The AMD/ATI Bar & Grill

Vi0L0 · 2013-12-13 09:19:40

@dfanz0r:
shot in the dark: maybe disabling xinerama and xrandr12 will help?

okraits · 2013-12-16 09:12:06

Hello,

i upgraded my system yesterday and after the reboot, i got complete graphic corruption after i issue startx (i can't read anything). I'm using catalyst 13.11-10 from the catalyst repo with a Asus Radeon HD 5770 and the i3 window manager. When using the vesa driver, the issue doesn't come up. I also tried downgrading linux, xorg-server and xorg-server-common, but that didn't solve the issue. The following packages were updated before the issue:

[2013-12-15 23:07] [PACMAN] upgraded linux-api-headers (3.10.6-1 -> 3.12.4-1)
[2013-12-15 23:07] [PACMAN] upgraded glibc (2.18-10 -> 2.18-11)
[2013-12-15 23:07] [PACMAN] upgraded binutils (2.23.2-3 -> 2.24-1)
[2013-12-15 23:07] [PACMAN] upgraded firefox (26.0-1 -> 26.0-2)
[2013-12-15 23:07] [PACMAN] upgraded gcc-libs (4.8.2-5 -> 4.8.2-6)
[2013-12-15 23:07] [PACMAN] upgraded gcc (4.8.2-5 -> 4.8.2-6)
[2013-12-15 23:07] [PACMAN] upgraded inetutils (1.9.1.341-1 -> 1.9.1.341-2)
[2013-12-15 23:07] [PACMAN] upgraded lib32-glibc (2.18-10 -> 2.18-11)
[2013-12-15 23:07] [PACMAN] upgraded lib32-gcc-libs (4.8.2-5 -> 4.8.2-6)
[2013-12-15 23:07] [PACMAN] upgraded linux (3.12.3-1 -> 3.12.5-1)
[2013-12-15 23:07] [PACMAN] upgraded make (4.0-1 -> 4.0-2)
[2013-12-15 23:07] [PACMAN] upgraded man-pages (3.54-1 -> 3.55-1)
[2013-12-15 23:07] [PACMAN] upgraded mesa (9.2.4-1 -> 9.2.5-1)
[2013-12-15 23:07] [PACMAN] upgraded mpfr (3.1.2.p4-1 -> 3.1.2.p5-1)
[2013-12-15 23:07] [PACMAN] upgraded python2-django (1.6-1 -> 1.6.1-1)
[2013-12-15 23:07] [PACMAN] upgraded util-linux (2.24-1 -> 2.24-2)
[2013-12-15 23:07] [PACMAN] upgraded x264 (20131030-1 -> 20131030-2)
[2013-12-15 23:07] [PACMAN] upgraded xorg-server-common (1.14.4-1 -> 1.14.5-1)
[2013-12-15 23:07] [PACMAN] upgraded xorg-server (1.14.4-1 -> 1.14.5-1)

Did anybody else experience that issue or has a hint for me?

Thanks in advance!

Kuci · 2013-12-16 13:17:17

People, is there any OpenCL implementation for ati-dri ? Doesn't matter in which stage it currently is. Thank you.

okraits · 2013-12-16 14:55:17

Ok, i solved my issue by rebuilding the fglrx module with catalyst-hook.

Vi0L0 · 2013-12-16 16:08:17

Kuci wrote:

People, is there any OpenCL implementation for ati-dri ? Doesn't matter in which stage it currently is. Thank you.

Take a closer look at openl-mesa-git package from lordheavy's mesa-git repo:
http://pkgbuild.com/~lcarlier/mesa-git/

Last edited by Vi0L0 (2013-12-16 16:08:42)

dmitriy · 2013-12-17 01:23:46

Hey, folks!
Had a long-time struggle with Catalyst drivers and got to testing it all out, still not having an ideal answer in my findings.

Basically, there are two common problems I experience:

Problem #1 - brightness is hard to control - I have to press a brigtness increase or decrease button (Fn+F2 or Fn+F3 in my case) multiple times until it gets changed, in turn, skipping many values.
Problem #2 - brightness controls doesn't work after suspend (even when echo'ing values to the proper files)

Test results:

13.1 - currently used - brightness control works, but the driver is unstable - wake from suspend is a lottery, with a chance of freeze with a black screen (requires xorg113)
13.4 - #1, tested with xorg113 and xorg114
xorg 114 required to all following
13.6 - #1
13.7 - #1
13.8 - #1
13.10 - #2
13.11 - #2

Used archived files for catalyst-hook and catalyst-utils from http://catalyst.wirephire.com/repo/catalyst/x86_64/

System:

