You are not logged in.

#5476 2013-08-02 18:13:03

Cdh
Member
Registered: 2009-02-03
Posts: 1,098

Re: The AMD/ATI Bar & Grill

sudo chmod -R a+rX /proc/ati


฿ 18PRsqbZCrwPUrVnJe1BZvza7bwSDbpxZz

Offline

#5477 2013-08-02 18:17:38

jfernyhough
Member
Registered: 2012-09-26
Posts: 5

Re: The AMD/ATI Bar & Grill

Spot on! I'd done a 'chmod -R +r /proc/ati' but I'd forgotten about +X for the directories! Nice!


1) Dell Studio 1749: i5 540 (2.5GHz), 8GB, ATi 5650, Intel 5300AGN, XFCE@U13.10x64
2) IBM X61: C2D T7200, 2GB, Intel Centrino, XFCE@U13.10x64
3) Compaq Mini 311c: Atom N270 (OC@2.0GHz), 3GB, Nvidia ION, Broadcom G+BT, KDE@U13.10x32
Comms: HTC Desire Z@JB4.2 (Flinny Andromadus CM10.1)

Offline

#5478 2013-08-02 18:51:03

Vi0L0
Member
From: Poland
Registered: 2009-06-24
Posts: 1,299
Website

Re: The AMD/ATI Bar & Grill

@jfernyhough:
flim's patch from here will fix this


I have updated catalyst-test@AUR with this new 13.8 beta release, soon I will update [catalyst] repo. For me it's working just like our "13.7-1", just with proper amd's 3.10 kernel support - I will try to backport it into older releases.

Offline

#5479 2013-08-02 19:38:39

terin7
Member
Registered: 2013-08-02
Posts: 1

Re: The AMD/ATI Bar & Grill

snowblind wrote:

Does anyone have any issues with their GPU periodically flipping between minimum and maximum clockspeeds with no apparent reason?

I think I'm getting this issue as well, but I'm not sure where I can see the current clock speeds, since the CCC on Linux isn't as fully featured as the Windows one. I definitely notice games, especially more graphically intense ones, are either at good framerates or low, with a clean break between them. On i3, switching between windowed and full screen of the program being affected will increase my framerates a bit, but after a little while it's no good.

Offline

#5480 2013-08-02 20:42:23

Roken
Member
From: UK
Registered: 2012-01-16
Posts: 652

Re: The AMD/ATI Bar & Grill

I'm not really a gamer, though I do stuff in wine that's heavily dependant on the GPU (3D gfx, and the only game I play, Trackmania) and I can't say that I've noticed any issues with GPU speed. However, and I could be listening to the wrong fan, but the GPU does seem more active since catalyst 12.9. I'm not the only one to notice this, but I can't verify without pulling my rig apart and connecting my Windows drive (I'm pretty much exclusively Arch, so my basis for comparison is close to NIL)


[img=Speedtest]http://www.speedtest.net/my-result/5145583518[/img]

Nvidia GTX 670 2Gb, AMD Phenom II X4 (965BE) @ 3.6 Ghz (Overclocked) 8GB RAM
Linux user #545703

Offline

#5481 2013-08-03 03:01:27

snowblind
Member
Registered: 2013-07-18
Posts: 34

Re: The AMD/ATI Bar & Grill

terin7 wrote:
snowblind wrote:

...

I think I'm getting this issue as well, but I'm not sure where I can see the current clock speeds, since the CCC on Linux isn't as fully featured as the Windows one. I definitely notice games, especially more graphically intense ones, are either at good framerates or low, with a clean break between them. On i3, switching between windowed and full screen of the program being affected will increase my framerates a bit, but after a little while it's no good.

You can use the following command to see your current clockspeed:

    aticonfig --odgc

I think part of the reason why the fans are lower in Windows is because, when you're just at your desktop, browsing the web, doing very low power stuff, it's using the Intel card. Windows' DWM is smart enough to be able to flip over to the AMD card whenever it needs more power; this is not something that X is very good at, certainly not X + Catalyst. (Though there's some progress in the form of the radeon xrandr offloadingsink thing, which I've fiddled a bit with, but I don't think that it works for fglrx).

So it could very well be that simply because the AMD card is doing *everything* when you have it enabled, it's just ramping up the fans. When my laptop is using the Intel card it makes relatively little noise.

EDIT: I should say, that's my experience based on AMD+Intel hybrid graphics in a laptop. I've never owned an AMD desktop card (likely never will) so I don't have any input in that regard.

