Skip to contentSkip to conversation

Badness in as_completed_request at drivers/block/as-iosched.c:919

DISCUSSION Oleg Drokin 3 participants 5 messages Oct 22, 2003 – Oct 23, 2003

lore
Share & tools
lore

Oleg Drokin

author
#0 lore raw

Hello!

2.6.0-test8 (current bk), greeted me with "Badness in as_completed_request" message on boot at smartd startup:

Oct 22 15:19:21 dol-guldur smartd[1225]: Device: /dev/sdb, is SMART capable. Adding to "monitor" list. 
Oct 22 15:19:21 dol-guldur kernel: Badness in as_completed_request at drivers/block/as-iosched.c:919
Oct 22 15:19:21 dol-guldur kernel: Call Trace:
Oct 22 15:19:21 dol-guldur kernel:  [<c01f7660>] as_completed_request+0x1d0/0x210
Oct 22 15:19:21 dol-guldur kernel:  [<c01efe30>] elv_completed_request+0x20/0x30
Oct 22 15:19:21 dol-guldur kernel:  [<c01f2246>] __blk_put_request+0x36/0xb0
Oct 22 15:19:21 dol-guldur smartd[1225]: Started monitoring 1 ATA and 2 SCSI devices 
Oct 22 15:19:22 dol-guldur kernel:  [<c01f3186>] end_that_request_last+0x56/0xe0
Oct 22 15:19:22 dol-guldur kernel:  [<c0215a74>] scsi_end_request+0xb4/0xe0
Oct 22 15:19:22 dol-guldur kernel:  [<c0215e28>] scsi_io_completion+0x1c8/0x4f0
Oct 22 15:19:22 dol-guldur kernel:  [<c0236149>] sd_rw_intr+0x89/0x270
Oct 22 15:19:22 dol-guldur kernel:  [<c0211024>] scsi_finish_command+0x84/0xf0
Oct 22 15:19:22 dol-guldur kernel:  [<c0210e0f>] scsi_softirq+0xdf/0x230
Oct 22 15:19:22 dol-guldur kernel:  [<c0124feb>] do_softirq+0xcb/0xd0
Oct 22 15:19:22 dol-guldur kernel:  [<c010bc05>] do_IRQ+0x105/0x130
Oct 22 15:19:22 dol-guldur kernel:  [<c0105000>] rest_init+0x0/0x60
Oct 22 15:19:22 dol-guldur kernel:  [<c0109f20>] common_interrupt+0x18/0x20
Oct 22 15:19:22 dol-guldur kernel:  [<c0107210>] default_idle+0x0/0x40
Oct 22 15:19:22 dol-guldur kernel:  [<c0105000>] rest_init+0x0/0x60
Oct 22 15:19:22 dol-guldur kernel:  [<c010723f>] default_idle+0x2f/0x40
Oct 22 15:19:22 dol-guldur kernel:  [<c01072cb>] cpu_idle+0x3b/0x50
Oct 22 15:19:22 dol-guldur kernel:  [<c0330947>] start_kernel+0x167/0x180
Oct 22 15:19:22 dol-guldur kernel:  [<c0330510>] unknown_bootoption+0x0/0x110
Oct 22 15:19:22 dol-guldur kernel: 
Oct 22 15:19:22 dol-guldur kernel: Badness in as_completed_request at drivers/block/as-iosched.c:919
Oct 22 15:19:22 dol-guldur kernel: Call Trace:
Oct 22 15:19:22 dol-guldur kernel:  [<c01f7660>] as_completed_request+0x1d0/0x210
Oct 22 15:19:22 dol-guldur kernel:  [<c01efe30>] elv_completed_request+0x20/0x30
Oct 22 15:19:22 dol-guldur kernel:  [<c01f2246>] __blk_put_request+0x36/0xb0
Oct 22 15:19:22 dol-guldur smartd[1225]: Device: /dev/sdb, SMART Failure: DATA CHANNEL IMPENDING FAILURE DATA ERROR RATE TOO HIGH 
Oct 22 15:19:22 dol-guldur kernel:  [<c01f3186>] end_that_request_last+0x56/0xe0
Oct 22 15:19:22 dol-guldur kernel:  [<c0215a74>] scsi_end_request+0xb4/0xe0
Oct 22 15:19:22 dol-guldur kernel:  [<c0215e28>] scsi_io_completion+0x1c8/0x4f0
Oct 22 15:19:22 dol-guldur kernel:  [<c0236149>] sd_rw_intr+0x89/0x270
Oct 22 15:19:22 dol-guldur kernel:  [<c0211024>] scsi_finish_command+0x84/0xf0
Oct 22 15:19:22 dol-guldur kernel:  [<c0210e0f>] scsi_softirq+0xdf/0x230
Oct 22 15:19:22 dol-guldur kernel:  [<c0124feb>] do_softirq+0xcb/0xd0
Oct 22 15:19:22 dol-guldur kernel:  [<c010bc05>] do_IRQ+0x105/0x130
Oct 22 15:19:22 dol-guldur kernel:  [<c0105000>] rest_init+0x0/0x60
Oct 22 15:19:22 dol-guldur kernel:  [<c0109f20>] common_interrupt+0x18/0x20
Oct 22 15:19:22 dol-guldur kernel:  [<c0107210>] default_idle+0x0/0x40
Oct 22 15:19:22 dol-guldur kernel:  [<c0105000>] rest_init+0x0/0x60
Oct 22 15:19:22 dol-guldur kernel:  [<c010723f>] default_idle+0x2f/0x40
Oct 22 15:19:22 dol-guldur kernel:  [<c01072cb>] cpu_idle+0x3b/0x50
Oct 22 15:19:22 dol-guldur kernel:  [<c0330947>] start_kernel+0x167/0x180
Oct 22 15:19:22 dol-guldur kernel:  [<c0330510>] unknown_bootoption+0x0/0x110
Oct 22 15:19:22 dol-guldur kernel: 
Oct 22 15:19:23 dol-guldur smartd: smartd startup succeeded

