Re: vinum and/or geom panic on alpha

On Sat, Jun 14, 2003 at 03:15:27PM +0930, Greg 'groggy' Lehey wrote:
> On Tuesday, 10 June 2003 at 14:05:11 +0200, Bernd Walter wrote:
> >
> > fatal kernel trap:
> >
> > Stopped at      g_dev_strategy+0x44:    stq     t0,0x20(v0) <0x20>      <t0=0x1a61da400,v0=0x0>
> > db> trace
> > g_dev_strategy() at g_dev_strategy+0x44
> > launch_requests() at launch_requests+0x390
> > prologue botch: displacement 128
> > frame size botch: adjust register offsets?
> > vinumstart() at vinumstart+0x250
> > prologue botch: displacement 64
> > frame size botch: adjust register offsets?
> > intr_n() at 0xccec340
> 
> Can you check the locals of launch_requests(), please?

Unfortunately it's currently not possible.
IIRC gdb -k doesn't work on alpha and I can't get a crashdump.
Last night I had the same panic on an i386 system, but it refused
to do a crashdump :(
Currently I'm seeing this problem on the following systems:
SMP alpha /  21164A CPUs / 5.1-RELEASE
UP alpha / 21164A CPU / 5.1-CURRENT (8th june)
UP i386 / AMD K6 CPU / 5.1-BETA (10th may with some NFS fixes)

I will arrange things for a remote gdb session and hope for the next
panic to happen on the prepared machine.

-- 
B.Walter                   BWCT                http://www.bwct.de
ticso_at_bwct.de                                  info_at_bwct.de