The AMD/ATI Bar & Grill

haagch · 2013-11-08 08:36:56

Yesterday I thought, wouldn't it be nice to use the "official" hybrid gpu support from amd without magical scripts for switching etc, with just one script to start a second X server?

It works and it is actually pretty simple.

First I removed /etc/profile.d/*catalyst*

I created /ati for this and symlinked all the fglrx stuff into /ati/fglrx (ln -s /usr/lib/xorg/modules/**/*fglrx* /ati/fglrx and ln -s /usr/lib32/xorg/modules/**/*fglrx* /ati/fglrx and ln -s /usr/lib/catalystpxp/* /ati/fglrx and so on).
I also have symlinks from /ati/fglrx/libGL.so.1.2, /ati/fglrx/libGL.so.1, /ati/fglrx/libGL.so to /usr/lib/catalystpxp/fglrx/fglrx-libGL.so.1.2 and libglx.so to fglrx-libglx.so.

So the contents are
fglrx_dri.so@
fglrx_drv.so@
fglrx-libglx.so@
libfglrxdrm.so@
libGL.so@
libGL.so.1@
libGL.so.1.2@
libglx.so@

The same in /ati/fglrx for all the 32 bit stuff.

Then I used aticonfig --initial to create a xorg.conf, moved it to /ati/xorg.conf and symlinked it to /etc/X11/xorg.conf.fglrx because later when you start X as root it can only use relative paths to the config under /etc/X11/ with no "..".

I also copied /etc/X11/xorg.conf.d/ to /ati/xorg.conf.d/ and symlinked it to /etc/X11/xorg.conf.d.fglrx/ because again, it'll only take relative paths.
I did this because I have fglrx's libglx.so named libglx.so in /usr/lib/xorg/modules/updates/extensions/fglrx/libglx.so and for some reason X would use this over the mesa's libglx.so in /usr/lib/xorg/modules/extensions/libglx.so so I have the file:
/etc/X11/xorg.conf.d/00-libglxModulePath.conf

Section "Files"
         ModulePath   "/usr/lib/xorg/modules/extensions"
         ModulePath   "/usr/lib/xorg/modules"
EndSection

Now /ati/xorg.conf is just the default aticonfig --initial created:

cat xorg.conf
Section "ServerLayout"
        Identifier     "aticonfig Layout"
        Screen      0  "aticonfig-Screen[0]-0" 0 0
EndSection

Section "Module"
EndSection

Section "Monitor"
        Identifier   "aticonfig-Monitor[0]-0"
        Option      "VendorName" "ATI Proprietary Driver"
        Option      "ModelName" "Generic Autodetecting Monitor"
        Option      "DPMS" "true"
EndSection

Section "Device"
        Identifier  "aticonfig-Device[0]-0"
        Driver      "fglrx"
        BusID       "PCI:1:0:0"
EndSection

Section "Screen"
        Identifier "aticonfig-Screen[0]-0"
        Device     "aticonfig-Device[0]-0"
        Monitor    "aticonfig-Monitor[0]-0"
        DefaultDepth     24
        SubSection "Display"
                Viewport   0 0
                Depth     24
        EndSubSection
EndSection

In /ati/xorg.conf.d/00-libglxModulePath.conf I have done the opposite thing than in my normal system, I put /ati/fglrx as the first module path, where all the fglrx is symlinked and only then the module path where all my system stuff is (because it needs to load the intel driver too, we need this)

Section "Files"
         ModulePath   "/ati/fglrx/"
         ModulePath   "/usr/lib/xorg/modules"
EndSection

Then I copied startx to /ati/startxati and edited some stuff in
Right at the top:

export LD_LIBRARY_PATH=/ati/fglrx:/ati/fglrx32
export LIBGL_DRIVERS_PATH=/ati/fglrx:/ati/fglrx32:/usr/lib/xorg/modules/dri/:/usr/lib32/xorg/modules/dri

Then edited serverargs and userclientrc so you can start a lightweight window manager and don't have to have it in ~/.xinitrc:

serverargs="-config xorg.conf.fglrx -configdir xorg.conf.d.fglrx/"
userclientrc=/ati/xinitrc

And that's about it.

