Red Hat Bugzilla – Bug 435907
X does not start using OpenChrome
Last modified: 2008-03-10 20:23:41 EDT
Description of problem:
I have a VN800 chipset in my laptop. While Openchrome worked fine for me in
Fedora 8, after having upgraded to rawhide, it does not start
Version-Release number of selected component (if applicable):
Latest Openchrome F9 package
Steps to Reproduce:
I am going to attach Xorg.0.log and xorg.conf
Created attachment 296731 [details]
Created attachment 296732 [details]
Just found out that it also happens using the vesa driver, but using it, I see
gdm but when I click some buttons in it, Xorg crashes, also.
Try to add Option "ActiveDevice" "LCD" in the device section of your xorg.conf.
When I use the Option, the screen stays black but I do not get any backtrace in
The vesa driver loads, but it crashes when I press any key.
(In reply to comment #5)
> When I use the Option, the screen stays black but I do not get any backtrace in
> my Xorg.0.log.
Can you please attach the log with the ActiveDevice option set ?
Also could you please try with Option "ForcePanel" (and attach the log with this
option too) ?
Created attachment 296755 [details]
Xorg.0.log with "ActiveDevice" "LCD"
When I use Option "ForcePanel" I get a picture, but when I type something, X
crashes (as with vesa) so probably an Xorg problem (and no openchrome one).
Created attachment 296758 [details]
Xorg.0.log with "ForcePanel"
A fixed package should appear in today's Rawhide push.
Here's the koji build link :
For the record, the evdev crash was fixed by an update to xorg-x11-server
*** Bug 435692 has been marked as a duplicate of this bug. ***
When I use the updated package without having "ForcePanel" in xorg.conf, the
same issue still appears.
Created attachment 297186 [details]
Xorg.0.log with latest package
error setting MTRR (base = 0xd1000000, size=0x00009000, type= 1) Invalid
X: symbol lockup error: /usr/lib/xorg/modules/drivers//openchrome_drv.so:
undefinied symbol: SUBVENDOR_ID
xinit: Connection refused [...]
# rpm -qa | grep openchrome
Created attachment 297188 [details]
Xorg.0.log with "ForcePanel" with -9 package
Even does not start with ForcePanel using the Koji package.
xorg-x11-drv-openchrome-0.2.901-8.fc9.i386 works using ForcePanel ( I rechecked )
Ok, there was a typo in the upstream commit.
New fixed build on it's way :
Created attachment 297299 [details]
Xorg.0.log with driver version -11
Created attachment 297302 [details]
Xorg.0.log with -12
Hopefully, this is the good one...
Take care the release is still -12.
Created attachment 297353 [details]
Xorg.0.log with -12 2
Does not work, but the log says something new which is goood ;)
ok, no need for ForcePanel anymore, but now we're hitting something new, the
VRAM size detection is busted :
(--) CHROME(0): Probed amount of VideoRAM = 1024 kB
You probably can workaround this with Option "VideoRAM" "65536"
What does happen if you change the amount of VRAM in BIOS ? Does the probed VRAM
value move proportionally ? Does it report 512K if you set 32 MB instead of 64 MB ?
> Does it report 512K if you set 32 MB instead of 64 MB ?
No it still reports 1024K
> No it still reports 1024K
( I changed "Framebuffer Size" in my BIOS, there aren't any other options like that)
Also, using "VideoRAM" "65536" fixes the issue: I have working Xv and do not
have to use ForcePanel.
Ok, thx again for all the tests. Latest Koji build should do the trick. Not as
clean has I'd like it to be, but we need something for F9 Beta.
Created attachment 297376 [details]
Xorg.1.log from /13
Martin, I'd be grateful if you can test this one asap tomorrow :
$ rpm -qa | grep openchrome
I am going to attach xorg.conf and Xorg.0.log
Created attachment 297429 [details]
Xorg.0.log with -14
Created attachment 297430 [details]
xorg.conf with -14 (WORKS,YAY!!)
Thanks you again, Martin :-) The beers are in the fridge...