0x28/0x38)
> [<c012fe30>] (xfs_buf_do_callbacks+0x28/0x38) from [<c012fffc>]
> (xfs_buf_iodone_callbacks+0x13c/0x164)
> [<c012fffc>] (xfs_buf_iodone_callbacks+0x13c/0x164) from [<c015ffc4>]
> (xfs_buf_iodone_work+0x1c/0x40)
> [<c015ffc4>] (xfs_buf_iodone_work+0x1c/0x40) from [<c0160194>]
> (xfs_bioerror+0x44/0x4c)
> [<c0160194>] (xfs_bioerror+0x44/0x4c) from [<c016075c>]
> (xfs_flush_buftarg+0xcc/0x148)
> [<c016075c>] (xfs_flush_buftarg+0xcc/0x148) from [<c01607f8>]
> (xfs_free_buftarg+0x20/0x5c)
> [<c01607f8>] (xfs_free_buftarg+0x20/0x5c) from [<c0167cd8>]
> (xfs_fs_fill_super+0x1cc/0x244)
> [<c0167cd8>] (xfs_fs_fill_super+0x1cc/0x244) from [<c00c05f4>]
> (mount_bdev+0x120/0x19c)
> [<c00c05f4>] (mount_bdev+0x120/0x19c) from [<c0166198>] (xfs_fs_mount+0x1=
0/0x18)
> [<c0166198>] (xfs_fs_mount+0x10/0x18) from [<c00bf39c>] (mount_fs+0x10/0x=
b8)
> [<c00bf39c>] (mount_fs+0x10/0xb8) from [<c00d6d50>] (vfs_kern_mount+0x50/=
0x88)
> [<c00d6d50>] (vfs_kern_mount+0x50/0x88) from [<c00d700c>]
> (do_kern_mount+0x34/0xc8)
> [<c00d700c>] (do_kern_mount+0x34/0xc8) from [<c00d8420>] (do_mount+0x620/=
0x688)
> [<c00d8420>] (do_mount+0x620/0x688) from [<c00d850c>] (sys_mount+0x84/0xc=
4)
> [<c00d850c>] (sys_mount+0x84/0xc4) from [<c0034260>] (ret_fast_syscall+0x=
0/0x30)
> Code: e3a03000 e1510003 e0832000 0a000003 (e7d0c003)
> ---[ end trace 9fae26d925820746 ]---
> note: mount[656] exited with preempt_count 2
> Segmentation fault
> #>
> #>
>
> Regards,
> Amit Sahrawat
>
> On Wed, Feb 15, 2012 at 5:26 PM, Amit Sahrawat
> <amit.sahrawat83@gmail.com> wrote:
>> Whenever there is a mount/unmount failure - there is a chance of calling=
 the
>> callbacks functions once - transaction ail mount pointer is destroyed. S=
o, it results
>> in NULL pointer exception followed by hang. So, before unmount of the lo=
g - flush all
>> the pending buffers.
>>
>> Signed-off-by: Amit Sahrawat <amit.sahrawat83@gmail.com>
>> Signed-off-by: Namjae Jeon <linkinjeon@gmail.com>
>> ---
>> =C2=A0fs/xfs/xfs_log.c =C2=A0 | =C2=A0 10 ++++++++++
>> =C2=A0fs/xfs/xfs_mount.c | =C2=A0 =C2=A09 ---------
>> =C2=A02 files changed, 10 insertions(+), 9 deletions(-)
>>
>> diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
>> index e2cc356..b924a5b 100644
>> --- a/fs/xfs/xfs_log.c
>> +++ b/fs/xfs/xfs_log.c
>> @@ -739,6 +739,16 @@ xfs_log_unmount_write(xfs_mount_t *mp)
>> =C2=A0void
>> =C2=A0xfs_log_unmount(xfs_mount_t *mp)
>> =C2=A0{
>> + =C2=A0 =C2=A0 =C2=A0 int error =3D 0;
>> + =C2=A0 =C2=A0 =C2=A0 /*
>> + =C2=A0 =C2=A0 =C2=A0 =C2=A0* Make sure all buffers have been flushed a=
nd completed before
>> + =C2=A0 =C2=A0 =C2=A0 =C2=A0* unmounting the log.
>> + =C2=A0 =C2=A0 =C2=A0 =C2=A0*/
>> + =C2=A0 =C2=A0 =C2=A0 error =3D xfs_flush_buftarg(mp->m_ddev_targp, 1);
>> + =C2=A0 =C2=A0 =C2=A0 if (error)
>> + =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 cmn_err(CE_WARN, "%d =
busy buffers during log unmount.", error);
>> + =C2=A0 =C2=A0 =C2=A0 xfs_wait_buftarg(mp->m_ddev_targp);
>> +
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0xfs_trans_ail_destroy(mp);
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0xlog_dealloc_log(mp->m_log);
>> =C2=A0}
>> diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c
>> index d06afbc..3bd2246 100644
>> --- a/fs/xfs/xfs_mount.c
>> +++ b/fs/xfs/xfs_mount.c
>> @@ -1519,15 +1519,6 @@ xfs_unmountfs(
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0"Freespace may not be correct on n=
ext mount.");
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0xfs_unmountfs_writesb(mp);
>>
>> - =C2=A0 =C2=A0 =C2=A0 /*
>> - =C2=A0 =C2=A0 =C2=A0 =C2=A0* Make sure all buffers have been flushed a=
nd completed before
>> - =C2=A0 =C2=A0 =C2=A0 =C2=A0* unmounting the log.
>> - =C2=A0 =C2=A0 =C2=A0 =C2=A0*/
>> - =C2=A0 =C2=A0 =C2=A0 error =3D xfs_flush_buftarg(mp->m_ddev_targp, 1);
>> - =C2=A0 =C2=A0 =C2=A0 if (error)
>> - =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 xfs_warn(mp, "%d busy=
 buffers during unmount.", error);
>> - =C2=A0 =C2=A0 =C2=A0 xfs_wait_buftarg(mp->m_ddev_targp);
>> -
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0xfs_log_unmount_write(mp);
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0xfs_log_unmount(mp);
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0xfs_uuid_unmount(mp);
>> --
>> 1.7.2.3
>>