View previous topic :: View next topic |
Author |
Message |
Herring42 Guru
Joined: 10 Mar 2004 Posts: 373 Location: Buckinghamshire
|
Posted: Tue Jan 19, 2010 9:11 pm Post subject: |
|
|
I have a feeling you might need to be on the git version of mesa as well. Certainly try the latest masked version media-libs/mesa-7.7-r1
I'm afraid I can't tell, as I just went crazy, and went for it.
As mentioned, the git version work for me! _________________ "The problem with quotes on the internet is that it is difficult
to determine whether or not they are genuine." -- Abraham Lincoln |
|
Back to top |
|
|
pingufunkybeat l33t
Joined: 01 Dec 2004 Posts: 610
|
Posted: Tue Jan 19, 2010 9:30 pm Post subject: |
|
|
Yeah, the latest mesa in portage should be plenty new for your card.
If that doesn't work, then you should try unmasking the latest libdrm too.
But with your card, you don't need all the latest and greatest. |
|
Back to top |
|
|
saiko n00b
Joined: 16 Apr 2007 Posts: 30 Location: Amsterdam
|
Posted: Tue Jan 19, 2010 11:02 pm Post subject: |
|
|
yep. I am trying to do that but get the same blocking packages libXext and xorg-server like in the previous time.
So following your post i emerged layman, added x11 and according to the ebuild instructions then added " source /usr/local/portage/layman/make.conf " in make.conf
I believe it went well as it pulls the right packages but it get messed up with the above ones.
Code: |
autounmask -p media-libs/mesa-9999
autounmask version 0.27 (using PortageXS-0.02.09 and portage-2.1.6.13)
* Using repositories:
/usr/portage
/usr/local/portage/layman/x11
/usr/local/portage
* Using package.keywords file: /etc/portage/package.keywords
* Using package.unmask file: /etc/portage/package.unmask
* Using package.use file: /etc/portage/package.use
* Unmasking media-libs/mesa-9999 and its dependencies.. this might take a while..
* Added '=media-libs/mesa-9999 ~x86' to /etc/portage/package.keywords
* Added '=media-libs/mesa-9999' to /etc/portage/package.unmask
* Added '=x11-proto/xextproto-9999 ~x86' to /etc/portage/package.keywords
* Added '=x11-proto/xextproto-9999' to /etc/portage/package.unmask
* Added '=x11-proto/xextproto-7.1.1 ~x86' to /etc/portage/package.keywords
* Added '=x11-libs/libdrm-9999 ~x86' to /etc/portage/package.keywords
* Added '=x11-libs/libdrm-9999' to /etc/portage/package.unmask
* Added '=x11-libs/libdrm-2.4.17 ~x86' to /etc/portage/package.keywords
* Added '=app-admin/eselect-opengl-1.1.1-r2 ~x86' to /etc/portage/package.keywords
* Added '=x11-proto/glproto-9999 ~x86' to /etc/portage/package.keywords
* Added '=x11-proto/glproto-9999' to /etc/portage/package.unmask
* Added '=x11-proto/glproto-1.4.11 ~x86' to /etc/portage/package.keywords
* Added '=x11-proto/dri2proto-9999 ~x86' to /etc/portage/package.keywords
* Added '=x11-proto/dri2proto-9999' to /etc/portage/package.unmask
* Added '=x11-proto/dri2proto-2.2 ~x86' to /etc/portage/package.keywords