/ati/startxati now starts an X server using fglrx without any "magic" switching scripts while a normally started X still uses the intel gpu only.
It still calls /lib/fglrx/switchlibGL but it doesn't really matter as long as there is /etc/X11/xorg.conf.d/00-libglxModulePath.conf and the environment variables that are set in in /etc/profile.d/*catalyst* are not set.

That's a very crude method and there is probably an easier way to do some of this stuff, but I just thought that could be a heads up that the switching scripts don't actually do any proprietary "magic" but only set a few variables and copy some stuff - which isn't really needed.

Also, if you have "hacked" the fglrx libraries for trying primus, put the original ones back, X will segfault otherwise.

ivdok · 2013-11-08 12:36:37

Vi0L0 wrote:

clfarron4 wrote:
Ok, so I am using catalyst-hook 13.11 as provided in Vi0L0's repository, and I'm not having any problems with the 3.12 kernel that's in [testing] at the moment.
Yeah, latest catalysts are working fine with 3.12, and the last one doesn't even need a patch. But everything older needs to be patched - hd234k, stable or 13.6 - i'm planning to update those asap, tomorrow or the day after (atm im exhausted).
I only want to remove catalyst package from these repos as soon as 3.12 kernel will land in [core], because i don't want to came up with fglrx module precompiled to present itself as GPL-licensed. So you will need to use hook or generator.

Could you please keep 13.6 in repos for a bit more time? Seems like XvBA is breaking apart with 13.7+, and I can't use radeonsi (which has OK XvBA support) + fglrx simultaneously. And I don't even know what's removed from xvba-open - nobody tells! Ok, I'm fine with "proprietary&l33t" stuff like DTS missing - but if MP4/MKV would stop playing at all - I'd be mad. My CPU is ridiculously slow, and only my powerful GPU lets me carry on.

Reded · 2013-11-10 02:43:22

Has anyone else had the issue of a solid black screen with gnome-shell 3.10 (And GDM) Using the legacy Catalyst drivers? I'm using an HD4350 with catalyst 13.1-16 from the repositories, and GDM loads up blank. If I start gnome-session from .xinitrc, I get an unhelpful error message (With a blank screen behind it).

I'm not sure where to actually ask about this issue... There was a thread on these forums about it, but it only got a couple of replies, all of which were people like me having the issue and not knowing what to do I guess there just aren't many gnome-shell users also using catalyst-legacy!

Jannis · 2013-11-10 03:30:26

Reded wrote:

Has anyone else had the issue of a solid black screen with gnome-shell 3.10 (And GDM) Using the legacy Catalyst drivers? I'm using an HD4350 with catalyst 13.1-16 from the repositories, and GDM loads up blank. If I start gnome-session from .xinitrc, I get an unhelpful error message (With a blank screen behind it).
I'm not sure where to actually ask about this issue... There was a thread on these forums about it, but it only got a couple of replies, all of which were people like me having the issue and not knowing what to do I guess there just aren't many gnome-shell users also using catalyst-legacy!

Have you tried running the open drivers instead? AFAIK they are superior to Catalyst legacy. They should work really nice with the HD4xxx series.

Vi0L0 · 2013-11-10 10:22:01

ivdok wrote:

Vi0L0 wrote:
clfarron4 wrote:
Ok, so I am using catalyst-hook 13.11 as provided in Vi0L0's repository, and I'm not having any problems with the 3.12 kernel that's in [testing] at the moment.
Yeah, latest catalysts are working fine with 3.12, and the last one doesn't even need a patch. But everything older needs to be patched - hd234k, stable or 13.6 - i'm planning to update those asap, tomorrow or the day after (atm im exhausted).
I only want to remove catalyst package from these repos as soon as 3.12 kernel will land in [core], because i don't want to came up with fglrx module precompiled to present itself as GPL-licensed. So you will need to use hook or generator.
Could you please keep 13.6 in repos for a bit more time? Seems like XvBA is breaking apart with 13.7+, and I can't use radeonsi (which has OK XvBA support) + fglrx simultaneously. And I don't even know what's removed from xvba-open - nobody tells! Ok, I'm fine with "proprietary&l33t" stuff like DTS missing - but if MP4/MKV would stop playing at all - I'd be mad. My CPU is ridiculously slow, and only my powerful GPU lets me carry on.

