rawhide->fc5; X won't start

[sean darcy]

I was running rawhide updated to March 10. Went on vacation, yesterday 
did a full upgrade to fc5. Used yum to get  updates released.  Now X 
won't start. I think it's some sort of  xfs problem.

with gnome, I get the fedora bubble logo, then a black screen, alt-f1 
gets me the console, which shows:

waiting for X server to shut down FreeFontPath: FPE "unix/7100" refcount 
is 2, should be 1, fixing

syslog shows:

gnome_segv2[3263]: segfault at 0000000000xxxxxxxx .........

kde also won't start. The console shows:

xset: bad font path element (#76), .......

and syslog shows kded, ksplash, kcminit, and ksmserver segfaulting.

I've restarted xfs ( no errors in syslog ). I've rebooted.

Do I need to do some mkfontdir voodoo in each font directory? What did 
the upgrade change? I've seen no similar problems posted, so I assume 
it's a rawhide->fc5 issue.

puzzled.

sean
sean

[sean darcy]

sean wrote:
>  I was running rawhide updated to March 10. Went on vacation,
> yesterday 
>  did a full upgrade to fc5. Used yum to get  updates released.  Now X 
>  won't start. I think it's some sort of  xfs problem.
>  
>  with gnome, I get the fedora bubble logo, then a black screen, alt-f1 
>  gets me the console, which shows:
>  
>  waiting for X server to shut down FreeFontPath: FPE "unix/7100" refcount 
>  is 2, should be 1, fixing
>  
>  syslog shows:
>  
>  gnome_segv2[3263]: segfault at 0000000000xxxxxxxx .........
>  
>  kde also won't start. The console shows:
>  
>  xset: bad font path element (#76), .......
>  
>  and syslog shows kded, ksplash, kcminit, and ksmserver segfaulting.
>  
>  I've restarted xfs ( no errors in syslog ). I've rebooted.
>  
>  Do I need to do some mkfontdir voodoo in each font directory? What did 
>  the upgrade change? I've seen no similar problems posted, so I assume 
>  it's a rawhide->fc5 issue.
>  
>  puzzled.
>  
>  sean
>  sean
>   
OK, so no one else sees this.

Any suggestions as to how to diagnose this? Any way to get more verbose 
or clearer error messages?

sean

[Ray Strode]

Hi, 

>  OK, so no one else sees this.
>  
>  Any suggestions as to how to diagnose this? Any way to get more verbose 
>  or clearer error messages? 
If you run fc-cache -f from your user account does your
problem go away?

--Ray

[sean darcy]

Ray Strode wrote:
>  Hi, 
>  
> > OK, so no one else sees this.
> >
> > Any suggestions as to how to diagnose this? Any way to get more verbose 
> > or clearer error messages?
>  If you run fc-cache -f from your user account does your problem go away?
>  
>  --Ray
>   
Thanks for the suggestion, but...(:

Here's the console message from trying kde.

>  (==) Using config file: "/etc/X11/xorg.conf"
>  xset:  bad font path element (#76), possible causes are:
>      Directory does not exist or has wrong permissions
>      Directory missing fonts.dir
>      Incorrect font server address or syntax
>  startkde: Starting up...
>  DCOP aborting call from 'anonymous-2482' to 'kded'
>  startkde: Shutting down...
>  klauncher: Exiting on signal 1
>   
All the /etc/X11/fs/config directories exist. I'm not sure what 
permissions they should have, but I've tried starting X as root, same 
problem.

[ could this be some selinux issue. It should be off.]

Here's the xorg.conf font server address:
>  
>  Section "Files"
>  
>  
>  	FontPath     "unix/:7100"
>  EndSection 
I've restarted xfs ( and rebooted), with no error messages.

I've run fc-cache -f

And all this ran fine under rawhide of March 10.