* Added '=media-fonts/font-util-9999 ~x86' to /etc/portage/package.keywords
* Added '=media-fonts/font-util-9999' to /etc/portage/package.unmask
* Added '=media-fonts/font-util-1.1.1-r1 ~x86' to /etc/portage/package.keywords
[ebuild R ] sys-apps/portage-2.1.6.13
[ebuild R ] virtual/libintl-0
[ebuild R ] sys-libs/zlib-1.2.3-r1
[ebuild R ] dev-libs/expat-2.0.1-r2
[ebuild R ] sys-devel/gnuconfig-20090203
[ebuild R ] virtual/libiconv-0
[ebuild R ] app-misc/pax-utils-0.1.19
[ebuild R ] app-arch/bzip2-1.0.5-r1
[ebuild R ] app-misc/mime-types-8
[ebuild R ] app-admin/python-updater-0.7
[ebuild R ] sys-devel/autoconf-wrapper-7
[ebuild R ] app-arch/cpio-2.9-r2
[ebuild R ] sys-devel/automake-wrapper-3-r1
[ebuild R ] dev-util/pkgconfig-0.23
[ebuild R ] app-arch/xz-utils-4.999.9_beta
[ebuild R ] sys-devel/m4-1.4.12
[ebuild R ] sys-apps/sandbox-1.6-r2
[ebuild R ] sys-kernel/linux-headers-2.6.27-r2
[ebuild R ] sys-apps/debianutils-3.1.3
[ebuild R ] virtual/acl-0
[ebuild R ] sys-devel/binutils-config-1.9-r4
[ebuild R ] virtual/libffi-0
[ebuild R ] virtual/perl-digest-base-1.16
[ebuild R ] virtual/perl-Module-Build-0.34.0201
[ebuild U ] app-admin/eselect-opengl-1.1.1-r2 [1.0.8-r1]
[ebuild R ] virtual/perl-Scalar-List-Utils-1.21
[ebuild R ] virtual/perl-ExtUtils-CBuilder-0.26.03
[ebuild R ] virtual/perl-IO-Compress-2.021
[ebuild R ] virtual/perl-Digest-MD5-2.39
[ebuild R ] virtual/perl-MIME-Base64-3.08
[ebuild R ] virtual/perl-libnet-1.22
[ebuild R ] virtual/perl-Archive-Tar-1.54
[ebuild R ] virtual/perl-Test-Harness-3.17
[ebuild R ] virtual/perl-ExtUtils-ParseXS-2.20.0401
[ebuild R ] virtual/perl-IO-Zlib-1.09
[ebuild R ] virtual/perl-Package-Constants-0.02
[ebuild R ] virtual/perl-Compress-Raw-Bzip2-2.021
[ebuild R ] virtual/perl-Compress-Raw-Zlib-2.021
[ebuild R ] sys-devel/autoconf-2.63-r1
[ebuild R ] sys-devel/automake-1.10.2
[ebuild R ] sys-devel/libtool-2.2.6b
[ebuild R ] dev-libs/libffi-3.0.8
[ebuild R ] x11-misc/util-macros-1.3.0
[ebuild R ] dev-lang/perl-5.8.8-r8
[ebuild R ] perl-core/Scalar-List-Utils-1.21
[ebuild R ] perl-core/MIME-Base64-3.08
[ebuild R ] perl-core/libnet-1.22
[ebuild R ] perl-core/Test-Harness-3.17
[ebuild R ] perl-core/Package-Constants-0.02
[ebuild R ] perl-core/Compress-Raw-Bzip2-2.021
[ebuild R ] perl-core/Compress-Raw-Zlib-2.021
[ebuild R ] dev-perl/yaml-0.68
[ebuild R ] perl-core/digest-base-1.16
[ebuild R ] perl-core/IO-Compress-2.021
[ebuild R ] perl-core/Digest-MD5-2.39
[ebuild R ] perl-core/IO-Zlib-1.09
[ebuild R ] dev-perl/Digest-SHA1-2.11
[ebuild R ] perl-core/Archive-Tar-1.54
[ebuild R ] dev-perl/Digest-HMAC-1.01-r1
[ebuild R ] dev-perl/Authen-SASL-2.12
[ebuild R ] perl-core/Module-Build-0.34.0201
[ebuild R ] perl-core/ExtUtils-CBuilder-0.26.03
[ebuild R ] dev-perl/Error-0.17.012
[ebuild R ] perl-core/ExtUtils-ParseXS-2.20.0401
[ebuild R ] sys-devel/binutils-2.18-r3
[ebuild R ] sys-libs/db-4.7.25_p4
[ebuild R ] x11-proto/xproto-7.0.16
