Hi all.
The first time since boot that I run e4defrag on a directory of
significant size, I get this warning in dmesg:
[ 199.076871]
[ 199.076876] =============================================
[ 199.076884] [ INFO: possible recursive locking detected ]
[ 199.076891] 2.6.31-rc7-red-debug #4
[ 199.076896] ---------------------------------------------
[ 199.076902] e4defrag/4399 is trying to acquire lock:
[ 199.076908] (&ei->i_data_sem){++++..}, at: [<ffffffffa0043839>]
ext4_move_extents+0x10e/0xa19 [ext4]
[ 199.076957]
[ 199.076958] but task is already holding lock:
[ 199.076964] (&ei->i_data_sem){++++..}, at: [<ffffffffa0043830>]
ext4_move_extents+0x105/0xa19 [ext4]
[ 199.077006]
[ 199.077007] other info that might help us debug this:
[ 199.077015] 3 locks held by e4defrag/4399:
[ 199.077019] #0: (&sb->s_type->i_mutex_key#7/1){+.+.+.}, at:
[<ffffffffa00437f3>] ext4_move_extents+0xc8/0xa19 [ext4]
[ 199.077069] #1: (&sb->s_type->i_mutex_key#7/2){+.+...}, at:
[<ffffffffa0043800>] ext4_move_extents+0xd5/0xa19 [ext4]
[ 199.077117] #2: (&ei->i_data_sem){++++..}, at: [<ffffffffa0043830>]
ext4_move_extents+0x105/0xa19 [ext4]
[ 199.077161]
[ 199.077162] stack backtrace:
[ 199.077170] Pid: 4399, comm: e4defrag Not tainted 2.6.31-rc7-red-debug #4
[ 199.077176] Call Trace:
[ 199.077188] [<ffffffff8105bcc4>] __lock_acquire+0x1467/0x14f5
[ 199.077200] [<ffffffff81059990>] ? mark_held_locks+0x52/0x70
[ 199.077212] [<ffffffff812a699a>] ? _spin_unlock_irq+0x2b/0x57
[ 199.077223] [<ffffffff8105bda7>] lock_acquire+0x55/0x71
[ 199.077261] [<ffffffffa0043839>] ? ext4_move_extents+0x10e/0xa19 [ext4]
[ 199.077275] [<ffffffff812a5b10>] down_read+0x40/0x4f
[ 199.077313] [<ffffffffa0043839>] ? ext4_move_extents+0x10e/0xa19 [ext4]
[ 199.077341] [<ffffffffa0043839>] ext4_move_extents+0x10e/0xa19 [ext4]
[ 199.077341] [<ffffffff8105af3b>] ? __lock_acquire+0x6de/0x14f5
[ 199.077341] [<ffffffffa002b44d>] ext4_ioctl+0x595/0x756 [ext4]
[ 199.077341] [<ffffffff810b1a91>] vfs_ioctl+0x1d/0x82
[ 199.077341] [<ffffffff812a6a00>] ? _spin_unlock_irqrestore+0x3a/0x67
[ 199.077341] [<ffffffff810b1ff0>] do_vfs_ioctl+0x483/0x4c9
[ 199.077341] [<ffffffff81059c30>] ? trace_hardirqs_on+0xd/0xf
[ 199.077341] [<ffffffff8113fa6f>] ? __up_write+0x115/0x124
[ 199.077341] [<ffffffff8100ba9c>] ? sysret_check+0x27/0x62
[ 199.077341] [<ffffffff810b2087>] sys_ioctl+0x51/0x74
[ 199.077341] [<ffffffff8113f97a>] ? __up_write+0x20/0x124
[ 199.077341] [<ffffffff8113f97a>] ? __up_write+0x20/0x124
[ 199.077341] [<ffffffff8100ba6b>] system_call_fastpath+0x16/0x1b
This might be happening when it hits files that don't use extents, but I
haven't tried e4defrag on a single non-extents file to verify that
theory. I've just so far only seen this happen when I defrag a directory
where I get the "operation not supported" message.
