139698 - Crash when resizing window containg directly loaded flash

Status: VERIFIED WORKSFORME

(External Software Affecting Firefox Graveyard :: Flash (Adobe), defect, P2)

Description

[Sitsofe Wheeler]

Steps to reproduce:
1. Go to http://www.groovechamber.com/sheeponcaffeine.swf
2. Resize the browser window a few times

Expected:
Flash animation to resize

Result:
Browser area may turn a light yellow, becomes unresponsive does not redraw
correctly and results in a crash.

Talkback 
TB5548819G
20020421 Linux

Comment 1

[Rene Pronk]

I can't reproduce this on RC1,WinNT4,Flash 5.0 r41.

Comment 2

[Erwan Hamon]

WFM Mozilla/5.0 (Windows; U; WinNT4.0; en-US; rv:1.0rc1) Gecko/20020417

Comment 3

[David Gerard]

WFM Win2k, 2002042206-1.0 branch.

Comment 4

[shrirang khanzode]

oh yeah..oh yeah, this does crash .Play with the sheep for a while, u guys ! 
10-15 resizes randomly(left/right/top/bottom) helped me reproduce this crash. 
talkback coming up...

Comment 5

[shrirang khanzode]

ok, talkback's not woken up yet. can't seem to find the trace..
anywho, another way of reproducing this easily is by resizing the window just 
once with the mouse and then clicking the 'resize' button on top right of the 
window, just toggle it 5-7 times and if u are lucky enough u might see the 
crash.

Comment 6

[David Gerard]

Nope, just doesn't seem to want to crash for me. Spent about five minutes
changing its size with the mouse and toggling the resize button.

Heck, it doesn't even seem to be leaking memory when I do so. Inconthievable!
:-)

Comment 7

[shrirang khanzode]

couldn't repro on windows ( earlier  crash was on linux branch 0423)

Comment 8

[rubydoo123]

I hate these linux only things ... off to serge

Comment 9

[serge (gone)]

it does exit on 
Gdk-ERROR **: BadValue (integer parameter out of range for operation)
  serial 2120323 error_code 2 request_code 148 minor_code 3

the stack trace does not look good:
(gdb) bt
#0  exit (status=1) at exit.c:40
#1  0x4045d533 in gdk_x_error () from /usr/lib/libgdk-1.2.so.0
#2  0x4050bd2e in _XError () from /usr/X11R6/lib/libX11.so.6
#3  0x40509552 in _XReadEvents () from /usr/X11R6/lib/libX11.so.6
#4  0x404f5b7b in XIfEvent () from /usr/X11R6/lib/libX11.so.6
#5  0x43547c0b in object.2 () from /u/serge/plugins/linux/libflashplayer5.so
#6  0x00000635 in ?? () at eval.c:41
#7  0x00000637 in ?? 
(Cannot access memory at address 0x4535746
(gdb)

Comment 10

[Sitsofe Wheeler]

Ok this bug is a bit more difficult to reproduce with Flash 5.0 r48 (previously
I was testing with 5.0 r47) but if you are dogged enough a crash eventually occurs.

Comment 11

[serge (gone)]

BTW: Flash 4.x seems more stable, at least I was unable to crash mozilla
on test case url with with old flash installed

Comment 12

[rubydoo123]

me

Comment 13

[Markus Multrus]

*** Bug 161070 has been marked as a duplicate of this bug. ***

Comment 14

[Daniel Schroeter]

Also crash with:
Shockwave Flash 5.0 r51

With the URL
http://home.t-online.de/home/webspace1/misc/cow.swf
it works very good. I didn't try the other URL. 

And the Talkback-IDs:
TB12700490G
TB12696590M

Comment 15

[Sitsofe Wheeler]

Hurah! This appears to be fixed with the shockwave flash 6 beta plugin
(www.macromedia.com/software/flashplayer/special/beta/) on my Moz 1.0
installation (I resized for about 30 seconds without a crash). Things would
temporarily become slow but there was no crash and eventually the updates caught
up. I would just like to take this moment to thank the folks at macromedia for
the improvements they have made to the latest flash. Hopefully the second class
treatment of Linux by flash is at an end...

Comment 16

[shrirang khanzode]

wow, this release has fixed quite some bugs !

Comment 17

[Ed Millard]

> wow, this release has fixed quite some bugs ! 
 
I'm hoping all the Flash related bugs are fixed in Flash 6, with the exception 
of new Flash 6 bugs and bugs involving focus, losing focus using the scroll 
wheel in particular.  The focus issue is not easily solved at the moment.  
 
If you have Flash bugs that are not fixed please make sure I'm cc'ed on them 
and make note that they persist in Flash 6 so I can investigate. We have very 
limited time left before the final beta. 

Comment 18

[Sitsofe Wheeler]

Verifying.

Comment 19

[Sitsofe Wheeler]

Er sorry for the spam but shouldn't this have been marked fixed rather than
worksforme?

Comment 20

[BMO Automation]

Version and milestone values are being reset to defaults as part of product refactoring.