Gentoo Forums :: View topic - HOWTO: Radeon 9500 (and newer) with DRI [Part I]

HOWTO: Radeon 9500 (and newer) with DRI [Part I]

View unanswered posts
View posts from last 24 hours

Goto page Previous 1, 2, 3 ... , 38, 39, 40 Next

This topic is locked: you cannot edit posts or make replies.

Gentoo Forums Forum Index

Unsupported Software

View previous topic :: View next topic

Author

Message

SuperDindon
Tux's lil' helper

Joined: 01 Sep 2005
Posts: 125

Posted: Mon Dec 25, 2006 1:33 pm Post subject:

You can but it's absolutely not necessary, the only important issue fixed since 1.1.99.903 is one of the EXA performance trap.

Phlogiston
Veteran
Veteran

Joined: 27 Jan 2004
Posts: 1925
Location: Europe, Swizerland

Posted: Mon Dec 25, 2006 2:09 pm Post subject:

SuperDindon wrote:

You can but it's absolutely not necessary, the only important issue fixed since 1.1.99.903 is one of the EXA performance trap.

And thats in mesa... ok

_________________
Workstation: 5.1 SurroundSound, LIRC remote control; Laptop [IBM-T43]: patched sources, s2disk/ram, fingerprint sensor

Phlogiston
Veteran
Veteran

Joined: 27 Jan 2004
Posts: 1925
Location: Europe, Swizerland

Posted: Mon Dec 25, 2006 2:32 pm Post subject:

Hmm by the way do you guys use XAA for beryl right? _________________ Workstation: 5.1 SurroundSound, LIRC remote control; Laptop [IBM-T43]: patched sources, s2disk/ram, fingerprint sensor

Primozic
Apprentice

Joined: 11 Jun 2004
Posts: 186
Location: Dover, NH, USA

Posted: Mon Dec 25, 2006 4:46 pm Post subject:

Again, sorry folks. The motherboard I received was dead on arrival. It was, admittedly, an open box item. I have an updated tarball is anyone would like to obtain and/or host it. Private message me your e-mail address if you're interested. _________________ Interceptor: Intel Core 2 Duo 2.4GHz \| AOpen i965GMt-LA Mini-ITX \| 4GB PC2-5300 DDR2 Pursuit: AMD Geode NX 1500 1.0GHz \| TYAN Tomcat K7M S2498AGNN FlexATX \| 2GB PC2700 DDR

_pF_
Apprentice

Joined: 22 Apr 2004
Posts: 195
Location: Central London

Posted: Mon Dec 25, 2006 5:08 pm Post subject:

Thank you SuperDindon, your overlay solved my long-standing X-deficiency problem mentioned in this thread and here.

Phlogiston
Veteran
Veteran

Joined: 27 Jan 2004
Posts: 1925
Location: Europe, Swizerland

Posted: Mon Dec 25, 2006 6:57 pm Post subject:

Primozic wrote:

Again, sorry folks. The motherboard I received was dead on arrival. It was, admittedly, an open box item.

I have an updated tarball is anyone would like to obtain and/or host it. Private message me your e-mail address if you're interested.

Here you go:
updated tarball

_________________
Workstation: 5.1 SurroundSound, LIRC remote control; Laptop [IBM-T43]: patched sources, s2disk/ram, fingerprint sensor

Last edited by Phlogiston on Mon Dec 25, 2006 7:43 pm; edited 1 time in total

R!tman
Veteran
Veteran

Joined: 18 Dec 2003
Posts: 1303
Location: Zurich, Switzerland

Posted: Mon Dec 25, 2006 7:28 pm Post subject:

Phlogiston wrote:

Primozic wrote:

Here you go:
updated tarball

The 'h' is missing in "http" ;-)

. Otherwise, thanks for hosting!

BTW, I could offer an rsync solution. As a short term solution on my own server (with a slow inet connection), and as of maybe tomorrow on another computer with a _really_ fast inet connection (_exactly_ as fast as Phlogiston's solution, but rsynced).

On the other hand, I doubt a fast internet connection is really needed...

Phlogiston
Veteran
Veteran

Joined: 27 Jan 2004
Posts: 1925
Location: Europe, Swizerland

Posted: Mon Dec 25, 2006 7:44 pm Post subject:

R!tman wrote:

The 'h' is missing in "http" ;-)

You're right thanks I corrected it. I don't think that so much people have a need for it.

Especially because the ebuilds are for svn, so you can update whenever you want. (if the compile doesn't fail of course

)
_________________
Workstation: 5.1 SurroundSound, LIRC remote control; Laptop [IBM-T43]: patched sources, s2disk/ram, fingerprint sensor

nigelhannam
n00b
n00b

Joined: 21 Feb 2003
Posts: 70

Posted: Mon Dec 25, 2006 9:46 pm Post subject:

txboy wrote:

tontudelbot wrote:

I see it's declarated in ativersion.h... (i don't know c++, sorry...) ativersion.h diff

Ok, I've checked my src/ativersion.h and that patch is included.

Does someone know other solution ?

