Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 3 Jul 2005 09:56:32 +0200
From:      Stefan Walter <sw@gegenunendlich.de>
To:        Hans Petter Selasky <hselasky@c2i.net>
Cc:        freebsd-usb@freebsd.org
Subject:   Re: usb/82660: EHCI: I/O stuck in state 'physrd'/panic
Message-ID:  <20050703075632.GA937@kyuzo.dunkelkammer.void>
In-Reply-To: <200507021553.49043.hselasky@c2i.net>
References:  <20050626091628.775DD3A1D@kyuzo.dunkelkammer.void> <200507011210.25250.hselasky@c2i.net> <20050701114207.GA826@kyuzo.dunkelkammer.void> <200507021553.49043.hselasky@c2i.net>

next in thread | previous in thread | raw e-mail | index | archive | help

--k+w/mQv8wyuph6w0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Hans Petter Selasky, 02.07.05, 15:53h CEST:

> > http://www.gegenunendlich.de/stuff/messages-umass-until-pullout.bz2
> > http://www.gegenunendlich.de/stuff/messages-umass-after-pullout.diff
>=20
> The end of the log is just before you pulled out your device, when the er=
ror=20
> happened, right?

Yes, but...

> I'm not sure what is wrong. From what I can see the problem is in an uppe=
r=20
> layer, hence UMASS's USB transfers should timeout at least, if they get=
=20
> stuck.

=2E..maybe I pulled the stick out too quickly until now. I just noticed
that, when I'm patient enough, the transfer is eventually resumed (after
about two minutes). Some data is transferred (size varying), then it
stalls again. I didn't notice that before. I've uploaded a message log of
a complete transfer of a 17 MB file (that is, the file was successfully
copied to the hard disk with mtools) now (with hw.usb.umass.debug=3D-1):

http://www.gegenunendlich.de/stuff/ehci/umass-messages-complete-transfer.bz2

The log also contains messages of sudo(8) being called. I ran sudo every
time I noticed that some data had been copied and the transfer stalled
again to mark those points of time in the log. You'll notice umass timeout
messages around those ones.

I hope this one is a bit more helpful than the others.

> While having [umass] debugging enabled, could you have tried playing=20
> around with "camcontrol" and see if it changes anything:
>=20
> man camcontrol
> camcontrol devlist
> camcontrol reset scbus0:0:0=20
>=20
> # replace "0:0" with "target" and "lun" for the device you are reading fr=
om

I did reset the device with another stalled transfer a few times. It
didn't resume the transfer, but here's a log:

http://www.gegenunendlich.de/stuff/ehci/umass-messages-bus-reset.bz2

> Could you have tried another brand of USB 2.0 flash disks and see if the=
=20
> problem is the same?

Currently not, no. I thought the other device I have was USB 2.0, too, but
I was wrong.

Stefan

--k+w/mQv8wyuph6w0
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (FreeBSD)
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=IvOx
-----END PGP SIGNATURE-----

--k+w/mQv8wyuph6w0--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050703075632.GA937>