Re: SYSLOG-Warning

New Message Reply Date view Thread view Subject view Author view

Rob Jenkins (robj++at++quid.csd.sgi.com)
Mon, 18 May 1998 10:35:23 -0700


That particular error just means that the gfx pipe timed out, the FIFOs allow
for parts of the pipe being a bottleneck and there's some time period after
which it decides that the pipe is blocked downstream and that the FIFOs aren't
going to drain, ie the pipe is hung. A gfx pipe hang can be the result of bad
HW ( including badly seated HW as Ong Tze Lin points out ) but it can also be
as a result of a SW bug. I tend to always go through the same few steps
initially with these problems:

o Checking Patches

Always double check that the machine with the problem has the latest iR gfx
patch installed. It's vital to make sure that you're not chasing a problem that
is already fixed. Currently the latest released iR gfx patches are:

patch 2922: Onyx2 6.4 graphics rollup #5 including GVO support or
patch 2327: Onyx (not Onyx2) InfiniteReality 6.2 Fifth Release

o Checking Hardware

Check that the latest Onyx Diagnostics patch is installed. The newer the
diagnostics patch you have, the more likely it is that any HW problem will get
detected. Passing diagnostics doesn't mean you can be 100% sure that there's no
HW problem but it at least gives some confidence that HW is OK. Currently the
latest iR diagnostics patches are:

patch 2795: Onyx2 Diagnostics 7th release or
patch 2371: InfiniteReality (Onyx) Diagnostics Fifth Release

o Debugging the Pipe

If you've checked the patches and checked the HW and you still have a problem
then here's some suggestions for narrowing down the problem.

Try and describe the symptoms, these problems generally fall into a couple of
loose areas, some useful things to establish are:

Does the problem happen on one machine ? If possible try and reproduce on other
similar machines. If it really happens on > 1 similar machines then the chances
of it being a HW problem are reduced.
Does the problem happen with one specific application ?
If it happens with the same application always, is it with the same combination
of actions ?
Does the pipe crash ( and then should restart, putting you back to the login
screen ) or does the pipe just hang in that the gfx locks up but the rest of
the machine is still fine ?
Did the same application used to run OK ? Did *any* gfx SW get changed ?

If you can check the above and log a support call at the same time ( support
should take you through at least the first 2 things anyway ) then you're well
on the way to getting this resolved. A majority now turn out to be known SW
problems fixed in the latest patches or some HW problem ( again sorted via a
support call ). If it turns out you have a new SW problem ( quite rare now )
then the fun begins :-)

The key to solving a new iR gfx pipe SW problem is a reproducible test case,
ideally that can be run at SGI. iR has a great utility 'Kona Post Mortem ( kpm
)' that dumps the state of the gfx pipe. There is a man page so you can look at
how to read the dumps yourself although they're really only useful to
engineering here so I won't go into great detail on what to look for. I can
give more detail on narrowing down a testcase though if needs be although
support can also help you do that.

I've dealt with 'a few' of these things :-) It seems a bore but it's never
worth skipping the first 2 stages fo checking patches then HW before embarking
down the path of debugging.

Cheers
Rob

On May 12, 2:38am, Ong Tze Lin wrote:
> Subject: Re: SYSLOG-Warning
> Hi
>
> Have you tried running irsaudit to pinpoint the problem? I have come across
> this kind of error message several times and more often than not its due to a
> bad (or badly-seated) RM board.
>
> Run irsaudit as root off a laptop or dumb terminal, but not on the graphics
> console:
>
> First, stop graphics with /usr/gfx/stopgfx.
>
> Run /usr/diags/IR/bin/irsaudit and sit back while it churns away. It will
take
> some time to complete and return, but it will display its test results along
> the way- you'll know if you have bad parts.
>
> Restart graphics with /usr/gfx/startgfx.
>
> Hope this helps.
>
> Tzelin
>
> On May 12, 5:33pm, Masahiko Yamanaka wrote:
> > Subject: SYSLOG-Warning
> > Hello all!
> >
> > Though this might be a simple H/W problem, I believe someone in this ML
> > know about this...
> >
> > One of my customers has Onyx2/iR and a Performer-based application.
> >
> > As running the app., sometimes the window disappears and SYSLOG says,
> >
> > WARNING:IR0:Pipe hung: on TBUS or ARM activity
> > WARNING:IR0:Timeout waiting for fifo to drain ( level == 0x70a)
> >
> > What does this mean, what's wrong?
> > Or should I explain more details? (Sorry, I have only above right now...)
> >
> > Any help will be greatly appreciated!
> >
> > Thank you,
> > --
> > M.Y.
> > =======================================================================
> > List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
> > Submissions: info-performer++at++sgi.com
> > Admin. requests: info-performer-request++at++sgi.com
> >-- End of excerpt from Masahiko Yamanaka
>
>
>
> --
> Ong Tze Lin Technical Consultant
> Silicon Graphics ASEAN HQ High-Performance Computer Graphics
> 89 Science Park Dr. #03-03 to 06
> The Rutherford, Singapore 118261 Tel: (065)777 3088 Fax: (065)779 3650
>
> =======================================================================
> List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
> Submissions: info-performer++at++sgi.com
> Admin. requests: info-performer-request++at++sgi.com
>-- End of excerpt from Ong Tze Lin

-- 
________________________________________________________________
Rob Jenkins mailto:robj++at++sgi.com
Silicon Graphics, Mtn View, California, USA
=======================================================================
List Archives, FAQ, FTP:  http://www.sgi.com/Technology/Performer/
            Submissions:  info-performer++at++sgi.com
        Admin. requests:  info-performer-request++at++sgi.com

New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:57:24 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.