On Tue, Jun 08, 2004 at 02:03:34PM -0400, [EMAIL PROTECTED] wrote: > Matt, > > Thanks for the info. I've recompiled the usb-storage module with verbose > debugging, and managed to trip the bug again. Unfortunately it look a > while to trigger, and now /var/log/messages is 74 megs. > > I'm attaching 2000 lines of usb-storage output from the end of the log. > If you end up needing more, let me know.
And here's where things to bad:
> Jun 8 13:56:59 cthulhu usb-storage: queuecommand called
> Jun 8 13:56:59 cthulhu usb-storage: *** thread awakened.
> Jun 8 13:56:59 cthulhu usb-storage: Command READ_10 (10 bytes)
> Jun 8 13:56:59 cthulhu usb-storage: 28 00 03 0a 56 8f 00 00 f0 00
> Jun 8 13:56:59 cthulhu usb-storage: Bulk Command S 0x43425355 T 0x12352 L 122880 F
> 128 Trg 0 LUN 0 CL 10
> Jun 8 13:56:59 cthulhu usb-storage: usb_stor_bulk_transfer_buf: xfer 31 bytes
> Jun 8 13:56:59 cthulhu usb-storage: Status code 0; transferred 31/31
> Jun 8 13:56:59 cthulhu usb-storage: -- transfer complete
> Jun 8 13:56:59 cthulhu usb-storage: Bulk command transfer result=0
> Jun 8 13:56:59 cthulhu usb-storage: usb_stor_bulk_transfer_sglist: xfer 122880
> bytes, 30 entries
> Jun 8 13:57:29 cthulhu usb-storage: command_abort called
> Jun 8 13:57:29 cthulhu usb-storage: usb_stor_stop_transport called
> Jun 8 13:57:29 cthulhu usb-storage: -- cancelling sg request
> Jun 8 13:57:29 cthulhu usb-storage: Status code -104; transferred 25344/122880
> Jun 8 13:57:29 cthulhu usb-storage: -- transfer cancelled
> Jun 8 13:57:29 cthulhu usb-storage: Bulk data transfer result 0x4
> Jun 8 13:57:29 cthulhu usb-storage: -- command was aborted
> Jun 8 13:57:29 cthulhu usb-storage: usb_stor_Bulk_reset called
> Jun 8 13:57:29 cthulhu usb-storage: usb_stor_control_msg: rq=ff rqtype=21
> value=0000 index=00 len=0
> Jun 8 13:57:35 cthulhu usb-storage: Soft reset: clearing bulk-in endpoint halt
> Jun 8 13:57:35 cthulhu usb-storage: usb_stor_control_msg: rq=01 rqtype=02
> value=0000 index=82 len=0
> Jun 8 13:57:35 cthulhu usb-storage: usb_stor_clear_halt: result = 0
> Jun 8 13:57:35 cthulhu usb-storage: Soft reset: clearing bulk-out endpoint halt
> Jun 8 13:57:35 cthulhu usb-storage: usb_stor_control_msg: rq=01 rqtype=02
> value=0000 index=01 len=0
> Jun 8 13:57:35 cthulhu usb-storage: usb_stor_clear_halt: result = 0
> Jun 8 13:57:35 cthulhu usb-storage: Soft reset done
> Jun 8 13:57:35 cthulhu usb-storage: scsi command aborted
> Jun 8 13:57:35 cthulhu usb-storage: *** thread sleeping.
Specifically, what's happening here is that a read command was sent to the
device, and the device gave us 25344 bytes (of 122880 requested). Then it
paused. Pausing is legal -- but it paused for about 90 seconds, so the
SCSI layer decided to abort the command. Abort processing kicked in,
The rest of the log shows that we reset the device, re-established
communication, and re-tried the command that caused the problem. That, in
turn, caused another failure. Lather, rince, repeat.
The filesystem code definately needs to handle this sort of thing.
Matt
--
Matthew Dharm Home: [EMAIL PROTECTED]
Maintainer, Linux USB Mass Storage Driver
I say, what are all those naked people doing?
-- Big client to Stef
User Friendly, 12/14/1997
pgpumcX8joM1U.pgp
Description: PGP signature