[ebuild R ] x11-proto/inputproto-1.5.1
[ebuild R ] x11-libs/xtrans-1.2.5
[ebuild R ] x11-proto/xf86vidmodeproto-2.2.2
[ebuild R ] x11-proto/kbproto-1.0.4
[ebuild R ] dev-libs/libpthread-stubs-0.1
[ebuild R ] x11-proto/xf86driproto-2.1.0
[ebuild R ] x11-proto/bigreqsproto-1.1.0
[ebuild R ] x11-proto/xf86bigfontproto-1.2.0
[ebuild R ] x11-proto/xcmiscproto-1.2.0
[ebuild R ] x11-proto/damageproto-1.2.0
[ebuild R ] x11-proto/fixesproto-4.0
[ebuild R ] sys-libs/gdbm-1.8.3-r4
[ebuild R ] x11-libs/libICE-1.0.6
[ebuild R ] x11-misc/makedepend-1.0.2
[ebuild R ] x11-libs/libXau-1.0.5
[ebuild R ] x11-libs/libXdmcp-1.0.3
[ebuild R ] sys-devel/libperl-5.8.8-r2
[ebuild R ] sys-devel/gettext-0.17
[ebuild R ] sys-devel/flex-2.5.35
[ebuild R ] dev-libs/popt-1.15
[ebuild R ] sys-apps/diffutils-2.8.7-r2
[ebuild R ] sys-apps/attr-2.4.43
[ebuild R ] sys-apps/findutils-4.4.0
[ebuild R ] dev-perl/Locale-gettext-1.05-r1
[ebuild R ] dev-libs/libgpg-error-1.7
[ebuild R ] sys-apps/acl-2.2.47
[ebuild R ] sys-apps/help2man-1.36.4-r1
[ebuild R ] dev-libs/libgcrypt-1.4.4
[ebuild R ] sys-apps/sed-4.2
[ebuild R ] net-misc/rsync-3.0.6
[ebuild R ] dev-libs/openssl-0.9.8l-r2
[ebuild R ] dev-perl/Net-SSLeay-1.35
[ebuild R ] dev-perl/IO-Socket-SSL-1.26
[ebuild R ] dev-perl/Net-SMTP-SSL-1.01
[ebuild R ] dev-util/git-1.6.4.4
[ebuild U ] media-fonts/font-util-9999 [1.1.1]
[ebuild U ] x11-proto/xextproto-9999 [7.0.5]
[ebuild U ] x11-libs/libdrm-9999 [2.4.15] USE="-static-libs%"
[ebuild U ] x11-proto/glproto-9999 [1.4.10]
[ebuild U ] x11-proto/dri2proto-9999 [2.1]
[ebuild R ] sys-libs/ncurses-5.7-r3
[ebuild R ] sys-apps/coreutils-7.5-r1
[ebuild R ] app-shells/bash-4.0_p35
[ebuild R ] sys-apps/texinfo-4.13
[ebuild R ] sys-apps/util-linux-2.16.1
[ebuild R ] sys-libs/gpm-1.20.5
[ebuild R ] sys-libs/readline-6.0_p4
[ebuild R ] app-misc/ca-certificates-20090709
[ebuild R ] app-admin/perl-cleaner-1.05
[ebuild R ] x11-libs/libSM-1.1.1
[ebuild R ] app-admin/eselect-1.2.8
[ebuild R ] app-admin/eselect-python-20090824
[ebuild R ] dev-lang/python-2.6.4
[ebuild R ] dev-libs/libxml2-2.7.3-r2
[ebuild R ] sys-apps/file-5.03
[ebuild R ] dev-libs/libxslt-1.1.26
[ebuild R ] x11-proto/xcb-proto-1.5
[ebuild R ] x11-libs/libxcb-1.4-r1
[ebuild R ] x11-libs/libX11-1.2.2
[ebuild R ] x11-libs/libXext-1.0.5
[ebuild R ] x11-libs/libXfixes-4.0.4
[ebuild R ] x11-libs/libXt-1.0.7-r1
[ebuild R ] x11-libs/libXmu-1.0.5
[ebuild R ] x11-libs/libXdamage-1.1.2
[ebuild R ] x11-libs/libXxf86vm-1.0.2
[ebuild R ] x11-libs/libXi-1.2.1
[ebuild U ] media-libs/mesa-9999 [7.5.2] USE="gallium%* xcb* (-selinux)" VIDEO_CARDS="-nouveau% -svga%"
[blocks B ] <x11-libs/libXext-1.0.99 ("<x11-libs/libXext-1.0.99" is blocking x11-proto/xextproto-9999)
[blocks B ] <x11-base/xorg-server-1.7 ("<x11-base/xorg-server-1.7" is blocking media-libs/mesa-9999)
* Error: The above package list contains packages which cannot be
* installed at the same time on the same system.
