Your problem *appears* to be a GART issue.
For some reason, and I've never experienced this particular error myself, the
X server is refusing to run more than once instance at the same time.
Can you try this:
1. Switch to Virtual Console 1 (VC 1) using CNTL-ALT-F1
2. Login as root.
3. If you're on a RedHat or similar box, drop OUT of any GUI runlevel:
# init 2
4. Kill any remaining kdm, gdm, or xdm display manager.
# killall gdm
# killall kdm
# killall xdm
5. Make sure your X11 server is not running.
# ps -axuw | grep X
If you see a XFree86 process running, we want to try and kill it.
Now that you have no X server running, we want to try and see if we can start
one manually:
# X :0 2>&1 >/tmp/error.log &
You should get a "grey stipple" background and a simple "X" cursor. Nothing
else will appear.
Did that work? If not, then I'm *very* confused at how your X desktop is
currently being started. Please send me the /tmp/error.log file in an email.
If it did work, try the following:
1. Use CNTL-ALT-F1 to return to your VC1 console.
4. Now, we want to start another X server on a second display:
# X :1 2>&1 > /tmp/error1.log &
Did a grey-stipple screen re-appear? If so, then you *can* have two X servers
running at the same time with your GART implementation. If not, there's a
weirdness there that I've yet to experience. To stop this test
Whenever you have a grey-stipple screen like that, you can either switch back
to a text mode VC using the CNTL-ALT-F[1-6] keystroke, or you can "kill" the
running X server using CNTL-ALT-BACKSPACE to kill it completely.
The easiest way to regain your desktop at this point is to reboot, but you may
also return to the running X runlevel or start kdm/gdm/xdm manually from the
command prompt.
- Ian
On Thursday 13 February 2003 09:29, John wrote:
> Hi Vernon! I only got to the first part of your test:
>
> -> First of all one thing to try as root is
> -> "startx -- :1"
> ->
> -> That should bring up kde in display 1.
> -> If root's home directory has an .xinitrc file all bets are
> off.
> -> If there is such a file delete it before trying the command
> above.
>
> Just so I haven't slipped up, let me tell you some details:
>
> First, last night I reinstalled RH8 so at this point in time
> there are NO files that have been modified for X forwarding,
> etc. Shouldn't need anything like that for your first test,
> correct?
>
> I checked the root directory for .xinitrc and there wasn't one.
>
> Then, I start with a machine that is freshly booted, sitting
> there with a graphic login screen.
>
> I hit ctrl-alt-F1, which gets me to first VC.
>
> Log in as root.
>
> Type: startx -- :1
>
> What I get is that same screen that I described earlier, PLUS
> some more stuff added at the bottom, now (hey, that's new!).
>
> I don't know how to capture a screen dump, so the "good stuff"
> (leaving out the things that I don't think are important) is:
>
> (EE)GARTInit: AGPIOC_INFO
> failed (invalid argument)
>
> (EE)I810(0): AGP GART support is not available. Make sure your
> kernal has agpgart support or that the agpgart kernel module is
> loaded.
>
> (EE)Screen(s) found, but none have usable configuration
>
> Fatal server error no screens found
>
> .......PLUS, here's the new stuff on the bottom of the screen:
>
> XIO: fatal IO error 104 (Connection reset by peer) on X server
> ":1.0" after 0 requests (0 known processed) with 0 events
> remaining.
>
> I'm hoping that you didn't need a +xhosts or anything enabled in
> some file for this to work, because this has nothing to do with
> a remote machine, correct?
>
> Thanks for the help!
>
> John
-- - Ian C. Blenke <icblenke@nks.net>(This message bound by the following: http://www.nks.net/email_disclaimer.html)
This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 15:18:15 EDT