It looks to me as though that patch is the cause of the problem. It changes the defines for ATI_VERSION_MAJOR, ATI_VERSION_MINOR and ATI_VERSION_PATCH so that instead of being set directly from PACKAGE_VERSION, they are set from PACKAGE_VERSION_MAJOR, PACKAGE_VERSION_MINOR and PACKAGE_VERSION_PATCHLEVEL (which never seem to be set themselves)

If you substitute these values for some unquoted numeric values, it seems to compile ok. (not finished compiling mesa yet, so will update when I've tested it)

argasek
Bodhisattva

Joined: 06 May 2004
Posts: 1120
Location: Sol [0,0], Poland, Kraków

Posted: Tue Dec 26, 2006 10:10 pm Post subject:

Here's a better solution.

Code:

--- xf86-video-ati-9999/work/xf86-video-ati-9999/configure.ac 2006-12-19 20:07:25.000000000 +0100
+++ xf86-video-ati-9999/work/xf86-video-ati-9999/configure.ac 2006-12-26 03:37:19.000000000 +0100
@@ -25,6 +25,15 @@
6.6.99,
[https://bugs.freedesktop.org/enter_bug.cgi?product=xorg],
xf86-video-ati)
+AC_DEFINE_UNQUOTED([PACKAGE_VERSION_MAJOR],
+       [$(echo $PACKAGE_VERSION | sed -e 's/^$[[0-9]]$\.[[0-9]]\.[[0-9]]/\1/')],
+       [Major version])
+AC_DEFINE_UNQUOTED([PACKAGE_VERSION_MINOR],
+       [$(echo $PACKAGE_VERSION | sed -e 's/^[[0-9]]\.$[[0-9]]$\.[[0-9]]/\1/')],
+       [Minor version])
+AC_DEFINE_UNQUOTED([PACKAGE_VERSION_PATCHLEVEL],
+       [$(echo $PACKAGE_VERSION | sed -e 's/^[[0-9]]\.[[0-9]]\.$[[0-9]]$/\1/')],
+       [Patch version])

AC_CONFIG_SRCDIR([Makefile.am])
AM_CONFIG_HEADER([config.h])

_________________
RLU #137109 | BYKOM STOP! | Chcesz do mnie napisać? | jog
"Mam plany gotowe i wszystko, no i wiesz, idę do tego, no, mebloroba..." (Raaf)

argasek
Bodhisattva

Joined: 06 May 2004
Posts: 1120
Location: Sol [0,0], Poland, Kraków

Posted: Tue Dec 26, 2006 11:44 pm Post subject:

Phlogiston wrote:

Hmm by the way do you guys use XAA for beryl right?

Well, both XAA and EXA can give good or bad results, it depends on the xorg-server version. In general, xorg-server-9999 will give you nice speed boost when using EXA (and some new bugs, probably

), but with Beryl you may need to experiment with render-to-texture / copy settings (sometimes one works, while the other doesn't, sometimes both work OK). For details of EXA improvements, refer to http://bugs.freedesktop.org/ (search for EXA). I'm currently experimenting with various sets of xorg-server / mesa / ati-drivers, will share my experiences if I find something interesting.

At the moment applying most new EXA patches from bugzilla to xorg-server-1.2.99.0 results in black cube (as opposed to "white cube") bug

, so I think one should stick to 1.1.99 version from overlay provided in this thread or may try xorg-server-9999 (it's in some overlay, Google for it).
_________________
RLU #137109 | BYKOM STOP! | Chcesz do mnie napisać? | jog
"Mam plany gotowe i wszystko, no i wiesz, idę do tego, no, mebloroba..." (Raaf)

tontudelbot
Tux's lil' helper

Joined: 08 Jun 2006
Posts: 77

Posted: Wed Dec 27, 2006 12:01 am Post subject:

Hi guys, how you get opengl applications working good with aiglx? I can't It must be an option in section device of xorg.conf, i think it should be easy

zxy
Veteran
Veteran

Joined: 06 Jan 2006
Posts: 1160
Location: in bed in front of the computer

Posted: Wed Dec 27, 2006 12:41 am Post subject:

Hi, to all of you. I confess that I didn't read through all the posts in this topic. I have a question. What drivers should I use on a machine with radeon 9600 (I think this is 350)? Should I try ati drivers or open source ones? _________________ Nature does not hurry, yet everything is accomplished. Lao Tzu

txboy
Tux's lil' helper

Joined: 01 Apr 2004
Posts: 86
Location: La Isla - Cádiz (Spain)

Posted: Wed Dec 27, 2006 9:44 am Post subject:

argasek wrote:

Here's a better solution.

Code:

Ok. I haven't time to try it now, so I'll try it later.
Thank you very much.

