Re: woes with 2.6.0-test1 and xscreensaver/xlock

[Edward Tandi <ed@efix.biz>]

On Wed, 2003-07-16 at 18:27, Nuno Monteiro wrote:
> Hi there people,
> 
> 
> Is anyone else having trouble with xscreensaver/xlock under 2.6.0-test1? 
> Whenever I lock my session using either "lock screen" from the menu (it 
> launches 'xscreensaver lock', afaik) or "xlock", I cant seem to ever get 
> my session back -- I type in the correct password, but they both just 
> hang there. The exact same setup works flawlessly in 2.4.21, and just for 
> the sake of curiosity I also tested 2.5.75, 2.5.74, 2.5.73, 2.5.72, 
> 2.5.71 and 2.5.70, they all exhibit the same behaviour as 2.6.0-test1. I 
> dont really have time to go on testing kernels to find out exactly where 
> it broke, so I'm hoping anyone else is experiencing these woes.

Yes, I reported the same problem (and others) on the 31st may and the
5th June.

It's good to see someone has found the gnome-terminal problem. The only
other biggie-ish is the rmmod issue.

Ed-T.

> Additionally, syslog spews out the following:
> 
> Jul 16 17:40:24 hobbes xscreensaver(pam_unix)[1501]: authentication 
> failure; logname= uid=501 euid=501 tty=:0.0 ruser= rhost=  user=nuno
> 
> I upgraded to the latest pam and xscreensaver packages from mandrake 
> cooker, but still no dice, it hangs exactly the same. And oh, this is 
> glibc 2.3.2, if thats interesting, and gcc 3.3.
> 
> Interestingly enough, sometimes killing the xscreensaver process leads to 
> a complete hang, although no oops is visible. With nmi_watchdog=1 it 
> doesnt hang, but it seems the keyboard is dead after killing xscreensaver 
> -- i see an error on console, something like 'xscreensaver: no interrupt 
> data for mouse/keyb on /proc/interrupt' (i didnt copy it down, sorry). 
> The keyboard is still useable from the console, but not from X.
> 
> Let me know if you need more info.