Last edited by snowblind (2013-08-03 03:02:24)

Offline

#5482 2013-08-03 07:39:25

Fregtux
Member
From: Italy
Registered: 2013-08-01
Posts: 55

Re: The AMD/ATI Bar & Grill

Hi, I've just update catalyst driver. Fglrxinfo is ok and and glxinfo show "direct rendering: yes". Hovewer, if I run verbose mode, I obtaine:

$ LIBGL_DEBUG=verbose fglrxinfo
libGL: AtiGetClientDriverName: 13.20.4 fglrx (screen 0)
libGL: OpenDriver: trying /usr/lib/xorg/modules/dri//fglrx_dri.so
ukiDynamicMajor: found major device number 249
ukiDynamicMajor: found major device number 249
ukiDynamicMajor: found major device number 249
ukiOpenDevice: node name is /dev/ati/card0
ukiOpenDevice: open result is 4, (OK)
ukiGetBusid returned 'PCI:1:0:0'
ukiOpenDevice: node name is /dev/ati/card1
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card2
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card3
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card4
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card5
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card6
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card7
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card8
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card9
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card10
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card11
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card12
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card13
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card14
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card15
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiDynamicMajor: found major device number 249
ukiDynamicMajor: found major device number 249
ukiOpenByBusid: Searching for BusID PCI:1:0:0
ukiOpenDevice: node name is /dev/ati/card0
ukiOpenDevice: open result is 5, (OK)
ukiOpenByBusid: ukiOpenMinor returns 5
ukiOpenByBusid: ukiGetBusid reports PCI:1:0:0
ukiDynamicMajor: found major device number 249
ukiDynamicMajor: found major device number 249
ukiDynamicMajor: found major device number 249
ukiOpenDevice: node name is /dev/ati/card0
ukiOpenDevice: open result is 11, (OK)
ukiGetBusid returned 'PCI:1:0:0'
ukiOpenDevice: node name is /dev/ati/card1
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card2
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card3
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card4
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card5
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card6
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card7
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card8
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card9
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card10
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card11
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card12
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card13
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card14
ukiOpenDevice: UKI_ERR_NOT_ROOT
ukiOpenDevice: node name is /dev/ati/card15
ukiOpenDevice: UKI_ERR_NOT_ROOT
display: :0  screen: 0
OpenGL vendor string: Advanced Micro Devices, Inc.
OpenGL renderer string: AMD Radeon HD 7900 Series
OpenGL version string: 4.3.12438 Compatibility Profile Context 13.15.100.1

What about this output?


ASUS P8P67 EVO | i7-2600k |RAM 16GB DDR 3 Vengeance@1600MHz | AMD 7950 Sapphire Dual-X | SSD: Samsung 830 250GB | HD: Western Digital Caviar Blue 1TB

Offline

#5483 2013-08-03 08:13:25

Shinto
Member
Registered: 2012-07-27
Posts: 67

Re: The AMD/ATI Bar & Grill

No more freezing with 13.8! smile

Offline

#5484 2013-08-03 08:16:08

zoqaeski
Member
From: /earth/australia/.
Registered: 2009-09-30
Posts: 124

Re: The AMD/ATI Bar & Grill

Fregtux wrote:
zoqaeski wrote:

I can’t upgrade from 12.8. Every single version higher than that has issues with the backlight, and none of the supposed fixes seems to work. Has anyone else had issues similar to mine?

What are your Xorg and kernel versions?

I'm with catalyst 13.7 and kernel stock 3.10 and catalyst work with no problem.

I was holding them all back to Xorg 1.12, Catalyst 12.8 and Kernel 3.9.9. Decided to try the open-source drivers today and my laptop nearly melted… the temperature went up to > 110°C then the computer abrubtly cut power. Removed open-source drivers, reinstalled latest (beta?) version of Catalyst (13.8?) with a full system update, and things seem to work, though I have to rely on a hack to get some brightness controls. The only thing that has improved is the temperature: on 12.8 the temp was often > 80°C, and with this latest driver it’s less than 60°C.

I’m starting to wonder why I went with the AMD system now.

Offline

#5485 2013-08-03 15:31:21

Anon11
Member
Registered: 2012-05-20
Posts: 33

Re: The AMD/ATI Bar & Grill

zoqaeski wrote:
Fregtux wrote:
zoqaeski wrote:

I can’t upgrade from 12.8. Every single version higher than that has issues with the backlight, and none of the supposed fixes seems to work. Has anyone else had issues similar to mine?