MADevil
n00b
n00b

Joined: 26 Nov 2006
Posts: 57

Posted: Wed Dec 27, 2006 2:19 pm Post subject:

I've some trouble when i try to compile xorg-server ,

Code:

< PATCHES="${FILESDIR}/${P}-mesa-git.patch"

is unneeded now ? (in the new ebuild)

Code:

i686-pc-linux-gnu-gcc -DHAVE_DIX_CONFIG_H -DNO_HW_ONLY_EXTS -DNO_MODULE_EXTS -DXFree86Server -Wall -Wpointer-arith -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations -Wnested-externs -fno-strict-aliasing -D_BSD_SOURCE -DHAS_FCHOWN -DHAS_STICKY_DIR_BIT -I/usr/include/freetype2 -I../../include -I../../include -I../../Xext -I../../composite -I../../damageext -I../../xfixes -I../../Xi -I../../mi -I../../miext/shadow -I../../miext/damage -I../../render -I../../randr -I../../fb -O2 -march=athlon-xp -pipe -Wl,-z -Wl,lazy -o Xvfb InitInput.o InitOutput.o dpmsstubs.o stubs.o miinitext.o fbcmap.o ../../dix/.libs/libdix.a ../../fb/.libs/libfb.a ../../mi/.libs/libmi.a ../../xfixes/.libs/libxfixes.a ../../Xext/.libs/libXext.a ../../dbe/.libs/libdbe.a ../../XTrap/.libs/libxtrap.a ../../record/.libs/librecord.a ../../GL/glx/.libs/libglx.a ../../GL/mesa/.libs/libGLcore.a -lpthread ../../render/.libs/librender.a ../../randr/.libs/librandr.a ../../damageext/.libs/libdamageext.a ../../miext/damage/.libs/libdamage.a ../../miext/cw/.libs/libcw.a ../../miext/shadow/.libs/libshadow.a ../../Xi/.libs/libXi.a ../../xkb/.libs/libxkb.a ../../xkb/.libs/libxkbstubs.a ../../composite/.libs/libcomposite.a ../../dix/.libs/libxpstubs.a ../../os/.libs/libcwrapper.a ../../os/.libs/libos.a /usr/lib/libXfont.so /usr/lib/libfreetype.so /usr/lib/libXau.so /usr/lib/libfontenc.so -lz /usr/lib/libXdmcp.so -lm -lrt
../../GL/mesa/.libs/libGLcore.a(fbobject.o): In function `_mesa_GenerateMipmapEXT':
fbobject.c:(.text+0x12a1): undefined reference to `_mesa_generate_mipmap'
../../GL/mesa/.libs/libGLcore.a(texstore.o): In function `_mesa_store_compressed_texsubimage2d':
texstore.c:(.text+0x10bf): undefined reference to `_mesa_generate_mipmap'
../../GL/mesa/.libs/libGLcore.a(texstore.o): In function `_mesa_store_compressed_teximage2d':
texstore.c:(.text+0x11c3): undefined reference to `_mesa_generate_mipmap'
../../GL/mesa/.libs/libGLcore.a(texstore.o): In function `_mesa_store_texsubimage3d':
texstore.c:(.text+0x1475): undefined reference to `_mesa_generate_mipmap'
../../GL/mesa/.libs/libGLcore.a(texstore.o): In function `_mesa_store_texsubimage2d':
texstore.c:(.text+0x1615): undefined reference to `_mesa_generate_mipmap'
../../GL/mesa/.libs/libGLcore.a(texstore.o):texstore.c:(.text+0x177f): more undefined references to `_mesa_generate_mipmap' follow
../../GL/mesa/.libs/libGLcore.a(texcompress_fxt1.o): In function `fxt1_encode':
texcompress_fxt1.c:(.text+0x1ed6): undefined reference to `_mesa_upscale_teximage2d'
../../GL/mesa/.libs/libGLcore.a(s_texstore.o): In function `_swrast_copy_texsubimage3d':
s_texstore.c:(.text+0x752): undefined reference to `_mesa_generate_mipmap'
../../GL/mesa/.libs/libGLcore.a(s_texstore.o): In function `_swrast_copy_texsubimage2d':
s_texstore.c:(.text+0x9d1): undefined reference to `_mesa_generate_mipmap'
../../GL/mesa/.libs/libGLcore.a(s_texstore.o): In function `_swrast_copy_texsubimage1d':
s_texstore.c:(.text+0xbf4): undefined reference to `_mesa_generate_mipmap'
../../GL/mesa/.libs/libGLcore.a(s_texstore.o): In function `_swrast_copy_teximage2d':
s_texstore.c:(.text+0xf04): undefined reference to `_mesa_generate_mipmap'
../../GL/mesa/.libs/libGLcore.a(s_texstore.o): In function `_swrast_copy_teximage1d':
s_texstore.c:(.text+0x1133): undefined reference to `_mesa_generate_mipmap'
collect2: ld returned 1 exit status
make[3]: *** [Xvfb] Erreur 1
make[3]: quittant le répertoire « /var/tmp/portage/x11-base/xorg-server-1.1.99.999/work/xorg-server-1.1.99.999/hw/vfb »
make[2]: *** [all] Erreur 2
make[2]: quittant le répertoire « /var/tmp/portage/x11-base/xorg-server-1.1.99.999/work/xorg-server-1.1.99.999/hw/vfb »
make[1]: *** [all-recursive] Erreur 1
make[1]: quittant le répertoire « /var/tmp/portage/x11-base/xorg-server-1.1.99.999/work/xorg-server-1.1.99.999/hw »
make: *** [all-recursive] Erreur 1

