Subject: Re: Machine hangs
To: Andrey Petrov <petrov@netbsd.org>
From: Sten Spans <sten@blinkenlights.nl>
List: port-sparc64
Date: 01/13/2004 20:57:48
On Mon, 12 Jan 2004, Andrey Petrov wrote:
> On Sat, Jan 10, 2004 at 04:26:39PM +0000, Steve Doyle wrote:
> >
> > Evening,
> >
> > I'm having a problem with the machine that I'd really appreciate
> > some assistance with. From time to time, normally after a couple of
> > weeks uptime, something seems to be killing the box stone dead without
> > any warning, I've been through the logfiles after the last crash and
> > so far it's turned up nothing that would shed some light on the cause.
> >
>
> Can you enter ddb? <Stop-A> from keyboard or 'break' from serial console.
> If you can, please post output from 't' and 'ps' commands.
I have an ultra5 ( 270mhz, 512mb, symbios scsi ) which seems to experience
this problem as well, its pretty easy to trigger. The machine hasn't
completed a ./build.sh distribution yet. This is with a recent current
kernel and softupdates enabled.
The sympthoms are that nbmake suddenly sleeps forever and the
load keeps going up ( ctrl-t ) until entering ddb breaks as well.
Here follows a paste from a serial console session ( the machine
has no configured network interface ).
cleandir ===> gnu/usr.bin/gcc3/gcc
rm -f gccspec.o gccspec.ln
rm -f a.out [Ee]rrs mklog core *.core .gdbinit gcc
rm -f gcc.cat1
rm -f .depend gccspec.d /usr/src/gnu/usr.bin/gcc3/gcc/tags
rm -f gcc.info
cleandir ===> gnu/usr.bin/gcc3/include
rm -f a.out [Ee]rrs mklog core *.core .gdbinit
cleandir ===> gnu/usr.bin/gcc3/protoize
load: 1.66 cmd: nbmake 9396 [running] 0.00u 16.10s 53% 0k
load: 1.77 cmd: nbmake 9396 [running] 0.00u 23.96s 68% 0k
load: 1.96 cmd: nbmake 9396 [running] 0.00u 29.76s 76% 0k
load: 2.19 cmd: nbmake 9396 [running] 0.00u 48.34s 90% 0k
load: 2.25 cmd: nbmake 9396 [running] 0.00u 53.51s 92% 0k
load: 3.02 cmd: nbmake 9396 [running] 0.00u 86.20s 98% 0k
kdb breakpoint at 12f8384
Stopped in pid 9396.1 (nbmake) at netbsd:cpu_Debugger+0x4: nop
db> t
sab_intr(2501700, 0, e0017ed0, 0, 12cbdf0, 0) at netbsd:sab_intr+0xa4
?(c67c840, 0, ffffffffffffffff, 1, c67bb80, 1000000) at 0x1008f94
ctx_alloc(c67c780, c67c780, 0, ffffffffffffffff, 0, da2b700) at netbsd:ctx_alloc+0x78
pmap_activate_pmap(c67c780, ffffffffffffffff, 1009d7c, b6e2000, 0, 80080d) at netbsd:pmap_activate_pmap+0x14
uvmspace_exec(c67bce0, 0, ffffffffffffffff, da2bab0, 0, 1364230) at netbsd:uvmspace_exec+0x44
sys_execve(0, da2bdd0, 0, 0, 267cb00, da36010) at netbsd:sys_execve+0x59c