Any help realllly appreciated.

sean

[Ray Strode]

Hi,

>  Ray Strode wrote:
>  > Hi, 
>  > 
>  >> OK, so no one else sees this.
>  >>
>  >> Any suggestions as to how to diagnose this? Any way to get more verbose 
>  >> or clearer error messages?
>  > If you run fc-cache -f from your user account does your problem go away?
>  > 
>  > --Ray
>  > 
>  
>  Thanks for the suggestion, but...(:
>  
>  Here's the console message from trying kde.
>  
>  > (==) Using config file: "/etc/X11/xorg.conf"
>  > xset:  bad font path element (#76), possible causes are:
>  >     Directory does not exist or has wrong permissions
>  >     Directory missing fonts.dir
>  >     Incorrect font server address or syntax
>  > startkde: Starting up...
>  > DCOP aborting call from 'anonymous-2482' to 'kded'
>  > startkde: Shutting down...
>  > klauncher: Exiting on signal 1
>  > 
What if you manually add some font paths to your Xorg.conf
Try adding some of the ones in 
/usr/share/X11/fonts/
and /usr/share/fonts

--Ray

[sean darcy]

Ray Strode wrote:
>  Hi,
>  
> > Ray Strode wrote:
> >> Hi, 
> >>
> >>> OK, so no one else sees this.
> >>>
> >>> Any suggestions as to how to diagnose this? Any way to get more verbose 
> >>> or clearer error messages?
> >> If you run fc-cache -f from your user account does your problem go away?
> >>
> >> --Ray
> >>
> > Thanks for the suggestion, but...(:
> >
> > Here's the console message from trying kde.
> >
> >> (==) Using config file: "/etc/X11/xorg.conf"
> >> xset:  bad font path element (#76), possible causes are:
> >>     Directory does not exist or has wrong permissions
> >>     Directory missing fonts.dir
> >>     Incorrect font server address or syntax
> >> startkde: Starting up...
> >> DCOP aborting call from 'anonymous-2482' to 'kded'
> >> startkde: Shutting down...
> >> klauncher: Exiting on signal 1
> >>
>  What if you manually add some font paths to your Xorg.conf
>  Try adding some of the ones in 
>  /usr/share/X11/fonts/
>  and /usr/share/fonts
>  
>  --Ray
>   
Great idea, but no joy.

Same result.

I changed xorg.conf to:

>  Section "Files"
>  
>  #
>  #	FontPath     "unix/:7100"
>    FontPath  "/usr/share/X11/fonts/misc:unscaled" 
>    FontPath  "/usr/share/X11/fonts/75dpi:unscaled"
>    FontPath  "/usr/share/X11/fonts/100dpi:unscaled"  
>  
>  EndSection 
And, FWIW,
>  find /usr/share/X11/fonts -name fonts.dir
>  /usr/share/X11/fonts/misc/fonts.dir
>  /usr/share/X11/fonts/75dpi/fonts.dir
>  /usr/share/X11/fonts/100dpi/fonts.dir
>  /usr/share/X11/fonts/TTF/fonts.dir
>  /usr/share/X11/fonts/Type1/fonts.dir 
Xorg.0.log shows no errors.

This is really frustrating. I no longer believe this a font issue. But I 
can't figure out how to get some log and error messages that show what's 
happening.

sean

[sean darcy]

sean wrote:
>  Ray Strode wrote:
> > Hi,
> >
> >> Ray Strode wrote:
> >>> Hi,
> >>>> OK, so no one else sees this.
> >>>>
> >>>> Any suggestions as to how to diagnose this? Any way to get more 
> >>>> verbose or clearer error messages?
> >>> If you run fc-cache -f from your user account does your problem go 
> >>> away?
> >>>
> >>> --Ray
> >>>
> >> Thanks for the suggestion, but...(:
> >>
> >> Here's the console message from trying kde.
> >>
> >>> (==) Using config file: "/etc/X11/xorg.conf"
> >>> xset:  bad font path element (#76), possible causes are:
> >>>     Directory does not exist or has wrong permissions
> >>>     Directory missing fonts.dir
> >>>     Incorrect font server address or syntax
> >>> startkde: Starting up...
> >>> DCOP aborting call from 'anonymous-2482' to 'kded'
> >>> startkde: Shutting down...
> >>> klauncher: Exiting on signal 1
> >>>
> > What if you manually add some font paths to your Xorg.conf
> > Try adding some of the ones in /usr/share/X11/fonts/
> > and /usr/share/fonts
> >
> > --Ray
> >
>  
>  Great idea, but no joy.
>  
>  Same result.
>  
>  I changed xorg.conf to:
>  
> > Section "Files"
> >
> > #
> > #    FontPath     "unix/:7100"
> >   FontPath  "/usr/share/X11/fonts/misc:unscaled"   FontPath  
> > "/usr/share/X11/fonts/75dpi:unscaled"
> >   FontPath  "/usr/share/X11/fonts/100dpi:unscaled" 
> > EndSection
>  
>  And, FWIW,
> > find /usr/share/X11/fonts -name fonts.dir
> > /usr/share/X11/fonts/misc/fonts.dir
> > /usr/share/X11/fonts/75dpi/fonts.dir
> > /usr/share/X11/fonts/100dpi/fonts.dir
> > /usr/share/X11/fonts/TTF/fonts.dir
> > /usr/share/X11/fonts/Type1/fonts.dir
>  
>  Xorg.0.log shows no errors.
>  
>  This is really frustrating. I no longer believe this a font issue. But I 
>  can't figure out how to get some log and error messages that show what's 
>  happening.
>  
>  sean
>   
I installed xfce to see if it was a gnome/kde problem. Here's what shows 
on the xfce console:

>  (==) Using config file: "/etc/X11/xorg.conf"
>  
>  /usr/bin/startxfce4: X server already running on display :0
>  Agent pid 4100
>  which: no xfce4-session in
> (/usr/kerberos/sbin:/usr/kerberos/bin:/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/usr/X11R6/bin:/usr/NX/bin:/root/bin)
>  
>  (xfce-mcs-manager:4103): libxfce4mcs-CRITICAL **: mcs_manager_add_channel_from_file:
> assertion `filename != NULL && strlen (filename) > 0' failed
>  /etc/xdg/xfce4/xinitrc: line 106:  4105 Segmentation fault      xfwm4 --daemon
>  ** Message: Cannot connect to session manager
>  /etc/xdg/xfce4/xinitrc: line 123:  4108 Segmentation fault      xfcalendar
>  /etc/xdg/xfce4/xinitrc: line 123:  4110 Segmentation fault      $panel 
So X appears to start - Fedora logo comes up; no error on Xorg.0.log.

But no gui's will start - not gnome,kde or xfce. Does this mean X isn't 
starting correctly so that the gui's can start? If not, what happens 
between X coming up and the gui's starting?

metacity?  Is there any way to cause metacity to do verbose logging?

Again, is there any way to get more verbose and clearer logs.

Or any other hints?

sean

[Kevin Fenzi]

> >>>> "sean" == sean 
> <seandarcy2(a)gmail.com&gt; writes: 
sean> I installed xfce to see if it was a gnome/kde problem. Here's
sean> what shows on the xfce console:

> > (==) Using config file: "/etc/X11/xorg.conf"
> /usr/bin/startxfce4: X
> > server already running on display :0 Agent pid 4100 which: no
> > xfce4-session in
> >
> (/usr/kerberos/sbin:/usr/kerberos/bin:/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/usr/X11R6/bin:/usr/NX/bin:/root/bin)
> > (xfce-mcs-manager:4103): libxfce4mcs-CRITICAL **:
> > mcs_manager_add_channel_from_file: assertion `filename != NULL &&
> > strlen (filename) > 0' failed /etc/xdg/xfce4/xinitrc: line 106:
> > 4105 Segmentation fault xfwm4 --daemon ** Message: Cannot connect
> > to session manager /etc/xdg/xfce4/xinitrc: line 123: 4108
> > Segmentation fault xfcalendar /etc/xdg/xfce4/xinitrc: line 123:
> > 4110 Segmentation fault $panel 
sean> So X appears to start - Fedora logo comes up; no error on
sean> Xorg.0.log.

sean> But no gui's will start - not gnome,kde or xfce. Does this mean
sean> X isn't starting correctly so that the gui's can start? If not,
sean> what happens between X coming up and the gui's starting?

sean> metacity?  Is there any way to cause metacity to do verbose
sean> logging?

sean> Again, is there any way to get more verbose and clearer logs.

sean> Or any other hints?

Well, not sure on gnome or KDE, but the above Xfce output looks like
the problem is the Seg faults when trying to start up. :( 

Several other things to try: 

- If you create a new fresh user, can they login ok?
(This would point to it being you users gtk theme, fonts setup or the
like). 

- Could be some package leftover from your rawhide install causing
issues, perhaps a 'yum list extras' to identify any packages that are
not in repositories you have currently enabled?

- Could you try and run xfwm4 under a debugger?
0) yum --enable-repo=core-debuginfo install xfwm4-debuginfo
1) Run X from a console, "xinit /usr/X11R6/bin/xterm &"
2) Switch to a console (Ctrl+Alt+F1)
3) Set the display "export DISPLAY=:0.0"
4) run "gdb /usr/bin/xfwm4"
5) if/when it crashes, run 'bt' and send me the backtrace. 

sean> sean
sean> -- fedora-devel-list mailing list fedora-devel-list(a)redhat.com
sean> https://www.redhat.com/mailman/listinfo/fedora-devel-list

kevin

[sean darcy]

Kevin Fenzi wrote:
> >>>>> "sean" == sean 
> <seandarcy2(a)gmail.com&gt; writes:
>  
>  sean> I installed xfce to see if it was a gnome/kde problem. Here's
>  sean> what shows on the xfce console:
>  
> >> (==) Using config file: "/etc/X11/xorg.conf" /usr/bin/startxfce4: X
> >> server already running on display :0 Agent pid 4100 which: no
> >> xfce4-session in
> >>
> (/usr/kerberos/sbin:/usr/kerberos/bin:/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/usr/X11R6/bin:/usr/NX/bin:/root/bin)
> >> (xfce-mcs-manager:4103): libxfce4mcs-CRITICAL **:
> >> mcs_manager_add_channel_from_file: assertion `filename != NULL &&
> >> strlen (filename) > 0' failed /etc/xdg/xfce4/xinitrc: line 106:
> >> 4105 Segmentation fault xfwm4 --daemon ** Message: Cannot connect
> >> to session manager /etc/xdg/xfce4/xinitrc: line 123: 4108
> >> Segmentation fault xfcalendar /etc/xdg/xfce4/xinitrc: line 123:
> >> 4110 Segmentation fault $panel
>  
>  sean> So X appears to start - Fedora logo comes up; no error on
>  sean> Xorg.0.log.
>  
>  sean> But no gui's will start - not gnome,kde or xfce. Does this mean
>  sean> X isn't starting correctly so that the gui's can start? If not,
>  sean> what happens between X coming up and the gui's starting?
>  
>  sean> metacity?  Is there any way to cause metacity to do verbose
>  sean> logging?
>  
>  sean> Again, is there any way to get more verbose and clearer logs.
>  
>  sean> Or any other hints?
>  
>  Well, not sure on gnome or KDE, but the above Xfce output looks like
>  the problem is the Seg faults when trying to start up. :( 
>  
>  Several other things to try: 
>  
>  - If you create a new fresh user, can they login ok?
>  (This would point to it being you users gtk theme, fonts setup or the
>  like). 
>   
Yes! I can login as a new user. Switched xorg.conf back to using xfs, 
and that worked to.

( Oddly, when I logout as the new user, I'm left at console 7 with a 
black screen and a blinking cursor at top left. console 1, witch I have 
to manually switch to, shows the usual X shutdown stuff.)

>  - Could be some package leftover from your rawhide install causing
>  issues, perhaps a 'yum list extras' to identify any packages that are
>  not in repositories you have currently enabled?
>  
>  - Could you try and run xfwm4 under a debugger?
>  0) yum --enable-repo=core-debuginfo install xfwm4-debuginfo 
yum
--enablerepo=extras-debuginfo ........  :)

>  1) Run X from a console, "xinit /usr/X11R6/bin/xterm
> &"
>  2) Switch to a console (Ctrl+Alt+F1)
>  3) Set the display "export DISPLAY=:0.0"
>  4) run "gdb /usr/bin/xfwm4"
>  5) if/when it crashes, run 'bt' and send me the backtrace. 
>   
Well, it appears to be font related. I hope you can figure out what the 
font problems is though. Here's the backtrace:

gdb /usr/bin/xfwm4
GNU gdb Red Hat Linux (6.3.0.0-1.122rh)
...................
(gdb) run
Starting program: /usr/bin/xfwm4

Program received signal SIGSEGV, Segmentation fault.
0x0000003fa28233ff in FcPatternUnserialize () from 
/usr/lib64/libfontconfig.so.1
(gdb)
(gdb) bt
#0  0x0000003fa28233ff in FcPatternUnserialize () from 
/usr/lib64/libfontconfig.so.1
#1  0x0000003fa282344e in FcPatternFindElt () from 
/usr/lib64/libfontconfig.so.1
#2  0x0000003fa2824b76 in FcPatternGet () from /usr/lib64/libfontconfig.so.1
#3  0x0000003fa2824e73 in FcPatternGetString () from 
/usr/lib64/libfontconfig.so.1
#4  0x0000003fa28136d1 in FcConfigBuildFonts () from 
/usr/lib64/libfontconfig.so.1
#5  0x0000003fa281b8bb in FcInitLoadConfigAndFonts () from 
/usr/lib64/libfontconfig.so.1
#6  0x0000003fa281b975 in FcInit () from /usr/lib64/libfontconfig.so.1
#7  0x0000003fa2812cab in FcConfigGetCurrent () from 
/usr/lib64/libfontconfig.so.1
#8  0x0000003fa2814d2b in FcConfigSubstituteWithPat () from 
/usr/lib64/libfontconfig.so.1
#9  0x0000003fa4f067d0 in pango_cairo_fc_font_map_get_type () from 
/usr/lib64/libpangocairo-1.0.so.0
#10 0x0000003fa4b098ea in pango_fc_font_map_get_type () from 
/usr/lib64/libpangoft2-1.0.so.0
#11 0x0000003fa4d14c1d in pango_context_get_font_description () from 
/usr/lib64/libpango-1.0.so.0
#12 0x0000003fa4d14f3d in pango_itemize_with_base_dir () from 
/usr/lib64/libpango-1.0.so.0
#13 0x0000003fa4d1c0e0 in pango_layout_iter_get_char_extents () from 
/usr/lib64/libpango-1.0.so.0
#14 0x0000003fa4d1cc28 in pango_layout_iter_get_char_extents () from 
/usr/lib64/libpango-1.0.so.0
#15 0x0000003fa4d1db66 in pango_layout_get_pixel_extents () from 
/usr/lib64/libpango-1.0.so.0
#16 0x00000000004203c2 in myScreenInit (display_info=0x560400, 
gscr=0x558120, event_mask=7995436) at screen.c:79
#17 0x000000000041a368 in main (argc=Variable "argc" is not available.
) at main.c:404
#18 0x0000003fa0f1d084 in __libc_start_main () from /lib64/libc.so.6
#19 0x0000000000408819 in _start ()
#20 0x00007fffffbdee78 in ?? ()
#21 0x0000000000000000 in ?? ()
(gdb)

  I really appreciate all this help. I'd never have figured out how to 
get this far.

sean

[Ray Strode]

Hi,
>  " == sean  <seandarcy2(a)gmail.com&gt; writes:
>  > 
>  > sean> I installed xfce to see if it was a gnome/kde problem. Here's
>  > sean> what shows on the xfce console:
>  > 
>  >>> (==) Using config file: "/etc/X11/xorg.conf" /usr/bin/startxfce4:
> X
>  >>> server already running on display :0 Agent pid 4100 which: no
>  >>> xfce4-session in
>  >>>
> (/usr/kerberos/sbin:/usr/kerberos/bin:/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/usr/X11R6/bin:/usr/NX/bin:/root/bin)
>  >>> (xfce-mcs-manager:4103): libxfce4mcs-CRITICAL **:
>  >>> mcs_manager_add_channel_from_file: assertion `filename != NULL &&
>  >>> strlen (filename) > 0' failed /etc/xdg/xfce4/xinitrc: line 106:
>  >>> 4105 Segmentation fault xfwm4 --daemon ** Message: Cannot connect
>  >>> to session manager /etc/xdg/xfce4/xinitrc: line 123: 4108
>  >>> Segmentation fault xfcalendar /etc/xdg/xfce4/xinitrc: line 123:
>  >>> 4110 Segmentation fault $panel
>  > 
>  > sean> So X appears to start - Fedora logo comes up; no error on
>  > sean> Xorg.0.log.
>  > 
>  > sean> But no gui's will start - not gnome,kde or xfce. Does this mean
>  > sean> X isn't starting correctly so that the gui's can start? If not,
>  > sean> what happens between X coming up and the gui's starting?
>  > 
>  > sean> metacity?  Is there any way to cause metacity to do verbose
>  > sean> logging?
>  > 
>  > sean> Again, is there any way to get more verbose and clearer logs.
>  > 
>  > sean> Or any other hints?
>  > 
>  > Well, not sure on gnome or KDE, but the above Xfce output looks like
>  > the problem is the Seg faults when trying to start up. :( 
>  > 
>  > Several other things to try: 
>  > 
>  > - If you create a new fresh user, can they login ok?
>  > (This would point to it being you users gtk theme, fonts setup or the
>  > like). 
>  > 
>  
>  Yes! I can login as a new user. Switched xorg.conf back to using xfs, 
>  and that worked to.
>  
>  ( Oddly, when I logout as the new user, I'm left at console 7 with a 
>  black screen and a blinking cursor at top left. console 1, witch I have 
>  to manually switch to, shows the usual X shutdown stuff.)
>  
>  > - Could be some package leftover from your rawhide install causing
>  > issues, perhaps a 'yum list extras' to identify any packages that are
>  > not in repositories you have currently enabled?
>  > 
>  > - Could you try and run xfwm4 under a debugger?
>  > 0) yum --enable-repo=core-debuginfo install xfwm4-debuginfo
>  yum --enablerepo=extras-debuginfo ........  :)
>  
>  > 1) Run X from a console, "xinit /usr/X11R6/bin/xterm &"
>  > 2) Switch to a console (Ctrl+Alt+F1)
>  > 3) Set the display "export DISPLAY=:0.0"
>  > 4) run "gdb /usr/bin/xfwm4"
>  > 5) if/when it crashes, run 'bt' and send me the backtrace. 
>  > 
>  
>  Well, it appears to be font related. I hope you can figure out what the 
>  font problems is though.  
Does 
rm -f ~/.fonts.cache-2 \
      ~/.rh-fontconfig/.fonts.cache-2 \
      ~/.fonts/fonts.cache-2

fix things up?

--Ray

[sean darcy]

Ray Strode wrote:
>  Hi,
> > " == sean  <seandarcy2(a)gmail.com&gt; writes:
> >> sean> I installed xfce to see if it was a gnome/kde problem. Here's
> >> sean> what shows on the xfce console:
> >>
> >>>> (==) Using config file: "/etc/X11/xorg.conf"
> /usr/bin/startxfce4: X
> >>>> server already running on display :0 Agent pid 4100 which: no
> >>>> xfce4-session in
> >>>>
> (/usr/kerberos/sbin:/usr/kerberos/bin:/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/usr/X11R6/bin:/usr/NX/bin:/root/bin)
> >>>> (xfce-mcs-manager:4103): libxfce4mcs-CRITICAL **:
> >>>> mcs_manager_add_channel_from_file: assertion `filename != NULL
> &&
> >>>> strlen (filename) > 0' failed /etc/xdg/xfce4/xinitrc: line 106:
> >>>> 4105 Segmentation fault xfwm4 --daemon ** Message: Cannot connect
> >>>> to session manager /etc/xdg/xfce4/xinitrc: line 123: 4108
> >>>> Segmentation fault xfcalendar /etc/xdg/xfce4/xinitrc: line 123:
> >>>> 4110 Segmentation fault $panel
> >> sean> So X appears to start - Fedora logo comes up; no error on
> >> sean> Xorg.0.log.
> >>
> >> sean> But no gui's will start - not gnome,kde or xfce. Does this mean
> >> sean> X isn't starting correctly so that the gui's can start? If not,
> >> sean> what happens between X coming up and the gui's starting?
> >>
> >> sean> metacity?  Is there any way to cause metacity to do verbose
> >> sean> logging?
> >>
> >> sean> Again, is there any way to get more verbose and clearer logs.
> >>
> >> sean> Or any other hints?
> >>
> >> Well, not sure on gnome or KDE, but the above Xfce output looks like
> >> the problem is the Seg faults when trying to start up. :( 
> >>
> >> Several other things to try: 
> >>
> >> - If you create a new fresh user, can they login ok?
> >> (This would point to it being you users gtk theme, fonts setup or the
> >> like). 
> >>
> > Yes! I can login as a new user. Switched xorg.conf back to using xfs, 
> > and that worked to.
> >
> > ( Oddly, when I logout as the new user, I'm left at console 7 with a 
> > black screen and a blinking cursor at top left. console 1, witch I have 
> > to manually switch to, shows the usual X shutdown stuff.)
> >
> >> - Could be some package leftover from your rawhide install causing
> >> issues, perhaps a 'yum list extras' to identify any packages that are
> >> not in repositories you have currently enabled?
> >>
> >> - Could you try and run xfwm4 under a debugger?
> >> 0) yum --enable-repo=core-debuginfo install xfwm4-debuginfo
> > yum --enablerepo=extras-debuginfo ........  :)
> >
> >> 1) Run X from a console, "xinit /usr/X11R6/bin/xterm &"
> >> 2) Switch to a console (Ctrl+Alt+F1)
> >> 3) Set the display "export DISPLAY=:0.0"
> >> 4) run "gdb /usr/bin/xfwm4"
> >> 5) if/when it crashes, run 'bt' and send me the backtrace. 
> >>
> > Well, it appears to be font related. I hope you can figure out what the 
> > font problems is though. 
>  
>  Does 
>  rm -f ~/.fonts.cache-2 \
>        ~/.rh-fontconfig/.fonts.cache-2 \
>        ~/.fonts/fonts.cache-2
>  
>  fix things up?
>   
I'll do that, but shouldn't fc-cache  -f ( which I've done) do he same 
thing?

In any case, I do NOT have a ~/.fonts dir. should I? If so, what do I 
install to get it, or is it just user fonts.

sean

[Ray Strode]

Hi,

>  I'll do that, but shouldn't fc-cache  -f ( which I've
> done) do he same 
>  thing? 
You'd think, but there might be a bug somewhere.  I'm really just
guessing at this point.  The thing is, for a while in rawhide fontconfig
was really broken and generated broken font cache files.  It seems like
you might be hitting that.

>  In any case, I do NOT have a ~/.fonts dir. should I? If so, what do I
>
>  install to get it, or is it just user fonts. 
it's just for user fonts, if you
don't have it no big deal.

--Ray

rawhide->fc5; X won't start : Solved

[sean darcy]

Ray Strode wrote:
>  Hi,
>  
>  
> >I'll do that, but shouldn't fc-cache  -f ( which I've done) do he same 
> >thing?
>  
>  You'd think, but there might be a bug somewhere.  I'm really just
>  guessing at this point.  The thing is, for a while in rawhide fontconfig
>  was really broken and generated broken font cache files.  It seems like
>  you might be hitting that.
>  
>   
Great guess. You're my vote for Fedora Man of the Year.

Thanks a lot to all who helped me with this.

sean

> >In any case, I do NOT have a ~/.fonts dir. should I? If so, what
> do I 
> >install to get it, or is it just user fonts.
>  
>  it's just for user fonts, if you don't have it no big deal.
>  
>  --Ray
>

[Jef Spaleta]

On 4/1/06, sean <seandarcy2(a)gmail.com&gt; wrote:
>  So X appears to start - Fedora logo comes up; no error on
> Xorg.0.log.
>
>  But no gui's will start - not gnome,kde or xfce. Does this mean X isn't
>  starting correctly so that the gui's can start? If not, what happens
>  between X coming up and the gui's starting? 
Look in the xsession error files. Which should be in /tmp with filenames like
xses-username.randomstring

There may be noise in there that from applications that is unrealted
with your current problem so don't freak out if you see gobs of error
outputs in that file.

-jef

[sean darcy]

Jeff Spaleta wrote:
>  On 4/1/06, sean <seandarcy2(a)gmail.com&gt; wrote:
> > So X appears to start - Fedora logo comes up; no error on Xorg.0.log.
> >
> > But no gui's will start - not gnome,kde or xfce. Does this mean X isn't
> > starting correctly so that the gui's can start? If not, what happens
> > between X coming up and the gui's starting?
>  
>  Look in the xsession error files. Which should be in /tmp with filenames like
>  xses-username.randomstring
>  
>  There may be noise in there that from applications that is unrealted
>  with your current problem so don't freak out if you see gobs of error
>  outputs in that file.
>   
I've been looking for the xsession error files, but I can't find them. 
They aren't in /tmp ( I looked, I tried find /tmp -name "xse*" and find 
/tmp -name ".xse*" ) or in /root.

locate doesn't find them. Is there some switch that causes them to be 
generated?

sean