!!! ERROR: x11-base/xorg-server-1.1.99.999 failed.
Call stack:
ebuild.sh, line 1580: Called dyn_compile
ebuild.sh, line 945: Called src_compile
ebuild.sh, line 1269: Called x-modular_src_compile
x-modular.eclass, line 333: Called x-modular_src_make
x-modular.eclass, line 328: Called die

!!! emake failed

If i roll back to the old xorg-server ebuild + patch , it's ok , it is the patch , or nouveau or xgi ? :wink:

_________________
Athlon XP-M @ 2.3 2x256 BH5 Dual 6600 GT Agp 8x NForce2 Ultra 400
Gentoo 2.6.22-r1 Gcc 4.2.0 xorg-server-1.2.0-r3;nvidia-drivers-1.0.9755-r1
kde 3.5.7 + compiz-fusion;atlantis;3d;screensaver;wallpaper;shift;ezoom...

Last edited by MADevil on Wed Dec 27, 2006 6:53 pm; edited 2 times in total

decoder
n00b
n00b

Joined: 27 Mar 2006
Posts: 7

Posted: Wed Dec 27, 2006 4:53 pm Post subject:

Hi all, I have an ATI X300 in my laptop, and I made it to get DRI working with the open source drivers. I emerged the xorg-server with aiglx, libdrm, and compiled the rest myself (Mesa and DRM). With glxgears I get 1800 FPS, glxinfo shows direct rendering. I even got beryl to work, but it is extremely slow (using XAA rendering, with EXA it freezes). I also tried to play bzflag (under normal KDE environment), it is unplayable even with the lowest details, slow like hell (like 5 FPS or less). With the binary ATI driver, this game, and others like UT2k4 worked like a charm, with the open source driver, glxgears is faster, but the games are slow like hell. Does anyone have an idea what could be going wrong or is this normal with the open source drivers? Chris

Phlogiston
Veteran
Veteran

Joined: 27 Jan 2004
Posts: 1925
Location: Europe, Swizerland

Posted: Wed Dec 27, 2006 8:10 pm Post subject:

argasek wrote:

Well, both XAA and EXA can give good or bad results, it depends on the xorg-server version. In general, xorg-server-9999 will give you nice speed boost when using EXA (and some new bugs, probably

, so I think one should stick to 1.1.99 version from overlay provided in this thread or may try xorg-server-9999 (it's in some overlay, Google for it).

But can you (or could you) use EXA stable? Here I can't even get to work properly: The content inside the windows is rendered very slowly and does not get refreshed correctly. Could you provide me a xorg.conf with options for working EXA? (I'm trying to get EXA to work because XAA is not stable here: It crashes under some load after a while)
_________________
Workstation: 5.1 SurroundSound, LIRC remote control; Laptop [IBM-T43]: patched sources, s2disk/ram, fingerprint sensor

Phlogiston
Veteran
Veteran

Joined: 27 Jan 2004
Posts: 1925
Location: Europe, Swizerland

Posted: Wed Dec 27, 2006 8:15 pm Post subject:

decoder wrote:

Hi all,

I have an ATI X300 in my laptop, and I made it to get DRI working with the open source drivers. I emerged the xorg-server with aiglx, libdrm, and compiled the rest myself (Mesa and DRM). With glxgears I get 1800 FPS, glxinfo shows direct rendering.

I even got beryl to work, but it is extremely slow (using XAA rendering, with EXA it freezes). I also tried to play bzflag (under normal KDE environment), it is unplayable even with the lowest details, slow like hell (like 5 FPS or less). With the binary ATI driver, this game, and others like UT2k4 worked like a charm, with the open source driver, glxgears is faster, but the games are slow like hell.

Does anyone have an idea what could be going wrong or is this normal with the open source drivers?

Chris

I have the same card. Did you try the options like pageflip oder colortiling in your xorg.conf? Beryl runs quite fast here but its not stable with XAA, and as you can get from my previous post I can't get EXA to work correctly...
I only tested quake3demo as a 3d game and it runs fast at 1024x768.
_________________
Workstation: 5.1 SurroundSound, LIRC remote control; Laptop [IBM-T43]: patched sources, s2disk/ram, fingerprint sensor

decoder
n00b
n00b

Joined: 27 Mar 2006
Posts: 7

Posted: Wed Dec 27, 2006 10:18 pm Post subject:

Hi yes, I had these options turned on in xorg.conf Option "AGPMode" "4" Option "AGPFastWrite" "yes" Option "EnablePageFlip" "on" Option "RenderAccel" "on" Option "AllowGLXWithComposite" "on" Option "AccelMethod" "XAA" # or EXA About the Color Tile Option, I had it turned on first, then turned it off to make sure this isnt an issue, but nothing changed, the problems didnt go away. Can you show/send me your xorg.conf so I can make sure it isnt a misconfiguration ? Thanks and regards, Chris P.S.: My email is decoder at own-hero dot net

