regular EXCEPTION_ACCESS_VIOLATIONs in jvm.dll

Haven't checked what the actual crash is caused by however on my W7 x64 workstation I am also having the latest EAP crashing relatively regularly... most disconcerting.

Andreas

Hello Oleg,

Please try to add the following to idea.exe.vmoptions:
-XX:-ReduceInitialCardMarks

See http://java.sun.com/javase/6/webnotes/6u18.html for more information.

I am expierencing regular EXCEPTION_ACCESS_VIOLATIONs in jvm.dll at
some random moments.

The VM crash logs start with the following text with some minor
changes from crush to crush:

===
#

1. A fatal error has been detected by the Java Runtime Environment:

#

1. EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x000000006dc87954,

pid=11740, tid=11552
#

1. JRE version: 6.0_18-b07

2. Java VM: Java HotSpot(TM) 64-Bit Server VM (16.0-b13 mixed mode

windows-amd64 )

1. Problematic frame:

2. V 

#

1. If you would like to submit a bug report, please visit:

2.   http://java.sun.com/webapps/bugreport/crash.jsp

#
---------------  T H R E A D  ---------------

Current thread (0x000000000045e800):  GCTaskThread [stack:
0x00000000351f0000,0x00000000352f0000]

siginfo: ExceptionCode=0xc0000005, reading address 0x00000000000000d9

===

The address of the problematic frame changes slightely and the siginfo
reading address changes but stays close to 0x0.

I have sent these crash logs to the supplied bug report address but it
is pretty useless.

There were similar problems some years ago
(http://www.jetbrains.net/devnet/message/5004875#5004875,
http://www.jetbrains.net/devnet/message/5048349#5048349,
http://youtrack.jetbrains.net/issue/IDEA-3444) that were related to
the JNI calls.

Has anybody seen something similar with the 9.0 builds?

For the record, i am on Windows 7 x64 and using JDK 1.6.0_18 x64.

---
Original message URL:
http://www.jetbrains.net/devnet/message/5258330#5258330

--
Dmitry Jemerov
Development Lead
JetBrains, Inc.
http://www.jetbrains.com/
"Develop with Pleasure!"

Dmitry, i am now trying to run Idea with -XX:-ReduceInitialCardMarks. Either it will help and i will not reply to this post anymore or i will return with new crash log. Hope for the former.

I was under the impression this was strictly a 32-bit issue. And heretofore, at least, only seen on Linux.


Randall Schulz

On Thu, 11 Mar 2010 17:57:08 +0000, Randall Schulz wrote:

I was under the impression this was strictly a 32-bit issue. And
heretofore, at least, only seen on Linux.

-XX:-ReduceInitialCardMarks certainly seems to have cured the problem I
was seeing on my 64-bit Linux box.  I've had no crashes at all in the
three weeks since I added it to my idea.vmoptions file.  They'd been
frustratingly common before that.

I meant to post a follow-up thanking you for suggesting it (it was you,
right?)...  So thank you :)

--
Mark Scott
mark@codebrewer.com

I may have been the first to suggest it here (in the IDEA forums), but I myself was put onto the problem and its solution by a person who frequents #scala and who keeps a close eye on HotSpot internals.


Randall Schulz

I just recently switched from Ubuntu Linux 09.10 64-bit to Windows 2008 R2 64-bit to run IntelliJ 9 since IntelliJ was crashing within 30-60 seconds on startup in Ubuntu.
Windows is working better, but it is crashing about once per hour with what looks like the same error mentioned in this thread.
I will will try the -XX:-ReduceInitialCardMarks and see if it fixes the problem.
Thanks -Alex

#
# A fatal error has been detected by the Java Runtime Environment:
#
#  EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x000000006dc87a7f, pid=4624, tid=4884
#
# JRE version: 6.0_18-b07
# Java VM: Java HotSpot(TM) 64-Bit Server VM (16.0-b13 mixed mode windows-amd64 )
# Problematic frame:
# V  [jvm.dll+0x3f7a7f]
#
# If you would like to submit a bug report, please visit:
#   http://java.sun.com/webapps/bugreport/crash.jsp
#

---------------  T H R E A D  ---------------

Current thread (0x0000000000384000):  GCTaskThread [stack: 0x0000000006290000,0x0000000006390000] [id=4884]

siginfo: ExceptionCode=0xc0000005, reading address 0x0000000000000018

I read the bug (http://bugs.sun.com/view_bug.do?bug_id=6896647) and it has a linked bug (http://bugs.sun.com/view_bug.do?bug_id=2186975) for java 6 update 19 which says it was fixed in 6u19(b02). Since released 6u19 says build b04, it should be fixed.


But I checked java 6 update 19 release notes, and I don't see the bugfix mentioned there.

http://java.sun.com/javase/6/webnotes/6u19.html

I will try java 6 update 19 first to see if that fixes the problem, and if that doesn't work I'll use the workaround.

Thanks,
-Alex

What you share is all my need and we all hope you can share us more info further time. And I like shopping online, for the coming Christmas Day ( http://www.overgift.com/ ) want to prepare beauty http://www.overgift.com** in that day, and search online, and find some info. More, used to buying kinds shoes online shop and yesterday I found unusual shoes called http://www.vibrambazaar.com/** which really tract my eyes. Then no words to say I bought a pair http://www.vibrambazaar.com/** not for myself, sent this as a gift for my father who love sport. So if you want to live a happy life you should concern more things around you, buy Christmas tree on http://www.overgift.com/** and http://www.vibrambazaar.com/** online shop. Come and more choices.

---
Original message URL: http://devnet.jetbrains.net/message/5278477#5278477

A round http://www.pandora-bracelets-australia.com/ cut has the http://www.pandora-bracelets-australia.com/ greatest http://www.pandora-bracelets-australia.com/pandora-charms/pandora-bangles/ number of facets, http://www.pandora-bracelets-australia.com/ and if properly proportioned, will appear http://www.pandora-bracelets-australia.com/ the most brilliant. An emerald http://www.pandora-bracelets-australia.com/pandora-earrings/ cut diamond has the fewest facets, http://www.pandora-bracelets-australia.com/pandora-beads/ so larger stones are need to make an impressive  display.Clarity: All diamonds, except the most rare, http://www.pandora-bracelets-australia.com/pandora-necklaces/ has some kind of mark,nick, or cloudiness to them http://www.pandora-bracelets-australia.com/ caused by flaws. http://www.pandora-bracelets-australia.com/ The clarity of the diamond is measured by an alphabetical and  numerical scale with the diamonds in the “F” category http://www.pandora-bracelets-australia.com/pandora-beads/ being completely http://www.pandora-bracelets-australia.com/pandora-beads/ flawless.

---
Original message URL: http://devnet.jetbrains.net/message/5279124#5279124