Kernel: 3.12.3-1-ARCH
Laptop: HP Pavilion G6-1201sy with AMD Dual-Core E2-3000M APU with Radeon HD 6380G Dual Graphics
I guess there'll be more info needed - I'll provide any you'd ask

What'll be the best way for me to get a more stable experience? Fixing the older ones, newer ones, installing something?
Hope there's anyone who can help with this kind of problem

Cheers,
Dmitriy

Last edited by dmitriy (2013-12-17 01:26:16)

clfarron4 · 2013-12-17 19:30:40

dmitriy wrote:

Hey, folks!
Had a long-time struggle with Catalyst drivers and got to testing it all out, still not having an ideal answer in my findings.
Basically, there are two common problems I experience:
Problem #1 - brightness is hard to control - I have to press a brigtness increase or decrease button (Fn+F2 or Fn+F3 in my case) multiple times until it gets changed, in turn, skipping many values.
Problem #2 - brightness controls doesn't work after suspend (even when echo'ing values to the proper files)
Test results:
13.1 - currently used - brightness control works, but the driver is unstable - wake from suspend is a lottery, with a chance of freeze with a black screen (requires xorg113)
13.4 - #1, tested with xorg113 and xorg114
xorg 114 required to all following
13.6 - #1
13.7 - #1
13.8 - #1
13.10 - #2
13.11 - #2
Used archived files for catalyst-hook and catalyst-utils from http://catalyst.wirephire.com/repo/catalyst/x86_64/
System:
Kernel: 3.12.3-1-ARCH
Laptop: HP Pavilion G6-1201sy with AMD Dual-Core E2-3000M APU with Radeon HD 6380G Dual Graphics
I guess there'll be more info needed - I'll provide any you'd ask
What'll be the best way for me to get a more stable experience? Fixing the older ones, newer ones, installing something?
Hope there's anyone who can help with this kind of problem
Cheers,
Dmitriy

On one of my rigs, I am exactly the same with an AMD A4-3400M APU with Radeon HD 6480G Graphics. Personally, I'd rather have #2 than #1, so I've just stuck with the latest beta release of Catalyst.

lano1106 · 2013-12-18 04:51:59

Vi0l0,

I just installed the newest version of catalyst-generator 13.11-11

What are the new additions in this new release?

Are you publishing a small release note when releasing a new version so I can check myself next time that I am curious?

thanks,

p.s.: for a fraction of second when I saw your repo updating, I hoped for catalyst 13.12!

Vi0L0 · 2013-12-18 07:23:27

I've added patch which brings 3.13 kernel support (tested on < rc5, should also work with final release) to the 13.11 packages, I will also add it to other pkgs/repos later.
I used to write about changes on the aur comments (this time in catalyst-test*), but I will think about some release notes.
Yeah, I also can't wait for 13.12. It would be nice if AMD would release it this week because in the next week I will be afk traveling to my family.

dmitriy · 2013-12-18 08:05:35

clfarron4 wrote:

dmitriy wrote:
Hey, folks!
Had a long-time struggle with Catalyst drivers and got to testing it all out, still not having an ideal answer in my findings.
Basically, there are two common problems I experience:
Problem #1 - brightness is hard to control - I have to press a brigtness increase or decrease button (Fn+F2 or Fn+F3 in my case) multiple times until it gets changed, in turn, skipping many values.
Problem #2 - brightness controls doesn't work after suspend (even when echo'ing values to the proper files)
Test results:
13.1 - currently used - brightness control works, but the driver is unstable - wake from suspend is a lottery, with a chance of freeze with a black screen (requires xorg113)
13.4 - #1, tested with xorg113 and xorg114
xorg 114 required to all following
13.6 - #1
13.7 - #1
13.8 - #1
13.10 - #2
13.11 - #2
Used archived files for catalyst-hook and catalyst-utils from http://catalyst.wirephire.com/repo/catalyst/x86_64/
System:
Kernel: 3.12.3-1-ARCH
Laptop: HP Pavilion G6-1201sy with AMD Dual-Core E2-3000M APU with Radeon HD 6380G Dual Graphics
I guess there'll be more info needed - I'll provide any you'd ask
What'll be the best way for me to get a more stable experience? Fixing the older ones, newer ones, installing something?
Hope there's anyone who can help with this kind of problem
Cheers,
Dmitriy
On one of my rigs, I am exactly the same with an AMD A4-3400M APU with Radeon HD 6480G Graphics. Personally, I'd rather have #2 than #1, so I've just stuck with the latest beta release of Catalyst.

Is there any other way to reason with this?

brittyazel · 2013-12-19 07:18:31

Catalyst 13.12 stable was released tonight.