MADevil
n00b
n00b

Joined: 26 Nov 2006
Posts: 57

Posted: Thu Dec 28, 2006 2:33 am Post subject:

decoder wrote:

Option "AGPMode" "4"
Option "AGPFastWrite" "yes"
Option "EnablePageFlip" "on"
Option "RenderAccel" "on"
Option "AllowGLXWithComposite" "on"
Option "AccelMethod" "XAA" # or EXA

Change your AGPFastWrite to false

RenderAccel and AllowGLXWithComposite are useless with radeon i think
_________________
Athlon XP-M @ 2.3 2x256 BH5 Dual 6600 GT Agp 8x NForce2 Ultra 400
Gentoo 2.6.22-r1 Gcc 4.2.0 xorg-server-1.2.0-r3;nvidia-drivers-1.0.9755-r1
kde 3.5.7 + compiz-fusion;atlantis;3d;screensaver;wallpaper;shift;ezoom...

argasek
Bodhisattva

Joined: 06 May 2004
Posts: 1120
Location: Sol [0,0], Poland, Kraków

Posted: Thu Dec 28, 2006 6:49 am Post subject:

MADevil wrote:

decoder wrote:

Option "AGPMode" "4"
Option "AGPFastWrite" "yes"
Option "EnablePageFlip" "on"
Option "RenderAccel" "on"
Option "AllowGLXWithComposite" "on"
Option "AccelMethod" "XAA" # or EXA

Change your AGPFastWrite to false

RenderAccel and AllowGLXWithComposite are useless with radeon i think

RenderAccel works for r200 cards, but r300+ are unsupported.
Second option is only for nVidia cards.
AGPFastWrite "yes" is asking for troubles.

_________________
RLU #137109 | BYKOM STOP! | Chcesz do mnie napisać? | jog
"Mam plany gotowe i wszystko, no i wiesz, idę do tego, no, mebloroba..." (Raaf)

infirit
l33t
l33t

Joined: 11 Jan 2003
Posts: 774
Location: Hoofddorp / The Netherlands

Posted: Thu Dec 28, 2006 3:36 pm Post subject:

argasek wrote:

AGPFastWrite "yes" is asking for troubles.

And mostly pointless in the performance department :wink:

_________________
EASY TO INSTALL = Difficult to install, but instruction manual has pictures.
Join the adopt an unanswered post initiative today

argasek
Bodhisattva

Joined: 06 May 2004
Posts: 1120
Location: Sol [0,0], Poland, Kraków

Posted: Thu Dec 28, 2006 6:28 pm Post subject:

Phlogiston wrote:

Here it is, but it's useless at the moment (read further why):

Code:

Section "ServerLayout"
Identifier "Standardowy"
# Screen 0 "FGLRX Screen 1" 0 0
# Screen 1 "FGLRX Screen 2" 0 0
Screen 0 "Radeon Screen" 0 0
InputDevice "A4TechMouse" "CorePointer"
InputDevice "A4TechKeyboard" "CoreKeyboard"
# InputDevice "Volito2" "SendCoreEvents"
Option "AIGLX" "true"
EndSection

Section "Module"
# SubSection "extmod"
# Option "omit xfree86-dga"
# EndSubSection
Load "dbe"
Load "freetype"
Load "dri"
Load "drm"
Load "glx"
Load "extmod"
EndSection

Section "InputDevice"
Identifier "Volito2"
Driver "wacom"
Option "Device" "/dev/input/wacom"
Option "Type" "stylus"
Option "USB" "on"
EndSection

Section "InputDevice"
Identifier "A4TechKeyboard"
Driver "kbd"
Option "AutoRepeat" "250 30"
Option "XkbRules" "xorg"
Option "XkbModel" "pc105"
Option "XkbLayout" "pl"
# Option "XkbOptions" "grp:menu_toggle,lv3:lwin_switch,ctrl:ctrl_aa,grp_led:scroll,caps:internal"
EndSection

Section "InputDevice"
Identifier "A4TechMouse"
Driver "mouse"
Option "Protocol" "IMPS/2"
Option "Device" "/dev/input/mice"
Option "ZAxisMapping" "4 5"
EndSection

Section "Modes"
Identifier "Rozdzielczosci"
############ WYSOKIE ROZDZIELCZOSCI ###############
# 640x400 @ 85 Hz, 37.86 kHz hsync
ModeLine "640x400" 31.5 640 672 736 832 400 401 404 445 -HSync +VSync
# 640x480 @ 85 Hz, 43.27 kHz hsync
ModeLine "640x480" 36 640 696 752 832 480 481 484 509 -HSync -VSync
# 800x600 @ 85 Hz, 55.84 kHz hsync
ModeLine "800x600" 60.75 800 864 928 1088 600 616 621 657 -HSync -VSync
# 1024x768 @ 85 Hz, 70.24 kHz hsync
ModeLine "1024x768" 98.9 1024 1056 1216 1408 768 782 788 822 -HSync -VSync
# 1152x864 @ 78 Hz, 70.8 kHz hsync
ModeLine "1152x864" 110 1152 1240 1324 1552 864 864 876 908
# 1280x960 @ 85Hz (VESA) hsync: 85.9kHz
ModeLine "1280x960" 148.5 1280 1344 1504 1728 960 961 964 1011 +hsync +vsync
# 1280x1024 @ 61 Hz, 64.2 kHz hsync
ModeLine "1280x1024" 110 1280 1328 1512 1712 1024 1025 1028 1054
# 1400x1050 @ 85Hz, 93.2 kHz hsync
Modeline "1400x1050" 184 1400 1464 1656 1960 1050 1051 1054 1100 +HSync +VSync

