crash after paste, reproducible always (for me)

do you have other protocols enabled? if so, does this crash if you paste into a conversation using a different account?

I don't use other protocols.
Tried IRC as I know it well, same crash on first try.

(Singoff message visible for others: "*** Signoff: PedX (Remote closed the connection)" ... just as I anticipated)

Thanks for your bug report. Could you please try to obtain a
backtrace by following the instructions on
       http://wiki.ubuntu.com/DebuggingProgramCrash.
This will greatly aid us in tracking down your problem.

Doesn't it crash elsewhere too, just my box?

Attaching backtrace file.

Your are the only one who opened a bug about that and copy to gaim works fine on my box.

The crash log has a:

"X Error: BadDevice, invalid or uninitialized input device 154
  Major opcode: 143
  Minor opcode: 3
  Resource id: 0x0
Failed to open device"

It looks like some xorg device is not correctly configured. Could you get a backtrace with libgtk2.0-0-dbg installed? If you want to try if gaim 2.0beta5 fixes that problem you can install the package from http://people.ubuntu.com/~seb128/gaim-edgy

Going wild in ideas (no facts involved):

I saw a bug (maybe closed one?) about crash with "empty" paste, I think these two may be related.

And some facts at last:

BTW, did you notice it's about Paste of result from SpeedCrunch?
It's important, because when I just select text from this page, I can paste it into gaim without crash.

Other way around I can paste result from SpeedCrunch into kopete (both in Gnome and in KDE), so maybe SpeedCrunch does put some weird data on clipboard which gaim can't handle, but kopete can.

Can somebody try it with SpeedCrunch (default kubuntu calculator), if it's still my box only?

Xorg Input device errors:

I saw something about wacom in my xorg.conf, but I don't use any tablet .. I'm not so experienced with linux to nail it down immediately, but I will investigate eventually how to get rid of those messages in correct way. But I think it may be connected to non-existent wacom tablet defined. And maybe not.
Those 2 XOrg errors are shown every time I run and X app, it's not gaim-only.

right, happens when using speedcrunch, gedit has no such problem with it

Debug backtrace:

 #0 paste_received_cb (clipboard=0x8587248, selection_data=0xbf8063f4,
     data=0x82089d8) at ../../gtk/gtkimhtml.c:1092
  text = 0x0
  imhtml = (GtkIMHtml *) 0x82089d8
 #1 0xb7d8f00b in selection_received (widget=0x85b8c60,
     selection_data=0xbf8063f4, time=2158134343) at gtkclipboard.c:854
  request_info = (RequestContentsInfo *) 0x83c1658
 #2 0xb7c61f9d in _gtk_marshal_VOID__BOXED_UINT (closure=0x85b4008,
     return_value=0x0, n_param_values=3, param_values=0xbf80619c,
     invocation_hint=0xbf8060ac, marshal_data=0xb7d8efb0)
     at gtkmarshalers.c:1457
  data1 = (gpointer) 0x85b8c60
  data2 = <value optimized out>
  __PRETTY_FUNCTION__ = "_gtk_marshal_VOID__BOXED_UINT"
 #3 0xb7a076ab in IA__g_closure_invoke (closure=0x85b4008, return_value=0x0,
     n_param_values=3, param_values=0xbf80619c, invocation_hint=0xbf8060ac)
     at gclosure.c:490
  marshal = (GClosureMarshal) 0xb7c61f40 <_gtk_marshal_VOID__BOXED_UINT>
  marshal_data = (gpointer) 0x0
  __PRETTY_FUNCTION__ = "IA__g_closure_invoke"
...

The backtrace is with gaim2.0beta5 on feisty

another crash

Brian, thank you for your interest to Ubuntu. Your crash has nothing to do with the one from that bug though, you should rather open a new bug when you are not sure of what you are doing the bugsquad will point you to the right direction then. If you think that's a duplicate without being sure feel free to point the bug number on the new bug then

On Thu, 2006-12-14 at 22:07 +0000, Sebastien Bacher wrote:
> Brian, thank you for your interest to Ubuntu. Your crash has nothing to
> do with the one from that bug though, you should rather open a new bug
> when you are not sure of what you are doing the bugsquad will point you
> to the right direction then. If you think that's a duplicate without
> being sure feel free to point the bug number on the new bug then

Fair enough. I guess the problem is that from apport, there is really
no way to compare given that you can't actually see the stacktrace.
This is one place where bug-buddy does a good job.

b.

--
My other computer is your Microsoft Windows server.

Brian J. Murrell

On jeu, 2006-12-14 at 22:18 +0000, Brian J. Murrell wrote:

> Fair enough. I guess the problem is that from apport, there is really
> no way to compare given that you can't actually see the stacktrace.
> This is one place where bug-buddy does a good job.

The apport window has an expander widget to displays details about the
bug and the backtrace is listed there

On Thu, 2006-12-14 at 22:29 +0000, Sebastien Bacher wrote:
> On jeu, 2006-12-14 at 22:18 +0000, Brian J. Murrell wrote:
>
> > Fair enough. I guess the problem is that from apport, there is really
> > no way to compare given that you can't actually see the stacktrace.
> > This is one place where bug-buddy does a good job.
>
> The apport window has an expander widget to displays details about the
> bug and the backtrace is listed there

Ahhh. Yes. I see. I guess the problem is that I don't have enough
-dbg packages loaded.

Are there any plans to have apport load relevant -dbg packages for the
user to see more meaningful stack traces?

b.

--
My other computer is your Microsoft Windows server.

Brian J. Murrell

Ped, can you please get gaim to crash again and this time tell gdb to print out the value of selection_data->length (and I suppose selection_data and *selection_data for good measure)?

On jeu, 2006-12-14 at 22:54 +0000, Brian J. Murrell wrote:

> Are there any plans to have apport load relevant -dbg packages for the
> user to see more meaningful stack traces?

apport already does that, you can run "apport-retrace --download-debug"
on the crash file, it'll download the package and use the dump from the
crash file to generate a debug backtrace. You can also use
"apport-retrace --gdb" to attach gdb to the dump then

Replying to the previous comment since the crash is easy to get and I've debug packages installed:

(gdb) p *selection_data
$2 = {selection = 0x45, target = 0x78, type = 0x78, format = 8,
  data = 0x83b0898 "", length = 0, display = 0x81050c8}

Does this crash happen only with Speedcrunch? Or also with other KDE applications? Other applications which use the same text widgets that Speedcrunch does? Things like that? I believe I have a fix but I'm trying to track down exactly what the root of the problem is.

Can't find any other application crashing gaim in the same way.
I'm not sure how to find same text widget, but IMHO for example in system settings the Search box is probably the same widget. (no crash)

Copy from Kate (text editor) works OK too, both copy, and copy as HTML.
Copy from gnome(?) "calculator" is OK too.

Maybe some SpeedCrunch creator can have idea what's so special about "Copy result".
(I'm not using gaim, I was just trying it out with my common way what I need from IM when I hit the crash. I tried some more [both KDE and non KDE] applications now (konqueror, text boxes from system settings, calculator, Kate, firefox, OpenOffice), and no more crashes, only from SpeedCrunch.)

So I'm afraid I can't help you any more.

Still you may also check the report about "empty paste" crash, that one looks very very similar to this, and also the "length = 0" sounds like it may be related.
#49691 but it looks like it's not reproducible anymore.

This bug still happens ?! and at feisty ?

the issue is fixed in the current pidgin version

Don't know if it is the same bug.

However, copying text from some PDF document and trying to paste it into some openoffice calc cell, systematically crashes X.

This on an Ubuntu maverick 64 bit machine, with intel graphics.