Hi All,
Is anyone else seeing problems with this graphics chipset? I have a test machine that runs fine for days and days until I try running my jogl app and then occasionally it will just freeze the complete system.
Kind Regards
Peter
What OS, JDK version and JOGL version are you running?
JSR-231 beta 3 on JDK 1.5.0_06 should work pretty well. There are some issues with the current JOGL nightly builds and when running with Mustang on X11 platforms that are being worked on.
Win2000 Profesional SP4
No updates from MB website so I have installed latest driver direct from Intel:
Intel 82852/82855 GM/GME Graphics Controller Driver version: 6.14.10.4497
Was running nightly build JSR-231 2006-03-02
Now running nightly build JSR-231 2006-03-10
was running:
java version “1.5.0_03”
Java™ 2 Runtime Environment, Standard Edition (build 1.5.0_03-b07)
Java HotSpot™ Client VM (build 1.5.0_03-b07, mixed mode)
now ‘running’:
java version “1.5.0”
Java™ 2 Runtime Environment, Standard Edition (build 1.5.0-b64)
Java HotSpot™ Client VM (build 1.5.0_06-b05, mixed mode, sharing)
But now the VM fails straight away for the jogl client (see below). The java server starts ok. I will roll back the JDK and try with just the update graphics driver and JOGL. I have to post now though in case of another crash 
Cheers Peter
An unexpected error has been detected by HotSpot Virtual Machine:
EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x6d76b5b6, pid=1332, tid=292
Java VM: Java HotSpot™ Client VM (1.5.0_06-b05 mixed mode, sharing)
Problematic frame:
V [jvm.dll+0x8b5b6]
An error report file with more information is saved as hs_err_pid1332.log
If you would like to submit a bug report, please visit:
http://java.sun.com/webapps/bugreport/crash.jsp
Continuing last post…
Ok, I just tried with the older JDK and it was okay for a few mins until I alt -tbbed back to send this post, Then I got:
An unexpected error has been detected by HotSpot Virtual Machine:
EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0xa0a6e800, pid=836, tid=1984
Java VM: Java HotSpot™ Client VM (1.5.0_03-b07 mixed mode)
Problematic frame:
C 0xa0a6e800
An error report file with more information is saved as hs_err_pid836.log
If you would like to submit a bug report, please visit:
http://java.sun.com/webapps/bugreport/crash.jsp
Could you please attach the hs_err log file to one of your posts?
As requested I had to rename it for the attach to work.
Cheers
Peter
The first crash looks very suspicious and unrelated to JOGL. Note that the crash occurs in Component.initIDs() which is called very early in JDK startup, definitely before JOGL has initialized. If this is reproducible with a stripped-down test case not involving JOGL then please file a bug with the Bug Tracker on java.sun.com, or post your test case here and I’ll file it. You may want to run with -Xcheck:jni to see if anything is caught.
The second crash again looks bad in that the stack is corrupted (or at least wasn’t walkable) and the error log is truncated around the top of the stack. I noted that that was generated with 5.0 update 3 – please continue to run with 5.0u6. We fix many bugs in every update release so running with an older version is not useful for the purpose of bug reporting.
After much testing, I think this issue is down to the motherboard and or the onboard graphics! I added an extra 512MB recently and without it the system seems to be stable again! Although I still don’t know why the updated JOGL or JDK made it soo much worse. I tried a new memory stick this weekend too, but with the extra 512MB I see crashes with opengl stuff, and without it I don’t. Very strange. Have emailed the MB company to see if it’s something wrong with the bios. (DFI 855GME-MGF if anyone else trips over this problem). So much for silent PC’s
I’ll post an update if I hear more.
Thanks.
Peter
update:
Well the system has been very stable with 1GB RAM until the moment I try OpenGL stuff and then it crashed with a couple of minutes.
So I just put a cheap NVidia card (GeForce6200) into the system and bingo, stability again!!!
So thats the crashing problem solved but it’s not the end of the story…
Please see new post ‘Two GLCanvas fighting for resources…’
Cheers
Peter
IIRC the Intel graphics shares memory with the main system. If your memory is slightly dodgy I wouldn’t be surprised if OpenGL shakes things up. You might want to try memtest86 www.memtest86.org and leave that soak testing. It usually finds problems. A while back I add some more RAM and had occasional BSDs every day until I ran memtest and found the ram was suspect. Different stick of ram solved the problem (both apparently identical spec tho different manufacturers).