I'm having occasional panics when doing disk-intensive activities like a buildworld or portupgrade, and I'm having trouble getting useful backtraces. For some reason I can't get the kernel to write a dump out. Most recently, while updating the pkgdb, my system did this:
Nov 19 14:47:19 basement syslogd: kernel boot file is /boot/kernel/kernel Nov 19 14:47:19 basement kernel: Slab at 0xc165bfd4, freei 1 = 0. Nov 19 14:47:19 basement kernel: panic: Duplicate free of item 0xc165b0cc from zone 0xc0a023c0(VMSPACE) Nov 19 14:47:19 basement kernel: Nov 19 14:47:19 basement kernel: Nov 19 14:47:19 basement kernel: syncing disks, buffers remaining... panic: bremfree: bp 0xc251d26c not locked Nov 19 14:47:19 basement kernel: Uptime: 1d1h20m15s Nov 19 14:47:19 basement kernel: Dumping 63 MB Nov 19 14:47:19 basement kernel: ata0: resetting devices .. Nov 19 14:47:19 basement kernel: panic: bremfree: bp 0xc250f3e8 not locked Nov 19 14:47:19 basement kernel: Uptime: 1d1h20m15s Nov 19 14:47:19 basement kernel: Terminate ACPI Nov 19 14:47:19 basement kernel: Automatic reboot in 15 seconds - press a key on the console to abort Nov 19 14:47:19 basement kernel: Rebooting... I have a dumpdev set up pointing to my swap space /dev/ad0s1b, I have DDB and DDB_UNATTENDED built into the kernel. I have a kernel with symbols in to use for getting a backtrace, if I could get a core dump from the kernel. What am I missing here? If neccessary, is there a way to force a core dump if I let it break into DDB and go sit at the console? --Mike Edenfield To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