('ebuild', '/', 'media-libs/mesa-9999', 'merge') pulled in by
=media-libs/mesa-9999
('installed', '/', 'x11-base/xorg-server-1.6.5-r1', 'nomerge') pulled in by
x11-base/xorg-server required by world
('ebuild', '/', 'x11-libs/libXext-1.0.5', 'merge') pulled in by
x11-libs/libXext required by ('ebuild', '/', 'x11-libs/libXxf86vm-1.0.2', 'merge')
x11-libs/libXext required by ('ebuild', '/', 'x11-libs/libXmu-1.0.5', 'merge')
x11-libs/libXext required by ('ebuild', '/', 'x11-libs/libXi-1.2.1', 'merge')
(and 1 more)
('ebuild', '/', 'x11-proto/xextproto-9999', 'merge') pulled in by
x11-proto/xextproto required by ('ebuild', '/', 'x11-libs/libX11-1.2.2', 'merge')
x11-proto/xextproto required by ('ebuild', '/', 'x11-libs/libXmu-1.0.5', 'merge')
x11-proto/xextproto required by ('ebuild', '/', 'x11-libs/libXfixes-4.0.4', 'merge')
(and 2 more)
For more information about Blocked Packages, please refer to the following
section of the Gentoo Linux x86 Handbook (architecture is irrelevant):
http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?full=1#blocked
* Restoring files.
* done!
|
What version of libXext and xorg-server do you guys use ? _________________ Dell Inspiron E1505 (Intel Core Duo 2GHz, 2GB RAM, SATA 120GB 7200rpm, the rest is standard ) |
|
Back to top |
|
|
saiko n00b
Joined: 16 Apr 2007 Posts: 30 Location: Amsterdam
|
Posted: Tue Jan 19, 2010 11:14 pm Post subject: |
|
|
I guess i am too tired to even look. You are talking about mesa-7.7-r1 and i am going for mesa-9999. Need to finally go to bed after this whole day's battle.
Still the output is the same though - same blocking packages when
Code: | autounmask -p media-libs/mesa-7.7-r1
.....
[blocks B ] <x11-libs/libXext-1.0.99 ("<x11-libs/libXext-1.0.99" is blocking x11-proto/xextproto-9999)
[blocks B ] <x11-base/xorg-server-1.7 ("<x11-base/xorg-server-1.7" is blocking media-libs/mesa-7.7-r1)
* Error: The above package list contains packages which cannot be
* installed at the same time on the same system.
('installed', '/', 'x11-base/xorg-server-1.6.5-r1', 'nomerge') pulled in by
x11-base/xorg-server required by world
('ebuild', '/', 'media-libs/mesa-7.7-r1', 'merge') pulled in by
=media-libs/mesa-7.7-r1
('ebuild', '/', 'x11-libs/libXext-1.0.5', 'merge') pulled in by
x11-libs/libXext required by ('ebuild', '/', 'x11-libs/libXxf86vm-1.0.2', 'merge')
x11-libs/libXext required by ('ebuild', '/', 'x11-libs/libXmu-1.0.5', 'merge')
x11-libs/libXext required by ('ebuild', '/', 'x11-libs/libXi-1.2.1', 'merge')
(and 1 more)
('ebuild', '/', 'x11-proto/xextproto-9999', 'merge') pulled in by
x11-proto/xextproto required by ('ebuild', '/', 'x11-libs/libX11-1.2.2', 'merge')
x11-proto/xextproto required by ('ebuild', '/', 'x11-libs/libXfixes-4.0.4', 'merge')
x11-proto/xextproto required by ('ebuild', '/', 'x11-libs/libXmu-1.0.5', 'merge')
(and 2 more)
|
_________________ Dell Inspiron E1505 (Intel Core Duo 2GHz, 2GB RAM, SATA 120GB 7200rpm, the rest is standard ) |
|
Back to top |
|
|
AmonAmarth l33t
Joined: 03 Mar 2005 Posts: 727
|
Posted: Tue Jan 19, 2010 11:35 pm Post subject: |
|
|
saiko wrote: | glxinfo still says Software Rasterising.
