[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #4 from Alessandro Polverini <alex@nibbles.it>  2010-03-20 21:53:31 ---
I updated to debian kernel 2.6.32-3-amd64 and I can reproduce the problem.
When trying to resize the fs I got this on the console:

nb1a kernel: [ 2518.347172] Oops: 0002 [#1] SMP
Message from syslogd@nb1a at Sat Mar 20 22:44:02 2010 ...
nb1a kernel: [ 2518.347195] last sysfs file:
/sys/devices/pci0000:00/0000:00:01.1/i2c-1/1-002c/fan8_input
nb1a kernel: [ 2518.348350] Stack:
nb1a kernel: [ 2518.348699] Call Trace:
nb1a kernel: [ 2518.348699] Code: 48 8b 74 24 50 48 8b 7c 24 50 8b 8b 40 03 00
00 8b 06 8b 57 28 d3 e8 89 c0 48 6b c0 0c 48 89 c1 48 03 8b 48 03 00 00 48 8d
71 04 <f0> 01 51 04 49 8b 95 80 02 00 00 48 03 83 48 03 00 00 8b 52 18
nb1a kernel: [ 2518.348699] CR2: 00000000000153d0

While this is what I can see in /var/log/messages:

Mar 20 22:44:02 nb1a kernel: [ 2518.347146] PGD 41d326067 PUD 41d5c5067 PMD 0 
Mar 20 22:44:02 nb1a kernel: [ 2518.347239] CPU 0 
Mar 20 22:44:02 nb1a kernel: [ 2518.347259] Modules linked in: tun
ip6table_filter ip6_tables iptable_raw xt_comment xt_recent xt_policy ipt_ULOG
ipt_REJECT ipt_REDIRECT ipt_NETMAP ipt_MASQUERADE ipt_LOG ipt_ECN ipt_ecn
ipt_CLUSTERIP ipt_ah ipt_addrtype nf_nat_tftp nf_nat_snmp_basic nf_nat_sip
nf_nat_pptp nf_nat_proto_gre nf_nat_irc nf_nat_h323 nf_nat_ftp nf_nat_amanda
ts_kmp nf_conntrack_amanda nf_conntrack_tftp nf_conntrack_sip
nf_conntrack_proto_sctp nf_conntrack_pptp nf_conntrack_proto_gre
nf_conntrack_netlink nf_conntrack_netbios_ns nf_conntrack_irc nf_conntrack_h323
nf_conntrack_ftp xt_tcpmss xt_pkttype xt_physdev xt_owner xt_NFQUEUE xt_NFLOG
nfnetlink_log xt_multiport xt_MARK xt_mark xt_mac xt_limit xt_length xt_iprange
xt_helper xt_hashlimit xt_DSCP xt_dscp xt_dccp xt_conntrack xt_CONNMARK
xt_connmark xt_CLASSIFY xt_tcpudp xt_state iptable_nat nf_nat nf_conntrack_ipv4
nf_defrag_ipv4 nf_conntrack iptable_mangle iptable_filter ip_tables x_tables
nfnetlink drbd cn autofs4 nfsd nfs lockd fscache nfs_acl a
Mar 20 22:44:02 nb1a kernel: th_rpcgss sunrpc xfs exportfs ext4 jbd2 crc16
w83793 w83627hf hwmon_vid ipmi_msghandler psmouse amd64_edac_mod evdev
serio_raw snd_pcm snd_timer edac_core edac_mce_amd k8temp shpchp snd soundcore
snd_page_alloc pci_hotplug button processor pcspkr i2c_nforce2 i2c_core ext3
jbd mbcache dm_mirror dm_region_hash dm_log dm_snapshot dm_mod ide_cd_mod cdrom
ide_pci_generic amd74xx ide_core sata_nv sd_mod crc_t10dif ata_generic floppy
r8169 mii 3w_9xxx ehci_hcd ohci_hcd libata forcedeth scsi_mod usbcore nls_base
thermal fan thermal_sys [last unloaded: scsi_wait_scan]
Mar 20 22:44:02 nb1a kernel: [ 2518.347914] Pid: 13590, comm: resize2fs Not
tainted 2.6.32-3-amd64 #1 H8DM8-2
Mar 20 22:44:02 nb1a kernel: [ 2518.347942] RIP: 0010:[<ffffffffa0310863>] 
