Hi.

I just did a fresh install of slack 11.0 yesterday. I made two changes from my previous installation(s):

1.) Using 2.6.17.13 SMP kernel.

2.) Installed NVIDIA driver (using a GeForce 8800 GTX), however I'm still using the xorg.conf that was generated when I did the slackware install as the NVIDIA xorg.conf file blackscreens me when I log out of X. But that's another problem entirely...


Now, the problem is that my utmp is somehow getting corrupted. I can't duplicate the problem. It's random, but occurs somewhat often. I believe it is X-related. When I log out of X sometimes, and do a "w", it'll show 2 users when only one is logged on, etc. Sometimes it'll work it's way up to 3 or 4 before I catch it...

Anyway, I know this isn't a slackware problem, but I'm hoping that another slack user has had the same problem and can tell me what it is that's corrupting my utmp. If not, I hope someone can tell me how I can figure out what it is that's corrupting my utmp.


Thanks,
Mike

If you think it's X, maybe you can look through the logs (/var/log/Xorg.0.log among others) and see if you see anything strange.

Can you try reinstalling the same exact way but without the nvidia drivers. (if you must know, I suspect these the most as the being the cause ... I can't imagine the smp kernel, especially the stock Slackware kernel doing this)

Thanks for the reply. I agree with you 100%. I've used slack since 3.5, and slack 11.0 since it came out, and never had a utmp problem in all these months with 11.0. I was just kinda throwing this out to see if anyone else had the problem (with nvidia) and figured out a fix for it. I can't find anything anywhere. Oh well, worth a shot...


Thanks again,
Mike

I have run in to the same issue and have for a long time. It seems to be related to logging into console while X it running. In fact i show 3 users logged in now.

is that even possible ? if I switch to a different console, and then switch back, X is no longer running.

Sure, ctrl-alt-f(whatever) from X to login to a text console, then alt-f7 from the console to get back to X.

So, if you boot up and login on tty1, and startx, you can ctrl-alt-f2 (f3, f4, etc) to login to another console, then alt-f7 to get back to X...etc, etc.



Regards,
Mike

Thank you sir. That may very well be the problem. I'm just surprised I didn't notice it all this time. Do you happen to have the nvidia driver installed too?


Thanks for your reply,
Mike

I do have the 1.0-9746 version installed but i have not had the issue you have. Two things to try make sure the driver you are using is the latest one and use the stock xorg.conf with the driver changed from vesa to nvidia. I don't really trust letting nvidia's installer edit the xorg.conf for me.

Check and check...

I didn't trust the nvidia config either. But, when I edit xorg.conf and change from vesa to nvidia, I get the "black screen of death", (not to be confused with the blue screen of death). I'm just going to back the driver out. I checked the nvidia linux forums and I'm not the only one having problems.

Realistically, I use X for browsing, viewing .pdf files, and maybe a game of Kolf here and there. I can do without the nvidia driver. It's not worth the trouble at this point.


Thanks for your help though.


Regards,
Mike

If you don't play games or use 3D apps of any form ... I highly recommend the free 'nv' driver (using 'xorgsetup' should set it up right).

Thanks, I'll try that one.

Update: I removed the nvidia module, and, lo and behold, I'm still getting a corrupt utmp. So, back to my original question, is there any way for me to isolate just what's causing it?



Thanks,
Mike


P.S. Sorry for blaming you, Mr. Nvidia.

What kind of programs or background processes are you running ? Try to check the integrity of important packages and see if it holds.