It's always good practice to get initial information. How long was the system up? In the case of a multiple CPU system, how many CPUs are on line and what where they doing? We will use some of the macros we wrote earlier. Hiya on p4d-2000a... cp $HOME/adb/mymacros/* . Hiya on p4d-2000a... adb -k unix.0 vmcore.0 physmem fd87 $<initial-info Initial Dump Information ======================== utsname+0x101: dragon Hostname 1994 Aug 30 09:08:12 Time of boot time: time: 1994 Sep 1 14:30:48 Time of crash $<msgbuf msgbuf: msgbuf: magic size bufx bufr 8724786 1fe8 16a 0 msgbuf+0x10: Y\ panic[cpu0]/thread=0xe18b1ec0: zero syncing file systems...panic[cpu0]/thread=0xe18d8ec0: panic sync timeout 12081 static and sysmap kernel pages 176 dynamic kernel data pages 565 kernel-pageable pages 0 segkmap kernel pages 0 segvn kernel pages 0 current user process pages 12822 total pages (12822 chunks) dumping to vp f584c804, offse *panicstr/s cpr_info+0x13c4: zero $<proconcpu ncpus: ncpus: Number of CPUS: 4 cpu0+8: e18d8ec0 Thread address 0xe18d8f60: e00ee110 Proc address p0+0x260: sched Next CPU... tcl_endptopen+0x50b0: e18f1ec0 Thread address 0xe18f1f60: e00ee110 Proc address p0+0x260: sched This CPU was idle Next CPU... tcl_endptopen+0x3cb0: e1914ec0 Thread address 0xe1914f60: e00ee110 Proc address p0+0x260: sched This CPU was idle Next CPU... tcl_endptopen+0x38b0: e1937ec0 Thread address 0xe1937f60: e00ee110 Proc address p0+0x260: sched This CPU was idle Based on what we see so far, the system was up for over two days when it was manually forced to panic. The "panic: zero" tells us this. |