Project

General

Profile

Actions

Bug #1446

closed

USB flash drives not working on one of my machines

Added by rumcic over 14 years ago. Updated over 9 years ago.

Status:
Closed
Priority:
Normal
Assignee:
-
Category:
Driver
Target version:
Start date:
Due date:
% Done:

0%

Estimated time:

Description

On master whenever I plug in an USB flash drive I get:
Aug 9 14:29:50 rumko kernel: usbd_transfer_cb: short transfer 0<8
Aug 9 14:29:50 rumko kernel: usbd_new_device: addr=2, getting first desc
failed
Aug 9 14:29:50 rumko kernel: uhub_explore: usb_new_device failed,
error=SHORT_XFER
Aug 9 14:29:50 rumko kernel: uhub1: device problem (SHORT_XFER), disabling
port 1

Found a freebsd PR (http://www.freebsd.org/cgi/query-pr.cgi?pr=103167&cat=)
that showed the same problem and a solution
(http://lists.freebsd.org/pipermail/cvs-all/2006-October/190784.html).
So I added the following patches:
http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/usb/Attic/ehci.c.diff?hideattic=0&r2=1.50&r1=1.49&f=u
http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/usb/Attic/ohci.c.diff?hideattic=0&r2=1.165&r1=1.164&f=u
http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/usb/Attic/usb_subr.c.diff?hideattic=0&r2=1.87&r1=1.86&f=u

But now I get (3 tries on 2 different ports):
Aug 9 15:47:13 rumko kernel: usbd_get_string: getting lang failed, using 0
Aug 9 15:47:39 rumko kernel: uhub_explore: usb_new_device failed,
error=TIMEOUT
Aug 9 15:47:39 rumko kernel: uhub1: device problem (TIMEOUT), disabling port 1
Aug 9 15:48:03 rumko kernel: usbd_get_string: getting lang failed, using 0
Aug 9 15:48:19 rumko kernel: uhub_explore: usb_new_device failed,
error=TIMEOUT
Aug 9 15:48:19 rumko kernel: uhub1: device problem (TIMEOUT), disabling port 8
Aug 9 15:52:17 rumko kernel: usbd_get_string: getting lang failed, using 0
Aug 9 15:52:33 rumko kernel: uhub_explore: usb_new_device failed,
error=STALLED
Aug 9 15:52:33 rumko kernel: uhub1: device problem (STALLED), disabling port 1

So it still doesn't work and am at a loss as to what to do next. I have 2 other
machines (another dfly and one freebsd), but the flash drives work there
without problems so it's fortunately not a big problem for me.
This machine is nforce4 based (recent dmesg can be seen at
http://www.rumko.net/dragonfly/dmesg.boot).
--
Regards,
Rumko

Actions #1

Updated by dillon over 14 years ago

Try this patch, Rumko, and tell me if it works.

fetch http://apollo.backplane.com/DFlyMisc/usb01.patch
The basic problem is that our ohci and ehci drivers have some old code
which plays tricks when dealing with the device address that makes
the FreeBSD patch simply not work.
-Matt
Actions #2

Updated by dillon over 14 years ago

:Just a quick question ... in the patch I see 2 blocks of "#if 1" ... "#endi=
:f"=20
:code ... on purpose or should it be #if 0?
:
:I'll test it in the evening ... in your timezone, morning I think
:
:=2D-=20
:Regards,
:Rumko

#if 1 ... the patch is correct.  It's a bit meaningless, I know.  It's
just a quick hack to see if it works or not.

-Matt

Actions #3

Updated by dillon over 14 years ago

I've committed a somewhat different version of the patch to master.

I also fixed some issues with EHCI probing during early boot that
were not in the patch.
-Matt
Actions #4

Updated by rumcic over 14 years ago

Matthew Dillon wrote:

I've committed a somewhat different version of the patch to master.

I also fixed some issues with EHCI probing during early boot that
were not in the patch.

-Matt

Unfortunately, still doesn't work (just a few mins ago updated to HEAD).

I turned up some debugging (set hw.usb.ehci.debug, hw.usb.debug and
hw.usb.uhub.debug to 4) and the output is at http://pastebin.com/m7055f929

Actions #5

Updated by tuxillo over 9 years ago

  • Description updated (diff)
  • Category set to Driver
  • Status changed from New to Feedback
  • Assignee deleted (0)
  • Target version set to 4.2

Hi,

Would it be possible to try reproducing this issue with latest master?

Cheers,
Antonio Huete

Actions #6

Updated by rumcic over 9 years ago

tuxillo wrote:

Hi,

Would it be possible to try reproducing this issue with latest master?

Cheers,
Antonio Huete

Not really, the only machine running dfly is 32bit, not supported on master.

Actions #7

Updated by tuxillo over 9 years ago

  • Status changed from Feedback to Closed

Not possible to reproduce it, 32-bit only machine.

Actions

Also available in: Atom PDF