Closed Bug 333506 Opened 18 years ago Closed 15 years ago

[Mac only] Crash when running Java applets from Sun Site

Categories

(Firefox :: General, defect)

Product:
Firefox
Component:
General
Version:
1.5.0.x Branch
Platform:
PowerPC
macOS
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED FIXED

People

(Reporter: marcia, Unassigned)

References

(
URL
)

Details

(Keywords: crash)

Attachments

(2 files)

Stack backtrace, 1.5.0.2 candidate + 10.4.6 + J2SE 1.4.2
37.69 KB, text/plain
Details
Apple crash report when running UB on 10.3.9
40.10 KB, text/plain
Details 
Seen running the UB on Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8.0.2) Gecko/20060328 Firefox/1.5.0.2.

Every time I visit http://java.sun.com/applets/jdk/1.4/index.html and run one of the tests I crash. Filed TB17419942H.
 
From Java Console: Java(TM) Plug-in: Version 1.4.2_09
Using JRE version 1.4.2_09 Java HotSpot(TM) Client VM
User home directory = /Users/knous
Proxy Configuration: No proxy


----------------------------------------------------
c:   clear console window
f:   finalize objects on finalization queue
g:   garbage collect
h:   display this help message
l:   dump classloader list
m:   print memory usage
o:   trigger logging
p:   reload proxy configuration
q:   hide console
r:   reload policy configuration
s:   dump system properties
t:   dump thread list
v:   dump thread stack
x:   clear classloader cache
0-5: set trace level to <n>
----------------------------------------------------
MRJ Plugin for Mac OS X v1.0.1
[starting up Java Applet Security @ Mon Apr 10 17:10:18 PDT 2006]
AppletHolderFactory: exception: unknown protocol: about
java.net.MalformedURLException: unknown protocol: about
	at java.net.URL.<init>(URL.java:544)
	at java.net.URL.<init>(URL.java:434)
	at java.net.URL.<init>(URL.java:383)
	at jep.AppletHolderFactory.createAppletHolder(AppletHolderFactory.java:101)
AppletHolderFactory: exception: unknown protocol: about
java.net.MalformedURLException: unknown protocol: about
	at java.net.URL.<init>(URL.java:544)
	at java.net.URL.<init>(URL.java:434)
	at java.net.URL.<init>(URL.java:383)
	at jep.AppletHolderFactory.createAppletHolder(AppletHolderFactory.java:101)


Here is my list of extensions: Chatzilla, TinyURL, Performancing, Foxytunes, Answers, Separe, delicious, tabbrowser preferences, tab mix plus
Macia - does a PPC build on the same machine under the same configuration/profile run ok?

schrep: I get the same crash when I run the PPC only build from ftp://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/1.5.0.2-candidates/mac-ppc, which preed indicates is the PPC only build.

(In reply to comment #1)
> Marcia - does a PPC build on the same machine under the same
> configuration/profile run ok?
> 


Safari seems to handle sites from the Sun Java URL listed above fine.
URL: http://java.sun.com/applets/jdk/1.4/i...
TB17419942H

That stack is useless.  Can you get something out of Apple's CrashReporter?

"Every time" makes this sound like something's weird with your Java configuration.  I don't see any crashes on the Sun demos with either the universal or ppc-only candies on a genuine ppc.  I'm set up to use J2SE 5, I'll test 1.4 too.
Aha!  I do crash with J2SE 1.4.2 in use.

With Firefox 1.5.0.1 and J2SE 1.4.2, I'm also hanging when attempting to visit the Sun demos.  This didn't used to happen.  It may be a result of the recent 10.4.6 update?

Marcia, can you check 10.3?  How about earlier 10.4 releases?
Mark, your crash log shows that you're loading libraries from both
Java 1.4.2 and Java 1.5.0 (aka J2SE 5.0).  The one other case of this
that I'm familiar with also caused a crash:

https://bugzilla.mozilla.org/show_bug.cgi?id=330956#c14

That crash was a self-inflicted wound -- the reporter had messed with
his "CurrentJDK" soft link.

You haven't done that, have you? :-)

If not, then this may conceivably be a browser problem.

(By the way, I agree with Mark that, where available (as they are on
Mac OS X), crashreporterd crash logs are _far_ superior to Talkback
ones.  _Please_ submit a crashreporterd crash log if you're going to
report a crash here.)

Here is my Apple stack when I crash. I am fairly certain I did not do anything like what Steven describes regarding messing with me CurrentJDK softlink.
You had Sun/Apple's Java Console open when this crash happened, I
think.  If so, this is bug 331190.  There's a workaround -- use the
JEP's "poor man's Java console".  But this bug is rather ugly, I
admit.

The JEP (actually the MRJ Plugin JEP) implements its "poor man's
console" by writing all Java error output to the "Java Console.log"
file in the Library/Logs/ directory under your home directory (i.e. to
~/Library/Logs/Java Console.log").  You can use the Console utility to
view this file (even in real time, as it's being written to) by
choosing "Open" from its "File" menu and browsing to
~/Library/Logs/Java Console.log.

The "poor man's console" has one very significant advantage over
Sun/Apple's Java Console -- you don't lose its contents when the
browser crashes.  But the "poor man's console" doesn't have the
Sun/Apple console's key commands.

I've actually got this bug fixed in my current version.  But it'll be
a few days before I can put out a new JEP release that contains this
fix.


I didn't touch that link, but the mtime on it looks about right for something an installer may have done for me...

Pointing the link to 1.4.2 to match my JDK selection alleviates the crash.  That puts me back into the "can't reproduce" camp.
Since no-one's responded, I assume I'm right about Marcia's crash
being bug 331190.