[<ffffffffa0310863>] ext4_group_add+0x136c/0x1451 [ext4]
Mar 20 22:44:02 nb1a kernel: [ 2518.347992] RSP: 0018:ffff88021e7c1c38  EFLAGS:
00010206
Mar 20 22:44:02 nb1a kernel: [ 2518.348017] RAX: 00000000000153cc RBX:
ffff88041d8e7800 RCX: 00000000000153cc
Mar 20 22:44:02 nb1a kernel: [ 2518.348045] RDX: 0000000000007dfe RSI:
00000000000153d0 RDI: ffff88021e7c1e48
Mar 20 22:44:02 nb1a kernel: [ 2518.348072] RBP: ffff8801e1ed1d20 R08:
ffff88041ee03380 R09: 0000000000000000
Mar 20 22:44:02 nb1a kernel: [ 2518.348100] R10: 000000501d59c868 R11:
0000000000000200 R12: 0000000000000000
Mar 20 22:44:02 nb1a kernel: [ 2518.348128] R13: ffff88041d8e1c00 R14:
ffff88021ed67600 R15: 0000000000000200
Mar 20 22:44:02 nb1a kernel: [ 2518.348156] FS:  00007f33572bc700(0000)
GS:ffff880008c00000(0000) knlGS:0000000000000000
Mar 20 22:44:02 nb1a kernel: [ 2518.348198] CS:  0010 DS: 0000 ES: 0000 CR0:
000000008005003b
Mar 20 22:44:02 nb1a kernel: [ 2518.348224] CR2: 00000000000153d0 CR3:
000000041db72000 CR4: 00000000000006f0
Mar 20 22:44:02 nb1a kernel: [ 2518.348251] DR0: 0000000000000000 DR1:
0000000000000000 DR2: 0000000000000000
Mar 20 22:44:02 nb1a kernel: [ 2518.348279] DR3: 0000000000000000 DR6:
00000000ffff0ff0 DR7: 0000000000000400
Mar 20 22:44:02 nb1a kernel: [ 2518.348307] Process resize2fs (pid: 13590,
threadinfo ffff88021e7c0000, task ffff8801ea533170)
Mar 20 22:44:02 nb1a kernel: [ 2518.348369]  00000000fcdf07e0 ffffffff8106beb9
ffff8802098927f8 ffff880209892808
Mar 20 22:44:02 nb1a kernel: [ 2518.348399] <0> ffff8802098927e8
0000000000000292 0000000000000292 ffffffff81098f06
Mar 20 22:44:02 nb1a kernel: [ 2518.348444] <0> 0000000000000000
000080000000fed6 ffff88021e7c1e48 0000000000001c51
Mar 20 22:44:02 nb1a kernel: [ 2518.348699]  [<ffffffff8106beb9>] ?
ktime_get_ts+0x68/0xb2
Mar 20 22:44:02 nb1a kernel: [ 2518.348699]  [<ffffffff81098f06>] ?
delayacct_end+0x74/0x7f
Mar 20 22:44:02 nb1a kernel: [ 2518.348699]  [<ffffffff81064a84>] ?
wake_bit_function+0x0/0x23
Mar 20 22:44:02 nb1a kernel: [ 2518.348699]  [<ffffffff8110ca96>] ?
sync_dirty_buffer+0x5b/0x93
Mar 20 22:44:02 nb1a kernel: [ 2518.348699]  [<ffffffffa02fb804>] ?
ext4_ioctl+0x5ee/0x753 [ext4]
Mar 20 22:44:02 nb1a kernel: [ 2518.348699]  [<ffffffff810f8fa2>] ?
vfs_ioctl+0x21/0x6c
Mar 20 22:44:02 nb1a kernel: [ 2518.348699]  [<ffffffff810f94f0>] ?
do_vfs_ioctl+0x48d/0x4cb
Mar 20 22:44:02 nb1a kernel: [ 2518.348699]  [<ffffffff810edc16>] ?
vfs_write+0xcd/0x102
Mar 20 22:44:02 nb1a kernel: [ 2518.348699]  [<ffffffff810f957f>] ?
sys_ioctl+0x51/0x70
Mar 20 22:44:02 nb1a kernel: [ 2518.348699]  [<ffffffff81010b42>] ?
system_call_fastpath+0x16/0x1b
Mar 20 22:44:02 nb1a kernel: [ 2518.348699]  RSP <ffff88021e7c1c38>
Mar 20 22:44:02 nb1a kernel: [ 2518.354183] ---[ end trace 9ee7dc16cc7e9d74
]---