If I use chattr +e to migrate a file to use extents, I get this:
[ 470.400044] ------------[ cut here ]------------
[ 470.400065] WARNING: at fs/inode.c:1210 generic_delete_inode+0x65/0x16a()
[ 470.400072] Hardware name: N/A
[ 470.400078] Modules linked in: hidp hid ecryptfs cpufreq_ondemand
binfmt_misc af_packet ipt_MASQUERADE iptable_nat nf_nat
nf_conntrack_ipv4 nf_defrag_ipv4 xt_state nf_conntrack ipt_REJECT
xt_tcpudp iptable_filter ip_tables x_tables rfcomm bridge stp llc bnep
sco l2cap kvm_intel kvm ipv6 container sbs sbshc microcode acpi_cpufreq
freq_table uinput sbp2 snd_hda_codec_si3054 snd_hda_codec_realtek
snd_hda_intel arc4 snd_hda_codec ecb snd_hwdep iwl3945 snd_pcm iwlcore
snd_seq_midi snd_rawmidi snd_seq_midi_event mac80211 snd_seq led_class
btusb joydev snd_timer sr_mod bluetooth snd_seq_device cdrom cfg80211
snd ata_generic tg3 soundcore pata_acpi ohci1394 uhci_hcd ehci_hcd
psmouse libphy snd_page_alloc ata_piix processor rfkill ieee1394 button
usbcore battery evdev serio_raw wmi ac sg sd_mod thermal fan fuse
dm_mirror dm_region_hash dm_log dm_mod ahci libata scsi_mod ext4 mbcache
jbd2 crc16
[ 470.400353] Pid: 4451, comm: chattr Not tainted 2.6.31-rc7-red-debug #4
[ 470.400359] Call Trace:
[ 470.400372] [<ffffffff81037771>] warn_slowpath_common+0x77/0x8f
[ 470.400385] [<ffffffff81037798>] warn_slowpath_null+0xf/0x11
[ 470.400395] [<ffffffff810b7f28>] generic_delete_inode+0x65/0x16a
[ 470.400405] [<ffffffff810b8044>] generic_drop_inode+0x17/0x1bd
[ 470.400413] [<ffffffff810b7083>] iput+0x61/0x65
[ 470.400455] [<ffffffffa003b229>] ext4_ext_migrate+0x5eb/0x66a [ext4]
[ 470.400492] [<ffffffffa002b1f8>] ext4_ioctl+0x340/0x756 [ext4]
[ 470.400507] [<ffffffff810b1a91>] vfs_ioctl+0x1d/0x82
[ 470.400517] [<ffffffff810b1ff0>] do_vfs_ioctl+0x483/0x4c9
[ 470.400527] [<ffffffff81059c30>] ? trace_hardirqs_on+0xd/0xf
[ 470.400537] [<ffffffff810b2087>] sys_ioctl+0x51/0x74
[ 470.400549] [<ffffffff8100ba6b>] system_call_fastpath+0x16/0x1b
[ 470.400557] ---[ end trace ab85723542352dac ]---
Aside from these messages, the system continues normally and I haven't
seen any data loss. After the chattr +e, I am able to defrag with
e4defrag just fine. This is with the latest kernel code today and with
the current ext4 next tree merged in. Same result also in both ordered
and writeback modes. I hope this info helps.
Brian
On Mon, Aug 24, 2009 at 03:39:24PM -0700, Brian Rogers wrote:
> Hi all.
>
> The first time since boot that I run e4defrag on a directory of
> significant size, I get this warning in dmesg:
>
>
> This might be happening when it hits files that don't use extents, but I
> haven't tried e4defrag on a single non-extents file to verify that
> theory. I've just so far only seen this happen when I defrag a directory
> where I get the "operation not supported" message.
>
> If I use chattr +e to migrate a file to use extents, I get this:
>
> [ 470.400044] ------------[ cut here ]------------
> [ 470.400065] WARNING: at fs/inode.c:1210 generic_delete_inode+0x65/0x16a()
Can you try this patch
commit 832ecbfe5f3cf16916552759fbeff085992a66c4
Author: Aneesh Kumar K.V <[email protected]>
Date: Tue Aug 25 15:14:11 2009 +0530
ext4: unlock the new inode before iput
We need to unlock the new inode before iput. This patch fix the below warning.
[ 470.400044] ------------[ cut here ]------------
[ 470.400065] WARNING: at fs/inode.c:1210 generic_delete_inode+0x65/0x16a()
[ 470.400072] Hardware name: N/A
.....
...