/dev/sdb is on aic7xxx-compatible scsi controller.

2.6.0-test7 have not shown anything like this.
Hope that helps

Bye,
Oleg

#1 lore raw
Quoted 3 lines · Oleg Drokin

Hello!

Hi Oleg,
The warning should be harmless. I'll remove it once I make sure. I don't think there have been any recent as-iosched changes, so something else must have just triggered it off.
Thanks.

#2 lore raw
Quoted 5 lines · Nick Piggin

Oleg Drokin wrote:

Quoted 8 lines
Quoted 3 lines

Hello!

Hi Oleg,
The warning should be harmless. I'll remove it once I make sure. I don't think there have been any recent as-iosched changes, so something else must have just triggered it off.

The smartd failure doesn't look harmless:

Device: /dev/sdb, SMART Failure: DATA CHANNEL IMPENDING FAILURE DATA ERROR RATE TOO HIGH

Or does this always happen?

I assume we're dealing with a non-fs request here.

#3 lore raw
Quoted 17 lines · Andrew Morton
Quoted 5 lines · Nick Piggin

Oleg Drokin wrote:

Quoted 9 lines
Quoted 4 lines

Hello!

Hi Oleg,
The warning should be harmless. I'll remove it once I make sure. I don't think there have been any recent as-iosched changes, so something else must have just triggered it off.

Quoted 6 lines

The smartd failure doesn't look harmless:

Device: /dev/sdb, SMART Failure: DATA CHANNEL IMPENDING FAILURE DATA ERROR RATE TOO HIGH

Or does this always happen?

I didn't see that. Try booting with elevator=deadline I guess.

author
#4 lore raw
In reply to Andrew Morton

Hello!

Quoted 6 lines · Andrew Morton
Quoted 3 lines

The warning should be harmless. I'll remove it once I make sure. I don't think there have been any recent as-iosched changes, so something else must have just triggered it off.

The smartd failure doesn't look harmless:
Device: /dev/sdb, SMART Failure: DATA CHANNEL IMPENDING FAILURE DATA ERROR RATE TOO HIGH Or does this always happen?

Yes, drive thinks it is. I am aware of this. But kernel looks surprised. I have another scsi drive in that system whose smart status is OK, and kernel does not produce this sort of output when smartd is adding /dev/sda to monitor list.

Quoted 1 line

I assume we're dealing with a non-fs request here.

I think you are right, though I think filesystems from that drive might be already mounted by the time smartd is started.

Bye,
Oleg