Messages in this thread | | | Date | Mon, 14 Jul 2008 11:57:33 -0400 (EDT) | From | Alan Stern <> | Subject | Re: 2.6.26-rc9-git12: Reported regressions from 2.6.25 |
| |
On Mon, 14 Jul 2008, Johannes Berg wrote:
> > > > Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11072 > > > > Subject : scsi-layer crash after usb storage device unplug > > > > Submitter : Johannes Berg <johannes@sipsolutions.net> > > > > Date : 2008-07-12 01:50 (2 days old) > > > > Handled-By : Alan Stern <stern@rowland.harvard.edu> > > > > Patch : http://bugzilla.kernel.org/show_bug.cgi?id=11072#c3 > > > > > > Alan, Greg? > > > > We haven't heard back yet from Johannes to confirm that the fix really > > works. If it does, IMO the patch should go into 2.6.26.1. > > > > (In fact it might not be so easy to tell, since the failure is caused > > be a race. Perhaps it would be best to submit the patch in any case.) > > I haven't had a chance to apply the patch, but given that to cause the > oops to happen I had to unplug/replug the device at random intervals for > probably 10 minutes (yes, it took me that long to figure out I had put > an SDHC card into an SD reader) I suppose it really won't be easy to > tell.
Add
printk(KERN_ALERT "Unplug now!\n"); ssleep(10);
to the start of slave_configure() in drivers/usb/storage/scsiglue.c. Then it should be very easy to trigger the failure. :-)
Assuming the failure does occur without the patch, see if it still occurs once the patch is in place.
Alan Stern
| |