i have a toshiba tecra 750cdt and two deskstations, one
at work and one at home. in the one at work, i have a scsi
drive internal to the deskstation. the scsi controller
is an AM53C974, which has worked fine as a module. usually,
i suspend and go home at the end of the day, then
resume and in the morning, i suspend to go to work.
usually, i get uptimes of several days or a week or more.
today, i had to reboot at work and when i came
home from work, i got this oops. it happened around the time
i tried to use ppp (also loaded as a module), right after a
return from a suspend (of course, no scsi drive was available
at this time, since i was at home). note that the drive
was not mounted at the time of the supend. also, after the
oops, the system was still useable for writing/reading files.
not sure what else it was good for.
this is my suspend command:
[20:39:25](2)bam:linux$ cat /usr/local/bin/goto-suspend
#!/bin/sh
sync
sync
rmmod AM53C974 >& /dev/null
rmmod floppy >& /dev/null
rmmod serial_cs &> /dev/null
sleep 4
sync
sync
sync
apm -s
sync
sync
modprobe -a floppy AM53C974 serial_cs >& /dev/null
soundoff
soundon
sync
[20:39:30](2)bam:linux$
here is the log:
Apr 9 19:31:55 bam ifup-ppp: pppd started for ppp1 on /dev/modem at 115200
Apr 9 19:31:56 bam kernel: divide error: 0000
Apr 9 19:31:56 bam kernel: CPU: 0
Apr 9 19:31:56 bam kernel: EIP: 0010:[<c01a6e3d>]
Apr 9 19:31:56 bam kernel: EFLAGS: 00010257
Apr 9 19:31:56 bam kernel: eax: 00000000 ebx: c6586000 ecx: 00000000
edx: 00000000
Apr 9 19:31:56 bam kernel: esi: c7953000 edi: c9c05ec0 ebp: 00000000
esp: c6587f4c
Apr 9 19:31:56 bam kernel: ds: 0018 es: 0018 ss: 0018
Apr 9 19:31:56 bam kernel: Process pppd (pid: 3226, process nr: 35,
stackpage=c6587000)
Apr 9 19:31:56 bam kernel: Stack: c6586000 c7953000 c6587f78 00005423
003337dd c0197903 c7953000 00000000
Apr 9 19:31:56 bam kernel: c7953000 08061bf8 c7953000 c6586000
c7953960 c01950a9 c7953000 00000000
Apr 9 19:31:56 bam kernel: 00000000 00005423 08061bf8 c90517e0
c012b976 c914e000 c9b33f00 00005423
Apr 9 19:31:56 bam kernel: Call Trace: [<c0197903>] [<c01950a9>] [<c012b976>]
[<c010994a>]
Apr 9 19:31:56 bam kernel: Code: f7 7f 10 89 c6 b9 05 00 00 00 31 d2 f7 f1 89
c6 85 f6 75 05
Apr 9 19:31:56 bam kernel: release_dev: ttyS1: read/write wait queue active!
Apr 9 19:32:27 bam last message repeated 31159 times
Apr 9 19:32:27 bam last message repeated 284 times
Apr 9 19:32:27 bam kernel: read/write wait queue active!
Apr 9 19:32:27 bam kernel: release_dev: ttyS1: read/write wait queue active!
Apr 9 19:32:57 bam last message repeated 32255 times
Apr 9 19:33:59 bam last message repeated 63876 times
Apr 9 19:35:00 bam last message repeated 64272 times
dmesg is full of these:
release_dev: ttyS1: read/write wait queue active!
release_dev: ttyS1: read/write wait queue active!
release_dev: ttyS1: read/write wait queue active!
top reports this
7:37pm up 9:25, 7 users, load average: 1.52, 1.18, 0.58
64 processes: 58 sleeping, 6 running, 0 zombie, 0 stopped
CPU states: 48.5% user, 51.4% system, 38.3% nice, 0.3% idle
Mem: 160508K av, 157100K used, 3408K free, 45656K shrd, 63928K buff
Swap: 130748K av, 0K used, 130748K free 36348K cached
PID USER PRI NI SIZE RSS SHARE STAT LIB %CPU %MEM TIME COMMAND
3226 root 12 5 0 0 0 RWN 0 38.3 0.0 2:16 pppd
147 root 8 0 472 472 392 R 0 37.7 0.2 2:18 syslogd
156 root 5 0 428 428 328 R 0 21.6 0.2 1:20 klogd
[20:27:37](2)bam:/tmp$ ./ksymoops /usr/src/linux/System.map < oopslog
Using `/usr/src/linux/System.map' to map addresses to symbols.
>>EIP: c01a6e3d <scsi_unregister_host+1d5/45c>
Trace: c0197903 <lf+3/64>
Trace: c01950a9 <n_tty_ioctl+451/52c>
Trace: c012b976 <sys_rename+62/28c>
Trace: c010994a <ret_from_sys_call+2/11>
Code: c01a6e3d <scsi_unregister_host+1d5/45c>
Code: c01a6e3d <scsi_unregister_host+1d5/45c> f7 7f 10 idivl
0x10(%edi),%eax
Code: c01a6e40 <scsi_unregister_host+1d8/45c> 89 c6 movl %eax,%esi
Code: c01a6e42 <scsi_unregister_host+1da/45c> b9 05 00 00 00 movl $0x5,%ecx
Code: c01a6e4d <scsi_unregister_host+1e5/45c> 31 d2 xorl %edx,%edx
Code: c01a6e4f <scsi_unregister_host+1e7/45c> f7 f1 divl %ecx,%eax
Code: c01a6e51 <scsi_unregister_host+1e9/45c> 89 c6 movl %eax,%esi
Code: c01a6e53 <scsi_unregister_host+1eb/45c> 85 f6 testl %esi,%esi
Code: c01a6e55 <scsi_unregister_host+1ed/45c> 75 05 jne
c01a6e56 <scsi_unregister_host+1ee/45c>
[20:27:37](2)bam:/tmp$
i hope this helps squash one more bug out.
while i'm at it, every time i compile a new kernel, i have to
uncomment this line from drivers/scsi/AM53C974.c:
#define AM53C974_SCSI_ID 7
could i ask that a config-time option be added for this
driver? any tips on how can i write this myself
and send a patch? (first timer)
thanks,
-- carlos
_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu