LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - Software (https://www.linuxquestions.org/questions/linux-software-2/)
-   -   Wine zombie processes completely kill sound (https://www.linuxquestions.org/questions/linux-software-2/wine-zombie-processes-completely-kill-sound-397842/)

Haiyadragon 12-30-2005 08:58 AM

Wine zombie processes completely kill sound
 
Wine has been leaving a zombie process called "wine-preloader" that, as the zombie part suggests, can't be killed even with a kill -9. That wouldn't be so bad if it weren't for the fact that sound is completely disabled. I usually use alsa for everything, and dmix should allow concurrent playback. Anyway, after Wine no sound can be played untill reboot.

Has anybody else encountered this problem and would somebody be able to tell me the root of the problem? I read somewhere it's a problem with kernels up until 2.6.7 but I'm running a self compiled 2.6.14.4. Is it a problem with this specific version? I really don't feel like recompiling kernel versions, rebooting every compile and every wine zombie occurence, until the problem disappears.

edit. So, it happens with vanilla self-compiled 2.6.14.4, 2.6.14.5 and 2.6.13.5. Also happens with 2.6.14.3 from slack testing. It happens when Wine crashes before it runs the desired game, but it also happens when I quit the game. So I can only play it once and then I have to reboot.

As an added treat my box is also unable to unmount any disks cleanly. So they are all nasty reboots. Also it happens with Wine 0.9.3 and 0.9.4.

Oh and I recently switched back from Arch Linux to Slack, they both experienced this problem.

edit2. Tried Slackware's stock kernel (2.4.32) and the process disappeared as it should. So now I'm going to try 2.6.12 and lower until I find a 2.6 that isn't broken. 2.4 kinda sucks without udev and with it's scsi emulation for everything.

mimithebrain 12-30-2005 12:19 PM

To what I research previously, zombie processes are left alive
for the parent process to clean after them. Zombie process
dies (in my case), when I terminate or kill the parent process

This should release your card as soon as the zombie process
killed (for real)

Haiyadragon 12-30-2005 12:23 PM

Problem is, these babies don't have a parent process.

Haiyadragon 12-30-2005 03:55 PM

Ok, after screwing around some more I finally find the origin of this monster. It wasn't a kernel thing after all (very glad about that). The reason it worked with 2.4.something and later 2.6.12 was because I was using the generic nv video drivers instead of the nvidia accelerated drivers. So I messed around with that and it seems to be a problem only with version 8174 of the nvidia drivers. I tried both 7676 and 8178, and they both don't leave that nasty zombie behind.

mimithebrain 01-01-2006 03:44 PM

very glad you fixed it... that's an interesting problem...


All times are GMT -5. The time now is 06:56 PM.