325987 - Unexpected Crash (Java refcounting bugs?)

Status: VERIFIED WORKSFORME

(Camino Graveyard :: General, defect)

Description

[ronald.gold]

User-Agent:       Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8.0.1) Gecko/20060203 Camino/1.0rc1
Build Identifier: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8.0.1) Gecko/20060203 Camino/1.0rc1

Camino 1.0rc1 unexpectedly crashed while downloading data from http://animal.discovery.com/cams/pandavidr.html

No problems with downloading webcam data from this site in the past with previous editions of Camino (both 1.0b1 and G4 Optimized build of 1.0b1).

Reproducible: Sometimes




Crash Report will be attached.

Comment 1

[ronald.gold]

Attachment: Crash Report

Comment 2

[Samuel Sidler (old account; do not CC)]

Ronald, that page loads fine for me. Can you repeat the crash?

That said, the video didn't load at all.

(Note for others, RealPlayer is required to view the video.)

Comment 3

[ronald.gold]

The page loaded normally. And I also got the Real Player notice, even though I have Real Player installed. The crash happened as the site began to download the video. Subsequently when I tried to access the videocam (of panda cub at National Zoo), I got pop-up message that the link was out-dated or down, but no crashes occurred. I was able to access other videocam at this site without any difficulty.

Comment 4

[Mike Pinkerton (not reading bugmail)]

if you look at the stack, it's from a system notification about a disk change, which caused some prefs to change. It doesn't appear to have anything to do with the page or movie in question.

Any ideas what's triggering a pref change when a drive is mounted/unmounted?

Comment 5

[Simon Fraser [no longer active]]

Maybe this is one of those bugs caused by Java (random Obj-C extre release or soemthing)?

Comment 6

[ronald.gold]

Attachment: Crash Report

Comment 7

[Simon Fraser [no longer active]]

Both crashes look like refcounting bugs related to Java.

Comment 8

[ronald.gold]

Attachment: Crash Log

Comment 9

[Steven Michaud [:smichaud] (Retired)]

The Java connection is very tenuous -- all I see is that the JVM was started
(because a Java applet was loaded) at some point before both of these crashes
happened.

On a possibly related note, I've noticed lots of crashes in Camino (1.0b2 and
later) in methods of the NSUserDefaults class, all in OS X 10.4.4.  Once I'm
able to discern a pattern, I'll open a bug here.  (I haven't yet tried 10.4.5
-- I'm holding off while I investigate the NSUserDefaults crashes.)

Comment 10

[Simon Fraser [no longer active]]

ronald: can you turn off Java, and see if you still crash? Also, if you can get a repeatable caes that crashes, that would be very interesting to us.

Comment 11

[ronald.gold]

The most unexpectod feature of all of these "Unexpected Crashes" is that they are not repeatable. For example,I have searched five times on the Bug List without causing a crash. Java on or off doesn't seem to make a difference. I will let you know if and when anything repeats.

Comment 12

[ronald.gold]

Attachment: Crash log

Comment 13

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

I still see JVM stuff in that last crash log.  Have you really disabled Java?

Comment 14

[Steven Michaud [:smichaud] (Retired)]

The JVM keeps running even after Java has been disabled -- though (unless
something is wrong) neither it nor the Java Embedding Plugin ever do anything
more than sit and wait for events that will never happen.

However, there's a (small) possibility that the JVM did something while it's
running that triggers a crash a _long_ time later, even after Java has been
disabled.

To be absolutely sure a given crash has nothing to do with Java, you need to
disable Java and restart the browser.

Comment 15

[Steven Michaud [:smichaud] (Retired)]

> However, there's a (small) possibility that the JVM did something while it's
> running that triggers a crash a _long_ time later, even after Java has been
> disabled.

I should rephrase this to make it clearer:

However, there's a (small) possibility that the JVM did something while it was
enabled that triggers a crash a _long_ time later, even after it was disabled.

Comment 16

[Steven Michaud [:smichaud] (Retired)]

It's possible (though unlikely) that the crashes reported here are related to
ones that I reported at bug 326195.

I just released a new version (0.9.5+d) of the Java Embedding Plugin that
fixes the crashes I reported at bug 326195 (which seem to be caused by Apple
bugs, and which I found a way to work around).

http://javaplugin.sourceforge.net/

Reporter:  Please try my new JEP version, and let us know if it makes any
difference to the problem you reported here (i.e. the crashes).

To use the new JEP version you will need to:

1) Install it in /Library/Internet Plug-Ins/ (according to the instructions in
   the JEP Readme).
2) Remove the old JEP files (JavaEmbeddingPlugin.bundle and MRJPlugin.plugin)
   from the Contents/MacOS/plugins/ directory of your Mozilla.org browser(s).

Comment 17

[ronald.gold]

(In reply to comment #16)
> It's possible (though unlikely) that the crashes reported here are related to
> ones that I reported at bug 326195.
> 
> I just released a new version (0.9.5+d) of the Java Embedding Plugin that
> fixes the crashes I reported at bug 326195 (which seem to be caused by Apple
> bugs, and which I found a way to work around).
> 
> http://javaplugin.sourceforge.net/
> 
> Reporter:  Please try my new JEP version, and let us know if it makes any
> difference to the problem you reported here (i.e. the crashes).
> 
> To use the new JEP version you will need to:
> 
> 1) Install it in /Library/Internet Plug-Ins/ (according to the instructions in
>    the JEP Readme).
> 2) Remove the old JEP files (JavaEmbeddingPlugin.bundle and MRJPlugin.plugin)
>    from the Contents/MacOS/plugins/ directory of your Mozilla.org browser(s).
> 

Holiday caused delay in following above instructions. Have done both 1 and 2 above. Will report if crashes recur. 

Comment 18

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

Ronald, has this been OK since you upgraded the JEP version?

Comment 19

[ronald.gold]

(In reply to comment #18)
> Ronald, has this been OK since you upgraded the JEP version?
> 

Since upgrading JEP on !7 March, I have not had any crashes. Does this help pinpoint source of problem?

Comment 20

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

I'm tempted to say that in the latest JEP, Steven fixed whatever was causing these odd crashes, and that we can close this bug WFM.  

I'm not sure how frequently you were experiencing these particular crashes in the past, so why don't we wait until the end of the week to be sure.

Comment 21

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

Closing WFM per discussion in the last few comments.