Closed Bug 257374 Opened 22 years ago Closed 17 years ago

Crash apparently when Java applet throws ArrayIndexOutOfBoundsException while not displayed on screen.

Categories

(Core Graveyard :: Java: OJI, defect)

PowerPC
macOS
defect
Not set
critical

Tracking

(Not tracked)

RESOLVED INCOMPLETE

People

(Reporter: gareth.randall, Assigned: alfred.peng)

References

()

Details

(Keywords: crash, Whiteboard: [needs simple Java applet created for retesting])

User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.2) Gecko/20040803 Build Identifier: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.2) Gecko/20040803 There is an applet (scroller.class) loaded by the supplied URL which seems to be able to crash the browser on its own. I conclude this because, when onscreen, the applet will display ArrayIndexOutOfBoundsException in the Mozilla status bar about two minutes after first loading the page. However, if the applet is not onscreen because it is another tab, the browser crashes after what seems to be about the same time delay. I think once it crashed Mozilla even though it was onscreen. My JRE is: java version "1.4.2_05" Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2_05-141) Java HotSpot(TM) Client VM (build 1.4.2-38, mixed mode) I am hoping this is not just because of a transient network connection (e.g. the applet cannot download data that it wants) so that you can see it happen too. (I will try to download the applet itself immediately after posting this report.) Reproducible: Always Steps to Reproduce: 1. Launch Mozilla. (I subsequently close the Mail client, should this be relevant.) 2. Open the URL http://www.nationalgrid.com/uk/ 3. Open another tab so that the Java applet is not on screen. 4. Wait something like 2 minutes (I inaccurately timed it at 130 secs.) Then the crash occurs. Actual Results: Crash resulting in both Apple bug submitter and Talkback coming up. Expected Results: Handled the Java exception (if this is what it actually is) without crashing. I have several Talkback crashes uploaded to you under the e-mail address: gareth.randall@wunger.com Feel free to examine them.
The applet is being loaded with the following parameters: <td width="400"><applet code="scroller.class" codebase="scroller" width="400" height="28" id="scroller"> <param name="feeder" value="scroller_ukhome.txt" /> <param name="font" value="name=Arial, style=BOLD, size=12" > <param name="direction" value="0" /> <param name="bg" value="r=0, g=102, b=153" /> <param name="fg" value="r=255, g=255, b=255"> <param name="hg" value="r=0, g=204, b=255"> <param name="fps" value="200" /> <param name="feedtime" value="60000" /> <param name="stopat" value="25" /> <param name="feedlength" value="10" /> <param name="stillfor" value="500" /> <param name="launchon" value="_top"> <param name="changecursor" value="1" /> </applet></td><!--news ticker end marker--> (I can't seem to download the applet itself yet. Maybe it is a common one in which case the source will be out there?)
-> oji
Assignee: blackconnect → kyle.yuan
Component: Java-Implemented Plugins → Java: OJI
QA Contact: avm
no crash with Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8a3) Gecko/20040828 scroller can be found at http://www.nationalgrid.com/uk/scroller/scroller.class Talkbacks 685194: SIGILL: Illegal Instruction: (signal 4) 685221,685222: SIGSEGV: Segmentation Violation: (signal 11) + stacktrace http://talkback-public.mozilla.org/talkback/fastfind.jsp?search=2&type=iid&id=685222 http://talkback-public.mozilla.org/talkback/fastfind.jsp?search=2&type=iid&id=685221 http://talkback-public.mozilla.org/talkback/fastfind.jsp?search=2&type=iid&id=685194
Keywords: helpwanted
Assignee: yuanyi21 → pete.zha
I've decompiled some of the Java classes on the website concerned and it is clear that their author has done some very amateur work with threading, including not synchronising on an object when calling methods in it, and calling AWT routines from outside of the event dispatcher thread! While the crashing classes are hopelessly wrong :-), it would be nice for Mozilla to be able to survive them. I used the Jad tool to decompile (http://kpdus.tripod.com/jad.html)
Still an issue on the latest release: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.11) Gecko/20050727 The original website have commented out the code that runs the applet, but if you download the page and replace their applet HTML line with one that specifies the codebase fully, then the crashing effect is still there: <applet code="scroller.class" codebase="http://www.nationalgrid.com/uk/scroller" width="400" height="28" id= "scroller">
mass reassign to Alfred
Assignee: zhayupeng → alfred.peng
(In reply to comment #3) > ... > scroller can be found at http://www.nationalgrid.com/uk/scroller/scroller.class > > Talkbacks > 685194: SIGILL: Illegal Instruction: (signal 4) > 685221,685222: SIGSEGV: Segmentation Violation: (signal 11) + stacktrace > http://talkback-public.mozilla.org/talkback/fastfind.jsp?search=2&type=iid&id=685222 > http://talkback-public.mozilla.org/talkback/fastfind.jsp?search=2&type=iid&id=685221 > http://talkback-public.mozilla.org/talkback/fastfind.jsp?search=2&type=iid&id=685194 Triage: Applet and talkback reports are no longer available.
Whiteboard: [needs simple Java applet created for retesting]
As long as no-one can supply further information or a working testcase we will close the bug as incomplete. Feel free to reopen when you have the requested information.
Severity: major → critical
Status: UNCONFIRMED → RESOLVED
Closed: 17 years ago
Keywords: crash
Resolution: --- → INCOMPLETE
Product: Core → Core Graveyard
You need to log in before you can comment on or make changes to this bug.