[16162] in Athena Bugs
Re: sun4 8.2.9: system panic
daemon@ATHENA.MIT.EDU (Greg Hudson)
Fri Aug 14 14:04:58 1998
Date: Fri, 14 Aug 1998 14:04:57 -0400
From: Greg Hudson <ghudson@MIT.EDU>
To: Katie Livingston Vale <katiel@MIT.EDU>
Cc: bugs@MIT.EDU
In-Reply-To: "[16161] in Athena Bugs"
> Machine freezes and displays error messages about memory allocation
It would be very helpful if, the next time this happens, you copy down
the exact panic message (mainly the explanatory message; the numbers
might be useful to jhawk but it's tedious to copy them down and I
doubt we'd do anything with them).
Also, if you could, at the prom, type "ctrace" and look at the stack
trace. The output of ctrace should look like:
PC: f0052794
Last leaf: jmpl f005b39c from 10006e38 client_handler+38
0 w %o0-%o5: (10000000 16 f0000000 0 3 301611e4 )
call 100349e4 p1275_sparc_cif_handler from 10031b24 prom_enter_mon+34
1 w %o0-%o5: (f005b39c 301612a0 e 104120c0 0 104126d0 )
call 10031af0 prom_enter_mon from 1001ea78 debug_enter+80
2 w %o0-%o5: (0 f 0 3 30 8 )
and so on. What I'm mainly interested is the last fields of the long
lines (client_handler+38, prom_enter_mon+34, debug_enter+80), but
unfortunately the top few usually aren't very interesting ("yup, it
panicked alright"). So if you could get the first few of those fields
after the functions stop having names like "debug_enter" or "do_panic"
or "cmn_err" or "cpu_log_err".
Thanks.
