613598 - Frequent browser crashing during normal usage

Status: RESOLVED INVALID

(Firefox :: General, defect)

Description

[rokeefe1]

User-Agent:       Mozilla/5.0 (Windows NT 5.1; rv:2.0b7) Gecko/20100101 Firefox/4.0b7
Build Identifier: Mozilla/5.0 (Windows NT 5.1; rv:2.0b7) Gecko/20100101 Firefox/4.0b7

Starting with a fresh install of Firefox Beta 7 on a Windows XP x64 machine the browser is frequently crashing during normal usage.

Reproducible: Sometimes

Steps to Reproduce:
1. Use the browser normally and it will just crash at random.
2.
3.
Actual Results:  
Browser crashes

Expected Results:  
Browser operates normally.

Comment 1

[(mostly gone) XtC4UaLL [:xtc4uall]]

Can you list some CrashIDs from about:crashes?
https://support.mozilla.com/en-US/kb/Mozilla Crash Reporter#Viewing_crash_reports

Comment 2

[(mostly gone) XtC4UaLL [:xtc4uall]]

Arrg, use this Link for Instructions:
https://developer.mozilla.org/En/How_to_get_a_stacktrace_for_a_bug_report

Comment 3

[rokeefe1]

When the browser crashes it does not bring up the crash reporter.  Is there any way to get additional information if the crash reporter doesn't launch after a crash?

Comment 4

[rokeefe1]

One correction to the bug I submitted... I'm using an x86 version of XP, not x64.  The browser is running in the virtual pc windows xp image that microsoft provides if you wish to be able to reproduce the environment.  I don't know if the hardware acceleration features that were previously turned off in prior builds is causing the problem or not.

Comment 5

[(mostly gone) XtC4UaLL [:xtc4uall]]

(In reply to comment #3)
> When the browser crashes it does not bring up the crash reporter.  Is there any
> way to get additional information if the crash reporter doesn't launch after a
> crash?
https://developer.mozilla.org/en/How_to_get_a_stacktrace_with_WinDbg

Comment 6

[rokeefe1]

I got one crash today that actually caused the the Mozilla crash reporter to come up.  The crash ID is: bp-ab7d078c-5686-42db-b895-00cf92101122

I installed WinDbg.  I was able to capture one crash.  The browswer crashed just by going to www.hotmail.com and logging in.  I only had one tab open.  I don't know if I captured the information correctly, however.  Firefox didn't actually close when it threw the exception.  I selected Break from WinDbg, but it still stayed open.  I had to actually detach the debugger before the browser actually crashed. I've attached the log.

Also, just an FYI, to help eliminate possible causes of this problem I have turned hardware acceleration off.

Please let me know if there is any additional information that I can provide.

Comment 7

[rokeefe1]

Attachment: FF Debug log

Comment 8

[timeless]

MOZCRT19!fastzero_I+0x20:
78147fde 660f7f07        movdqa  xmmword ptr [edi],xmm0 ds:0023:0627d000=????????????????????????????????
0:000> |* ~* kp

.  0  Id: ee8.b04 Suspend: 1 Teb: 7ffde000 Unfrozen
Could not allocate memory for stack trace

Now that's an interesting failure. It looks like you ran out of memory.

Comment 9

[rokeefe1]

I noticed that too.  Is there anything I can try???  I have the virtual memory settings set to have Windows manage it.  I can now consistently crash the browser.  All I have to do is open it up, go to Gmail and log in.  The browser will crash the majority of the time.  Is the new JavaScript engine causing this???

Comment 10

[timeless]

i haven't spent enough time looking through windbg to see what sort of memstats stuff it has.

some memory stuff comes from our graphics layer behavior which changed for 4.0.

how much free space do you have on the drive w/ your swap file?

Comment 11

[rokeefe1]

I give this VPC 512Mb of memory.  There was nothing running in it other than the virus scanner.  The C drive shows 121 GB free on the C drive.  That's not accurate since the machine it is running on only has 20.5GB free.  I found a virtual memory setting messed up in the Virtual PC.  I fixed that.  I hope that is all it takes to fix this problem.  I'll watch it for a day or two.  If the problem seems to go away then I'll close the ticket.

Comment 12

[rokeefe1]

After changing some of the memory settings in Windows this issue appears to have gone away.  It appears that the root cause was a Windows configuration issue so I'm closing this ticket.