I accidentally found out that the change which caused problems with 3.12 kernel support was removed. EXPORT_SYMBOL_GPL(acpi_bus_get_device) was reverted, even after Rafael's words that he won't revert the change... So no patch is needed for catalyst to support 3.12 kernel any more.
It's good to see how it ends, it's bad that i wasted my and [catalyst-stable] repo and aur users time on the unneeded update - sorry for this.

woomia · 2013-11-10 15:49:23

The Catalyst legacy drivers work well up until a certain version of Xorg, so I've stopped upgrading until I can get a new card. I have a HD 4X and the open source drivers aren't great with WoW.

Reded · 2013-11-10 17:33:14

Jannis wrote:

Reded wrote:
Has anyone else had the issue of a solid black screen with gnome-shell 3.10 (And GDM) Using the legacy Catalyst drivers? I'm using an HD4350 with catalyst 13.1-16 from the repositories, and GDM loads up blank. If I start gnome-session from .xinitrc, I get an unhelpful error message (With a blank screen behind it).
I'm not sure where to actually ask about this issue... There was a thread on these forums about it, but it only got a couple of replies, all of which were people like me having the issue and not knowing what to do I guess there just aren't many gnome-shell users also using catalyst-legacy!
Have you tried running the open drivers instead? AFAIK they are superior to Catalyst legacy. They should work really nice with the HD4xxx series.

Yes I've tried them, they run gnome-shell fantastically However game performance is... Sub-par, to say the least... They're getting there though, and will be a viable solution soon I hope!

kpy · 2013-11-11 16:35:39

Hi everyone!
I've got some trouble with catalyst-test-pxp 13.11. There are some applications which won't start. Also they were not giving me error messages when I start them from command line. Examples are Firefox and Texmaker. It's strange because with the 13.10 catalyst-test-pxp everything was working fine. I use the catalyst-test-pxp repo because my backlight control was broken with catalyst-total-pxp. Any ideas what I can do or can anyone give me the catalyst-test-pxp 13.10 tarball?

Vi0L0 · 2013-11-11 17:50:27

kpy wrote:

Hi everyone!
I've got some trouble with catalyst-test-pxp 13.11. There are some applications which won't start. Also they were not giving me error messages when I start them from command line. Examples are Firefox and Texmaker. It's strange because with the 13.10 catalyst-test-pxp everything was working fine. I use the catalyst-test-pxp repo because my backlight control was broken with catalyst-total-pxp. Any ideas what I can do or can anyone give me the catalyst-test-pxp 13.10 tarball?

Here are archived tarballs. Let us know if it helped.

Kuci · 2013-11-12 15:03:57

These Catalyst drivers gets even worse...With the latest version, screen brightness still does not work for me and now I am even not able to start GDM (X log does not contain information about any problems...). I use AMD Radeon HD 7600G (integrated) and AMD Radeon HD 8550 (discrete).

Last edited by Kuci (2013-11-12 15:07:38)

MRWITEK · 2013-11-12 16:35:12

Hello.
After upgrade to the latest catalyst-generator today I encountered very strange bug (aside all the usual). After starting awesome WM (tested with openbox too) kworker process uses 100% of one of CPU as it said by top, iotop does not tell any unusual behaviour. At this time some programs can't run e. g. something called like 'pgp2-signature' evaluated by 'pacman -Syuu', and sometimes termit (VTE-based terminal) window is stuck in gray state and don't respond to any actions about it even killing. After 10 minutes of waiting everything gets normal and all programs and signals gets going and everything works like a charm (performance is higher than before). Except backlight control is still broken (12.8 is last with it) and I can startx before reboot only once.
I use samsung laptop NP355V5X with ATI HD7520G+7670M muxless.

Last edited by MRWITEK (2013-11-12 17:15:43)

AuHop · 2013-11-12 17:14:35

Did AMD kill the service for pulling drivers directly from their server? Running:
wget http://www2.ati.com/drivers/beta/amd-ca … x86_64.zip
just grabs a 47k broken zip file. Navigating there directly with a browser just redirects me to an AMD page complaining about direct links lacking a referral.
The latest catalyst-test-pxp package (at least) also seems to suffer from this. At the very least this will make it cumbersome to install/update from the AUR, and at worst impossible to rollback without archived driver packs.
The latter is a problem since the current drivers are _still_ lacking backlight brightness control, which several people have mentioned on this page alone of the thread.

Last edited by AuHop (2013-11-12 17:39:30)

kpy · 2013-11-12 18:24:59

Thanks Vi0L0 for your help.
It seems that the old beta driver isn't available. I've got the same problem than AuHop. The downloaded file has only 47k.

Vi0L0 · 2013-11-12 19:29:21

AuHop wrote:

Did AMD kill the service for pulling drivers directly from their server? Running:
wget http://www2.ati.com/drivers/beta/amd-ca … x86_64.zip
just grabs a 47k broken zip file. Navigating there directly with a browser just redirects me to an AMD page complaining about direct links lacking a referral.
The latest catalyst-test-pxp package (at least) also seems to suffer from this. At the very least this will make it cumbersome to install/update from the AUR, and at worst impossible to rollback without archived driver packs.
The latter is a problem since the current drivers are _still_ lacking backlight brightness control, which several people have mentioned on this page alone of the thread.

Yes, the impossibility to downgrade is the worst... What have you done AMD? We need to do something with it...
I have got the backup of catalyst' installers (i guess that almost if not all releases of the last 3-4 years maybe ;P), i think i will upload at least releases from the last year into some free net disk (and provide sha sums for security reasons), but don't expect this before saturday or sunday.

BTW. because of this i've updated catalyst-total@AUR (other will be updated later, i want to update sdk and codexl at first) and noted that amd has changed 13.4 a little bit - there's a lot of new entries in amdpcsdb.default file.

kpy wrote:

Thanks Vi0L0 for your help.
It seems that the old beta driver isn't available. I've got the same problem than AuHop. The downloaded file has only 47k.

BTW The 47k file is actually a html page.
You can still use catalyst-hook + catalyst-utils-pxp (and additional lib32-catalyst-utils-pxp if needed) from the [catalyst-13.10-beta2] repo:

[catalyst-13.10-beta2]
Server = http://catalyst.wirephire.com/repo/catalyst/$arch

Last edited by Vi0L0 (2013-11-12 19:35:03)

clfarron4 · 2013-11-14 20:22:45

Hi, so I have the following question:

If I am using AMD Catalyst, does passing radeon.dpm = 1 through the kernel parameters actually do anything beneficial for me?

ChemBro · 2013-11-14 22:35:09

clfarron4 wrote:

Hi, so I have the following question:
If I am using AMD Catalyst, does passing radeon.dpm = 1 through the kernel parameters actually do anything beneficial for me?

No, that is radeon (driver) only. Catalyst has its own dynamic power management.

lano1106 · 2013-11-15 04:08:57

Hi,

I'm currently building multilib version of

xvba-va-driver
libvdpau-va-gl-git

to be able to enable video acceleration from the 32bit flash plugin used by the steam client.

(I have no intention to support and maintain the PKGFILEs I create but I'll share them if someone express the interest)

I have stumbled into a small problem with lib32-catalyst-utils

A symbolic link is missing to be able to link with it:

lrwxrwxrwx 1 root root 15 Nov 14 23:01 /usr/lib32/libXvBAW.so -> libXvBAW.so.1.0

cabutan · 2013-11-15 19:56:48

I have a laptop with Radeon HD 6620G and xorg-server 1.14.4-1 ..is there a proprietary driver available? if so, which one? I was happily using the open source one but I need a 3D software and some things are not displaying properly. Thanks

clfarron4 · 2013-11-15 20:30:11

cabutan wrote:

I have a laptop with Radeon HD 6620G and xorg-server 1.14.4-1 ..is there a proprietary driver available? if so, which one? I was happily using the open source one but I need a 3D software and some things are not displaying properly. Thanks

Read on: https://wiki.archlinux.org/index.php/AMD_Catalyst

cabutan · 2013-11-15 20:36:16

clfarron4 wrote:

cabutan wrote:
I have a laptop with Radeon HD 6620G and xorg-server 1.14.4-1 ..is there a proprietary driver available? if so, which one? I was happily using the open source one but I need a 3D software and some things are not displaying properly. Thanks
Read on: https://wiki.archlinux.org/index.php/AMD_Catalyst

what kind of lead is that? that document is for the open source driver which I am currently using. I am trying to figure out whether the catalyst total or the catalyst alone. Anyways, good try.

clfarron4 · 2013-11-15 20:46:23

cabutan wrote:

clfarron4 wrote:
Read on: https://wiki.archlinux.org/index.php/AMD_Catalyst
what kind of lead is that? that document is for the open source driver which I am currently using. I am trying to figure out whether the catalyst total or the catalyst alone. Anyways, good try.