Nevertheless now Xorg.0.log is with no mismatches and feel much more like it should . So that's good. Only this software rasterizing...
glxinfo is still at 170 FPS
Here is Xorg.0.log |
is your user account member of the video group?
about mesa: if you want to use the newest git ati drivers (9999), i recommend to use mesa-9999 too and not to downgrade to mesa-7.7.
i guess you blocker problem derives from the installed x11-base/xorg-server-1.6.5-r1. you have to add the ~arch keywords for it to demask it. and sometimes you have to unmerge it first with emerge -C xorg-server, then upgrade to the newer one.
BTW: your problems are not xorg bleeding edge problems. you are struggling with problems how to demask your unstables versions properly. try do read the documentation and learn how to do this actions. |
|
Back to top |
|
|
rufnut Apprentice
Joined: 16 May 2005 Posts: 252
|
Posted: Thu Jan 21, 2010 12:40 am Post subject: |
|
|
Hi ppls
Any word whether these drivers support XvBA ?
I have the proprietary drivers and tried mplayer-9999-r21 with the arcon overlay which didn't compile.
There is also an ebuild here of which I am yet to try.
http://www.gentoo-quebec.org/index.php?ID=2429
maybe the open source driver can get it running first?
|
|
Back to top |
|
|
pingufunkybeat l33t
Joined: 01 Dec 2004 Posts: 610
|
Posted: Thu Jan 21, 2010 2:20 am Post subject: |
|
|
No.
XvBA is only working in the proprietary driver, and will probably remain that way. Thank the movie industry for this.
Open drivers will be getting decode acceleration via shaders, but that is waiting for Gallium3D, so don't expect it for at least half a year, I'd say. |
|
Back to top |
|
|
VinzC Watchman
Joined: 17 Apr 2004 Posts: 5098 Location: Dark side of the mood
|
Posted: Thu Jan 21, 2010 8:59 am Post subject: |
|
|
pingufunkybeat wrote: | XvBA is only working in the proprietary driver, and will probably remain that way. Thank the movie industry for this. |
Could you expand this? (Off topic, I know.) _________________ Gentoo addict: tomorrow I quit, I promise!... Just one more emerge...
1739! |
|
Back to top |
|
|
pingufunkybeat l33t
Joined: 01 Dec 2004 Posts: 610
|
Posted: Thu Jan 21, 2010 1:41 pm Post subject: |
|
|
The HD line of Radeons have specialised video decoding hardware, like later nVidia cards. They are called UVD/UVD2.
This is the last part of the hardware that hasn't been opened in terms of specifications. The reason why the specs haven't been opened is that these chips also do DRM on Windows hardware, and there is great fear within AMD that opening too much information will help someone crack the BluRay protection. If this happens, AMD is in deep shit, because they would likely lose the license to make this stuff, which would hurt them dearly (imagine if nVidia were the only manufacturer who could speed up playing HD content, and AMD couldn't sell this stuff anymore).
So the IP related to video decoding is being double-triple-quadruple checked by both the technical and legal guys before it is opened. It likely never will be opened.
If it's not opened, you'll never see open source support for it, obviously. Even the AMD guys don't know if and when these specs will be released, but John Bridgman said that they would try to examine it when more basic stuff (power saving) is working, and see how much can be opened.
A more generic acceleration based on OpenCL and shaders will definitely be implemented in the open drivers (using the 3D engine to accelerate many stages of the decoding process, but not using the specialised hardware). These will speed up decoding considerably, but will probably not bring as much powersaving and speed improvement as the dedicated hardware solution. |
|
Back to top |
|
|
Torangan Apprentice
Joined: 21 Mar 2003 Posts: 178
|
Posted: Thu Jan 21, 2010 6:13 pm Post subject: |
|
|
Using KMS and all tools from ~AMD64 I get about half the framerate in glxgears compared to not using KMS. Is there something badly configured or is it just a problem of the drivers? Before the latest -r reversions I had a bug with KMS and before that 1/3 of the framerate.