[ 470.400353] Pid: 4451, comm: chattr Not tainted 2.6.31-rc7-red-debug #4
[ 470.400359] Call Trace:
[ 470.400372] [<ffffffff81037771>] warn_slowpath_common+0x77/0x8f
[ 470.400385] [<ffffffff81037798>] warn_slowpath_null+0xf/0x11
[ 470.400395] [<ffffffff810b7f28>] generic_delete_inode+0x65/0x16a
[ 470.400405] [<ffffffff810b8044>] generic_drop_inode+0x17/0x1bd
[ 470.400413] [<ffffffff810b7083>] iput+0x61/0x65
[ 470.400455] [<ffffffffa003b229>] ext4_ext_migrate+0x5eb/0x66a [ext4]
[ 470.400492] [<ffffffffa002b1f8>] ext4_ioctl+0x340/0x756 [ext4]
[ 470.400507] [<ffffffff810b1a91>] vfs_ioctl+0x1d/0x82
[ 470.400517] [<ffffffff810b1ff0>] do_vfs_ioctl+0x483/0x4c9
[ 470.400527] [<ffffffff81059c30>] ? trace_hardirqs_on+0xd/0xf
[ 470.400537] [<ffffffff810b2087>] sys_ioctl+0x51/0x74
[ 470.400549] [<ffffffff8100ba6b>] system_call_fastpath+0x16/0x1b
[ 470.400557] ---[ end trace ab85723542352dac ]---
Signed-off-by: Aneesh Kumar K.V <[email protected]>
diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c
index 313a50b..05361ad 100644
--- a/fs/ext4/migrate.c
+++ b/fs/ext4/migrate.c
@@ -618,7 +618,7 @@ err_out:
tmp_inode->i_nlink = 0;
ext4_journal_stop(handle);
Aneesh Kumar K.V wrote:
> On Mon, Aug 24, 2009 at 03:39:24PM -0700, Brian Rogers wrote:
>
>> Hi all.
>>
>> The first time since boot that I run e4defrag on a directory of
>> significant size, I get this warning in dmesg:
>>
>>
>> This might be happening when it hits files that don't use extents, but I
>> haven't tried e4defrag on a single non-extents file to verify that
>> theory. I've just so far only seen this happen when I defrag a directory
>> where I get the "operation not supported" message.
>>
>> If I use chattr +e to migrate a file to use extents, I get this:
>>
>> [ 470.400044] ------------[ cut here ]------------
>> [ 470.400065] WARNING: at fs/inode.c:1210 generic_delete_inode+0x65/0x16a()
>>
>
>
> Can you try this patch
>
> commit 832ecbfe5f3cf16916552759fbeff085992a66c4
> Author: Aneesh Kumar K.V <[email protected]>
> Date: Tue Aug 25 15:14:11 2009 +0530
>
> ext4: unlock the new inode before iput
>
> We need to unlock the new inode before iput. This patch fix the below warning.
>
> [ 470.400044] ------------[ cut here ]------------
> [ 470.400065] WARNING: at fs/inode.c:1210 generic_delete_inode+0x65/0x16a()
> [ 470.400072] Hardware name: N/A
> .....
> ...
> [ 470.400353] Pid: 4451, comm: chattr Not tainted 2.6.31-rc7-red-debug #4
> [ 470.400359] Call Trace:
> [ 470.400372] [<ffffffff81037771>] warn_slowpath_common+0x77/0x8f
> [ 470.400385] [<ffffffff81037798>] warn_slowpath_null+0xf/0x11
> [ 470.400395] [<ffffffff810b7f28>] generic_delete_inode+0x65/0x16a
> [ 470.400405] [<ffffffff810b8044>] generic_drop_inode+0x17/0x1bd
> [ 470.400413] [<ffffffff810b7083>] iput+0x61/0x65
> [ 470.400455] [<ffffffffa003b229>] ext4_ext_migrate+0x5eb/0x66a [ext4]
> [ 470.400492] [<ffffffffa002b1f8>] ext4_ioctl+0x340/0x756 [ext4]
> [ 470.400507] [<ffffffff810b1a91>] vfs_ioctl+0x1d/0x82
> [ 470.400517] [<ffffffff810b1ff0>] do_vfs_ioctl+0x483/0x4c9
> [ 470.400527] [<ffffffff81059c30>] ? trace_hardirqs_on+0xd/0xf
> [ 470.400537] [<ffffffff810b2087>] sys_ioctl+0x51/0x74
> [ 470.400549] [<ffffffff8100ba6b>] system_call_fastpath+0x16/0x1b
> [ 470.400557] ---[ end trace ab85723542352dac ]---
>
> Signed-off-by: Aneesh Kumar K.V <[email protected]>
>
> diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c
> index 313a50b..05361ad 100644
> --- a/fs/ext4/migrate.c
> +++ b/fs/ext4/migrate.c
> @@ -618,7 +618,7 @@ err_out:
> tmp_inode->i_nlink = 0;
>
> ext4_journal_stop(handle);
> -
> + unlock_new_inode(tmp_inode);
> iput(tmp_inode);
>
> return retval;
>
That fixed the chattr-triggered warning. Thanks!
Brian