syscall(da2bed0, 3b, 4043f45c, da2bdd0, 0, 4043f460) at netbsd:syscall+0xcc
?(232ef0, ffffffffffffce80, ffffffffffffd848, 0, 0, 0) at 0x1008cb8
db> ps
PID PPID PGRP UID S FLAGS LWPS COMMAND WAIT
>9396 5609 284 0 2 0x100012 1 nbmake
5609 4259 284 0 2 0x4002 1 nbmake ppwait
4259 5479 284 0 2 0x4002 1 sh wait
5479 12977 284 0 2 0x4002 1 nbmake piperd
12977 12511 284 0 2 0x4002 1 sh wait
12511 6090 284 0 2 0x4002 1 nbmake wait
6090 8749 284 0 2 0x4002 1 sh wait
8749 5213 284 0 2 0x4002 1 nbmake wait
5213 7953 284 0 2 0x4002 1 sh wait
7953 8016 284 0 2 0x4002 1 nbmake wait
8016 2263 284 0 2 0x4002 1 sh wait
2263 2742 284 0 2 0x4002 1 nbmake wait
2742 2346 284 0 2 0x4002 1 sh wait
2346 2571 284 0 2 0x4002 1 nbmake wait
2571 2613 284 0 2 0x4002 1 sh wait
2613 284 284 0 2 0x4002 1 nbmake wait
284 282 284 0 2 0x4002 1 sh wait
282 1 282 0 2 0x4002 1 csh pause
271 1 271 0 2 0 1 cron
254 1 254 0 2 0 1 inetd pause
123 1 123 0 2 0 1 syslogd
9 0 0 0 2 0x20200 1 aiodoned aiodone
8 0 0 0 2 0x20200 1 ioflush
7 0 0 0 2 0x20200 1 pagedaemon pgdaemo
6 0 0 0 2 0x20200 1 lfs_writer lfswrit
5 0 0 0 2 0x20200 1 atapibus0 sccomp
4 0 0 0 2 0x20200 1 scsibus0 sccomp
3 0 0 0 2 0x20200 1 atabus1 atath
2 0 0 0 2 0x20200 1 atabus0 atath
1 0 1 0 2 0x4000 1 init wait
0 -1 0 0 2 0x20200 1 swapper
db> continue
load: 3.10 cmd: nbmake 9396 [running] 0.00u 89.62s 98% 0k
load: 3.10 cmd: nbmake 9396 [running] 0.00u 94.36s 98% 0k
load: 6.84 cmd: nbmake 9396 [running] 0.00u 107.68s 99% 0k
load: 6.84 cmd: nbmake 9396 [running] 0.00u 109.32s 99% 0k
kdb breakpoint at 12f8384
Stopped in pid 9396.1 (nbmake) at netbsd:cpu_Debugger+0x4: nop
db> t
sab_intr(2501700, 0, e0017ed0, 0, 12cbdf0, 0) at netbsd:sab_intr+0xa4
?(c67c840, 0, ffffffffffffffff, 1, c67bb80, 1000000) at 0x1008f94
ctx_alloc(c67c780, c67c780, 0, ffffffffffffffff, 0, da2b700) at
netbsd:ctx_alloc+0x78
pmap_activate_pmap(c67c780, ffffffffffffffff, 1009d7c, b6e2000, 0, 80080d)
at netbsd:pmap_activate_pmap+0x14
uvmspace_exec(c67bce0, 0, ffffffffffffffff, da2bab0, 0, 1364230) at
netbsd:uvmspace_exec+0x44
sys_execve(0, da2bdd0, 0, 0, 267cb00, da36010) at netbsd:sys_execve+0x59c
syscall(da2bed0, 3b, 4043f45c, da2bdd0, 0, 4043f460) at netbsd:syscall+0xcc
?(232ef0, ffffffffffffce80, ffffffffffffd848, 0, 0, 0) at 0x1008cb8
db> continue
load: 8.87 cmd: nbmake 9396 [running] 0.00u 115.62s 99% 0k
load: 11.33 cmd: nbmake 9396 [running] 0.00u 133.18s 99% 0k
load: 12.03 cmd: nbmake 9396 [running] 0.00u 134.78s 99% 0k
etc.
--
Sten Spans
"There is a crack in everything, that's how the light gets in."
Leonard Cohen - Anthem