Port-sparc64 archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: SMP bugs ( was Re: Severe deadlock issues with 5.0/MP )



Hello,

On Feb 4, 2009, at 10:41 AM, Julian Coleman wrote:

I see them under heavy load, but I'm not so sure anymore.
This time I got a PC address which was nowhere near locore.s and an  
nPC that didn't make any sense at all ( being PC + 0x10 with no branch  
anywhere near that could do this )
I need to read more about OpenFirmware's debugging stuff, maybe I was  
on the wrong CPU.

I was tending toward hardware, after reading messages similar to:

 http://markmail.org/message/irop2we4mclu7eas
 http://markmail.org/message/iqalxxkv4bowksk2

I haven't tried the "f8002010 wector p" command mentioned yet.  Other
messages also suggest memory problems.  ".trap-registers" was mentioned as
a maybe useful OFW command.

I'll look at those.

Also, is what OF reports as 'Watchdog Reset' always an SIR?

In this case "Externally Initiated Reset" is an XIR.  Not sure when it can
happen on a U60 though - hardware watchdog and CPU trap when traps are
disabled - maybe others?

I don't think the U60 has a watchdog.
Maybe it's an ECC error? Somehow I doubt that though, booting with diag-switch? true and diag-level max didn't find anything funny. Too bad /memory doesn't have a selftest method anymore.

have fun
Michael

Attachment: PGP.sig
Description: This is a digitally signed message part



Home | Main Index | Thread Index | Old Index