What are your Xorg and kernel versions?

I'm with catalyst 13.7 and kernel stock 3.10 and catalyst work with no problem.

I was holding them all back to Xorg 1.12, Catalyst 12.8 and Kernel 3.9.9. Decided to try the open-source drivers today and my laptop nearly melted… the temperature went up to > 110°C then the computer abrubtly cut power. Removed open-source drivers, reinstalled latest (beta?) version of Catalyst (13.8?) with a full system update, and things seem to work, though I have to rely on a hack to get some brightness controls. The only thing that has improved is the temperature: on 12.8 the temp was often > 80°C, and with this latest driver it’s less than 60°C.

I’m starting to wonder why I went with the AMD system now.

In case you did not know , check this : http://www.phoronix.com/scan.php?page=a … _dpm&num=1

With kernel 3.11rc3 + firmware update + pass radeon.dpm=1 to kernel
I get temperature as good as catalyst with open source drivers and its working great so far (except sleep issues with my laptop model)

With mesa 9.2 performance will improve further, Happy with amd now i am smile

Offline

#5486 2013-08-03 16:45:00

Cdh
Member
Registered: 2009-02-03
Posts: 1,098

Re: The AMD/ATI Bar & Grill

snowblind wrote:

this is not something that X is very good at, certainly not X + Catalyst

Yes, there is the simple fact that there simply isn't an abstraction of moving 3d programs to other hardware. But then it's really hard. Imagine a window manager that uses OpenGL 4.4 for effects. You can run that on your nvidia gpu. Now you want to move it to your intel GPU which is currently at OpenGL 3.2. How do you do that?

bumblebee sorta works for catalyst now, by the way. Maybe I'll try to make a package in the AUR because it needs a bit of manual setup... But it's possible to use intel for everything and only run programs with optirun on the radeon gpu. (Maybe soon with primus and better performance.)

Last edited by Cdh (2013-08-03 16:47:25)


฿ 18PRsqbZCrwPUrVnJe1BZvza7bwSDbpxZz

Offline

#5487 2013-08-03 16:49:30

Jannis
Member
Registered: 2013-06-03
Posts: 32

Re: The AMD/ATI Bar & Grill

13.8 is looking good so far but vsync seems to still be broken.

Without vsync I get close to 300fps in TF2 with occasional drops to 60fps but with vsync on fps is constantly 30 and below while the controls feel to be delayed by at least 700ms.

Also is there a way to enable vsync on desktop (KWin) without using Catalyst's "Tear Free"? Because that setting forces vsync for all applications and then TF2 gets unplayable.

Oh, and by the way: @Vi0L0 do you accept small donations? I'd like to spend a coffee/pizza/beer/something for your ongoing work!

EDIT: Okay, random freezes now and after a reboot KDE desktop effects don't work anymore. Thinking about downgrading to 13.7 beta. Why do they always screw up the "stable" releases?

Last edited by Jannis (2013-08-04 14:26:13)

Offline

#5488 2013-08-04 17:20:52

abbradar
Member
Registered: 2010-06-10
Posts: 34

Re: The AMD/ATI Bar & Grill

The latest 13.3 with 3.10 kernel segfaults for me. Xorg.0.log: http://codepad.org/qk460rNf
I don't want to stay with 3.9 and can't use newer Catalyst because of this bug: http://ati.cchtml.com/show_bug.cgi?id=831
Any hope for me?


Sorry for my poor English.

Offline

#5489 2013-08-04 17:56:42

Vi0L0
Member
From: Poland
Registered: 2009-06-24
Posts: 1,299
Website

Re: The AMD/ATI Bar & Grill

Cdh wrote:

bumblebee sorta works for catalyst now, by the way

So there are some pros, good smile

Jannis wrote:

13.8 is looking good so far but vsync seems to still be broken.

Without vsync I get close to 300fps in TF2 with occasional drops to 60fps but with vsync on fps is constantly 30 and below while the controls feel to be delayed by at least 700ms.

Yeah, broken tear-free is also pissing me off...

Jannis wrote:

Also is there a way to enable vsync on desktop (KWin) without using Catalyst's "Tear Free"? Because that setting forces vsync for all applications and then TF2 gets unplayable.

There's a v-sync option in systemettings's "Desktop Effects", but it's not working well.

Jannis wrote:

Oh, and by the way: @Vi0L0 do you accept small donations? I'd like to spend a coffee/pizza/beer/something for your ongoing work!

