We haven't forgotten about it, but I haven't been able to reproduce it yet, and it's not the only fire right now. Could one of you try issuing the command /gpf in your client, and tell me if the bug reporter comes up for that?

Yep, it works with a /gpf. Presuming you got a bug report from me just now, that is.

Incidentally, yay, now we have a new command to teach noobs!

Heh, yea, only amd64-using noobs, though, and only until we've got the crash reporter/vc issues worked out.

Poked around a bit with this found that this only happened to me using the newer 2.6.27 kernels. rolled back to a 2.6.22 and everything was fine. It may have to do with the newer HAL that most distro's are using with the newer kernels.

Wow- thanks Chana! I'll verify, see what more I can dig up, and pass it on to teamspeak or maybe alsa or lkml. For some reason I thought I was current on this thread, but I only just read your post. I've been working on bigger problems since this came up, and I just came back to review the situation and take another stab at it. Seems you did the work for me.

Hi Mom did some more testing on sunday a week ago. With several Disto's and even a vanillia build and had no problem at all with the new hal and xorg input so figured it got fixed some how.

Still segfaults for me.

% uname -r
2.6.27-gentoo

% X -version

X.Org X Server 1.5.2
Release Date: 10 October 2008
X Protocol Version 11, Revision 0
Build Operating System: Linux 2.6.25-gentoo-r6 x86_64
Current Operating System: Linux CaveOfBirds 2.6.27-gentoo #1 SMP PREEMPT Wed Oct 22 23:11:35 PDT 2008 x86_64
Build Date: 22 October 2008 08:57:57AM

% hald --version
HAL package version: 0.5.11

And I've been using X with input hotplugging since before voice chat came out, so I kinda doubt that that is it.

Still segfaults for me as well.

Everything is the same version as roguelazer, although it is Ubuntu centric. The notable difference is that CONFIG_PREEMPT is not set in the kernel.

I'm running the same versions of hald and x, and 2.6.27-gentoo-r4 with preempt enabled (low-latency-desktop), and haven't yet managed to segfault it. Any other variables we might consider? What sound hardware do you have roguelazer? Are you using the in-kernel version of alsa?

Here is my output from the alsa-info.sh script:

http://www.alsa-project.org/db/?f=bb4e3c17d43d5a339a63d5c901c7ffb45c95bee0

I've got a SoundBlaster Audigy 2 and a HeadRoom Mobile. The audigy is configured to be the default input device, and the headroom the default output device.

Interesting. You both have 2 devices. Are you both using alsa to set one card's in and the other's out as defaults? Could you try a more normal alsa config and see if it still segfaults?

I get the segfault on my laptop as well. It has only one device.