On Sat, 2009-08-15 at 10:54 +0400, Alexey Tarasov wrote: > Hello. > Here is core.txt: > http://lexasoft.ru/metamphetamine/Panic/core.txt.0 > > 'All versions' means all versions before and including 7-STABLE. > This servers were used before on Linux. There were no problems. This is odd. For the bug to have existed since 6.0 without being seen before, and given you are seeing this issue across 30 servers, it suggests that this may be triggered by something unusual in your setup. Are you doing anything unusual on these machines? For example, strange NFS mount options, unusual filesystems mounted, or anything? Are you using jails or similar? Is there anything else you can think of that you may be doing on these servers that is in any way unusual? Do you know if the panics seen on 6.x are the same as the panics seen on 7.x? There's always a chance you're seeing entirely different problems. Also, in the first email in this thread, you posted some pictures of the debugger output, but there isn't actually a photo of the panic text itself. Is there any chance you still have this information? Lastly, if you still have the kernel that was running during the first panic, can you do the following please: objdump -t /usr/obj/usr/src/sys/GENERIC/kernel | grep witness_lock there should be one ".text" entry for "witness_lock". Add 0x1c7 to it, then: addr2line -e /usr/obj/usr/src/sys/GENERIC/kernel 0xADDRESS (with address replaced by what you calculated above). Thanks, GavinReceived on Mon Aug 17 2009 - 12:01:09 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:54 UTC