Vi0L0 · 2013-12-19 07:55:50

Fix list looks good. If nothing will change I will update later today.

brittyazel · 2013-12-19 08:25:25

"[384194]: Tear-Free Desktop sets V-Sync to 30Hz instead of 60Hz" I wonder if this bug is what was causing the laggyness when tear free is enabled.

Webmaniac2 · 2013-12-19 14:12:30

Vi0L0 wrote:

Fix list looks good. If nothing will change I will update later today.

It says that only Linux kernel 2.6 to 3.11 are supported. Does it mean it doesn't run on 3.12 at all?

snack · 2013-12-19 14:16:51

Webmaniac2 wrote:

Vi0L0 wrote:
Fix list looks good. If nothing will change I will update later today.
It says that only Linux kernel 2.6 to 3.11 are supported. Does it mean it doesn't run on 3.12 at all?

Have faith in Vi0l0, man. I'm still using legacy drivers version 13.1 (yes, January 2013) and thanks to Vi0l0's patch they work flawlessly with kernel 3.12. I bet that also 13.12 will run without any problem.

Vi0L0 · 2013-12-19 17:17:51

yeah, the bug is trivial, amd already fixed it in ie. 13.11 beta 9.4 (maybe also in older releases, edit: also in 13.11 beta 6) so i borrowed a fix from out there, LOL ;P...
let me just eat something ;P, then some testing etc...

Last edited by Vi0L0 (2013-12-19 17:54:10)

clfarron4 · 2013-12-19 17:30:08

dmitriy wrote:

clfarron4 wrote:
dmitriy wrote:
Snip
On one of my rigs, I am exactly the same with an AMD A4-3400M APU with Radeon HD 6480G Graphics. Personally, I'd rather have #2 than #1, so I've just stuck with the latest beta release of Catalyst.
Is there any other way to reason with this?

Aside from telling the AMD developers themselves, not really.

haagch · 2013-12-19 18:45:07

Vi0L0 wrote:

let me just eat something ;P, then some testing etc...

Relevant? http://ati.cchtml.com/show_bug.cgi?id=958#c1
edit: I probably misread that comment. I haven't tested anything with 3.12 yet.

Anyway, is there a patch for 3.13 rc- already?

Last edited by haagch (2013-12-19 18:46:25)

Vi0L0 · 2013-12-19 18:57:22

haagch wrote:

Vi0L0 wrote:
let me just eat something ;P, then some testing etc...
Relevant? http://ati.cchtml.com/show_bug.cgi?id=958#c1

Indeed

haagch wrote:

Anyway, is there a patch for 3.13 rc- already?

you may grab it from catalyst-test@AUR, or:

--- 13.11/common/lib/modules/fglrx/build_mod/kcl_acpi.c	2013-11-22 19:52:37.000000000 +0100
+++ 13.11/common/lib/modules/fglrx/build_mod/kcl_acpi.c	2013-11-25 11:48:51.227678982 +0100
@@ -778,7 +778,9 @@
 unsigned int ATI_API_CALL KCL_ACPI_GetHandles(kcl_match_info_t *pInfo)
 {
 #if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,12)
-    #if LINUX_VERSION_CODE >= KERNEL_VERSION(3,8,0)
+    #if LINUX_VERSION_CODE >= KERNEL_VERSION(3,13,0)
+        pInfo->video_handle = pInfo->pcidev->dev.acpi_node.companion;
+    #elif LINUX_VERSION_CODE >= KERNEL_VERSION(3,8,0)
         pInfo->video_handle = pInfo->pcidev->dev.acpi_node.handle;
     #elif LINUX_VERSION_CODE > KERNEL_VERSION(2,6,19)
         pInfo->video_handle = pInfo->pcidev->dev.archdata.acpi_handle;

clfarron4 · 2013-12-19 20:13:25

@Vi0L0: What does the catalyst-daemon package do (and why is it not mentioned on the Wiki)? I understand the others (hook, generator)

Vi0L0 · 2013-12-19 21:00:01

it was on the wiki but someone removed it, but i don't mind since arch sysvinit is... well... dead. I'm still supporting and rebuilding this package in case of some hardcore users .
Anyhow it is a sysvinit daemon able to build fglrx kernel module whenever it's necessary.

BTW:
[catalyst] and [catalyst-stable] will soon be equal, serving catalyst 13.12, same as AUR's main and test packages.
Current [catalyst-stable] with catalyst 13.4 inside will be moved to [catalyst-stable-13.4].
Current [catalyst] with catalyst 13.11 beta 9.4 inside will be moved to [catalyst-13.11-beta94].
In 3... 2... 1... Energize!

