kernel panic, 2.6.39, hot pullout of usb wireless data modem
Posted: Jul 10, 11, 22:48 falcon74
 Status: Interested
Joined: 08 Jul 2011
Posts: 12
Reply Quote
Hi,

Currently running 2.6.39-2.dmz.1-liquorix-amd64, and I was using a Huawei EC1262 CDMA EV-DO data-modem (USB), with a connection in established mode. In this state, when I pulled out the modem, I got a kernel panic and here's what I captured (manually, so slight errors may be there), based on what was displayed on screen.

Since this data-modem has a usb-storage function (set using modeswitch), I guess during the pullout operation, something related to storage-mgt went boo-boo. But then, it's a wild guess (almost -- I see lot of those scsi_mod in the stack trace).

:: Code ::
FS: 00007f679a7bd740(0000) GS: ffff88003f480000(0000) knlGS: 0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000000000000640 CR3: 00000000284d7000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR4: 00000000ffff0ff0 DR7: 0000000000000400
Process umount (pid: 7832, threadinfo ffff88001eb1a000, task ffff88003ed3b180)
Stack:
 ffff88002f698380 ffffffffa009f223 ffff88002f640760 ffff88002fa669a8
 ffff88002f640760 ffff88002fa669a8 ffff88001eb1nfd8 ffffffffa009f3a5
 ffff880000000008 ffff88002fa669a8 ffff88002f640760 ffffffff911e443e
Call Trace:
[<ffffffffa009f223>] ? scsi_setup_blk_pc_cmnd+0x23/0x160 [scsi_mod]
...                    scsi_pre_fn+0x45/0x50 [scsi_mod]
...                    blk_peek_request+0xce/0x250
...                    scsi_request_fn+0x3c6/0x490 [scsi_mod]
...                    blk_execute_rq_nowait+0x61/0xb0
...                    blk_execute_rq+0x73/0xe0
...                    do_lookup+0x1e1/0x2a0
...                    ext4_lookup+0xb8/0x120 [ext4]
...                    scsci_execute+0xfd/0x170 [scsi_mod]
...                    scsci_execute_req+0xf5/0x170 [scsi_mod]
...                    ioctl_internal_command.clone.4+x0x64/0x1c0 [scsi_mod]
...                    try_preempt+0xcb/0x120
...                    sched_clock_cpu+0x65/0xf0
...                    scsi_set_medium_removal+0x82/0xb0 [scsi_mod]
...                    cdrom_release+0x155/0x270 [cdrom]
...                    sr_block_release+0x2e/0x60 [sr_mod]
...                    _blkdev_put+0x17c/0x1c0
[<ffffffff811245b3>] ? deactivate_locked_super+0x43/0x70
[<ffffffff81140a87>] ? sys_umount+0x77/0x3d0
[<ffffffff813c8252>] ? system_call_fastpath+0x16/0x1b
Code: 14 24 48 8b 7a 08 e8 bc 13 14 e1 48 8b 7b 58 e8 03 90 32 e1 48 8b 14 24 48 89 44 24 08 48 89 d7 e8 02 88 ff ff eb a0 48 83 ec 08 <8b> 87 40 06 00 00 83 f8 02 75 07 31 c0 48 83 c4 08 c3 83 e8 04
RIP [<ffffffffa009e794>] scsi_prep_state_check+0x4/0x90 [scsi_mod]
 RSP <ffff88001eb1bab8>
CR2: 0000000000000640


Note that the "..." for the addresses are where I got really tired of manually copying them.

PS> I am not very familiar with kernel panics or debugging them, so if there is a better, more reliable way to capture this information, please let me know.

cheers,
falcon74
Back to top
Posted: Jul 11, 11, 21:07 damentz
 Status: Assistant
Joined: 09 Sep 2008
Posts: 1135
Reply Quote
Try 2.6.39.3* - there are some USB eject fixes.
Back to top
Posted: Jul 12, 11, 4:34 falcon74
 Status: Interested
Joined: 08 Jul 2011
Posts: 12
Reply Quote
Okay.
Will it be rolled into a precompiled .deb with header ?
Not very savvy with kernel compilation.
Back to top
Posted: Jul 12, 11, 18:53 damentz
 Status: Assistant
Joined: 09 Sep 2008
Posts: 1135
Reply Quote
Just upgrade your packages with apt-get, you should know how to do that.
Back to top
Display posts from previous:   

All times are GMT - 8 Hours
Contact Us
Hosting: Pair Networks: 0.033
Forum Style © techpatterns.com
info

tech forums