Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 19 Aug 2004 12:12:05 +0200 (CEST)
From:      Barry Bouwsma <freebsd-misuser@remove-NOSPAM-to-reply.NOSPAM.dyndns.dk>
To:        freebsd-current@freebsd.org
Subject:   Re: Unloading USB driver while device is attached. 
Message-ID:  <200408191012.i7JAC5t01192@Mail.NOSPAM.DynDNS.dK>
References:  <200407200045.aa99979@salmon.maths.tcd.ie> <200407271843.aa10377@salmon.maths.tcd.ie> <200408102244.i7AMid614188@Mail.NOSPAM.DynDNS.dK> <200408142159.i7ELxdH00871@Mail.NOSPAM.DynDNS.dK>

next in thread | previous in thread | raw e-mail | index | archive | help
Once again, I carry on a conversation with myself, concerning
the observation that the USB2 umass code from -current ported
to my -stable machine seemed to work fine with one set of
hardware (although problematic on different hardware):

> Now, the same from the machine where it works as expected:

Sigh, too good to be true.

While pounding on the disk (cvsup doing a tagging operation),
the USB2-attached disk went south, with the `need toggle update'
message, from which it never recovered.

So I guess things aren't all rosy, although I was able to do
far more before the failure with this one controller card than
with my other one (where I see that message within seconds or
minutes).

I guess I need to verify that NetBSD on the same hardware
can handle more intense disk activity than I had tried, to
assure myself that the code there is fine...


Failure was like this:

Aug 15 13:45:21 dastardly /kernel-BEAST: ehci_idone: need toggle update status=4df30d00 nstatus=2a008d80
Aug 15 13:45:21 dastardly /kernel-BEAST: umass0: Invalid CSW: sig 0x6574726f should be 0x53425355

which resulted in everything heading south:

Aug 15 13:45:24 dastardly /kernel-BEAST: spec_getpages:(#da/0x20004) I/O read failure: (error=5) bp 0xc2139d20 vp 0xc547d0c0
Aug 15 13:45:24 dastardly /kernel-BEAST: size: 31232, resid: 31232, a_count: 30744, valid: 0x0
Aug 15 13:45:24 dastardly /kernel-BEAST: nread: 0, reqpage: 0, pindex: 0, pcount: 8
Aug 15 13:45:24 dastardly /kernel-BEAST: vm_fault: pager read error, pid 1621 (cvsup-FreeBSD)


For reference, the messages I see during the boot, concerning the
toggle update, are as follows, should there be any clues in there.
Unfortunately, I think these include not only the disk which had
problems, but also some multi-card-readers:

Aug 14 23:56:31 dastardly /kernel-BEAST: ehci_idone: need toggle update status=80018d40 nstatus=80008c80
Aug 14 23:56:31 dastardly /kernel-BEAST: ehci_idone: need toggle update status=80028d40 nstatus=80008c80
Aug 14 23:56:31 dastardly /kernel-BEAST: ehci_idone: need toggle update status=00080248 nstatus=80008d80
Aug 14 23:56:34 dastardly /kernel-BEAST: ehci_idone: need toggle update status=00080248 nstatus=80008d80
Aug 14 23:56:34 dastardly /kernel-BEAST: ehci_idone: need toggle update status=00080248 nstatus=80008d80
Aug 14 23:56:34 dastardly /kernel-BEAST: ehci_idone: need toggle update status=00080248 nstatus=80008d80
Aug 14 23:56:35 dastardly /kernel-BEAST: ehci_idone: need toggle update status=00080248 nstatus=80008d80
Aug 14 23:56:36 dastardly /kernel-BEAST: ehci_idone: need toggle update status=00080248 nstatus=80008d80



thanks
barry bouwsma



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200408191012.i7JAC5t01192>