Last edited by Vi0L0 (2013-12-19 21:53:11)

life_enjoyer · 2013-12-20 02:22:38

I am not sure if someone already answered this question but why isnt this thread in hardware?

zyth88 · 2013-12-20 06:26:59

Is there a way to improve Team Fortress 2 performance with the Tear Free option on? I tried Serious Sam 3 and it ran well (better than the open source), although in TF2 I get more fps with the proprietary driver, but it feels sluggish. TF2 is much better with the open source driver.

Robsteady · 2013-12-20 16:22:24

Hey all, I'm new to Arch and I'm loving it so I'm trying to get myself figured out. I downloaded the AMD driver (13.11 v9.4) from the AMD site and I installed that package as I would previously with Ubuntu, Mint, Debian and Manjaro. I know the driver was only listing compatibility to kernel 3.11 but I went through with it anyway. Upon reboot it seemed like X didn't want to start so I switched terminals and ran the ATI uninstaller then ran pacman -S xorg. Once I rebooted again the screen would flicker to black then go to a distorted graphics and back to black. From there I am unable to switch to a terminal I can log in to and run any commands. Would it be worth taking the time to change grub to boot to the CLI and fix the installation or would it be better off doing a fresh install using Vi0L0's repositories for the drivers? I would have no problem doing a full system wipe, I'm anal about backing up my /home and I don't mind spending the time putting everything back together, I just don't want to run into a headache chasing the problem or any instability once I get X working again.

clfarron4 · 2013-12-20 16:45:25

Robsteady wrote:

Hey all, I'm new to Arch and I'm loving it so I'm trying to get myself figured out. I downloaded the AMD driver (13.11 v9.4) from the AMD site and I installed that package as I would previously with Ubuntu, Mint, Debian and Manjaro. I know the driver was only listing compatibility to kernel 3.11 but I went through with it anyway. Upon reboot it seemed like X didn't want to start so I switched terminals and ran the ATI uninstaller then ran pacman -S xorg. Once I rebooted again the screen would flicker to black then go to a distorted graphics and back to black. From there I am unable to switch to a terminal I can log in to and run any commands. Would it be worth taking the time to change grub to boot to the CLI and fix the installation or would it be better off doing a fresh install using Vi0L0's repositories for the drivers? I would have no problem doing a full system wipe, I'm anal about backing up my /home and I don't mind spending the time putting everything back together, I just don't want to run into a headache chasing the problem or any instability once I get X working again.

I would at least try booting into target=multi-user.target and trying to install one of the packages from Vi0l0's repository (I would recommend the catalyst-hook package and enable the service),

Arch Linux

#5976 2013-12-13 09:19:40

Re: The AMD/ATI Bar & Grill

#5977 2013-12-16 09:12:06

Re: The AMD/ATI Bar & Grill

#5978 2013-12-16 13:17:17

Re: The AMD/ATI Bar & Grill

#5979 2013-12-16 14:55:17

Re: The AMD/ATI Bar & Grill

#5980 2013-12-16 16:08:17

Re: The AMD/ATI Bar & Grill

#5981 2013-12-17 01:23:46

Re: The AMD/ATI Bar & Grill

#5982 2013-12-17 19:30:40

Re: The AMD/ATI Bar & Grill

#5983 2013-12-18 04:51:59

Re: The AMD/ATI Bar & Grill

#5984 2013-12-18 07:23:27

Re: The AMD/ATI Bar & Grill

#5985 2013-12-18 08:05:35

Re: The AMD/ATI Bar & Grill

#5986 2013-12-19 07:18:31

Re: The AMD/ATI Bar & Grill

#5987 2013-12-19 07:55:50

Re: The AMD/ATI Bar & Grill

#5988 2013-12-19 08:25:25

Re: The AMD/ATI Bar & Grill

#5989 2013-12-19 14:12:30

Re: The AMD/ATI Bar & Grill

#5990 2013-12-19 14:16:51

Re: The AMD/ATI Bar & Grill

#5991 2013-12-19 17:17:51

Re: The AMD/ATI Bar & Grill

#5992 2013-12-19 17:30:08

Re: The AMD/ATI Bar & Grill

#5993 2013-12-19 18:45:07

Re: The AMD/ATI Bar & Grill

#5994 2013-12-19 18:57:22

Re: The AMD/ATI Bar & Grill

#5995 2013-12-19 20:13:25

Re: The AMD/ATI Bar & Grill

#5996 2013-12-19 21:00:01

Re: The AMD/ATI Bar & Grill

#5997 2013-12-20 02:22:38

Re: The AMD/ATI Bar & Grill

#5998 2013-12-20 06:26:59

Re: The AMD/ATI Bar & Grill

#5999 2013-12-20 16:22:24

Re: The AMD/ATI Bar & Grill

#6000 2013-12-20 16:45:25

Re: The AMD/ATI Bar & Grill

Board footer