############ NISKIE ROZDZIELCZOSCI ################
# 320x200 @ 70 Hz, 31.5 kHz hsync, 8:5 aspect ratio
ModeLine "320x200" 12.588 320 336 384 400 200 204 205 225 Doublescan
# 320x240 @ 72 Hz, 36.5 kHz hsync
ModeLine "320x240" 15.750 320 336 384 400 240 244 246 262 Doublescan
# 400x300 @ 72 Hz, 48.0 kHz hsync
ModeLine "400x300" 25 400 424 488 520 300 319 322 333 Doublescan
# 480x300 @ 72 Hz, 48.0 kHz hsync
ModeLine "480x300" 29.952 480 504 584 624 300 319 322 333 Doublescan
# 512x384 @ 85 Hz, 34.38 kHz hsync
ModeLine "512x384" 22 512 528 592 640 384 385 388 404 -HSync -VSync
EndSection

Section "Monitor"
Identifier "LG Flatron F700P FGLRX"
Gamma 1.16 1.18 1.16
EndSection

Section "Monitor"
Identifier "LG Flatron F700P Radeon"
Gamma 1.16 1.18 1.16

# FGLRX 8.32.5 musi mieæ off :-(
UseModes "Rozdzielczosci"

# Te opcje s± tylko dla sterowników nVidii
# Option "IgnoreDisplayDevices" "DFP, TV"
# Option "ConnectedMonitor" "CRT"
EndSection

Section "Device"
Identifier "Radeon"
Driver "radeon"
BusID "PCI:1:0:0"
Screen 0
Option "AccelMethod" "EXA"
Option "AccelDFS" "on"
Option "BackingStore" "off"
Option "ColorTiling" "on"
Option "EnablePageFlip" "on"
#Option "FBTexPercent" "50"
Option "GartSIZE" "128"
Option "RenderAccel" "off"
Option "MigrationHeuristic" "greedy" # smart=default, always

# XAA buggy rendering workarounds
#Option "XaaNoScanlineImageWriteRect"
#Option "XaaNoScanlineCPUToScreenColorExpandFill"
#Option "XAANoOffscreenPixmaps" "true"
EndSection

Section "Device"
Identifier "FGLRX 1"
Driver "fglrx"
BusID "PCI:1:0:0"
Option "VideoOverlay" "on"
Option "DPMS" "true"
Option "backingstore" "true"
EndSection

Section "Device"
Identifier "FGLRX 2"
Driver "fglrx"
BusID "PCI:1:0:1"
Option "VideoOverlay" "on"
Option "DPMS" "true"
Option "backingstore" "true"
EndSection

Section "Screen"
Identifier "Radeon Screen"
Device "Radeon"
Monitor "LG Flatron F700P Radeon"
DefaultDepth 24
SubSection "Display"
Depth 16
Modes "1152x864" "1024x768" "800x600" "640x480"
EndSubSection
SubSection "Display"
Depth 24
Modes "1152x864" "1024x768" "800x600" "640x480"
EndSubSection
EndSection

Section "Screen"
Identifier "FGLRX Screen 1"
Device "FGLRX 1"
Monitor "LG Flatron F700P FGLRX"
DefaultDepth 24
SubSection "Display"
Depth 24
Modes "1152x864" "1024x768" "800x600" "640x480"
EndSubSection
SubSection "Display"
Depth 32
Modes "1152x864" "1024x768" "800x600" "640x480"
EndSubSection
EndSection

Section "Screen"
Identifier "FGLRX Screen 2"
Device "FGLRX 2"
Monitor "LG Flatron F700P FGLRX"
DefaultDepth 32
SubSection "Display"
Depth 32
Modes "1152x864" "1024x768" "800x600" "640x480"
EndSubSection
EndSection

Section "Files"
FontPath "/usr/share/fonts/100dpi"
FontPath "/usr/share/fonts/75dpi"
FontPath "/usr/share/fonts/misc"
FontPath "/usr/share/fonts/Type1"
FontPath "/usr/share/fonts/local"
FontPath "/usr/share/fonts/default/ghostscript"
FontPath "/usr/share/fonts/Adobe"
FontPath "/usr/share/fonts/Fontic"
FontPath "/usr/share/fonts/GUST"
FontPath "/usr/share/fonts/Microsoft"
FontPath "/usr/share/fonts/terminus"
FontPath "/usr/share/fonts/dejavu"

FontPath "unix/:-1"
EndSection

Section "DRI"
# Yeah, full rights is quite a satanic number ;-) -- argasek
Mode 0666
EndSection

Section "Extensions"
Option "Composite" "Enable"
EndSection

But as I wrote earlier, it's useless. It's no magic here, maybe with exception of new option "MigrationHeuristic" - which is only valid for GIT current branch (this one which is going to be X.org 7.3, I suppose). More useful (if You're brave one) could be my current xorg-server-9999.ebuild and mesa-9999.ebuild + patches. Notice new USE options I've added for convinience: beryl, exa - they activate various fixes for these.

