farestam@ULTIMA.CERFACS.FR (Stefan Farestam) (11/30/90)
A graphics program of mine occasionally makes the window server
go down. Although exhaustive searches for bugs, I have not been
able to locate the cause. The program runs on a 4D25GT under
system release 3.2.1. The following are some excerpts from
/usr/adm/SYSLOG taken right after the crashes:
grcond[15409]: Child process /etc/gl/pandora terminated by signal 15
grcond[15409]: Restoring PROM textport microcode
grcond[15866]: In limbo
grcond[15866]: Alive
grcond[15866]: CIO: UT (FIFO still > half full)!
grcond[15866]: CIO:
grcond[15866]: CIO: WARNING: Graphics error: GE PC = 2672
grcond[15866]: CIO:
grcond[15866]: Child process /etc/gl/pandora terminated by signal 15
grcond[15866]: Restoring PROM textport microcode
grcond[16883]: In limbo
grcond[16883]: Alive
grcond[16883]: CIO: UT (FIFO still > half full)!
grcond[16883]: CIO:
grcond[16883]: CIO: WARNING: Graphics error: GE PC = 2672
grcond[16883]: CIO:
grcond[16883]: CIO: gr1_fifo_intr: TIMEOUT (FIFO still > half full)!
grcond[16883]: CIO:
grcond[16883]: CIO: WARNING: Graphics error: GE PC = 2678
grcond[16883]: CIO:
grcond[14580]: CIO: gr1_fifo_intr: TIMEOUT (FIFO still > half full)!
grcond[14580]: CIO:
grcond[14580]: CIO: WARNING: Graphics error: GE PC = 2672
grcond[14580]: CIO: gr1_fifo_intr: TIMEOUT (FIFO still > half full)!
grcond[14580]: CIO:
grcond[14580]: CIO: WARNING: Graphics error: GE PC = 2680
grcond[16883]: Child process /etc/gl/pandora terminated by signal 15
grcond[16883]: Restoring PROM textport microcode
grcond[17241]: In limbo
grcond[17241]: Alive
grcond[17241]: CIO: UT (FIFO still > half full)!
grcond[17241]: CIO:
grcond[17241]: CIO: WARNING: Graphics error: GE PC = 2672
grcond[17241]: CIO:
grcond[17241]: CIO: gr1_fifo_intr: TIMEOUT (FIFO still > half full)!
grcond[17241]: CIO:
grcond[17241]: CIO: WARNING: Graphics error: GE PC = 2678
grcond[17241]: CIO:
grcond[17241]: CIO: gr1_fifo_intr: TIMEOUT (FIFO still > half full)!
grcond[17241]: CIO:
grcond[17241]: CIO: WARNING: Graphics error: GE PC = 2672
grcond[17241]: CIO:
I wonder if someone might have a clue as to how to interpret
these error messages, and consequently being able to hint me
what to look for in my program.
Any suggestions appreciated.
Thanks,
Stefan Farestam
--
.................................................................
. Stefan Farestam <farestam@orion.cerfacs.fr> .
. __ __ __ _ _ _ .
. / |_ )|_ /_\/ ( European Centre for Research and .
. \_ |__\| / \__) Advanced Training in Scientific Computation .
................................................................. amys@TheImprov.ncsa.uiuc.edu (Amy Swanson) (12/05/90)
farestam@ULTIMA.CERFACS.FR (Stefan Farestam) writes: > A graphics program of mine occasionally makes the window server > go down. Although exhaustive searches for bugs, I have not been > able to locate the cause. The program runs on a 4D25GT under > system release 3.2.1. The following are some excerpts from > /usr/adm/SYSLOG taken right after the crashes: < excerpts from SYSLOG deleted > We get these messages, as well, along with a lot of dead window managers that have a "SIGNAL 15" error message on the console just before dying. All our problems started after our systems were upgraded from 4D/20Gs to 4D/25TGs. An initial problem was caused by a graphics timing chip that wasn't upgraded to run with the faster processor, but we still see these messages after the chips have been replaced. The only answer I have received from SGI is that upgrading to IRIX 3.3.1 will fix the problem. We will be upgrading over the Christmas break, so I'll see what happens then. If anyone at SGI has a "technical" answer to this question, please post it. I'd really like to know the reason behind these error messages. Amy Amy K. Swanson SGI Systems Administrator NCSA - National Center for Supercomputing Applications University of Illinois @ Urbana-Champaign amys@ncsa.uiuc.edu