The size of the partition:
# df /mnt/disk2/
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/mapper/vg1-backup
                     935361592 878396176  56965416  94% /mnt/disk2

I'll now try to reboot the machine and collect to e2image -h data you
requested.

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #5 from Alessandro Polverini <alex@nibbles.it>  2010-03-21 08:32:35 ---
The -h option of e2image is not recognised on my tool, sorry, I tried e2image
1.41.10.
The bzip2 compressed output of e2image of my partition is around 200Mb long,
please everybody who is interested in it contact me so I can prepare a download
link for it.

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #6 from Eric Sandeen <sandeen@redhat.com>  2010-03-22 16:51:25 ---
I'm very sorry, my fingers translated my brain wrong ;)  Ideally we can start
with dumpe2fs -h (not e2image -h, sorry) to quickly recreate a similar
filesystem, and try expanding it.

Keep that e2image around though just in case (although we would need one
created with -r to use it)

Thanks,

-Eric

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #7 from Alessandro Polverini <alex@nibbles.it>  2010-03-22 17:00:22 ---
Hello Eric,
here is the output of dumpe2fs (the file system was mounted when running it).
Please note that now the FS has been expanded, but the operation has been done
(like the other times) offline (unmounted). The kernel hangs when trying to do
it online (mounted).

dumpe2fs 1.41.3 (12-Oct-2008)
Filesystem volume name:   <none>
Last mounted on:          /mnt/disk2
Filesystem UUID:          b0e46f62-5ac9-48f5-8582-f602652385d3
Filesystem magic number:  0xEF53
Filesystem revision #:    1 (dynamic)
Filesystem features:      has_journal ext_attr resize_inode dir_index filetype
needs_recovery extent flex_bg sparse_super large_file huge_file uninit_bg
dir_nlink extra_isize
Filesystem flags:         signed_directory_hash
Default mount options:    (none)
Filesystem state:         clean
Errors behavior:          Continue
Filesystem OS type:       Linux
Inode count:              62259200
Block count:              249036800
Reserved block count:     0
Free blocks:              23664521
Free inodes:              55316695
First block:              0
Block size:               4096
Fragment size:            4096
Reserved GDT blocks:      964
Blocks per group:         32768
Fragments per group:      32768
Inodes per group:         8192
Inode blocks per group:   512
Filesystem created:       Sun Nov 16 15:50:37 2008
Last mount time:          Mon Mar 22 07:22:00 2010
Last write time:          Mon Mar 22 07:22:00 2010
Mount count:              2
Maximum mount count:      32
Last checked:             Sun Mar 21 09:28:05 2010
Check interval:           15552000 (6 months)
Next check after:         Fri Sep 17 10:28:05 2010
Lifetime writes:          7204 GB
Reserved blocks uid:      0 (user root)
Reserved blocks gid:      0 (group root)
First inode:              11
Inode size:               256
Required extra isize:     28
Desired extra isize:      28
Journal inode:            8
Default directory hash:   half_md4
Directory Hash Seed:      7aad3588-6b5c-4bcb-b60f-0227bef217b7
Journal backup:           inode blocks
Journal size:             128M

Please tell me if I can be of further assistance,
Alex

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #8 from Eric Sandeen <sandeen@redhat.com>  2010-03-22 19:08:48 ---
Hm, so now that dumpe2fs output is already expanded ... ideally we'd have had
the pre-resize geometry to work with ... but I'll see what I can do.

