216881 - crash upon pasting (Ctrl+v) into new composer window

Status: RESOLVED INVALID

(SeaMonkey :: Composer, defect)

Description

[hugo van woerkom]

User-Agent:       Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.5a) Gecko/20030820
Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.5a) Gecko/20030820

Open the URL above. Select all. Copy. Open new composer window. Paste. Boom!

Reproducible: Always

Steps to Reproduce:
1.Open URL mentioned above.
2.Edit->Select All
3.Edit->Copy
4.Open new composer window
5.Ctrl+v

Actual Results:  
Mozilla dies and leaves this message in a text window, when invoked from there:
"pure virtual method called"

Expected Results:  
Shown the page pasted onto the new composer window

This is happening on 1.5a that I pulled from MOZILLA_1_5a_RELEASE yesterday.
It does not happen on Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4)
Gecko/20030819 which is pulled from MOZILLA_1_4_BRANCH.
I am pulling MOZILLA_1_4_RELEASE and we will see if it happens there.

Comment 1

[Mats Palmgren (inactive)]

*** Bug 216882 has been marked as a duplicate of this bug. ***

Comment 2

[Hermann Schwab]

wfm Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.5b) Gecko/20030820
so it is a linux only bug?

Comment 3

[hugo van woerkom]

Actually, no need to use my URL from the bug description:
you cannot paste anything in that new composer window.

Comment 4

[Adam Hauner]

Could you provide stacktrace or (if you can retest it with actual nightbuild)
TalkBack incident ID?

Comment 5

[hugo van woerkom]

Just pulled cvs MOZILLA_1_4_RELEASE and the bug does not happen there, I guess
we knew that.

Comment 6

[hugo van woerkom]

I'm about to compile the source again because I compiled with gcc 2.95 before
and now the java plugin (3.2) doesn't match.
What option(s) do you enable to get a stack trace?
Or what do you do to get a stack trace? gdm?

Comment 7

[Andrew Schultz]

> and now the java plugin (3.2) doesn't match.

just use the plugin in the ns610 directory (instead of ns610-gcc32)

> What option(s) do you enable to get a stack trace?

don't include --disable-debug or
include --enable-optimize="-g -foo"  (-foo = -O or whatever other optimization
you want)

> Or what do you do to get a stack trace? gdm?

gdm (Gnome Display Manager) won't help.  try gdb (GNU DeBugger).  or
mozilla -g -d gdb

see http://www.mozilla.org/unix/debugging-faq.html

pulling mozilla 1.5a isn't too useful for testing/debugging (too old).  pull
trunk instead.

Comment 8

[Olivier Cahagne]

wfm debug build 20030819 on Linux (gcc 3.2, default optimization flag -O). No
assertions.
Make sure you can reproduce this with a fresh profile.

Comment 9

[hugo van woerkom]

I installed Mozilla 1.5b Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.5b)
Gecko/20030821 from the sea.targz of the latest trunk.
It does not fail.
I will now recompile the Xft enabled 1.5a release cvs, even though too old.
Sorry for trying to debug with the gnome display manager, must be the reason I
don't favor Gnome ;-b

Comment 10

[hugo van woerkom]

Bug is gone. I apologize for having wasted you guys' time. I recompiled the same
source (MOZILLA_1_5a_RELEASE) that I did get the bug on, using the same options
but with gcc 3.2. The problem is gone.