I'm using 780G chipset graphic and radeon driver. |
|
Back to top |
|
|
saiko n00b
Joined: 16 Apr 2007 Posts: 30 Location: Amsterdam
|
Posted: Thu Jan 21, 2010 11:03 pm Post subject: |
|
|
What a noob i am ... Yes the group was the problem. Never though about it as i was using ati-drivers before with no problem and then just upgraded. Obviously they worked without my account being in the video group. Still i don't think this exercise was pointless as i leaned new things for which i want to thank you all. I appreciate all the help. _________________ Dell Inspiron E1505 (Intel Core Duo 2GHz, 2GB RAM, SATA 120GB 7200rpm, the rest is standard ) |
|
Back to top |
|
|
pingufunkybeat l33t
Joined: 01 Dec 2004 Posts: 610
|
Posted: Thu Jan 21, 2010 11:49 pm Post subject: |
|
|
glxgears is not a benchmark.
The fact that you're getting lower framerates means that there is an extra memory copy involved (since DRI2 enables redirected direct rendering -- i.e. things are rendered into a buffer and then blitted onto the final image). This makes it possible to use several 3D windows on a composited desktop like KWin or Compiz, and it's a good thing. Glxgears is so ridiculously easy to render on today's hardware that the memory copy becomes a bottleneck.
It does NOT mean that your card is rendering at half the speed. You will notice a much smaller difference when using more complex scenes, like modern 3d games.
The optimisation of the driver performance will likely come as the last step, once everything is working, so expect the performance to change from time to time depending on the latest features being added. |
|
Back to top |
|
|
VinzC Watchman
Joined: 17 Apr 2004 Posts: 5098 Location: Dark side of the mood
|
Posted: Fri Jan 22, 2010 9:46 pm Post subject: |
|
|
pingufunkybeat wrote: | glxgears is not a benchmark.
The fact that you're getting lower framerates means that there is an extra memory copy involved (since DRI2 enables redirected direct rendering -- i.e. things are rendered into a buffer and then blitted onto the final image). This makes it possible to use several 3D windows on a composited desktop like KWin or Compiz, and it's a good thing. Glxgears is so ridiculously easy to render on today's hardware that the memory copy becomes a bottleneck.
It does NOT mean that your card is rendering at half the speed. You will notice a much smaller difference when using more complex scenes, like modern 3d games.
The optimisation of the driver performance will likely come as the last step, once everything is working, so expect the performance to change from time to time depending on the latest features being added. |
Thanks for your explanation. I've waited long before I understood why not to use glxgears to check the framerate. So basically that means glxgears is not fitted to the recent enhancements brought to the open source driver, right? _________________ Gentoo addict: tomorrow I quit, I promise!... Just one more emerge...
1739! |
|
Back to top |
|
|
rufnut Apprentice
Joined: 16 May 2005 Posts: 252
|
Posted: Sat Jan 23, 2010 1:16 am Post subject: |
|
|
VinzC wrote: | So basically that means glxgears is not fitted to the recent enhancements brought to the open source driver, right? |
It does let you know that your 3d is working.
The application belongs to Mesa-Progs.
|
|
Back to top |
|
|
pingufunkybeat l33t
Joined: 01 Dec 2004 Posts: 610
|
Posted: Sat Jan 23, 2010 2:55 am Post subject: |
|
|
The glxgears frame rate is simply measuring the wrong thing.
If you are rendering 1800 frames each second, you are making 1800 extra memory copies each second. This is a relatively quick operation, but there's many of them and they make a large proportion of what the driver is doing. With a game like OpenArena or Nexuiz, which render 100 frames per second, the difference in performance should be much smaller.
That, and the optimisations are still coming. |
|
Back to top |
|
|
VinzC Watchman
Joined: 17 Apr 2004 Posts: 5098 Location: Dark side of the mood
|
Posted: Thu Feb 04, 2010 9:26 pm Post subject: |
|
|
I have synced today and upgraded quite a few packages. Now I'm running mesa-7.7-r1, xorg-server-1.7.4.901 and xorg-x11-7.4-r1. All I had to do was save xf86-video-ati from the first page in this thread. I have 3D working! Yessssssssssss...