-Eric

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #9 from Alessandro Polverini <alex@nibbles.it>  2010-03-22 20:16:20 ---
The problem is reproducible at will, this is not the first time I resized the
partition and everytime there was a kernel oops.

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #10 from Eric Sandeen <sandeen@redhat.com>  2010-03-22 20:26:22 ---
ah so you've been incrementally resizing it and it oopses each time?  hmmm

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #11 from Alessandro Polverini <alex@nibbles.it>  2010-03-22 20:41:25 ---
Yes, a kernel oops each time.
And each time I made a new test I upgraded the kernel, so I verified the
problem on all versions from 2.6.29 to 2.6.32, if I remember well.

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #12 from Eric Sandeen <sandeen@redhat.com>  2010-03-22 21:51:35 ---
I'm afraid I've not been able to reproduce it:

create a sparse loopback filesystem:

[root@inode test]# touch fsfile
[root@inode test]# truncate --size 1020054732800 fsfile 
[root@inode test]# ls -lh fsfile 
-rw-r--r--. 1 root root 950G Mar 22 16:28 fsfile
[root@inode test]# mkfs.ext4 fsfile 
mke2fs 1.41.10 (10-Feb-2009)

<at this point I checked that the filesystem size and features are similar to
your fs>

[root@inode test]# ls -lh fsfile 
-rw-r--r--. 1 root root 950G Mar 22 16:28 fsfile

Grow the container by about 50G:

[root@inode test]# truncate --size 1099511627776 fsfile 
[root@inode test]# ls -lh fsfile 
-rw-r--r--. 1 root root 1.0T Mar 22 16:32 fsfile

Resize while mounted:

[root@inode test]# mount -o loop fsfile  mnt/
[root@inode test]# resize2fs /dev/loop0 
resize2fs 1.41.10 (10-Feb-2009)
Filesystem at /dev/loop0 is mounted on /mnt/test/mnt; on-line resizing required
old desc_blocks = 60, new_desc_blocks = 64
Performing an on-line resize of /dev/loop0 to 268435456 (4k) blocks.
The filesystem on /dev/loop0 is now 268435456 blocks long.

No oops ... so I'm not sure what is going on here.

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #13 from Alessandro Polverini <alex@nibbles.it>  2010-03-23 20:14:38 ---
I don't know if this can help but my file system is very "complex", with 15+
millions of files/directories most of which are hard links (it's a backuppc
volume).

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #14 from Eric Sandeen <sandeen@redhat.com>  2010-03-23 20:50:32 ---
Well, I hate to say it, but it may be time for a compressed e2image -r to try
to replicate the problem.

-Eric

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #15 from Alessandro Polverini <alex@nibbles.it>  2010-03-23 21:19:30 ---
Hello Eric and thanks, I've sent you the link to download it.
If someone else is interested please ask me in private.

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549


Christoph Biedl <bugzilla.kernel.bpeb@manchmal.in-ulm.de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |bugzilla.kernel.bpeb@manchm
                   |                            |al.in-ulm.de




--- Comment #16 from Christoph Biedl <bugzilla.kernel.bpeb@manchmal.in-ulm.de>  2010-03-27 17:48:48 ---
The last days I ran into the same problems on several machines and was finally
able to reproduce the BUG in 2.6.32.10, 2.6.33, 2.6.34-rc1.  The trick is to
resize a file system that once was ext3.



