It crashed again this morning/last night (not sure when)
Symptoms are different.  Traceback different.  Smells like
HW.  Reseat memory, CPU, ROM - lather, rinse, repeat.

Sorry to bother everyone....

-dgl-

At 9:26 AM -0500 7/21/03, Donald Lee wrote:
>I have been trying to get a new server (beige G3, OF 2.4, 500 Mhz
>upgrade G3 CPU, 256 Mb mem) tested and into production.
>
>Unlike all my previous NetBSD machines, this one is not stable.
>Every couple of weeks, it falls over with a panic of some sort.
>The last one was a "corrupted inode".  This latest one is detailed
>below.
>
>I am hoping that someone can help with some suggestions, or at least
>some info on how I can best track this down.  I can use a debugger, but
>am not familiar with many of the internal structures of NetBSD.
>I would very much
>like to go production with this one, but I have to get the MTTI up
>into the same range as my other machines (2 years between reboots.)
>
>Thanks,
>
>-dgl-
>
>
>(below is the screen output at the time of the crash...)
>
>trap: kernel read DSI @ 0xdd48f014 by 0x258dbc (DSISR 0x40000000, err=14)       
>Press a key to panic.
>tlp0: receive ring overrun
>panic: trap
>Stopped in pid 13316 (find) at cpu_Debugger+0x10:   lwz   r0, r1, 0x14
>db> tr
>0xdd48ee00: at panic+174
>0xdd48eec0: at trap+908
>0xdd48ef30: kernel DSI read trap @ 0xdd48f014 by cache_lookup + 4a8: srr1=0x9032
>        r1=0xdd48efe0 cr=0x8000f038 xer=0x20000000 ctr=0x226ff8 dsisr=0x4000
>0000
>  <????>  : at cache_lookup+34c
>0xdd48efe0: at 0x418761c4
>db>