You are not logged in.

#1 2017-04-05 22:27:29

kutabe
Member
From: Ukraine, Kiev
Registered: 2017-04-05
Posts: 3
Website

catalyst-hd234k Error: unable to open display (null)

Good evening.
I've been trying to install proprietary drivers on my integrated Radeon HD3000 and encountered some issues.
What I've done:
I've installed xorg < 1.13

xorg112/xorg-server
xorg112/xorg-server-utils
xorg112/xf86-input-evdev
xorg-xintit

And drivers

catalyst-hd234k/catalyst-hook
catalyst-hd234k/catalyst-utils
catalyst-hd234k/catalyst-libgl
catalyst-hd234k/opencl-catalyst
catalyst-hd234k/lib32-catalyst-utils
catalyst-hd234k/lib32-catalyst-libgl
catalyst-hd234k/lib32-opencl-catalyst
catalyst-hd234k/mesa-noglvnd

Added `nomodeset` kernel parameter @ /boot/syslinux/syslinux.cfg

-- APPEND root=/dev/sda1 rw
++ APPEND root=/dev/sda1 rw nomodeset

And rebooted.
Next, I've generated xorg.conf with

aticonfig --initial

And tried to start X

startx

But what I've got was black screen and lockup.
After forced reboot I've tried

fglrxinfo

And got this

Error: unable to open display (null)

And

cat /var/log/Xorg.0.log

was empty.
My Kernel version is

4.10.6-1-ARCH

EDIT:
Same with

4.10.8-1-ARCH

LOGS:
So here is dmesg and journalctl -b
before I start X. I've tried to capture logs after start X
but I can't, 'coz I get screenlock every time I try. I even
tried to do the trick, in one session I did

sleep 10 && dmesg > dmesg_after.log && journalctl -b > journalctl_after.log

but file was empty. Same with delayed `pkill X`, it didn't werk.

Last edited by kutabe (2017-04-07 01:43:27)

Offline

#2 2017-04-06 13:00:50

Lone_Wolf
Member
From: Netherlands, Europe
Registered: 2005-10-04
Posts: 11,867

Re: catalyst-hd234k Error: unable to open display (null)

AUR catalyst-total-hd234k is at version 13.1-42 and was updated just 3 days ago. What version do you get from the repo ?


Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.


(A works at time B)  && (time C > time B ) ≠  (A works at time C)

Online

#3 2017-04-06 21:50:36

kutabe
Member
From: Ukraine, Kiev
Registered: 2017-04-05
Posts: 3
Website

Re: catalyst-hd234k Error: unable to open display (null)

Thanks for reply, my catalyst-hook version is `13.1-42`.
Somehow I managed to start xorg-twm after I nuked `/etc/X11/xorg.conf`
So I think the root of my problems somewhere there, maybe.
Here is what I get from

aticonfig --initial

=> xorg.conf
EDIT:
Nope, I use vesa right now, that's why xorg-twm started.
I've tried

aticonfig --acpi-services=off

and it didn't help.

Last edited by kutabe (2017-04-07 00:34:58)

Offline

#4 2017-04-07 07:03:17

Vi0L0
Member
From: Poland
Registered: 2009-06-24
Posts: 1,349
Website

Re: catalyst-hd234k Error: unable to open display (null)

It's probably fglrx module problem which seems to not work well with kernels newer than 4.7. I don't have legacy hardware to check it and untill now none of the users gave me any additional information about it. It's also probable that I may be too weak in c and kernel itself to figure it out, but I would like to try anyway.
I was hoping to see something in dmesg but it, strangely, is clear...
Maybe you will be able to get more info from X using something like `startx &> logfile` or maybe try something like `strace startx &> logfile`.
Additionally you may debug deeper using caredumpctl, it remember segfaults so you won't have to lockup your pc to check it, maybe there you will find some info:
coredumpctl list
coredumpctl gdb <pid>
Also: is it hard lock? did you try to ssh into system after lock?
Can you check if downgrading kernel to 4.7.x is helping, just to be sure that it is module-related?

Last edited by Vi0L0 (2017-04-07 07:06:27)

Offline

#5 2017-04-09 03:48:21

kutabe
Member
From: Ukraine, Kiev
Registered: 2017-04-05
Posts: 3
Website

Re: catalyst-hd234k Error: unable to open display (null)

Vi0L0 wrote:

It's probably fglrx module problem which seems to not work well with kernels newer than 4.7. I don't have legacy hardware to check it and untill now none of the users gave me any additional information about it. It's also probable that I may be too weak in c and kernel itself to figure it out, but I would like to try anyway.
I was hoping to see something in dmesg but it, strangely, is clear...
Maybe you will be able to get more info from X using something like `startx &> logfile` or maybe try something like `strace startx &> logfile`.
Additionally you may debug deeper using caredumpctl, it remember segfaults so you won't have to lockup your pc to check it, maybe there you will find some info:
coredumpctl list
coredumpctl gdb <pid>
Also: is it hard lock? did you try to ssh into system after lock?
Can you check if downgrading kernel to 4.7.x is helping, just to be sure that it is module-related?

I confirm that it works with kernel version 4.7.6, however, I'm not satisfied with result.
Thanks for your work, and sorry, I can't investigate what exactly causes lockup.
I rolled back to free drivers (god bless btrfs) and decided that it's better to either stick with free drivers either buy damn nvidia video card.

Offline

Board footer

Powered by FluxBB