Steps to reproduce:

# create a volume group vg_test. (lvm was mainly used for convenience)

# create a logical volume
lvcreate -n test -L 128m vg_test

DEV=/dev/vg_test/test
# create an ext3 filesystem
mke2fs -j $DEV

# convert to ext4
tune2fs -O extents,uninit_bg,dir_index,flex_bg,huge_file,dir_nlink,extra_isize
$DEV
e2fsck -yDf -C0 $DEV

# mount
mkdir /tmp/text
mount -o noatime $DEV /tmp/test

# resize LV
lvresize -L +4m $DEV

# online resize ext4
resize2fs -p $DEV



Observed behaviour:


kernel: BUG: unable to handle kernel NULL pointer dereference at 00000184
kernel: IP: [<c10c9f66>] ext4_group_add+0xf8f/0x104d
kernel: *pde = 00000000
kernel: Oops: 0002 [#1]
kernel: last sysfs file:
/sys/devices/pci0000:00/0000:00:01.1/host0/target0:0:1/0:0:1:0/block/sdb/sdb1/dev
kernel:
kernel: Pid: 1302, comm: resize2fs Not tainted 2.6.34-rc2 #7 /VirtualBox
kernel: EIP: 0060:[<c10c9f66>] EFLAGS: 00010202 CPU: 0
kernel: EIP is at ext4_group_add+0xf8f/0x104d
kernel: EAX: 00000180 EBX: cfba8200 ECX: 00007dfe EDX: 00000180
kernel: ESI: ce03def0 EDI: 00000000 EBP: 00100001 ESP: ce03de40
kernel:  DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0069
kernel: Process resize2fs (pid: 1302, ti=ce03c000 task=cf8e4280
task.ti=ce03c000)
kernel: Stack:
kernel:  00100202 00000000 ffffffff cfba8400 cf4ca3ec 00000000 ce03def0
cfba8200
kernel: <0> 00000000 00100202 00000000 00000002 00000000 cfba8200 cdc4b400
00100001
kernel: <0> 00000000 00000000 00000020 00000080 cfba8304 00000000 cfba8304
00000020
kernel: Call Trace:
kernel:  [<c10c0c96>] ? ext4_ioctl+0x57a/0x674
kernel:  [<c114e1c9>] ? do_output_char+0x84/0x191
kernel:  [<c10c071c>] ? ext4_ioctl+0x0/0x674
kernel:  [<c10753e6>] ? vfs_ioctl+0x12/0x42
kernel:  [<c10758dc>] ? do_vfs_ioctl+0x438/0x47c
kernel:  [<c106c83d>] ? vfs_write+0xf7/0x131
kernel:  [<c107594d>] ? sys_ioctl+0x2d/0x44
kernel:  [<c1237f75>] ? syscall_call+0x7/0xb
kernel: Code: 00 59 8b 40 38 f6 40 61 02 74 38 8b 5c 24 34 8b 74 24 18 8b 8b b0
01 00 00 8b 06 8b 93 b4 01 00 00 d3 e8 8b 4e 24 6b c0 0c 01 c2 <01> 4a 04 8b 4c
24 0c 03 83 b
kernel: EIP: [<c10c9f66>] ext4_group_add+0xf8f/0x104d SS:ESP 0069:ce03de40
kernel: CR2: 0000000000000184
kernel: ---[ end trace a1e9f008f870cb3b ]---

The code is (let's see whether bugzilla preserves the formatting):

        if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FLEX_BG)) {
                ext4_group_t flex_group;
                flex_group = ext4_flex_group(sbi, input->group);
                atomic_add(input->free_blocks_count,
                           &sbi->s_flex_groups[flex_group].free_blocks);
c10c9f46:       8b 5c 24 34             mov    0x34(%esp),%ebx
c10c9f4a:       8b 74 24 18             mov    0x18(%esp),%esi
c10c9f4e:       8b 8b b0 01 00 00       mov    0x1b0(%ebx),%ecx
c10c9f54:       8b 06                   mov    (%esi),%eax
c10c9f56:       8b 93 b4 01 00 00       mov    0x1b4(%ebx),%edx
c10c9f5c:       d3 e8                   shr    %cl,%eax
 *
 * Atomically adds @i to @v.
 */
static inline void atomic_add(int i, atomic_t *v)
{
        asm volatile(LOCK_PREFIX "addl %1,%0"
c10c9f5e:       8b 4e 24                mov    0x24(%esi),%ecx
c10c9f61:       6b c0 0c                imul   $0xc,%eax,%eax
c10c9f64:       01 c2                   add    %eax,%edx
c10c9f66:       01 4a 04                add    %ecx,0x4(%edx)
c10c9f69:       8b 4c 24 0c             mov    0xc(%esp),%ecx
                atomic_add(EXT4_INODES_PER_GROUP(sb),
                           &sbi->s_flex_groups[flex_group].free_inodes);
c10c9f6d:       03 83 b4 01 00 00       add    0x1b4(%ebx),%eax
c10c9f73:       8b 91 5c 01 00 00       mov    0x15c(%ecx),%edx
c10c9f79:       8b 52 0c                mov    0xc(%edx),%edx
c10c9f7c:       01 10                   add    %edx,(%eax)
        }


