You are not logged in.
Hi Guys,
I am hoping one of you can help as I am running out of ideas as to what could be wrong,
Here is some background on the problem I have run into,
I seem unable to perform a system wide upgrade from kernel 2.6.35 running xbmc 9.11-18 without it breaking XBMC to the point where it won't start. I know that the xbmc (9.11-19) package was recently updated to include a dependency on the the mesa-demos package (now containing glxinfo) due to a xbmc having a dependency on glxinfo but xbmc still refuses to start and instead simply leaves a blank screen at the console.
Looking at the Xorg.0.log X seems to start ok as it did before the upgrade and it recognises all the hardware including my Nvidia MSI G210 grahics card and correctly scans the various resolutions of my Samsung TV. I see no errors of any kind in Xorg.0.log
The kernel gets upgraded to 2.6.36 and the Nvidia driver package was also upgraded during the system wide upgrade along with xbmc and X. I did not encounter any errors during the upgrade.
After system upgrade dmesg shows no errors or failures, xbmc will simply not start and as I say just gives a blank screen at console. I can ssh into the box as normal without issue.
I don't run any kind of desktop with X, it is configured simply to boot straight into xbmc using the following /etc/inittab and .xinitrc in my home directory which has worked perfectly for the last 18 months or more following regular system wide upgrades.
Contents of /etc/inittab
## Only one of the following two lines can be uncommented!
# Boot to console
#id:3:initdefault:
# Boot to X11
id:5:initdefault:
rc::sysinit:/etc/rc.sysinit
rs:S1:wait:/etc/rc.single
rm:2345:wait:/etc/rc.multi
rh:06:wait:/etc/rc.shutdown
su:S:wait:/sbin/sulogin -p
# -8 options fixes umlauts problem on login
c1:2345:respawn:/sbin/agetty -8 38400 tty1 linux
c2:2345:respawn:/sbin/agetty -8 38400 tty2 linux
c3:2345:respawn:/sbin/agetty -8 38400 tty3 linux
c4:2345:respawn:/sbin/agetty -8 38400 tty4 linux
c5:2345:respawn:/sbin/agetty -8 38400 tty5 linux
c6:2345:respawn:/sbin/agetty -8 38400 tty6 linux
# Hypervisor Virtual Console for Xen and KVM
#h0:2345:respawn:/sbin/agetty -8 38400 hvc0 linux
ca::ctrlaltdel:/sbin/shutdown -t3 -r now
# Example lines for starting a login manager
x:5:once:/bin/su lee -l -c "/bin/bash --login -c startx >/dev/null 2>&1"
#x:5:respawn:/usr/bin/xdm -nodaemon
#x:5:respawn:/usr/sbin/gdm -nodaemon
#x:5:respawn:/usr/bin/kdm -nodaemon
#x:5:respawn:/usr/bin/slim >& /dev/null
# End of file
Contents of .xinitrc
exec /usr/bin/ck-launch-session /usr/bin/dbus-launch --exit-with-session /usr/bin/xbmc --standalone -fs
I notice this morning that the xbmc package has now been upgraded to the official Dharma final release of 10.0 I have not tried upgrading this package yet as I have currently rolled back to my working backup running kernel 2.6.35 and xbmc 9.11-18
Am I missing a significant change in the way X windows or xbmc starts up? I am really stuck on this as I can't find any significant error in /var/log and as xbmc won't start at all I have nothing to review in ~/.xbmc/temp/xbmc.log
Any ideas on this are gratefully received.
Last edited by adr3nal1n (2010-12-22 13:10:08)
Offline
I'd be inclined to comment out the line in the /etc/inittab that does the boot into XBMC (or change the default runlevel to 3), get to a command line and run XBMC manually, possibly you'd get some output to the console that might help.
Offline
Sounds like a plan, I'll give that a go this evening and report back my findings. As you say this will hopefully provide some clues.
Thanks again.
Offline
Welcome, it's a method worked for me when the glxinfo/mesa-demos problem cropped up.
Offline
Same problem here after updating XBMC today.
Error message on boot:
init: Id "x" respawning too fast: diabled for 5 minutes
I did a complete removal of XBMC, then reinstalled. When pacman finishes the install, the following error message appears when XBMC is installed:
error: command failed to execute correctly
I've already tried booting direct to console and booting without xorg.conf. Still no go.
Offline
'Seems to start okay' is not something we can work with. You can assume a lot, but without testing or verifying, nobody can tell what your problem is. So post that Xorg.0.log.
I have had my Xbmc setup break a few times myself, it invariably was Xorg that broke, not Xbmc. Xbmc just runs on top of X, and whether or not you run a WM/DE, that doesn't make a single difference. If X breaks, X breaks.
As velophile said, you can always just run startx from runlevel 3 and see what happens, or call xbmc directly (not the script, the binary itself). The latter gets my vote, the wrapper script in /usr/bin isn't always handy.
Got Leenucks? :: Arch: Power in simplicity :: Get Counted! Registered Linux User #392717 :: Blog thingy
Offline
Solved my problem by installing missing library libass 0.9.11-1
sudo pacman -S libass
My HTPC once again boots direct to XBMC, and all is well.
Offline
Ok .:B:., will launch from the xbmc binary directly and will post Xorg.0.log this evening.
Will also try installing libass too just in case my setup has developed the same quirk as Skruf. Although the xbmc package does not list libass as a dependency?
Thanks very much for your help on this.
Offline
Usually tells you when you try and start it from the shell if it's missing a dependancy...thats why I suggested trying it without the runlevel5/autologon, that's how I spotted I was missing the glxinfo stuff.
Offline
Hi Guys,
Ok, I ran the system wide upgrade again and got the following results,
If I try and start xbmc from the console I get the following error,
Error: unable to open display
XBMC needs hardware accelerated OpenGL rendering.
Install an appropriate graphics driver.
Please consult XBMC Wiki for supported hardware
http://xbmc.org/wiki/?title=Supported_hardware
xdpyinfo: unable to open display "".
XBMC cannot run unless the
screen color depth is atleast 24 bit.
Please reconfigure your screen.
Does this mean my xorg.conf is no good?
Offline
It looks like it, or maybe the nvidia drivers aren't installed or installed properly. I'd start with the nvidia wiki and see if that helps : https://wiki.archlinux.org/index.php/NVIDIA (save you looking it up), running
nvidia-xconfig
might be a good first step
Offline
Ok .:B:., will launch from the xbmc binary directly and will post Xorg.0.log this evening.
Will also try installing libass too just in case my setup has developed the same quirk as Skruf. Although the xbmc package does not list libass as a dependency?
Thanks very much for your help on this.
The PKGBUILD (I just checked it when I saw mentioned Arch bumped to 10.0) has --disable-external-libass, but the problem is, if the maintainer has built the package on a contaminated system (with libass being installed), then the external libass will be picked up no matter what, leaving you indeed with a broken Xbmc binary. I have had that problem myself a few times, building Xbmc from SVN. It's the main reason why I now build in a chroot.
Check dmesg as well and lsmod output to see if the nvidia binary driver is loaded (dmesg will often be enough to debug kernel module errors).
Got Leenucks? :: Arch: Power in simplicity :: Get Counted! Registered Linux User #392717 :: Blog thingy
Offline
Still no luck with the Nvidia driver. Have re-installed it using pacman and it installs without error.
dmesg shows the following entries for nvidia:
nvidia: module license 'NVIDIA' taints kernel.
nvidia 0000:01:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
nvidia 0000:01:00.0: setting latency timer to 64
NVRM: loading NVIDIA UNIX x86_64 Kernel Module 260.19.29 Wed Dec 8 12:08:56 P ST 2010
Have tried nvidia-xconfig to generate a new xorg.conf but this makes no difference. If I run xbmc from the console I get the same error as in my previous post.
Have noticed from trying to re-install xbmc using pacman i get "error: command failed to execute correctly"
Any ideas?
Offline
Thanks for your continued help on this guys,
'lsmod |grep nvidia' provides the following output:
nvidia 10081406 0
i2c_core 18726 2 nvidia,i2c_i801
Please let me know if I can provide any other info.
Offline
Well nvidia loads fine as you can see, what about that Xorg log? Please pastebin it.
Got Leenucks? :: Arch: Power in simplicity :: Get Counted! Registered Linux User #392717 :: Blog thingy
Offline
I've had the "command failed to execute correctly" error on mine when I updated XBMC last night, it's still working ok though.
It looks as though possibly the DISPLAY variables not set: http://forums.debian.net/viewtopic.php?f=5&t=15400
As .:B:. says best chance is posting the xorg.0.log and probably the xorg.conf, I doubt it'll be me but someone might be able to make sense of it.
Offline
@Velophile, you're right, the DISPLAY variable is not set on my 2.6.35 system. If I run glxinfo I get an error of "Error: unable to open display" and the same with xdpyinfo I also get the same error.
If I then "export DISPLAY=:0" using the user account I use to start X both glxinfo and xdpyinfo work correctly.
How do I set the above variable to be permanent across reboots?
One other thing I noticed whilst looking through the X11 files:
My 2.6.35 working system /etc/X11/xinit/xserverrc contains the following:
exec /usr/bin/X -nolisten tcp
I am fairy sure after I tried the upgrade again last night this file had some extra characters at the end of the line, something like "@:$" ? Could you check your /etc/X11/xinit/xserverrc please to see if it is different to my 2.6.35 working system as I feel you may be onto something with the display variable not being set and X now depending on it.
Once I have the performed the system wide upgrade again this evening (as I have currently rolled back to my working 2.6.35 system) I will try again and then send my Xorg.0.log into pastebin
Thanks again to both you and .:B:. for your continued help on this, it's much appreciated.
Offline
BTW, here is the /etc/X11/xorg.conf from my working 2.6.35 system:
Section "Device"
Identifier "nvidia"
Driver "nvidia"
Option "NoLogo" "true"
Option "DynamicTwinView" "false"
Option "NoFlip" "false"
Option "FlatPanelProperties" "Scaling = Native"
Option "ModeValidation" "NoVesaModes, NoXServerModes"
Option "UseDisplayDevice" "DFP-1"
Option "ModeDebug" "true"
Option "HWCursor" "false"
EndSection
Section "Screen"
Identifier "screen"
Device "nvidia"
EndSection
Section "Extensions"
Option "Composite" "false"
EndSection
Offline
I have seen someone else mentione those extra "$@" characters did in break Xorg for him. So you might want to take those out (don't remember the exact sequence or characters).
Got Leenucks? :: Arch: Power in simplicity :: Get Counted! Registered Linux User #392717 :: Blog thingy
Offline
You're welcome.
I'll have a look if I get chance tonight and post my /etc/X11/xinit/xserverrc
Offline
Solved my problem by installing missing library libass 0.9.11-1
sudo pacman -S libass
My HTPC once again boots direct to XBMC, and all is well.
I made a new install of xbmc on a new arch x64 installation and confirm that libass is a missing dependency in the xbmc package. xbmc will not start unless I also install libass in addition to the xbmc package.
Last edited by ronnylov (2010-12-22 12:19:56)
Offline
Offline
OMG its working!
I couldn't wait to try again this evening so I ssh'd in to the box, did the upgrade and then started trying stuff to fix it.
Setting the display variable had no effect and so, before delving too deep into xorg.conf troubleshooting and Xorg logs, I then tried the earlier suggestion of adding libass with 'pacman -S libass' rebooted and BAM, it's now working! I'm sure I tried that last night but maybe I installed libass before I did the upgrade last night whereas this time I installed libass after the upgrade. Apologies for the Schoolboy error.
I can't verify it has worked physically as I am not at home at the moment to try out XBMC on the TV, but I can see from ssh-ing into the box and running top that xbmc.bin and X are running and the ~/.xbmc/temp/xbmc.log now shows xbmc has started successfully and without error. So I am hopeful that all will be well when I pick up the remote when I get home! :-)
At last I may just be able to try out XBMC Dharma! :-)
Massive thanks to Velophile, .:B:. and Skruf for all their help and advice.
Happy Christmas and all that!
Offline
Whoops! took too long to post my message to one of the earlier posts before I saw the post from ronnylov.
Nice to know I wasn't the only one with this issue.
Thanks again guys, these forums are fantastic for support.
Offline
Winner! good to know it's sorted. I tried to raise a bug on the tracker about this but it needs me to create an account as I've not got one, if any of you have an account already could you do raise a bug please? I'm not a fan of creating accounts that'll probably only get used once.
Offline