I think you might find that this page is for the Open Source xf86-intel-ati driver, where the link I posted above is for the proprietary AMD Catalyst driver.

Anyways, if you are looking for advice as to which package out of Vi0L0's repository you should install, I would go with the catalyst-hook package from the [catalyst] repository, run sudo aticonfig --initial to get the main screen running (refer to my above post if you have multiple screens, or an odd setup) and enable the catalyst-hook.service and be done with it.

I'm not entirely sure what the other catalyst-[name] packages do, but I do know the catalyst itself is packaged without any means for generating kernel modules, which is inconvenient if you run custom kernels (like me), run [testing] or just so happen to update the kernel before Vi0L0 updates the package with new kernel modules.

Last edited by clfarron4 (2013-11-15 20:50:07)

Lockheed · 2013-11-15 22:00:40

I have a problem which apparently is related to graphic drivers. I am using hd234k version from Server = http://catalyst.wirephire.com/repo/cata … 234k/$arch (by the way - when can we expect update of this line?)

I installed Civilization 4 on wine, and it runs fine - I can hear Intro movie and then Menu music. I can even see the mouse curour ingame. However, the picture is one scrambled static image, from start to finish (till I press Alt+F4).

While it runs, the terminal keeps returning this:

fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
err:d3d_draw:drawStridedFast >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION (0x506) from glDrawElementsBaseVertex @ ../../../wine/dlls/wined3d/drawprim.c / 73
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
err:d3d_draw:drawStridedFast >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION (0x506) from glDrawElementsBaseVertex @ ../../../wine/dlls/wined3d/drawprim.c / 73
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
err:d3d_draw:drawStridedFast >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION (0x506) from glDrawElementsBaseVertex @ ../../../wine/dlls/wined3d/drawprim.c / 73
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
err:d3d_draw:drawStridedFast >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION (0x506) from glDrawElementsBaseVertex @ ../../../wine/dlls/wined3d/drawprim.c / 73
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
err:d3d_draw:drawStridedFast >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION (0x506) from glDrawElementsBaseVertex @ ../../../wine/dlls/wined3d/drawprim.c / 73
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
err:d3d_draw:drawStridedFast >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION (0x506) from glDrawElementsBaseVertex @ ../../../wine/dlls/wined3d/drawprim.c / 73
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
err:d3d_draw:drawStridedFast >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION (0x506) from glDrawElementsBaseVertex @ ../../../wine/dlls/wined3d/drawprim.c / 73
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
err:d3d_draw:drawStridedFast >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION (0x506) from glDrawElementsBaseVertex @ ../../../wine/dlls/wined3d/drawprim.c / 73
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER (0x8cdb)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x1cd39530) WINED3DFMT_D24_UNORM_S8_UINT 1024x768 2 samples.
err:d3d_draw:drawStridedFast >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION (0x506) from glDrawElementsBaseVertex @ ../../../wine/dlls/wined3d/drawprim.c / 73
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_MISSING_ATTACHMENT (0x8cd7)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_MULTISAMPLE (0x100000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_MISSING_ATTACHMENT (0x8cd7)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_RESOLVED (0x200000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
err:d3d_surface:surface_blt_fbo >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION (0x506) from glBlitFramebuffer() @ ../../../wine/dlls/wined3d/surface.c / 1237
fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_INCOMPLETE_MISSING_ATTACHMENT (0x8cd7)
fixme:d3d:context_check_fbo_status 	Location SFLAG_INRB_RESOLVED (0x200000).
fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1cd392a0) WINED3DFMT_B8G8R8X8_UNORM 1024x768 2 samples.
err:d3d:swapchain_blit >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION (0x506) from Swapchain present blit(EXT_framebuffer_blit)

I installed the game and configured it the same way as another guy configured it and run successfully on an Nvidia card, but this is my result.

Does anyone know how could this be resolved?

