core dump

jack_zheng

when trying to load the block driver, the system crashed and gave a core dump.

use dumpadm to show where the dump file is,
-bash-3.00# dumpadm
      Dump content: kernel pages
       Dump device: /dev/dsk/c0d0s1 (swap)
Savecore directory: /var/crash/sunhost
  Savecore enabled: yes

but on solaris 6 or earlier release, modify /etc/inid.d/syssetup to change the directory.

-bash-3.00# ls -l
total 117202
-rw-r--r--   1 root     root           2 May 16 17:41 bounds
-rw-r--r--   1 root     root     1241958 May 16 17:41 unix.0
-rw-r--r--   1 root     root     58707968 May 16 17:41 vmcore.0

Fortunately, there are some clues in the system log. So no need to do analysis of the dump file.

May 16 17:38:02 jackv genunix: [ID 335743 kern.notice] BAD TRAP: type=e (#pf Page fault) rp=d6270c30 addr=c occu
rred in module "block_1" due to a NULL pointer dereference
May 16 17:38:02 jackv unix: [ID 100000 kern.notice]
May 16 17:38:02 jackv unix: [ID 839527 kern.notice] sched:
May 16 17:38:02 jackv unix: [ID 753105 kern.notice] #pf Page fault
May 16 17:38:02 jackv unix: [ID 532287 kern.notice] Bad kernel fault at addr=0xc
May 16 17:38:02 jackv unix: [ID 243837 kern.notice] pid=0, pc=0xfa452482, sp=0xfa4521d0, eflags=0x10286
find the bug in the code. It was a NULL pointer.


本文来自ChinaUnix博客，如果查看原文请点：http://blog.chinaunix.net/u1/38597/showart_302708.html