Greg Harrison (harrisog++at++geocities.com)
Thu, 10 Dec 1998 21:33:45 -0500
By rejected they mean crash, because it will. For as much as SGI says on
their own site (http://www.sgi.com/developers/marketing/forums/multi_thd)
that pthreads are better (they are), everything in Performer is implemented
with sproc() and therefore pthreads can't be linked in the same app, the
only way to use them would be through a separate program and shared memory.
OpenGL 1.1 is pthread compliant, but Performer will not allow them.
Hopefully SGI will listen to their own PR about threads ("think pthreads!")
and make Performer 2.3+ out of pthreads instead of sprocs.
Greg Harrison
Air Force Institute of Technology
-----Original Message-----
From: Jason Leigh [mailto:spiff++at++evl.uic.edu]
Sent: Thursday, December 10, 1998 10:30 AM
To: 'info-performer++at++sgi.com'
Subject: re: pthread and perfly
Hi Chiron
I saw your post regarding the pthreads/performer problem. Were you able to
find a solution. We are experiencing problems with pthreads/performer too
whereby the application always crashes immediately in IRIX 6.5 on an
Onyx-class
machine just by linking pthreads but not making a single call.
We are primarily using pthreads with Performer and the CAVE library so it
has
been difficult isolating where the problem is. We wrote an exact same kind
of
perfly-like application using both sproc vs pthreads and the sproc program
worked flawlessly. The pthreads program crashed instantaneously.
As far as kernel panics go, I have been getting those while running
pthreads/CAVE/performer applications on my O2 under IRIX 6.3.
Jason
=============================================================
Jason Leigh, PhD
Senior Scientist
spiff++at++evl.uic.edu Electronic Visualization Lab (M/C 154)
EVL Phone (312) 996-3002 University of Illinois at Chicago
EVL FAX (312) 413-7585 851 S. Morgan St. Room 1120 SEO
http://www.evl.uic.edu/spiff Chicago, IL 60607-7053
"... Science in those days worked in broad strokes. They got right to
the point. Nowadays, it's all just molecule, molecule,
molecule. Nothing ever happens big!"
=======================================================================
List Archives, FAQ, FTP: http://www.sgi.com/software/performer/
Submissions: info-performer++at++sgi.com
Admin. requests: info-performer-request++at++sgi.com
This archive was generated by hypermail 2.0b2 on Thu Dec 10 1998 - 18:50:34 PST