Glad to be helpfull, but - unfortunatelly - I don't accept any donations.
Well... maybe except thousands of bitcoins big_smile
my_precious_bitcoin.jpg

abbradar wrote:

The latest 13.3 with 3.10 kernel segfaults for me. Xorg.0.log: http://codepad.org/qk460rNf
I don't want to stay with 3.9 and can't use newer Catalyst because of this bug: http://ati.cchtml.com/show_bug.cgi?id=831
Any hope for me?

Was it working well before the update?
I have updated [catalyst-stable],[catalyst-hd234k] and [catalyst-13.3-beta3] with gentoo 3.10 patch. It is much better than the one we were previously using. It's not just a workaround, it's working and handling proc well

Offline

#5490 2013-08-04 18:38:36

abbradar
Member
Registered: 2010-06-10
Posts: 34

Re: The AMD/ATI Bar & Grill

Vi0L0 wrote:

Was it working well before the update?
I have updated [catalyst-stable],[catalyst-hd234k] and [catalyst-13.3-beta3] with gentoo 3.10 patch. It is much better than the one we were previously using. It's not just a workaround, it's working and handling proc well

Well, I had my own package till 3.10 came (with DKMS support, hasn't updated it from May or so). It worked nicely for 3.9, and I performed full system upgrade just today and replaced my own package with yours. I had checked your package, too, and it had worked, but it was a long-long time ago. If you have old version without 3.10 patches, I could test it on 3.9 kernel. AFAIR, your version has some patches, and in my package I've used vanilla fglrx source.
I have also checked new 13.8 beta1, but it doesn't work for me.

Last edited by abbradar (2013-08-04 18:39:37)


Sorry for my poor English.

Offline

#5491 2013-08-04 19:42:48

Vi0L0
Member
From: Poland
Registered: 2009-06-24
Posts: 1,299
Website

Re: The AMD/ATI Bar & Grill

abbradar wrote:
Vi0L0 wrote:

Was it working well before the update?
I have updated [catalyst-stable],[catalyst-hd234k] and [catalyst-13.3-beta3] with gentoo 3.10 patch. It is much better than the one we were previously using. It's not just a workaround, it's working and handling proc well

Well, I had my own package till 3.10 came (with DKMS support, hasn't updated it from May or so). It worked nicely for 3.9, and I performed full system upgrade just today and replaced my own package with yours. I had checked your package, too, and it had worked, but it was a long-long time ago. If you have old version without 3.10 patches, I could test it on 3.9 kernel. AFAIR, your version has some patches, and in my package I've used vanilla fglrx source.
I have also checked new 13.8 beta1, but it doesn't work for me.

13.8 returns same error msg?
ok, so here you can find 13.3 beta2 and beta3 tarballs of the packages. You will have to build on yor own.
You should install latest catalyst-utils and lib32-catalyst-utils if needed. Now take ie. catalyst-hook-13.3, there' a lots of version there:
-13 comes with gentoo 3.10 patch
-11 comes with kolasa's 3.10 patch
-8 comes clean
Choose your destiny, maybe kolasa's patch will work on your 3.10 kernel

Offline

#5492 2013-08-04 21:15:56

gotjazz
Member
Registered: 2013-06-06
Posts: 11

Re: The AMD/ATI Bar & Grill

hrm anyone else suffering from memory leaks in xbmc since updating to 13.8beta?
in phoronix forum someone seems to have the same problem with xbmc and some (not all) valve games...

Offline

#5493 2013-08-04 21:16:35

Jannis
Member
Registered: 2013-06-03
Posts: 32

Re: The AMD/ATI Bar & Grill

Anyone else having problems with high RAM usage?

Seems like the freezes I'm experiencing since the upgrade to 12.8 are caused by having no more memory left. Since I've got 12gb RAM I don't have a swap file.
Weird thing is, I can't figure out which process is leaking as the sum of all processes' virtual memory and the buffers doesn't come close to what's reported as actually used.

I know, empty memory is wasted memory but before the update to 12.8 my system never used above 2gb without a reason, now it's going beyond 10gb and it stays used even if I stop KDM.

EDIT: @gotjazz I'm not using XBMC but playing TF2. There's something going severely wrong with memory usage.

EDIT²: Okay, so in the thread over at phoronix people are having similar problems. Memory is leaking while some OpenGL applications are running and stuff is getting swapped to disk. The system is slowing down to unusability and killing the OpenGL application doesn't get the memory back; you'll have to reboot.