Moreover I'm running X without xorg.conf, i.e. no xorg configuration file at all. I'll see tomorrow how it is working with an external monitor.
I just have a little problem with Blender-2.49a and menus/buttons. The focus in menus doesn't follow the pointer and no drop down list seems to work. I'll check that a little later. _________________ Gentoo addict: tomorrow I quit, I promise!... Just one more emerge...
1739! |
|
Back to top |
|
|
pingufunkybeat l33t
Joined: 01 Dec 2004 Posts: 610
|
Posted: Fri Feb 05, 2010 3:28 am Post subject: |
|
|
It's great to see that people need much less effort nowadays to get free 3d acceleration on newer Radeons.
Once the DDX driver is updated, you'll be able to run an accelerated system directly from portage. |
|
Back to top |
|
|
VinzC Watchman
Joined: 17 Apr 2004 Posts: 5098 Location: Dark side of the mood
|
Posted: Fri Feb 05, 2010 7:41 am Post subject: |
|
|
pingufunkybeat wrote: | It's great to see that people need much less effort nowadays to get free 3d acceleration on newer Radeons.
Once the DDX driver is updated, you'll be able to run an accelerated system directly from portage. |
And I forgot to mention KMS also works... So I have a console screen with native resolution, working Xorg without config file, multiple displays working and functional 3D -- I could even play UT2004 in 1280x800; a tiny little bit jerky but playable! Man, a beer to anyone here!...
Cheers!
EDIT: LOL! I've just hit my first "weird" bug. While I was reading mails in thunderbird, Xorg stopped responding. The pointer was still moving but clicks were blind and I couldn't switch to a TTY. I used ACPI and the power button to power off my laptop. Also strange is that the screen started to blur and fade while powering down; it reminds me of a known bug with with xf86-video-ati prior to 6.12 IIRC.
I'll investigate this later on when I have some time. But in general, it's good to see it work. I suspect dual screen to be still a little buggy. I'll check once Gentoo sources 2.6.33 are available.
As far as I can tell, it looks like only the kernel driver is responsible for allowing Xorg to run without a configuration file. I've switched back to 2.6.31-r6 and I've had to restore my latest xorg.conf. _________________ Gentoo addict: tomorrow I quit, I promise!... Just one more emerge...
1739! |
|
Back to top |
|
|
Sannin Tux's lil' helper
Joined: 09 Jan 2006 Posts: 85 Location: Greece
|
Posted: Sat Feb 06, 2010 7:29 pm Post subject: |
|
|
Hey
Do you know how radeon + KMS compare to proprietary driver at Opengl games (like World of Warcraft in wine ofc)? |
|
Back to top |
|
|
sinister n00b
Joined: 30 Sep 2004 Posts: 44
|
Posted: Sat Feb 06, 2010 7:35 pm Post subject: |
|
|
Sannin wrote: | Do you know how radeon + KMS compare to proprietary driver at Opengl games (like World of Warcraft in wine ofc)? |
http://www.x.org/wiki/RadeonProgram |
|
Back to top |
|
|
Sannin Tux's lil' helper
Joined: 09 Jan 2006 Posts: 85 Location: Greece
|
Posted: Sat Feb 06, 2010 8:33 pm Post subject: |
|
|
sinister wrote: | Sannin wrote: | Do you know how radeon + KMS compare to proprietary driver at Opengl games (like World of Warcraft in wine ofc)? |
http://www.x.org/wiki/RadeonProgram |
Thank you, it seems i won' t keep the 4890 after all...