e2fsprogs is from Debian lenny (1.41.3-1).

Let me know if you're interested in the kernel .config.

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #17 from Eric Sandeen <sandeen@redhat.com>  2010-03-27 18:26:26 ---
Great, thanks for the reproducer, I'll give that a shot.

Alessandro, I have your image but still haven't made time to test with it,
hopefully soon (maybe today) - did you start your filesystem out as ext3?

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #18 from Eric Sandeen <sandeen@redhat.com>  2010-03-27 19:16:09 ---
Reproducer works perfectly, thanks.

So here's the issue; sbi->flex_groups[] doesn't get filled out in
ext4_fill_flex_info() because:

        if (groups_per_flex < 2) {
                sbi->s_log_groups_per_flex = 0;
                return 1;
        }

but resize is unconditionally doing this in ext4_group_add as long as the
FLEX_BG feature is set:

                atomic_add(input->free_blocks_count,
                           &sbi->s_flex_groups[flex_group].free_blocks);

so for a NULL s_flex groups it went boom.

Every other access to ->s_flex_groups checks s_log_groups_per_flex first, so
this should be the proper fix:

Index: linux-2.6/fs/ext4/resize.c
===================================================================
--- linux-2.6.orig/fs/ext4/resize.c
+++ linux-2.6/fs/ext4/resize.c
@@ -930,7 +930,8 @@ int ext4_group_add(struct super_block *s
     percpu_counter_add(&sbi->s_freeinodes_counter,
                EXT4_INODES_PER_GROUP(sb));

-    if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FLEX_BG)) {
+    if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FLEX_BG) &&
+        sbi->s_log_groups_per_flex) {
         ext4_group_t flex_group;
         flex_group = ext4_flex_group(sbi, input->group);
         atomic_add(input->free_blocks_count,

This fixes the reproducer, need to double check it on Alessandro's image.

-Eric

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13549


Eric Sandeen <sandeen@redhat.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|fs_ext4@kernel-bugs.osdl.or |sandeen@redhat.com
                   |g                           |




--- Comment #19 from Eric Sandeen <sandeen@redhat.com>  2010-03-27 19:28:11 ---
Alessandro, your image also has groups_per_flex == 1 so it is surely the same
thing.

-Eric

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

http://bugzilla.kernel.org/show_bug.cgi?id=13549


Eric Sandeen <sandeen@redhat.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sandeen@redhat.com




--- Comment #3 from Eric Sandeen <sandeen@redhat.com>  2010-03-15 19:34:54 ---
Alessandro, apologies for this bug sitting for so long.

Is this still something you can reproduce?

If so, an e2image -h of the original filesystem, and the size of the block
device you're resizing to, might help us reproduce the problem.

Thanks,
-Eric

-- 
Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

http://bugzilla.kernel.org/show_bug.cgi?id=13549





--- Comment #2 from Alessandro Polverini <alex@nibbles.it>  2009-09-18 19:03:11 ---
not it's not a drbd device but a lvm one

-- 
Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.

[Bug 13549] Kernel oops while online resizing of an ext4 filesystem

[bugzilla-daemon]

http://bugzilla.kernel.org/show_bug.cgi?id=13549


Roland Kletzing <devzero@web.de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |devzero@web.de




--- Comment #1 from Roland Kletzing <devzero@web.de>  2009-09-18 15:25:00 ---
this isn`t a drbd device (as that out of tree module seems to be loaded) - is
it ?

-- 
Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.