Last edited by Jannis (2013-08-04 22:22:55)

Offline

#5494 2013-08-04 23:28:49

abbradar
Member
Registered: 2010-06-10
Posts: 34

Re: The AMD/ATI Bar & Grill

Vi0L0 wrote:

13.8 returns same error msg?
ok, so here you can find 13.3 beta2 and beta3 tarballs of the packages. You will have to build on yor own.
You should install latest catalyst-utils and lib32-catalyst-utils if needed. Now take ie. catalyst-hook-13.3, there' a lots of version there:
-13 comes with gentoo 3.10 patch
-11 comes with kolasa's 3.10 patch
-8 comes clean
Choose your destiny, maybe kolasa's patch will work on your 3.10 kernel

No, it's another message. Anyway, I've made some builds and experimented a bit and found the reason! Seems that I just needed to lower my Virtual screen setting in xorg.conf, silly me. Thanks, I have your latest package working now! Maybe though one day people from AMD will finally get my bug fixed in newer versions.


Sorry for my poor English.

Offline

#5495 2013-08-04 23:39:19

abbradar
Member
Registered: 2010-06-10
Posts: 34

Re: The AMD/ATI Bar & Grill

Also, quick question: why does your packages conflict with ati-dri and xf86-video-ati? I had them installed side by side with my own package and used them from time to time to check progress of OSS driver.
Another request is to make /etc/profile.d/xvba.sh a "backup" file in xvba-video. The reason is for -pxp users to use custom environment settings for Intel or AMD vaapi drivers working when needed. For example, I use this script (called xvba2.sh to force it to run after yours), it works on muxed switching: http://codepad.org/U5pWT0c3 The same request goes for /etc/modules-load.d/catalyst.conf in catalyst-utils-pxp, because users of old muxed switching (like me) need to change between i915 and radeon side-by-side and fglrx (with changing BIOS settings). I handle this with special script which checks on current hardware, loads modules and performs other configuration, so it would be good to allow changing this file so it would keep itself after updates.


Sorry for my poor English.

Offline

#5496 2013-08-05 18:11:35

cpuddle
Member
Registered: 2013-08-05
Posts: 2

Re: The AMD/ATI Bar & Grill

Jannis wrote:

Anyone else having problems with high RAM usage?

Seems like the freezes I'm experiencing since the upgrade to 12.8 are caused by having no more memory left. Since I've got 12gb RAM I don't have a swap file.
Weird thing is, I can't figure out which process is leaking as the sum of all processes' virtual memory and the buffers doesn't come close to what's reported as actually used.

I know, empty memory is wasted memory but before the update to 12.8 my system never used above 2gb without a reason, now it's going beyond 10gb and it stays used even if I stop KDM.

EDIT: @gotjazz I'm not using XBMC but playing TF2. There's something going severely wrong with memory usage.

EDIT²: Okay, so in the thread over at phoronix people are having similar problems. Memory is leaking while some OpenGL applications are running and stuff is getting swapped to disk. The system is slowing down to unusability and killing the OpenGL application doesn't get the memory back; you'll have to reboot.

Edit: Solved it by following the guide at https://wiki.archlinux.org/index.php/Do … g_Packages. Downgraded to 13.6.

Been having problems with video since two updates ago on my HD6870 (using Vi0l0's repo).

I have 8GB RAM. System was sitting on 9GB (1GB swapped). Killing the XBMC process did nothing for restoring the memory.

Video is not working at all. I get artifacts on the screen if using VAAPI. As in the pic below. The artificacts are actually over my whole screen and not just the XBMC window. I get the same thing no matter which video player I try, VLC, SMPlayer, etc.

Anyone have advice on how this can be fixed or how I can roll back the driver?

Screenshot

Last edited by cpuddle (2013-08-05 18:41:06)

Offline

#5497 2013-08-05 21:36:48

clfarron4
Member
From: London, UK
Registered: 2013-06-28
Posts: 2,162
Website

Re: The AMD/ATI Bar & Grill

Jannis wrote:

Anyone else having problems with high RAM usage?
...
EDIT: @gotjazz I'm not using XBMC but playing TF2. There's something going severely wrong with memory usage.

EDIT²: Okay, so in the thread over at phoronix people are having similar problems. Memory is leaking while some OpenGL applications are running and stuff is getting swapped to disk. The system is slowing down to unusability and killing the OpenGL application doesn't get the memory back; you'll have to reboot.

And you're not the only one here. I, like cpuddle, downgraded to Catalyst 13.6 and all is working again. I remember 13.7 being a little bit laggy, which is why I went all the way back to 13.6.