But let's make it clear: it's untested & buggy, may even not compile for You, can kill Your cat, etc.

At the moment this ebuild with USE="beryl exa" accelerates EXA a lot, but trying to launch Beryl crashes X with

Code:

Backtrace:
0: /usr/bin/X(xf86SigHandler+0x82) [0x80cbd52]

Fatal server error:
Caught signal 11. Server aborting

which I have to investigate yet. With USE="-beryl exa" EXA is of course accelerated, Beryl runs but with "white cube" bug (see ebuild and bugs.freedesktop.org for details) and is useless.

I will post (for sure!) first working xorg-server/mesa/patches combination in this thread, but I guess until it happens X.org 7.3 will be realeased and everybody gets happier

(By "working" I mean working EXA+GLX+Beryl).

xorg-server-9999.ebuild:

Code:

# Copyright 1999-2006 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# This ebuild is based on spyderous' one, patches from bugs.freedesktop.org

# Must be before x-modular eclass is inherited
SNAPSHOT="yes"

inherit x-modular multilib cvs git

EGIT_REPO_URI="git://anongit.freedesktop.org/git/xorg/xserver"
SRC_URI=""

OPENGL_DIR="xorg-x11"

MESA_PN="Mesa"
MESA_P="${MESA_PN}"

DESCRIPTION="X.Org X servers"
# It's suid and has lazy bindings, so FEATURES="stricter" doesn't work
RESTRICT="stricter"
KEYWORDS="-*"
IUSE_INPUT_DEVICES="
input_devices_acecad
input_devices_aiptek
input_devices_calcomp
input_devices_citron
input_devices_digitaledge
input_devices_dmc
input_devices_dynapro
input_devices_elo2300
input_devices_elographics
input_devices_evdev
input_devices_fpit
input_devices_hyperpen
input_devices_jamstudio
input_devices_joystick
input_devices_keyboard
input_devices_magellan
input_devices_microtouch
input_devices_mouse
input_devices_mutouch
input_devices_palmax
input_devices_penmount
input_devices_spaceorb
input_devices_summa
input_devices_tek4957
input_devices_ur98
input_devices_vmmouse
input_devices_void

input_devices_synaptics
input_devices_wacom"
IUSE_VIDEO_CARDS="
video_cards_apm
video_cards_ark
video_cards_chips
video_cards_cirrus
video_cards_cyrix
video_cards_dummy
video_cards_epson
video_cards_fbdev
video_cards_glint
video_cards_i128
video_cards_i740
video_cards_i810
video_cards_impact
video_cards_imstt
video_cards_mach64
video_cards_mga
video_cards_neomagic
video_cards_newport
video_cards_nsc
video_cards_nv
video_cards_r128
video_cards_radeon
video_cards_rendition
video_cards_s3
video_cards_s3virge
video_cards_savage
video_cards_siliconmotion
video_cards_sis
video_cards_sisusb
video_cards_sunbw2
video_cards_suncg14
video_cards_suncg3
video_cards_suncg6
video_cards_sunffb
video_cards_sunleo
video_cards_suntcx
video_cards_tdfx
video_cards_tga
video_cards_trident
video_cards_tseng
video_cards_v4l
video_cards_vesa
video_cards_vga
video_cards_via
video_cards_vmware
video_cards_voodoo