WoW doesnt render at all. In games that render perfectly like UT2004 do we know how the drivers compare performance-wise? |
|
Back to top |
|
|
derbrain Apprentice
Joined: 09 Apr 2005 Posts: 178
|
Posted: Thu Feb 18, 2010 5:38 pm Post subject: |
|
|
I was using the packages from the x11-overlay, but they got masked by EAPI 3 some time ago. I didn't realize this, and they got automatically downgraded to the main tree unstable drivers (xf86-video-ati-6.12.4-r1, mesa-7.7-r1, libdrm-2.4.18 ), but I still have 3D hardware acceleration (Radeon HD 4200 onboard graphics)
But now my problem is that after resuming from standby (using the hibernate-ram script), the driver gets really slow and causes high CPU load, so I guess it falls back to software acceleration somehow. But the strange thing is that I don't see any errors, compiz still works (unlike before moving to the new drivers).
Can anyone confirm this behavior? |
|
Back to top |
|
|
earendilion Tux's lil' helper
Joined: 26 Nov 2007 Posts: 115 Location: Meudon (92), France
|
Posted: Sat Feb 27, 2010 5:44 pm Post subject: |
|
|
Hi,
I do not succeed in compiling a kernel with extra-firware support...
I get this error at compilation :
Code: | ithil linux # make
CHK include/linux/version.h
CHK include/generated/utsrelease.h
CALL scripts/checksyscalls.sh
CHK include/generated/compile.h
make[1]: *** Pas de règle pour fabriquer la cible « radeon/R600_rlc.bin », nécessaire pour « firmware//lib/firmware.gen.o ». Arrêt.
make: *** [firmware] Erreur 2
ithil linux # grep firmware .config |
Here is a part of my kernel config :
Code: | CONFIG_FIRMWARE_IN_KERNEL=y
CONFIG_EXTRA_FIRMWARE="/lib/firmware"
CONFIG_EXTRA_FIRMWARE_DIR="radeon/R600_rlc.bin radeon/R700_rlc.bin"
CONFIG_DRM=y
CONFIG_DRM_KMS_HELPER=m
CONFIG_DRM_TTM=m
CONFIG_DRM_RADEON=m
CONFIG_DRM_RADEON_KMS=y |
Any idea where the error come from ? _________________ Laptop : XPS M1330/Core2 T9300 - 2.50Ghz/2Go Ram - DDR2/250Go 7200TPM/nVidia 9200GS/Led Screen
-----
Desktop : Gygabyte EX58-UD3R/Core i7 920 2.67Ghz/3Go Ram - DDR3/WD 1To - 7200TPM - 32Mo Cache/ATI HD4850 - 1Go Ram |
|
Back to top |
|
|
Herring42 Guru
Joined: 10 Mar 2004 Posts: 373 Location: Buckinghamshire
|
Posted: Sat Feb 27, 2010 7:12 pm Post subject: |
|
|
Your need to ensure the firmware files R600_rlc.bin and R700_rlc.bin are in /lib/firmware/radeon
They don't come with the kernel! _________________ "The problem with quotes on the internet is that it is difficult
to determine whether or not they are genuine." -- Abraham Lincoln |
|
Back to top |
|
|
earendilion Tux's lil' helper
Joined: 26 Nov 2007 Posts: 115 Location: Meudon (92), France
|
Posted: Sat Feb 27, 2010 7:18 pm Post subject: |
|
|
They are in there :
Code: | lib/ lib32/ lib64/
earendil@ithil ~/www $ ls /lib/firmware/radeon/
R100_cp.bin R300_cp.bin R520_cp.bin R600_pfp.bin R700_rlc.bin RS690_cp.bin RS780_pfp.bin RV610_pfp.bin RV620_pfp.bin RV630_pfp.bin RV635_pfp.bin RV670_pfp.bin RV710_pfp.bin RV730_pfp.bin RV770_pfp.bin
R200_cp.bin R420_cp.bin R600_me.bin R600_rlc.bin RS600_cp.bin RS780_me.bin RV610_me.bin RV620_me.bin RV630_me.bin RV635_me.bin RV670_me.bin RV710_me.bin RV730_me.bin RV770_me.bin |
_________________ Laptop : XPS M1330/Core2 T9300 - 2.50Ghz/2Go Ram - DDR2/250Go 7200TPM/nVidia 9200GS/Led Screen
-----
Desktop : Gygabyte EX58-UD3R/Core i7 920 2.67Ghz/3Go Ram - DDR3/WD 1To - 7200TPM - 32Mo Cache/ATI HD4850 - 1Go Ram |
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|