On Wednesday, 3 May 2000 at 21:24:47 -0400, Howard Leadmon wrote:
>
>>> OK, well as I don't remember what options had been in what kernel
>>> from the old crashes, I just setup the machine to generate more
>>> crash dumps and sure enough it was willing to give me one
>>> quickly.. :)
>>>
>>> Here is a backtrace done on the dump I got only a few minutes ago,
>>> also please note that currently I am using a DEC based network card
>>> instead of the EEpro adapter as I had both sitting around. If it
>>> would be better to try and get the dumps with the EEpro I am sure it
>>> can be arranged. Also note that I am currently running SMP, but did
>>> remove one CPU and built a non-SMP kernel to see what happened, and
>>> still the machine dies..
>>
>>> ---
>>> #0 0xc013abd1 in boot ()
>>> (kgdb) back
>>> #0 0xc013abd1 in boot ()
>>> #1 0xc013af94 in poweroff_wait ()
>>> #2 0xc02283cc in trap_fatal ()
>>> #3 0xc022805d in trap_pfault ()
>>> #4 0xc0227c57 in trap ()
>>> #5 0xc01cdca5 in acquire_lock ()
>>> #6 0xc01d2f7c in softdep_count_dependencies ()
>>> #7 0xc01d624c in ffs_fsync ()
>>> #8 0xc01d4d66 in ffs_sync ()
>>> #9 0xc01673ef in sync ()
>>> #10 0xc013a9b3 in boot ()
>>> #11 0xc013af94 in poweroff_wait ()
>>> #12 0xc02283cc in trap_fatal ()
>>> #13 0xc022805d in trap_pfault ()
>>> #14 0xc0227c57 in trap ()
>>> #15 0xc017246b in bpfioctl ()
>>> #16 0xc01c19 in ?? ()
>>> cannot read proc at 0
>>> (kgdb)
>>
>> Interesting. That explains the splbio, anyway. The real problem is
>> at frame 15, in bpfioctl, but the system trapped while trying to sync
>> before dumping.
>>
>> As mentioned in the handbook, you need symbols in your kernel in order
>> to find out any more information. Build a kernel with the -g option
>> and try again.
>
> Umm, the kernel was built with the -g option, as I knew I needed to be
> able to debug it when it crashed. I looked at the handbook, but I don't
> personally see a clear direction as to where I need to go from here. If
> you can tell me what to look for I am more than willing to go digging..
Ah, then you're in better shape. I assume you installed the stripped
version of the kernel (it's default, despite my protests), so that's
what savecore saved for you. You'll have the original still in
/usr/src/sys/compile/<MYKERNEL>/kernel.debug. Copy it to /var/crash,
overwriting your kernel.2 or whatever, and the backtrace will be much
more verbose.
Greg
--
Finger [EMAIL PROTECTED] for PGP public key
See complete headers for address and phone numbers
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message
