Closed Bug 337438 Opened 19 years ago Closed 13 years ago

java error message INTERNAL ERROR on Browser End: Did not understand spontaneous code 20

Categories

(Core Graveyard :: Java: OJI, defect)

1.8 Branch
x86
Linux
defect
Not set
critical

Tracking

(Not tracked)

RESOLVED INCOMPLETE

People

(Reporter: u123541, Unassigned)

References

()

Details

(Keywords: crash)

Attachments

(1 file)

User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.0.2) Gecko/20060308 Firefox/1.5.0.2 Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.0.2) Gecko/20060308 Firefox/1.5.0.2 Not sure if this site is the source of the problem, or if it's due to random wireless link drops; but this has happened several times today (thanks for crash session recovery)... The console where I started FF gives these messages at the time of the crashes: INTERNAL ERROR on Browser End: Did not understand spontaneous code 20 System error?:: Success Reproducible: Didn't try Steps to Reproduce: Unknown as problem is random and only started today -- my system is heavily loaded and was paging a lot at the time, though it pages quite a bit normally (768MB).
Assignee: nobody → yuanyi21
Component: General → Java: OJI
Product: Firefox → Core
QA Contact: general → zhayupeng
Summary: Crash → java gave a random error message to a confused user
Version: unspecified → 1.8 Branch
Do you have any Talkback IDs? (http://kb.mozillazine.org/Talkback)
No; but I can now reproduce it... I have several tabs open to several pages such as this one: http://rodweb02.co.wake.nc.us/books/genext/genextimages.asp?docid=100543443 More pages can be opened by searching here: http://rodweb01.co.wake.nc.us/books/genext/genextsearch.asp and entering "swans mill" in Grantee field and selecting BOTH in Grantee Type. On the results page, I opened several images in new tabs (right-click on "X"). Now, if I try to close one of these image pages, FF crashes. It seemed random at first; but is now reproducible -- however, I've been using the crash restore... on next crash, I'll try a clean restart of FF.
OK... sessions restored after a crash will continue to crash immediately upon tab close. After a fresh restart of FF, the crash is no longer easily reproducible.
Can you reproduce this in safe mode? (http://kb.mozillazine.org/Safe_Mode) If you're running an official Mozilla build of Firefox, you should have some Talkback IDs. (http://kb.mozillazine.org/Talkback#Linux)
"If you think that it might be installed, you can search for "talkback" and see if any results come up." Can this be any more vague...? :^) Can I get a better clue, as in 'where'..? OK... found /usr/bin/firefox/extensions/talkback@mozilla.org/components/talkback/*. While I've reported a number of problems via talkback (now that I realize that's how those were done :) , these crashes never produced the talkback window... just sudden death with the 2 lines of output on the console. So far, after a clean restart, rather than crash recovery, I still haven't triggered it... Once I can reproduce it again, if I can restart in safe-mode plus crash recovery, it might reproduce... here's hoping.
Just managed to reproduce the crash... though only after I used the java app's Save to save one of the tif images. This reminds me... all of the crashes I've had over the past weeks, and which were reported via the talkback screen, were all a direct result of trying to do a "Save as..." (coincidence?). Right now, the FF session that is the focus of this bug, is hung... I started it it in safe-mode, selected restore crashed sessions; but it ignored this. So I started opening Deed images; when I switched to one of those tabs, FF hung with the Java "sun" sitting there with its progress bar at 100% and my CPU pinned at 100%: 1061 pierre 25 0 110m 32m 15m R 87.2 4.3 7:06.80 /usr/bin/firefox/firefox-bin -safe-mode strace shows it spinning on this: select(49, [48], NULL, NULL, {0, 0}) = 0 (Timeout) The only parts of this window that are being re-displayed are the Java segment and that app's Search, etc buttons... weird... kill'ed it via the Close button and "non-responsive app..." dialog -- 3 processes left, killed those. OK... fired up fine this time... opened a couple of Deed images, closed one and it closed fine. Then, I opened another Deed image and did a Save (Java app's Save) on the first image. As soon as I closed the 2nd image's tab, FF crashed with: INTERNAL ERROR on Browser End: Did not understand spontaneous code 20 System error?:: Success [6]+ Exit 255 /usr/bin/firefox/firefox -safe-mode and no talkback dialog... What next?
upgraded to 1.5.0.3 and still have the same crash... Also, if I try loading the images into separate tabs in quick succession, I can beat the java app's "old signature" dialog and end up with the signature dialog overlapped with the right-click menu and only the mouse appearing to work (no clicks). I switched to a virtual console, and killed FF to restore access to my desktop. I probably should file a "race condition" bug; but I'll document here until I get a better handle on that one...
the problem you're having is with java. fix your java.
To what future version...? :} libjavaplugin_oji.so -> /usr/java/jre1.5.0_06/plugin/i386/ns7/libjavaplugin_oji.so
(In reply to comment #9) > To what future version...? :} > libjavaplugin_oji.so -> > /usr/java/jre1.5.0_06/plugin/i386/ns7/libjavaplugin_oji.soP Pierre, how did you do with the latest version? As of June the latest is 1.5.0_07.
Keywords: crash
Blocks: 353557
Summary: java gave a random error message to a confused user → java error message INTERNAL ERROR on Browser End: Did not understand spontaneous code 20
No longer blocks: 353557
Assignee: yuanyi21 → nobody
QA Contact: zhayupeng → java.oji
Just tried and in FF3 it's horrendous... (I'll upload photo of screen) When I opened the second search result, all I had left was mouse cursor movement; all keyboard and mouse clicks were unresponsive. Logged in from another system and killed FF3 which restored everything... without a second system to log in from, the system would need rebooting. HTH.
only mouse cursor was operational when this photo was taken.
I think this is the same as this bug: https://bugzilla.mozilla.org/show_bug.cgi?id=500772 I confirm it happens in late builds of Firefox trunk to-day (July 1st and about a week earlier), I also believe it is not only Sun's java to blame it also happens with this plugin: libgcjwebplugin.so (The GCJ Web Browser Plugin executes Java applets)
Product: Core → Core Graveyard
OJI has been discontinued and Java now runs out-of-process, so this stuff has probably changed a lot. This bug has no info about current software versions, please file a new bug for new issues or reopen this one with current info, including a crash signature if it still happens and move it to a component outside of graveyard.
Status: UNCONFIRMED → RESOLVED
Closed: 13 years ago
Resolution: --- → INCOMPLETE
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: