From: piaojun <piaojun@huawei.com>
To: Changwei Ge <ge.changwei@h3c.com>,
Changwei Ge <gechangwei@live.cn>,
Andrew Morton <akpm@linux-foundation.org>,
"ocfs2-devel@oss.oracle.com" <ocfs2-devel@oss.oracle.com>,
"junxiao.bi@oracle.com" <junxiao.bi@oracle.com>,
"jiangqi903@gmail.com" <jiangqi903@gmail.com>
Cc: "stable@vger.kernel.org" <stable@vger.kernel.org>
Subject: Re: [Ocfs2-devel] [PATCH RESEND] ocfs2: fix a misuse a of brelse after failing ocfs2_check_dir_entry
Date: Sat, 10 Nov 2018 16:04:31 +0800 [thread overview]
Message-ID: <5BE6910F.1090304@huawei.com> (raw)
In-Reply-To: <63ADC13FD55D6546B7DECE290D39E37301277E2772@H3CMLB12-EX.srv.huawei-3com.com>
Hi Changwei,
Sorry for the oversight for your patch. There is nothing wrong.
Thanks,
Jun
On 2018/11/10 15:58, Changwei Ge wrote:
> Hi Jun,
>
> Thanks for looking into this, my code from Linux mainline is not looked like yours...
>
> On 2018/11/10 14:36, piaojun wrote:
>> Hi Changwei,
>>
>> On 2018/5/28 22:40, Changwei Ge wrote:
>>> From: Changwei Ge <ge.changwei@h3c.com>
>>>
>>> Somehow, file system metadata was corrupted, which causes
>>> ocfs2_check_dir_entry() to fail in function ocfs2_dir_foreach_blk_el().
>>>
>>> According to the original design intention, if above happens we should
>>> skip the problematic block and continue to retrieve dir entry. But there
>>> is obviouse misuse of brelse around related code.
>>>
>>> After failure of ocfs2_check_dir_entry(), currunt code just moves to next
>>> position and uses the problematic buffer head again and again during
>>> which the problematic buffer head is released for multiple times. I
>>> suppose, this a serious issue which is long-lived in ocfs2. This may
>>> cause other file systems which is also used in a the same host insane.
>>>
>>> So we should also consider about bakcporting this patch into
>>> linux -stable.
>>>
>>> Suggested-by: Changkuo Shi <shi.changkuo@h3c.com>
>>> Cc: stable@vger.kernel.org
>>> Signed-off-by: Changwei Ge <ge.changwei@h3c.com>
>>> ---
>>> fs/ocfs2/dir.c | 3 +--
>>> 1 file changed, 1 insertion(+), 2 deletions(-)
>>>
>>> diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c
>>> index b048d4f..c121abb 100644
>>> --- a/fs/ocfs2/dir.c
>>> +++ b/fs/ocfs2/dir.c
>>> @@ -1897,8 +1897,7 @@ static int ocfs2_dir_foreach_blk_el(struct inode *inode,
>>> /* On error, skip the f_pos to the
>>> next block. */
>>> ctx->pos = (ctx->pos | (sb->s_blocksize - 1)) + 1;
>>> - brelse(bh);
>>> - continue;
>>> + break;
>>
>> I guess return is more appropriate than break here as it will cause double
>> free buffer:
>>
>> "
>> ctx->pos = (ctx->pos | (sb->s_blocksize - 1)) + 1;
>> brelse(bh);
>> break;
>> "
>
> Linux mainline should not have 'brelse(bh)', could your please rebase your tree?
>
> '''
> ccd979bdb (Mark Fasheh 2005-12-15 14:31:24 -0800 1897) /* On error, skip the f_pos to the
> ccd979bdb (Mark Fasheh 2005-12-15 14:31:24 -0800 1898) next block. */
> 3704412bd (Al Viro 2013-05-22 21:06:00 -0400 1899) ctx->pos = (ctx->pos | (sb->s_blocksize - 1)) + 1;
> 29aa30167 (Changwei Ge 2018-11-02 15:48:15 -0700 1900) break;
> '''
>>
>> "
>> brelse(bh);
>> bh = NULL;
>> if (!persist && stored)
>> break;
>> "
>>
>> Thanks,
>> Jun
>>
>>> }
>>> if (le64_to_cpu(de->inode)) {
>>> unsigned char d_type = DT_UNKNOWN;
>>>
>>
>> _______________________________________________
>> Ocfs2-devel mailing list
>> Ocfs2-devel@oss.oracle.com
>> https://oss.oracle.com/mailman/listinfo/ocfs2-devel
>>
> .
