On Wednesday, October 13, 2010 05:33:19 Claude Jones wrote: > On 10/12/2010 1:23 PM, Marko Vojinovic wrote: > > 1) remove/rename any existing /etc/X11/xorg.conf > > 2) create a new dummy user > > 3) do a clean boot into runlevel 3 > > 4) login as the dummy user, do a startx > > 5) see X fail, copy /var/log/Xorg.0.log > > 6) post the whole file for us to see > > > here you go - I followed your suggested steps - on reboot after creating > the new user, the bootup stalled - I did a Ctl-F2 to get a login prompt > and logged into the new user, then ran startX, which failed - here's the > Xorg.0.log (I deleted the old Xorg.0.log before rebooting but the new > one is really long) I did disconnect the old keyboard and mouse cable, > which connects to a KVM, and plugged in a different mouse and keyboard > directly to the PC Sorry for the late response, I wasn't around yesterday... You didn't do quite what I suggested --- the idea was not to run a normal bootup and then do a ctrl-shift-f2 to go to a console, because that way the system is already trying to load one instance of X, and you are then issuing startx for another. The idea was to boot into runlevel 3, so that the system does not even attempt to load X until you say so. In principle the log files in two cases are different. But in this case I have a feeling it doesn't matter. Now, about the log... > [ 34.779] > X.Org X Server 1.8.2 > Release Date: 2010-07-01 [snip] > [ 35.965] (**) Option "xkb_rules" "evdev" > [ 35.965] (**) Option "xkb_model" "pc105+inet" > [ 35.965] (**) Option "xkb_layout" "us" > [ 35.965] (**) Option "xkb_options" "terminate:ctrl_alt_bksp," [I guess this is the 1-second blank spot where something goes wrong] > [ 36.682] (II) Power Button: Close > [ 36.682] (II) UnloadModule: "evdev" > [ 36.686] (II) Power Button: Close > [ 36.686] (II) UnloadModule: "evdev" > [ 36.697] (II) Microsoft Microsoft 5-Button Mouse with > IntelliEye(TM): Close > [ 36.697] (II) UnloadModule: "evdev" > [ 36.705] (II) USB Gaming Keyboard: Close > [ 36.705] (II) UnloadModule: "evdev" > [ 36.712] (II) USB Gaming Keyboard: Close > [ 36.712] (II) UnloadModule: "evdev" > [ 36.712] (II) NOUVEAU(0): NVLeaveVT is called. > [ 36.713] (II) NOUVEAU(0): Closed GPU channel 2 There are no errors in the log (apart from the irrelevant one for the absent Gallium library). Apparently X goes up cleanly, and then is immediately instructed to shut down (again, cleanly). Is there any process that might issue an X shutdown? A cron job running every second? Some other service? If I didn't know it's impossible, I'd say ctrl-alt-backspace is stuck on your keyboard. ;-) Seriously, the log looks completely sane and regular up to [35.965]. And after less than one missing second, from [36.682] it looks like a regular X shutdown procedure. From the point of view of X, nothing is broken or wrong. It is just being instructed to start and then to shutdown immediately after. The mystery is _why_. The only idea I am able to think of is to investigate further (ie. read more logs). You can do a startx with the "--logverbose 4" option (or higher than 4), and try to find out what happens in that blank second (between [35.965] and [36.682] in the above example). Also, monitor /var/log/messages and dmesg for the same period. If there is no hint in any of those logs, I'd submit a bug against nouveau, or ask on their mailing list (they seem responsive afaict). HTH, :-) Marko -- users mailing list users@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe or change subscription options: https://admin.fedoraproject.org/mailman/listinfo/users Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
