Am Freitag, 5. August 2005 14:42 schrieb [EMAIL PROTECTED]:
> Pete Zaitcev wrote:
> > On Fri, 05 Aug 2005 11:00:49 +0200, [EMAIL PROTECTED] wrote:
> >
> >
> >>>Please file a bug with Red Hat.
> >
> >
> >>May be I'm wrong, but I've opened a service request (600548) and I
> >>expected to be part of the internal "escalation" to open a bug.
> >
> >
> > That's even better. It'll get to me when I get from vacation,
> > then I'll ask what you actually did to produce the oops
> > (you forgot to mention it in the original report).
>
> You are for sure referring to my post on linux-usb-devel (because in the
> service request there is a ton of thing...)
>
> Anyway, to produce the panics I follow this procedure:
>
> 1) Install a USB modem (Multitech and Zyxel, at the moment)
> 2) Installa hylafax (an open source fax server)
> 3) Configure the modem to send/receive faxes
> 4) Send a fax to a busy line
> 5) Wait
>
> Hylafax, after first failure, tries to send the fax in about 4 minutes.
> At the third (or at maximum the fourth tentative), I get a kernel panic.
>
> For what I can "see", when the timeout of re-send the fax comes, the
> process that send faxes requests to lock the device and from time to
> time the request raises the panic.
>
> Just to say, if the line is free and the call do happen, it works
> normally (and at the moment I never experienced kernel panic unless
> after a failed call). I can add that no panic come from received calls.
>
> In the meantime that you are on vacation, may you (or some one else in
> the list) lend me a hint on how to compile the redhat kernel with debug
> information? (So that I can use the crash utility with the vmcores?)
>
> I've tried to put -g in the CFLAGS but nothing happens. (tried also
> CFLAGS_KERNEL).
>
> I suspect I have to switch something in the menuconfig (perhaps kernel
> debugging/frame pointers?) and do some other additional tricks.
You can compile the acm driver with debugging by altering cdc-acm.c
in the kernel sources. It has a line "#undef DEBUG" which needs to be removed.
As acm at least triggers it, knowing what acm does is important.
HTH
Oliver
-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
_______________________________________________
linux-usb-devel@lists.sourceforge.net
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
