33766 - Open In New Window Crash

Status: VERIFIED WORKSFORME

(SeaMonkey :: General, defect, P3)

Description

[Gervase Markham [:gerv]]

From Bugzilla Helper:
User-Agent: Mozilla/4.7 [en] (Win95; U)
BuildID:    20000328

If you go to the above URL, and do the following as quickly as possible:

Reproducible: Sometimes
Steps to Reproduce:
1. Right click www.microsoft.com | Open In New Window
2. Alt-Tab back to this window
3. Right click on www.excite.com/travel | Open In New Window
4. Alt-Tab back to this window

It very often crashes.

I get this maybe one time in two - fairly reproducible. I can't get it at all on 
20000327. 

MOZILLA caused an invalid page fault in
module KERNEL32.DLL at 014f:bff7926e.
Registers:
EAX=0a4ca388 CS=014f EIP=bff7926e EFLGS=00010206
EBX=0a4ca38c SS=0157 ESP=0068f8d4 EBP=0a1c6330
ECX=00002148 DS=0157 ESI=00000000 FS=0d37
EDX=2d2d6c6d ES=0157 EDI=006a0000 GS=0000
Bytes at CS:EIP:
89 6a 08 8b 2b 8b 00 89 45 04 68 00 02 00 00 51 
Stack dump:
0068f92c 0000111c 0a4cb4ac 0bab9da4 0a4ca384 78001436 006a0000 00000000 0a4cb4ac 
0000111c 0babcf10 0bab9da4 00000005 00000008 00000009 00000000 

And another:
MOZILLA caused an invalid page fault in
module KERNEL32.DLL at 014f:bff782af.
(I have the whole thing if you want it).

One time in three it locks the machine totally solid, without warning. Other 
times I can do stuff after I've dismissed the crash dialog (but not before).

I hate to file against Browser-General, but I'm hoping someone else can 
determine the component.

Gerv

Comment 1

[Gervase Markham [:gerv]]

Adding crash, smoketest, regression keywords. CC'ing granrose.

BTW, the above "don't get it at all" should read 20000328.

If it doesn't crash, try opening a few more windows. I have no reports of this 
yet on other platforms.

Gerv

Comment 2

[Donnie Cranford]

NT 4.0 SP 4, Tried bug and nothing happened, tried to reprocduce 4 times...no
go. I would say that this is not a blocker...proceed.

Comment 3

[John Morrison]

I have seen this happen, and there is a talkback at
    http://cyclone/reports/SingleIncidentInfo.cfm?dynamicBBID=7588274
However, the stack trace is non-existent. 

I wouldn't take this as a blocker based on my experience, as this was not easy 
to reproduce (for me). But this may depend on some other factors, such as win95 
vs. win98, network speed and cpu speed (i.e., "as quickly as possible").

Comment 4

[Gervase Markham [:gerv]]

Thank goodness I wasn't hallucinating... [e-mail] - have you been able 
to reproduce on yesterday's nightly? If not, great. We can narrow the blame :-)

Gerv

Comment 5

[Jon Granrose]

since this isn't impacting a major portion of the product, I'm removing 
smoketest blocker status.

Comment 6

[John Morrison]

No, I couldn't reproduce it with mar28 win98, but I'm not sure that is proof
of much -- I only hit it once, by chance for mar29. Gerv -- is this happening
to you with regularity (like 10 of 10, or 5 of 10, or ...). 

Comment 7

[Gervase Markham [:gerv]]

About one time in two, perhaps. I e-mailed my raw talkback data to 
[e-mail] (Namachivayam Thirumazhusai) before my PC crashed and 
Talkback along with it. The data is still there, but I don't know how to rerun 
Talkback to get it to retry the send; and anyway, I think our network config 
here prevents it getting through.

Gerv

Comment 8

[Asa Dotzler [:asa]]

[e-mail] (Asa Dotzler) taking ownership of Browser General bugs. 
 New QA Contact is [e-mail] (Joseph Elwell).  Sorry for the spam.

Comment 9

[Asa Dotzler [:asa]]

Are any of you still seeing this?  if so, any ideas about where it should go.

Comment 10

[Asa Dotzler [:asa]]

I am unable to reproduce this with win32 build from 4/19.  Are you guys still 
seeing this?

Comment 11

[Gervase Markham [:gerv]]

I won't be back on my W95 machine for a week; I say we close this one and I'll 
reopen if I find the problem still happening. I think it was probably a 
transient anyway.

Gerv

Comment 12

[Asa Dotzler [:asa]]

OK.  resolved WORKSFORME

Comment 13

[Asa Dotzler [:asa]]

Sorry for the spam.  New QA Contact for Browser General.  Thanks for your help
Joseph (good luck with the new job) and welcome aboard Doron Rosenberg

Comment 14

[Doron Rosenberg (IBM)]

verified