>
WARNING: multiple messages have this Message-ID (diff)
From: piaojun <piaojun@huawei.com>
To: Changwei Ge <ge.changwei@h3c.com>,
Changwei Ge <gechangwei@live.cn>,
Andrew Morton <akpm@linux-foundation.org>,
"ocfs2-devel@oss.oracle.com" <ocfs2-devel@oss.oracle.com>,
"junxiao.bi@oracle.com" <junxiao.bi@oracle.com>,
"jiangqi903@gmail.com" <jiangqi903@gmail.com>
Cc: "stable@vger.kernel.org" <stable@vger.kernel.org>
Subject: [Ocfs2-devel] [PATCH RESEND] ocfs2: fix a misuse a of brelse after failing ocfs2_check_dir_entry
Date: Sat, 10 Nov 2018 16:04:31 +0800 [thread overview]
Message-ID: <5BE6910F.1090304@huawei.com> (raw)
In-Reply-To: <63ADC13FD55D6546B7DECE290D39E37301277E2772@H3CMLB12-EX.srv.huawei-3com.com>
Hi Changwei,
Sorry for the oversight for your patch. There is nothing wrong.
Thanks,
Jun
On 2018/11/10 15:58, Changwei Ge wrote:
> Hi Jun,
>
> Thanks for looking into this, my code from Linux mainline is not looked like yours...
>
> On 2018/11/10 14:36, piaojun wrote:
>> Hi Changwei,
>>
>> On 2018/5/28 22:40, Changwei Ge wrote:
>>> From: Changwei Ge <ge.changwei@h3c.com>
>>>
>>> Somehow, file system metadata was corrupted, which causes
>>> ocfs2_check_dir_entry() to fail in function ocfs2_dir_foreach_blk_el().
>>>
>>> According to the original design intention, if above happens we should
>>> skip the problematic block and continue to retrieve dir entry. But there
>>> is obviouse misuse of brelse around related code.
>>>
>>> After failure of ocfs2_check_dir_entry(), currunt code just moves to next
>>> position and uses the problematic buffer head again and again during
>>> which the problematic buffer head is released for multiple times. I
>>> suppose, this a serious issue which is long-lived in ocfs2. This may
>>> cause other file systems which is also used in a the same host insane.
>>>
>>> So we should also consider about bakcporting this patch into
>>> linux -stable.
>>>
>>> Suggested-by: Changkuo Shi <shi.changkuo@h3c.com>
>>> Cc: stable at vger.kernel.org
>>> Signed-off-by: Changwei Ge <ge.changwei@h3c.com>
>>> ---
>>> fs/ocfs2/dir.c | 3 +--
>>> 1 file changed, 1 insertion(+), 2 deletions(-)
>>>
>>> diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c
>>> index b048d4f..c121abb 100644
>>> --- a/fs/ocfs2/dir.c
>>> +++ b/fs/ocfs2/dir.c
>>> @@ -1897,8 +1897,7 @@ static int ocfs2_dir_foreach_blk_el(struct inode *inode,
>>> /* On error, skip the f_pos to the
>>> next block. */
>>> ctx->pos = (ctx->pos | (sb->s_blocksize - 1)) + 1;
>>> - brelse(bh);
>>> - continue;
>>> + break;
>>
>> I guess return is more appropriate than break here as it will cause double
>> free buffer:
>>
>> "
>> ctx->pos = (ctx->pos | (sb->s_blocksize - 1)) + 1;
>> brelse(bh);
>> break;
>> "
>
> Linux mainline should not have 'brelse(bh)', could your please rebase your tree?
>
> '''
> ccd979bdb (Mark Fasheh 2005-12-15 14:31:24 -0800 1897) /* On error, skip the f_pos to the
> ccd979bdb (Mark Fasheh 2005-12-15 14:31:24 -0800 1898) next block. */
> 3704412bd (Al Viro 2013-05-22 21:06:00 -0400 1899) ctx->pos = (ctx->pos | (sb->s_blocksize - 1)) + 1;
> 29aa30167 (Changwei Ge 2018-11-02 15:48:15 -0700 1900) break;
> '''
>>
>> "
>> brelse(bh);
>> bh = NULL;
>> if (!persist && stored)
>> break;
>> "
>>
>> Thanks,
>> Jun
>>
>>> }
>>> if (le64_to_cpu(de->inode)) {
>>> unsigned char d_type = DT_UNKNOWN;
>>>
>>
>> _______________________________________________
>> Ocfs2-devel mailing list
>> Ocfs2-devel at oss.oracle.com
>> https://oss.oracle.com/mailman/listinfo/ocfs2-devel
>>
> .
>
next prev parent reply other threads:[~2018-11-10 17:49 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-05-28 14:40 [PATCH RESEND] ocfs2: fix a misuse a of brelse after failing ocfs2_check_dir_entry Changwei Ge
2018-05-28 14:40 ` [Ocfs2-devel] " Changwei Ge
2018-11-10 6:33 ` piaojun
2018-11-10 6:33 ` piaojun
2018-11-10 7:58 ` Changwei Ge
2018-11-10 7:58 ` Changwei Ge
2018-11-10 8:04 ` piaojun [this message]
2018-11-10 8:04 ` piaojun
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=5BE6910F.1090304@huawei.com \
--to=piaojun@huawei.com \
--cc=akpm@linux-foundation.org \
--cc=ge.changwei@h3c.com \
--cc=gechangwei@live.cn \
--cc=jiangqi903@gmail.com \
--cc=junxiao.bi@oracle.com \
--cc=ocfs2-devel@oss.oracle.com \
--cc=stable@vger.kernel.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.