Vi0L0 · 2013-11-17 21:22:19

lano1106 wrote:

Hi,
I'm currently building multilib version of
xvba-va-driver
libvdpau-va-gl-git
to be able to enable video acceleration from the 32bit flash plugin used by the steam client.
(I have no intention to support and maintain the PKGFILEs I create but I'll share them if someone express the interest)

May i suggest throwing PKGBUILDs into AUR?

lano1106 wrote:

I have stumbled into a small problem with lib32-catalyst-utils
A symbolic link is missing to be able to link with it:
lrwxrwxrwx 1 root root 15 Nov 14 23:01 /usr/lib32/libXvBAW.so -> libXvBAW.so.1.0

Thanks, i've added symlinks to all (i think) my packages on AUR and updated repos: [catalyst],[catalyst-stable],[catalyst-hd234k] and [catalyst-13.6]

Vi0L0 wrote:

AuHop wrote:
...
Yes, the impossibility to downgrade is the worst... What have you done AMD? We need to do something with it...
I have got the backup of catalyst' installers (i guess that almost if not all releases of the last 3-4 years maybe ;P), i think i will upload at least releases from the last year into some free net disk (and provide sha sums for security reasons), but don't expect this before saturday or sunday.

I'm testing mega now ;P, looks good, its speed is impressive. Here are the amds archive installers :
https://mega.co.nz/#F!J4E2zIRD!ODSaaCP1cP9FRp-RLMm_Ig
and here are md5/sha sums : https://docs.google.com/document/d/1Mdk … sp=sharing

Feel free to check it out, download archives and ask for more

Last edited by Vi0L0 (2013-11-17 22:17:52)

kpy · 2013-11-17 22:22:55

Thanks Vi0L0! After building and installing catalyst 13.10 beta2 again I had a problem while building the modules, but installing catalyst-hook, catalyst-utils-pxp and lib32-catalyst-utils-pxp from your repo solves the problem and everything works perfect again.

lano1106 · 2013-11-18 01:08:01

Vi0L0 wrote:

May i suggest throwing PKGBUILDs into AUR?

Sure you can and I will do that. However I should have tested my native 64-bits xvba-va-driver/libvdpau-va-gl setup with the flash plugin before going through the trouble to create the multilib setup because on my first attempt to play a youtube video, my flash plugin sig faulted.

I have good knowledge of libva, I will try to nail down my problem before offering my PGKBUILDs (unless someone would like to have them ASAP...)

Arch Linux

#5851 2013-11-08 08:36:56

Re: The AMD/ATI Bar & Grill

#5852 2013-11-08 12:36:37

Re: The AMD/ATI Bar & Grill

#5853 2013-11-10 02:43:22

Re: The AMD/ATI Bar & Grill

#5854 2013-11-10 03:30:26

Re: The AMD/ATI Bar & Grill

#5855 2013-11-10 10:22:01

Re: The AMD/ATI Bar & Grill

#5856 2013-11-10 15:49:23

Re: The AMD/ATI Bar & Grill

#5857 2013-11-10 17:33:14

Re: The AMD/ATI Bar & Grill

#5858 2013-11-11 16:35:39

Re: The AMD/ATI Bar & Grill

#5859 2013-11-11 17:50:27

Re: The AMD/ATI Bar & Grill

#5860 2013-11-12 15:03:57

Re: The AMD/ATI Bar & Grill

#5861 2013-11-12 16:35:12

Re: The AMD/ATI Bar & Grill

#5862 2013-11-12 17:14:35

Re: The AMD/ATI Bar & Grill

#5863 2013-11-12 18:24:59

Re: The AMD/ATI Bar & Grill

#5864 2013-11-12 19:29:21

Re: The AMD/ATI Bar & Grill

#5865 2013-11-14 20:22:45

Re: The AMD/ATI Bar & Grill

#5866 2013-11-14 22:35:09

Re: The AMD/ATI Bar & Grill

#5867 2013-11-15 04:08:57

Re: The AMD/ATI Bar & Grill

#5868 2013-11-15 19:56:48

Re: The AMD/ATI Bar & Grill

#5869 2013-11-15 20:30:11

Re: The AMD/ATI Bar & Grill

#5870 2013-11-15 20:36:16

Re: The AMD/ATI Bar & Grill

#5871 2013-11-15 20:46:23

Re: The AMD/ATI Bar & Grill

#5872 2013-11-15 22:00:40

Re: The AMD/ATI Bar & Grill

#5873 2013-11-17 21:22:19

Re: The AMD/ATI Bar & Grill

#5874 2013-11-17 22:22:55

Re: The AMD/ATI Bar & Grill

#5875 2013-11-18 01:08:01

Re: The AMD/ATI Bar & Grill

Board footer