xfs_file_last_byte() needs to acquire ilock
A message that is not in the archive (sent off-list or not yet received): 344266684.4811240544710893.JavaMail.root@zmail05.collab.prod.int.phx2.redhat.com
Quoted 35 lines · ----- "Eric Sandeen
Quoted 13 lines · Lachlan McIlroy
We had some systems crash with this stack:
[<a00000010000cb20>] ia64_leave_kernel+0x0/0x280 [<a00000021291ca00>] xfs_bmbt_get_startoff+0x0/0x20 [xfs] [<a0000002129080b0>] xfs_bmap_last_offset+0x210/0x280 [xfs] [<a00000021295b010>] xfs_file_last_byte+0x70/0x1a0 [xfs] [<a00000021295b200>] xfs_itruncate_start+0xc0/0x1a0 [xfs] [<a0000002129935f0>] xfs_inactive_free_eofblocks+0x290/0x460 [xfs] [<a000000212998fb0>] xfs_release+0x1b0/0x240 [xfs] [<a0000002129ad930>] xfs_file_release+0x70/0xa0 [xfs] [<a000000100162ea0>] __fput+0x1a0/0x420 [<a000000100163160>] fput+0x40/0x60Quoted 2 lines
The problem here is that
xfs_file_last_byte()does not acquire the inode lock and can therefore race with another thread that ismodifying
Quoted 2 lines
the extext list. While
xfs_bmap_last_offset()is trying to lookup what was the last extent some extents were merged and the extentlist
Quoted 1 line
shrunk so the index we lookup is now beyond the end of the extent
list
fs/xfs/xfs_inode.c xfs_file_last_byte(
Quoted 7 lines
and potentially in a freed buffer.
fs/xfs/xfs_inode.c +0
-0@@ -1258,8 +1258,10 @@ xfs_file_last_byte(/*
* Only check for blocks beyond the EOF if the extents have
* been read in. This eliminates the need for the inode
lock,
* and it also saves us from looking when it really isn't
Quoted 6 lines
fs/xfs/xfs_inode.c xfs_file_last_byte( · line 1259
Quoted 2 lines
fs/xfs/xfs_inode.c +0
-0* necessary. */ I suppose that comment should be modified too, and maybe the commit log
should say why, exactly, it was wrong? :)
Ha, I didn't even read the comment! It's still kind of correct in that we wont have to get the inode lock if the extents have not been read in.
Quoted 12 lines
-Eric
Quoted 9 lines
fs/xfs/xfs_inode.c +2
-0if (ip->i_df.if_flags &XFS_IFEXTENTS ) { +xfs_ilock (ip,XFS_ILOCK_SHARED ); error =xfs_bmap_last_offset (NULL , ip, &last_block,XFS_DATA_FORK ); +xfs_iunlock (ip,XFS_ILOCK_SHARED );if (error) { last_block =0 ; }
Quoted 10 lines
Quoted 5 lines
_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
Quoted 41 lines · Lachlan McIlroy
Quoted 38 lines · ----- "Eric Sandeen
Quoted 13 lines · Lachlan McIlroy
We had some systems crash with this stack:
[<a00000010000cb20>] ia64_leave_kernel+0x0/0x280 [<a00000021291ca00>] xfs_bmbt_get_startoff+0x0/0x20 [xfs] [<a0000002129080b0>] xfs_bmap_last_offset+0x210/0x280 [xfs] [<a00000021295b010>] xfs_file_last_byte+0x70/0x1a0 [xfs] [<a00000021295b200>] xfs_itruncate_start+0xc0/0x1a0 [xfs] [<a0000002129935f0>] xfs_inactive_free_eofblocks+0x290/0x460 [xfs] [<a000000212998fb0>] xfs_release+0x1b0/0x240 [xfs] [<a0000002129ad930>] xfs_file_release+0x70/0xa0 [xfs] [<a000000100162ea0>] __fput+0x1a0/0x420 [<a000000100163160>] fput+0x40/0x60Quoted 2 lines
The problem here is that
xfs_file_last_byte()does not acquire the inode lock and can therefore race with another thread that ismodifying
Quoted 2 lines
the extext list. While
xfs_bmap_last_offset()is trying to lookup what was the last extent some extents were merged and the extentlist
Quoted 1 line
shrunk so the index we lookup is now beyond the end of the extent
list
fs/xfs/xfs_inode.c xfs_file_last_byte(
Quoted 7 lines
and potentially in a freed buffer.
fs/xfs/xfs_inode.c +0
-0@@ -1258,8 +1258,10 @@ xfs_file_last_byte(/*
* Only check for blocks beyond the EOF if the extents have
* been read in. This eliminates the need for the inode
lock,
* and it also saves us from looking when it really isn'tQuoted 2 lines
fs/xfs/xfs_inode.c +0
-0* necessary. */
Quoted 7 lines
Quoted 3 lines
I suppose that comment should be modified too, and maybe the commit log
should say why, exactly, it was wrong? :)Ha, I didn't even read the comment! It's still kind of correct in that we wont have to get the inode lock if the extents have not been
read in.
I'd still think the comments could be made less confusing if we're adding the inode lock here.
Felix
Quoted 25 lines
Quoted 14 lines
-Eric
Quoted 9 lines
fs/xfs/xfs_inode.c +2
-0if (ip->i_df.if_flags &XFS_IFEXTENTS ) { +xfs_ilock (ip,XFS_ILOCK_SHARED ); error =xfs_bmap_last_offset (NULL , ip, &last_block,XFS_DATA_FORK ); +xfs_iunlock (ip,XFS_ILOCK_SHARED );if (error) { last_block =0 ; }Quoted 10 lines
Quoted 5 lines
_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
Quoted 4 lines
_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs