|
|
Linux Technical issues under Linux platform |
|
Thread Tools | Display Modes |
|
05-11-2009, 08:12 AM | #1 |
Master
Join Date: Jul 2008
Location: South Central USA
Posts: 260
|
A Possible workaround to the SEGFAULT issue..
This thread refers to the issues discussed here:
http://regnumonlinegame.com/forum/sh...ad.php?t=36774 ******* THIS METHOD IS NOW OBSOLETE ************ it occurred to me that we should be able to get around this using a chroot environment and I have begun some loose scripting to build a hardy chroot, configure it, install RO and get it to run from your current desktop. This means your PC would NOT be locked into using Hardy (2 versions old now) you would just simply need a very basic install of hardy.. just afew hundred megs.. debootstraped into a local directory or partition and properly setup... then a wrapper script to run RO bottled in this environment. This is the same way 64bit users once solved the problem of needing to run 32bit apps... and Sounds "simple" enough but I have hit some snags that I am hoping you guys here can help with.... On my test platform (Ubuntu Intrepid) I tried this with a Hardy chroot and was excited to see RO run..RO was started like this: Code:
sudo chroot ./test-root/ /opt/regnum/rolauncher Code:
libGL: Display claims not to support XXXX visual So I decided I would start a seperate X session inside the chroot on :1. Desktop starts as normal but RO gives the old "Your video card is too old" crap. Upon further investigation I found that ' glxinfo |grep direct ' says I have no direct rendering... and this is where Im stuck. I followed these instructions on setting up the environment which suggests that DRI should "just work" https://wiki.ubuntu.com/DebootstrapChroot also as a side note.. I followed the examples above to create a hardy chroot inside hardy.. installed xorg, mesa, openal, gtk and RO and running RO worked perfectly as I tried 1st above (sudo chroot ./test-root/ /opt/regnum/rolauncher). Im guessing the graphics issues I experienced on the Intrepid <==> Hardy chroot is due to incompatibilities between the mesa libs being used and the running X server. ... Also, starting a X session under this chroot also showed that DRI was NOT available. So this is where I will start trying to solve as I feel at this point running it in chroot with the Hardy libs on a Intrepid X server will be a lost cause. Froste? imefisto? Xephandor? anyone? Can this be solved? if so I can provide a nice little wrapper to work around this issue.. thanks, -onemyndseye *** EDIT: Issues to overcome ------------------------------------------------- [SOLVED] 1. More secure xhost incantation [SOLVED] 2. Solid way to move away from running as root 3. ?? Last edited by onemyndseye; 10-29-2009 at 08:49 PM. Reason: To clarify some points |
05-11-2009, 08:35 AM | #2 |
Pledge
Join Date: Dec 2007
Location: Behind you
Posts: 17
|
It might be due to incompatibilities between the DRI version used in the kernel and the x-driver used in your chroot.
Could you post the output of Code:
grep "EE" /var/log/Xorg.0.log
__________________
Speed kills... Accelerate! |
05-11-2009, 08:59 AM | #3 |
Master
Join Date: Jul 2008
Location: South Central USA
Posts: 260
|
Just thought of of posting that myself .. however as stated above DRI does not enable on a Hardy chroot inside a Hardy install.. all versions all the same -- this was my "control" for testing and I will not move further until DRI will function on this control... once this works it should work on the Inptrepid ==> Hardy chroot unless as you stated we run into module problems
I believe the related portion is: Code:
drmOpenDevice: node name is /dev/dri/card0 drmOpenDevice: open result is 8, (OK) drmOpenDevice: node name is /dev/dri/card0 drmOpenDevice: open result is 8, (OK) drmOpenByBusid: Searching for BusID pci:0000:01:00.0 drmOpenDevice: node name is /dev/dri/card0 drmOpenDevice: open result is 8, (OK) drmOpenByBusid: drmOpenMinor returns 8 drmOpenByBusid: drmGetBusid reports pci:0000:01:00.0 (II) [drm] DRM interface version 1.0 (EE) [drm] Could not set DRM device bus ID. (EE) RADEON(0): [dri] DRIScreenInit failed. Disabling DRI. Last edited by onemyndseye; 05-11-2009 at 09:14 AM. |
05-11-2009, 11:01 AM | #4 |
Duke
Join Date: Nov 2006
Location: 0x00CAFE
Posts: 3,366
|
Wow, this is a great idea! Using chroot to run Regnum on new Ubuntu releases, without having to keep your old release just to play. Nice workaround. :P
Could you please post the kernel version numbers for both Hardy and Intrepid? I don't know if chroot may have problems if you're loading modules from older kernel releases, as qwasd said.* Maybe this can help: http://www.phoronix.com/forums/showp...18&postcount=4 *In fact, I only use chroot to build packages, and it's been a long time since the last time. =/
__________________
I don't have a solution, but I admire the problem. |
05-11-2009, 11:22 AM | #5 |
Master
Join Date: Jul 2008
Location: South Central USA
Posts: 260
|
Thanks ArticWolf
Pretty big difference in versions: Hardy: 2.6.24-24 Intrepid: 2.6.27-7 Jaunty: 2.6.28-11 (The finial target) However, again we need a proof of concept.. this means a Hardy <==> Hardy chroot should have Direct Rendering.. and at this point it does not. This control setup is identical to its host **** EDIT: Xeph, according to the link you provided Direct Rendering on a secondary desktop is not implemented in DRI bummer --- Hopes of running RO in a seperate chrooted X session are dashed, atleast for now That being the case I suppose it wouldnt hurt to take a solid pass at running it chrooted without the 2nd X session but it doesnt look good. I'll post back with some debug info Last edited by onemyndseye; 05-14-2009 at 02:33 PM. |
05-11-2009, 08:57 PM | #6 | ||
Initiate
Join Date: Aug 2008
Location: San Luis - Argentina
Posts: 213
|
Quote:
Quote:
Considering your info about your kernel's version, maybe that's the problem.
__________________
Inek Bárbaro en desarrollo. Sólo sangre azul. Sombras de Alsius (todavía me queda grande el cartelito) Hasta siempre Feo...G* |
||
Tags |
linux, linux misc. |
|
|