There are two basic possibilities: 1) A device problem 2) Some sort of very subtle HCD bug.
I can't help you with either... folks on linux-usb-devel might be able to
help if it's an HCD bug. But that's kinda unlikely.
Matt
On Tue, Jun 08, 2004 at 02:16:48PM -0400, [EMAIL PROTECTED] wrote:
> Any theories what is causing the pauses in the first place? Perhaps
> something I can change on my side?
>
> On Tue, 8 Jun 2004, Matthew Dharm wrote:
>
> > 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
> >
--
Matthew Dharm Home: [EMAIL PROTECTED]
Maintainer, Linux USB Mass Storage Driver
Hey Chief. We've figured out how to save the technical department. We
need to be committed.
-- The Techs
User Friendly, 1/22/1998
pgpFY3eVzLlKw.pgp
Description: PGP signature
