Hello,
Am Samstag, 20. September 2003 15:57 schrieb Ernie Schroder:
> On Saturday 20 September 2003 06:08 am, Michael Schreckenbauer wrote:
> > Hello,
> >
> > Am Samstag, 20. September 2003 00:47 schrieb Ernie Schroder:
> > > On Friday 19 September 2003 06:21 pm, Ernie Schroder wrote:
> > > > On Friday 19 September 2003 05:50 pm, Michael Schreckenbauer
>
> wrote:
> > > > > Hello,
> > > > >
> > > > > Am Freitag, 19. September 2003 22:34 schrieb Ernie Schroder:
> > > > > > Interestingly, When the XF86Config contains "nvidia"
> > > > > > attempting to start X produces NO log output. The machine
> > > > > > just locks up hard. The "nvidia" module can be successfully
> > > > > > loaded prior to starting X but still, the damned thing
> > > > > > locks solid. This has got to be a kernel issue. I looked
> > > > > > for the obvious by diff'ing the old and new .configs and
> > > > > > removed AGP and DRM and rebuilt it but I'm still in the
> > > > > > same spot. Next I'll rebuild with the old config if I can.
> > > > > > In the mean time, I'm using the "nv" module which logs
> > > > > > with no errors and only one warning about APM.
> > > > > >
> > > > > > Ernie
> > > > > >
> > > > > > On Friday 19 September 2003 02:36 pm, Michael
> > > > > > Schreckenbauer
> > > >
> > > > wrote:
> > > > > > > Hi Ernie,
> > > > > > >
> > > > > > > Am Freitag, 19. September 2003 20:05 schrieb Ernie
>
> Schroder:
> > > > > > > > Hmmmm is right! yes, the module is loading in the r7
> > > > > > > > kernel but startx completely hangs the box. I'm going
> > > > > > > > to rebuild the nvidia pkgs. on the r1 kernel and, if
> > > > > > > > that works I;ll assume I've left something out of the
> > > > > > > > kernel
> > > > > > > > Ernie
> > > > > > >
> > > > > > > is it possible to post your /var/log/XFree86.0.log after
> > > > > > > trying to startx with the nvidia-driver? Maybe there are
> > > > > > > some useful hints in it.
> > > > > > > Michael
> > > > >
> > > > > very weird :-( Perhaps you can get some information when
> > > > > running strace X 1>out.txt 2>&1. And then have a look into
> > > > > out.txt, at which point the server hangs.
> > > > >
> > > > > Michael
> > > >
> > > > I'm not familiar with strace. I'm assuming I need to back out
> > > > of X, reset the Xconfig to the nvidia driver and THEN run the
> > > > command. Right?
> > >
> > > I did as above and I got a zero bit file. I did:
> > > # emerge strace
> > > # strace X 1>out.txt 2>&1
> >
> > your commands look ok. A zero bit file is very strange. strace
> > traces system calls and signals. I cant imagine, how this could be
> > empty and the kernel locks. Could you try this as root, and add -f
> > to strace? This allowes tracing forked processes.
> > #strace -f X 1>out.txt 2>&1
> >
> > Michael
>
> I ran strace with the "f" flag and do get some output (see below)
> At a quick glance, nothing jumps out at me. Perhaps you can make some
> sense of it.
>
> # cat out.txt
> <snipped a lot of output>
> geteuid32() = 0
> execve("/etc/X11/X", ["/etc/X11/X"], [/* 36 vars */]
>
> I really appreciate the help, Michael. Thanks
the last command executed (/etc/X11/X) does not return. This is a symlink to /
usr/X11R6/bin/XFree86. So maybe you can get further information by running:
#strace -f /usr/X11R6/bin/XFree86 1>out.txt 2>&1
On my laptop I see things like opening the config and loading the drivers
specified there in the output.
I'm not very familiar with debugging such things, but I hope, we can get it
running this way.
Michael
--
[EMAIL PROTECTED] mailing list