Mark's crash is (then) completely unrelated.  I'd be interested to
hear, Mark, when/if you find out which installer changed your
"CurrentJDK" soft link.

> I've actually got this bug fixed in my current version.  But it'll
> be a few days before I can put out a new JEP release that contains
> this fix.

So ... would Mozilla.org like me to do a JEP release (a new "nightly")
in the next few days that contains the fix for bug 331190?

That's not my current plan.  Instead I intend to release a new JEP
"nightly" a few weeks from now, and have it also (if possible) include
fixes for three additional problems -- the LiveConnect side of bug
329457, bug 333003 and
http://sourceforge.net/tracker/index.php?func=detail&aid=1456457&group_id=107955&atid=649116.


Steven: I actually may have had it open, but if memory serves me correct the first time I tried to run a page with Java Applets the Console launched on its own.  I can verify that now on my machine.

(In reply to comment #11)
> Since no-one's responded, I assume I'm right about Marcia's crash
> being bug 331190.
> 
> Mark's crash is (then) completely unrelated.  I'd be interested to
> hear, Mark, when/if you find out which installer changed your
> "CurrentJDK" soft link.
> 

(In reply to comment #9)
> You had Sun/Apple's Java Console open when this crash happened, I
> think.  If so, this is bug 331190.  There's a workaround -- use the
> JEP's "poor man's Java console".  But this bug is rather ugly, I
> admit.
> 
> The JEP (actually the MRJ Plugin JEP) implements its "poor man's
> console" by writing all Java error output to the "Java Console.log"
> file in the Library/Logs/ directory under your home directory (i.e. to
> ~/Library/Logs/Java Console.log").  You can use the Console utility to
> view this file (even in real time, as it's being written to) by
> choosing "Open" from its "File" menu and browsing to
> ~/Library/Logs/Java Console.log.
> 
> The "poor man's console" has one very significant advantage over
> Sun/Apple's Java Console -- you don't lose its contents when the
> browser crashes.  But the "poor man's console" doesn't have the
> Sun/Apple console's key commands.
> 
> I've actually got this bug fixed in my current version.  But it'll be
> a few days before I can put out a new JEP release that contains this
> fix.
> 


If the Java Console has been "turned on", it does pop up whenever you
load a Java applet.  But by default it's off -- the setting is "Use
Java console" in the Java Control Panel.

Rather than have you rely on your memory, I'd like you to try an
experiment.  Try turning the Java Console off.  Load a bunch of
applets and see if you crash -- I expect that you won't.  Then try
turning it back on, and see if you start crashing again.  In
particular, observe if what triggers your crashes matches these
descriptions:

https://bugzilla.mozilla.org/show_bug.cgi?id=331190#c0
https://bugzilla.mozilla.org/show_bug.cgi?id=329457#c9


Steven,

Mozilla is just about to release 1.5.0.2.  The next opportunity to get an updated JEP into the wild would be 1.5.0.3 which will code freeze May 1:

http://wiki.mozilla.org/Firefox:1.5.0.3:Schedule

Best,

Mike


(In reply to comment #12)
> > I've actually got this bug fixed in my current version.  But it'll
> > be a few days before I can put out a new JEP release that contains
> > this fix.
> 
> So ... would Mozilla.org like me to do a JEP release (a new "nightly")
> in the next few days that contains the fix for bug 331190?
> 
> That's not my current plan.  Instead I intend to release a new JEP
> "nightly" a few weeks from now, and have it also (if possible) include
> fixes for three additional problems -- the LiveConnect side of bug
> 329457, bug 333003 and
> http://sourceforge.net/tracker/index.php?func=detail&aid=1456457&group_id=107955&atid=649116.
> 


> Mozilla is just about to release 1.5.0.2.  The next opportunity to
> get an updated JEP into the wild would be 1.5.0.3 which will code
> freeze May 1

"Into the wild"? :-)

But I know what you mean ... I think.

I will try to release a JEP 0.9.5+e no later than a week before May
1st -- with the expectation that, if no new problems are found, it
should make it into the 1.5.0.3 release.


Steven: I have tried various combinations trying to reproduce the crash, both with the Java Console on and off.  I have only succeeded in crashing one with the console on - in that instance I did press the delete key and did get instances of the Apple crash report.  I also followed the steps in the two bugs and was not able to reliably crash.  I am running 10.3.9 on the same machine with the same UB build: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8.0.2) Gecko/20060328 Firefox/1.5.0.2,

(In reply to comment #14)
> If the Java Console has been "turned on", it does pop up whenever you
> load a Java applet.  But by default it's off -- the setting is "Use
> Java console" in the Java Control Panel.
> 
> Rather than have you rely on your memory, I'd like you to try an
> experiment.  Try turning the Java Console off.  Load a bunch of
> applets and see if you crash -- I expect that you won't.  Then try
> turning it back on, and see if you start crashing again.  In
> particular, observe if what triggers your crashes matches these
> descriptions:
> 
> https://bugzilla.mozilla.org/show_bug.cgi?id=331190#c0
> https://bugzilla.mozilla.org/show_bug.cgi?id=329457#c9
> 


Marcia, thanks for your tests.  Since you didn't see crashes with the
Java Console off, and since I'm able to reproduce them reliably with
the Java Console on (and to generate crash logs basically identical to
the one you posted), I'm going to continue to assume that the crashes
you did see were due to bug 331190.


I've just released a new version (0.9.5+e) of the Java Embedding
Plugin that fixes bug 331190.  See bug 331190 comment 4.
Blocks: 353557
No longer blocks: 353557
Marking FIXED on the belief that bug 331190 fixed this.  Marcia, please reopen if you're still seeing this.
Status: NEW → RESOLVED
Closed: 15 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: