[see evaluation - on Solaris at least,
reproducing this leads to a segv in Java2d native code.]

Name: mc57594 Date: 09/30/99
Verified on Solaris. With
java version "1.3.0"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.3.0-I)
Java HotSpot (TM) Client VM (build 1.3-I, interpreted mode)

I get the exception:

#
# HotSpot Virtual Machine Error, Unexpected Signal 11
#
# Error ID: 4F533F534F4C415249530E4350500790 01
#
[chamness]
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-==-=-=--

I sometimes get this error:

#
# HotSpot Virtual Machine Error, EXCEPTION_ACCESS_VIOLATION
#
# Error ID: 4F533F57494E13120E43505002B0
#

abnormal program termination


It is not very reproducable, but I think it is caused by something in the Graphics2D image rendering code. My program is over 60000 lines long so I decided not to post it here, I can send it to you if you want it.

===============================================
Date: Thu, 30 Sep 1999 20:46:05 -0600
From: David Daney <###@###.###>
To: chamness <###@###.###>
Subject: Re: (Review ID: 95994) EXCEPTION_ACCESS_VIOLATION # Error ID: 4F533F57494E13120E43505002B0

Here is the test case. Since it is a problem in the VM you will not be
needing any source code, the raw class files should be sufficient.

To run it set your CLASSPATH to be the attached pixtran.jar then run it
with the following command:

java Viewer

Using the jdk1.3beta VM running on Windows98 (I have 128MB ram and a
24bpp display)

It should open a Window on the screen. Next select from the menu bar
File->Open File ... and select the castle6.tif file which is also
attached.

Now comes the impressive part. Scrollbars should appear in the window
but it will probably remain white with no images in it.

Scroll the scroll bars around some by both dragging and clicking in the
area above and below the bar part. You should get alot of messages
printing to the console when you do this.

Now from the Scaling menu choose 35%. If this does not crash it scroll
it around somemore and try different Orientations and Scaling settings
or resizing the window.

I can usually produce the crash in less the 15 seconds by doing this.

If you run with java -classic Viewer you should not get the crash.

Let me know if this helps any. There is a lot of drawing going on here
so I suspect it may be an issue with the Graphics2D image rendering code
interacting with the HotSpot VM. (If you can find out why nothing shows
on the screen let me know).


(Review ID: 95994)
======================================================================