Nekonomicon - Indigo2 Impact R10K crashing

SiliconClassics
Who joined April 11, 2004, 4:07 p.m.
and authored 1374 notes

Wrote the following at Nov. 17, 2010, 6:25 p.m...

I've got an R10K Impact I2 system that's having issues. Frequent, replicable crashes while browsing certain directory windows on the desktop (non-fatal, the system stays up but the desktop, icon catalog, and file manager windows disappear, and logging-out and back in again restores them). Also, it's suffering segmentation faults and similar errors when starting Maya, the same errors every time. These problems occur identically with fresh installs of both IRIX 6.5.13 and 6.5.22. I've tried several different video board sets, from Solid to High to Max Impact, with no effect.

Oddly enough, the system seems pretty stable otherwise: no kernel panics or crashes when running typical programs, and running the IDE from the boot menu does not reveal any obvious errors. I'd like to read the detailed results of the IDE - is there a way to read them from within IRIX by opening a text file?

Any ideas what this might be?

SAQ
Who joined July 19, 2006, 7:37 a.m.
and authored 5590 notes

Wrote the following at Nov. 17, 2010, 9 p.m...

ajerimez wrote:

I've got an R10K Impact I2 system that's having issues. Frequent, replicable crashes while browsing certain directory windows on the desktop (non-fatal, the system stays up but the desktop, icon catalog, and file manager windows disappear, and logging-out and back in again restores them). Also, it's suffering segmentation faults and similar errors when starting Maya, the same errors every time. These problems occur identically with fresh installs of both IRIX 6.5.13 and 6.5.22. I've tried several different video board sets, from Solid to High to Max Impact, with no effect.

Oddly enough, the system seems pretty stable otherwise: no kernel panics or crashes when running typical programs, and running the IDE from the boot menu does not reveal any obvious errors. I'd like to read the detailed results of the IDE - is there a way to read them from within IRIX by opening a text file?

Any ideas what this might be? The crashes seem graphics-related but not video-board specific, so I'll try reseating the midplane, but other than that I'm stumped.

You can run IDE from a serial console and save the output (typescript on UNIX with cu/tip, a number of other options with other terminal emulators).

_________________
Damn the torpedoes, full speed ahead!

Systems available for remote access on request.

:Indigo:

(single-CM)

guardian452
Who joined Aug. 21, 2007, 9:12 p.m.
and authored 2551 notes

Wrote the following at Nov. 18, 2010, 6:39 a.m...

Is this my old machine? Check the memory. I was unable to get irix to install until I removed one bank of memory and never had a problem since. Try reseating the memory, it's been over two years since I've powered it on.

It never had any 'repeatable' errors or any errors at all once I got it working. I put a 20" box fan blowing on the memory to install irix but memtest 86 in a PC confirmed those sticks were bad (the other two banks checked out OK), so I threw them out. It could be something on the motherboard itself, memory errors aren't typically *that* repeatable as the system will give a program (i.e. maya) different chunks of memory each time the program runs. Right?

_________________
:Onyx:

(Maradona) :Octane:

(DavidVilla) A1186 (Xavi)
A1370 (Messi) dp43tf (Puyol) A1387 (Abidal) A1408 (Guardiola)

"InfiniteReality Graphics - Power Through Complexity"

SiliconClassics
Who joined April 11, 2004, 4:07 p.m.
and authored 1374 notes

Wrote the following at Nov. 18, 2010, 7:43 a.m...

Haha, good guess Sybr! Shuffling the memory sticks around seems to alter the nature of the crashes, though they still occur in similar ways. For example, with the memory in one configuration, Maya crashes with a segmentation fault; in another configuration, Maya crashes with a bad system call. Both of these crashes occur upon trying to rotate the viewport by holding ALT+mouse button.

At this point, I'm concerned that the problem lies somewhere on the motherboard - a memory controller perhaps, or somewhere in the CPU cache memory, though IDE reports no memory-related errors.

SiliconClassics
Who joined April 11, 2004, 4:07 p.m.
and authored 1374 notes

Wrote the following at Dec. 3, 2010, 2:30 p.m...

Update: The mystery deepens. I replaced the motherboard and even the midplane, and the same errors still occur - random crashes in X and crashing of Maya upon orbiting the viewport. I've now swapped video boards, RAM, motherboard, and midplane, with absolutely no effect. At this point, what else could it be other than either the CPU or the power supply?

guardian452
Who joined Aug. 21, 2007, 9:12 p.m.
and authored 2551 notes

Wrote the following at Dec. 3, 2010, 3 p.m...

ajerimez wrote:

At this point, what else could it be other than either the CPU or the power supply?

Your neighbor's HAMster radio set?

I would swap the processor out, I presume you have a complete guts machine to work with; if that doesn't do it, the power supply. If THAT doesn't do it, then it must have been possesed by the antichrist as it was working when I had it (though I never ran maya on it)

Seriously, I would check out the CPU. Same program crashing in the same operation... sadly I think that processor is pretty rare (but I've never looked into the indigo2 market).

Poor unloved indigo2. The guy before me (I am pretty sure) got it for free, I got it for free, you got it for free... is she still sexy?

_________________
:Onyx:

(Maradona) :Octane:

(DavidVilla) A1186 (Xavi)
A1370 (Messi) dp43tf (Puyol) A1387 (Abidal) A1408 (Guardiola)

"InfiniteReality Graphics - Power Through Complexity"

Stoo
Who joined July 26, 2010, 8:13 a.m.
and authored 134 notes

Wrote the following at Dec. 4, 2010, 3:54 a.m...

195mhz 10k's are available, I've seen a fair few on ebay recently in the uk - I'd have thought there would be even more on .com

_________________
:Octane2:

Octane2 Dual R14K 600Mhz, 4GB RAM, VPro V12, DCD, PCI Card Cage, GBit Ethernet
:Indigo2IMP:

Indigo2 R10K 195MHz, 1GB RAM, Maximum Impact, I2V
The boxen are breeding... help!

SiliconClassics
Who joined April 11, 2004, 4:07 p.m.
and authored 1374 notes

Wrote the following at Jan. 4, 2011, 4:43 p.m...

UPDATE: Just to put a cap on this thread, it turns out that it was a bad CPU module after all. I replaced it with an equivalent R10K 195MHz and all symptoms have disappeared. Unfortunately I swapped out almost every single other component before trying the CPU!

Not sure what exactly was wrong with the processor module, but it's going to the recycler. Thanks to all who offered assistance, and to Mike K for sending me parts until we found the culprit!

foetz
Who joined April 14, 2003, 3:34 a.m.
and authored 4683 notes

Wrote the following at Jan. 6, 2011, 6:56 a.m...

ajerimez wrote:

UPDATE: Just to put a cap on this thread, it turns out that it was a bad CPU module after all. I replaced it with an equivalent R10K 195MHz and all symptoms have disappeared. Unfortunately I swapped out almost every single other component before trying the CPU!

sometimes it's the only way but now you got a nice (and working) thingy after all

_________________
r-a-c.de

jan-jaap
Who joined June 17, 2004, 10:35 a.m.
and authored 3603 notes

Wrote the following at Jan. 6, 2011, 7:54 a.m...

foetz wrote:

now you got a nice (and working) thingy after all

Or maybe it just surfaced on eBay as yet another "Studio System" ?

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum : almost every MIPS/IRIX system.
Wanted : GM1 board for Professional Series GT graphics (030-0076-003, 030-0076-004)

SGI: Hardware

Indigo2 Impact R10K crashing