Skip to contentSkip to conversation

ocfs2: fix a misuse a of brelse after failing ocfs2_check_dir_entry

PATCH RESEND New Changwei Ge 3 participants 4 messages May 28, 2018 – Nov 10, 2018

Subsystems (2)
Stable branch OCFS2
lore
Share & tools
lore

Changwei Ge

author
#0 lore raw

PATCH RESEND 1/1 ocfs2: fix a misuse a of brelse after failing ocfs2_check_dir_entry

From: Changwei Ge <[email protected]>

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 <[email protected]>
Cc stable at vger.kernel.org
Signed-off-by Changwei Ge <[email protected]>
---
 fs/ocfs2/dir.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/ocfs2/dir.c +1 -2
index b048d4f..c121abb 100644
@@ -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;
 			}
 			if (le64_to_cpu(de->inode)) {
 				unsigned char d_type = DT_UNKNOWN;
Signature · 1 line
2.7.4
#1 lore raw

PATCH RESEND 1/1 ocfs2: fix a misuse a of brelse after failing ocfs2_check_dir_entry

Hi Changwei,

Quoted 30 lines · Changwei Ge

From: Changwei Ge <[email protected]>

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 <[email protected]>
Cc stable at vger.kernel.org
Signed-off-by Changwei Ge <[email protected]>
---
 fs/ocfs2/dir.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/ocfs2/dir.c +0 -0
index b048d4f..c121abb 100644

fs/ocfs2/dir.c static int ocfs2_dir_foreach_blk_el(struct inode *inode, · line 1900

Quoted 6 lines
fs/ocfs2/dir.c +1 -2
@@ -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;
"

"
brelse(bh);
bh = NULL;
if (!persist && stored)

	break;

"

Thanks,
Jun

Quoted 4 lines
fs/ocfs2/dir.c +0 -0
 			}
 			if (le64_to_cpu(de->inode)) {
 				unsigned char d_type = DT_UNKNOWN;
#2 lore raw

PATCH RESEND 1/1 ocfs2: fix a misuse a of brelse after failing ocfs2_check_dir_entry

Hi Jun,

Thanks for looking into this, my code from Linux mainline is not looked like yours...

Quoted 43 lines · piaojun

Hi Changwei,

Quoted 30 lines · Changwei Ge

From: Changwei Ge <[email protected]>

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 <[email protected]>
Cc stable at vger.kernel.org
Signed-off-by Changwei Ge <[email protected]>
---
  fs/ocfs2/dir.c | 3 +--
  1 file changed, 1 insertion(+), 2 deletions(-)
fs/ocfs2/dir.c +0 -0
index b048d4f..c121abb 100644

fs/ocfs2/dir.c static int ocfs2_dir_foreach_blk_el(struct inode *inode, · line 1900

Quoted 6 lines
fs/ocfs2/dir.c +1 -2
@@ -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:

Quoted 5 lines

"
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;

'''

Quoted 21 lines

"
brelse(bh);
bh = NULL;
if (!persist && stored)

	break;

"

Thanks,
Jun

fs/ocfs2/dir.c static int ocfs2_dir_foreach_blk_el(struct inode *inode, · line 1903

Quoted 4 lines
fs/ocfs2/dir.c +0 -0
  			}
  			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

#3 lore raw

PATCH RESEND 1/1 ocfs2: fix a misuse a of brelse after failing ocfs2_check_dir_entry

Hi Changwei,

Sorry for the oversight for your patch. There is nothing wrong.

Thanks,
Jun

Quoted 85 lines · Changwei Ge

Hi Jun,

Thanks for looking into this, my code from Linux mainline is not looked like yours...

Quoted 43 lines · piaojun

Hi Changwei,

Quoted 30 lines · Changwei Ge

From: Changwei Ge <[email protected]>

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 <[email protected]>
Cc stable at vger.kernel.org
Signed-off-by Changwei Ge <[email protected]>
---
  fs/ocfs2/dir.c | 3 +--
  1 file changed, 1 insertion(+), 2 deletions(-)
fs/ocfs2/dir.c +0 -0
index b048d4f..c121abb 100644

fs/ocfs2/dir.c static int ocfs2_dir_foreach_blk_el(struct inode *inode, · line 1900

Quoted 6 lines
fs/ocfs2/dir.c +1 -2
@@ -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:

Quoted 5 lines

"
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;

'''

Quoted 16 lines

"
brelse(bh);
bh = NULL;
if (!persist && stored)

	break;

"

Thanks,
Jun

fs/ocfs2/dir.c static int ocfs2_dir_foreach_blk_el(struct inode *inode, · line 1903

Quoted 4 lines
fs/ocfs2/dir.c +0 -0
  			}
  			if (le64_to_cpu(de->inode)) {
  				unsigned char d_type = DT_UNKNOWN;
Quoted 5 lines

_______________________________________________
Ocfs2-devel mailing list
Ocfs2-devel at oss.oracle.com
https://oss.oracle.com/mailman/listinfo/ocfs2-devel

.