Claire is fine.
Problems? I have dysgraphia, so clear and concise please.
My public GPG key for package signing
My x86_64 package repository

Offline

#5498 2013-08-06 14:27:18

gotjazz
Member
Registered: 2013-06-06
Posts: 11

Re: The AMD/ATI Bar & Grill

clfarron4 wrote:
Jannis wrote:

Anyone else having problems with high RAM usage?
...
EDIT: @gotjazz I'm not using XBMC but playing TF2. There's something going severely wrong with memory usage.

EDIT²: Okay, so in the thread over at phoronix people are having similar problems. Memory is leaking while some OpenGL applications are running and stuff is getting swapped to disk. The system is slowing down to unusability and killing the OpenGL application doesn't get the memory back; you'll have to reboot.

And you're not the only one here. I, like cpuddle, downgraded to Catalyst 13.6 and all is working again. I remember 13.7 being a little bit laggy, which is why I went all the way back to 13.6.


same here. I was actually kind of suprised finding that 13.6 worked with kernel 3.10 considering 3.10 support was in the 13.7 driver changelog IIRC.

Offline

#5499 2013-08-06 20:01:54

Vi0L0
Member
From: Poland
Registered: 2009-06-24
Posts: 1,299
Website

Re: The AMD/ATI Bar & Grill

I also tested l4d and xbmc, leaks as hell, added my report here: http://ati.cchtml.com/show_bug.cgi?id=866

abbradar wrote:

Also, quick question: why does your packages conflict with ati-dri and xf86-video-ati? I had them installed side by side with my own package and used them from time to time to check progress of OSS driver.

-pxp packages does not conflict with those packages. "normal" packages are not messing with system's variables, their libraries are in conflict with ie. mesa's libgl libraries - so they are in conflict also with ati-dri or intel-dri

abbradar wrote:

Another request is to make /etc/profile.d/xvba.sh a "backup" file in xvba-video. The reason is for -pxp users to use custom environment settings for Intel or AMD vaapi drivers working when needed. For example, I use this script (called xvba2.sh to force it to run after yours), it works on muxed switching: http://codepad.org/U5pWT0c3 The same request goes for /etc/modules-load.d/catalyst.conf in catalyst-utils-pxp, because users of old muxed switching (like me) need to change between i915 and radeon side-by-side and fglrx (with changing BIOS settings). I handle this with special script which checks on current hardware, loads modules and performs other configuration, so it would be good to allow changing this file so it would keep itself after updates.

But it won't work on muxless systems. The script is good though, could be used by many - maybe it will be a good idea to paste it on the wiki?

gotjazz wrote:

I was actually kind of suprised finding that 13.6 worked with kernel 3.10 considering 3.10 support was in the 13.7 driver changelog IIRC.

You shouldn't be wink
kolasa's 3.10 kernel patch did it

Offline

#5500 2013-08-06 21:33:44

abbradar
Member
Registered: 2010-06-10
Posts: 34

Re: The AMD/ATI Bar & Grill

Vi0L0 wrote:

-pxp packages does not conflict with those packages. "normal" packages are not messing with system's variables, their libraries are in conflict with ie. mesa's libgl libraries - so they are in conflict also with ati-dri or intel-dri

Not only -pxp packages, but catalyst-hook too --- maybe only catalyst-utils (not -pxp) should conflict? I'm talking about this (it's for catalyst-hook 13.3-13):

➜  ~  LANG=en_US.UTF-8 sudo pacman -S xf86-video-ati 
resolving dependencies...
looking for inter-conflicts...
:: xf86-video-ati and catalyst-hook are in conflict. Remove catalyst-hook? [y/N]
Vi0L0 wrote:

But it won't work on muxless systems. The script is good though, could be used by many - maybe it will be a good idea to paste it on the wiki?

I've fixed script to work on muxless systems too: http://codepad.org/S4fUGdOe
The script needs a patch for /usr/lib/fglrx/switchlib*: http://codepad.org/7LjjVQZM
But I was asking not to replace script, but to make it (and /etc/modules-load.d/catalyst.conf) "backup" in PKGBUILD, so someone like me with custom scripts will be able to replace this files and they will not be overwritten on update. Though, my new script might be useful for all -pxp users, you can include it into your package if you wish.
Thanks again for backporting 13.3 for 3.10, it's working like a charm. ^_^


Sorry for my poor English.

Offline

Board footer

Powered by FluxBB