video_cards_fglrx
video_cards_nvidia"
IUSE_SERVERS="dmx kdrive xorg"
IUSE="${IUSE_VIDEO_CARDS}
${IUSE_INPUT_DEVICES}
${IUSE_SERVERS}
3dfx
aiglx
beryl
exa
dri ipv6 minimal nptl sdl xprint"
RDEPEND="x11-libs/libXfont
x11-libs/xtrans
x11-libs/libXau
x11-libs/libXext
x11-libs/libX11
x11-libs/libxkbfile
x11-libs/libXdmcp
x11-libs/libXmu
x11-libs/libXrender
x11-libs/libXi
media-libs/freetype
>=media-libs/mesa-6.5-r2
media-fonts/font-adobe-75dpi
media-fonts/font-misc-misc
media-fonts/font-cursor-misc
x11-misc/xbitmaps
|| ( x11-misc/xkeyboard-config x11-misc/xkbdata )
x11-apps/iceauth
x11-apps/rgb
x11-apps/xauth
x11-apps/xinit
app-admin/eselect-opengl
x11-libs/libXaw
x11-libs/libXpm
x11-libs/libXxf86misc
x11-libs/libXxf86vm
dmx? ( x11-libs/libdmx )
!minimal? ( x11-libs/libXtst
   x11-libs/libXres )
>=x11-libs/libxkbui-1.0.2
x11-libs/liblbxutil
kdrive? ( sdl? ( media-libs/libsdl ) )"
# Xres is dmx-dependent, xkbui is xorgcfg-dependent
# Xaw is dmx- and xorgcfg-dependent
# Xpm is dmx- and xorgcfg-dependent, pulls in Xt
# Xxf86misc and Xxf86vm are xorgcfg-dependent
# liblbxutil is lbx- dependent
DEPEND="${RDEPEND}
x11-proto/randrproto
x11-proto/renderproto
>=x11-proto/fixesproto-4
x11-proto/damageproto
x11-proto/xextproto
x11-proto/xproto
x11-proto/xf86dgaproto
x11-proto/xf86miscproto
x11-proto/xf86rushproto
x11-proto/xf86vidmodeproto
x11-proto/xf86bigfontproto
>=x11-proto/compositeproto-0.3
x11-proto/recordproto
x11-proto/resourceproto
x11-proto/videoproto
>=x11-proto/scrnsaverproto-1.1.0
x11-proto/evieext
x11-proto/trapproto
>=x11-proto/xineramaproto-1.1-r1
x11-proto/fontsproto
>=x11-proto/kbproto-1.0.3
>=x11-proto/inputproto-1.4
x11-proto/bigreqsproto
x11-proto/xcmiscproto
>=x11-proto/glproto-1.4.8
dmx? ( x11-proto/dmxproto )
dri? ( x11-proto/xf86driproto
   >=x11-libs/libdrm-2.2 )
xprint? ( x11-proto/printproto
   x11-apps/mkfontdir
   x11-apps/mkfontscale
   x11-apps/xplsprinters )"

# Drivers
PDEPEND="
xorg? (
   input_devices_acecad? ( >=x11-drivers/xf86-input-acecad-1.1.0 )
   input_devices_aiptek? ( >=x11-drivers/xf86-input-aiptek-1.0.1 )
   input_devices_calcomp? ( >=x11-drivers/xf86-input-calcomp-1.1.0 )
   input_devices_citron? ( >=x11-drivers/xf86-input-citron-2.2.0 )
   input_devices_digitaledge? ( >=x11-drivers/xf86-input-digitaledge-1.1.0 )
   input_devices_dmc? ( >=x11-drivers/xf86-input-dmc-1.1.0 )
   input_devices_dynapro? ( >=x11-drivers/xf86-input-dynapro-1.1.0 )
   input_devices_elo2300? ( >=x11-drivers/xf86-input-elo2300-1.1.0 )
   input_devices_elographics? ( >=x11-drivers/xf86-input-elographics-1.1.0 )
   input_devices_evdev? ( >=x11-drivers/xf86-input-evdev-1.1.1 )
   input_devices_fpit? ( >=x11-drivers/xf86-input-fpit-1.1.0 )
   input_devices_hyperpen? ( >=x11-drivers/xf86-input-hyperpen-1.1.0 )
   input_devices_jamstudio? ( >=x11-drivers/xf86-input-jamstudio-1.1.0 )
   input_devices_joystick? ( >=x11-drivers/xf86-input-joystick-1.1.0 )
   input_devices_keyboard? ( >=x11-drivers/xf86-input-keyboard-1.1.0 )
   input_devices_magellan? ( >=x11-drivers/xf86-input-magellan-1.1.0 )
   input_devices_microtouch? ( >=x11-drivers/xf86-input-microtouch-1.1.0 )
   input_devices_mouse? ( >=x11-drivers/xf86-input-mouse-1.1.0 )
   input_devices_mutouch? ( >=x11-drivers/xf86-input-mutouch-1.1.0 )
   input_devices_palmax? ( >=x11-drivers/xf86-input-palmax-1.1.0 )
   input_devices_penmount? ( >=x11-drivers/xf86-input-penmount-1.1.0 )
   input_devices_spaceorb? ( >=x11-drivers/xf86-input-spaceorb-1.1.0 )
   input_devices_summa? ( >=x11-drivers/xf86-input-summa-1.1.0 )
   input_devices_tek4957? ( >=x11-drivers/xf86-input-tek4957-1.1.0 )
   input_devices_ur98? ( >=x11-drivers/xf86-input-ur98-1.1.0 )
   input_devices_vmmouse? ( >=x11-drivers/xf86-input-vmmouse-12.4.0 )
   input_devices_void? ( >=x11-drivers/xf86-input-void-1.1.0 )

   input_devices_synaptics? ( x11-drivers/synaptics )
   input_devices_wacom? ( x11-drivers/linuxwacom )

   video_cards_apm? ( >=x11-drivers/xf86-video-apm-1.1.1 )
   video_cards_ark? ( >=x11-drivers/xf86-video-ark-0.6.0 )
   video_cards_chips? ( >=x11-drivers/xf86-video-chips-1.1.1 )
   video_cards_cirrus? ( >=x11-drivers/xf86-video-cirrus-1.1.0 )
   video_cards_cyrix? ( >=x11-drivers/xf86-video-cyrix-1.1.0 )
   video_cards_dummy? ( >=x11-drivers/xf86-video-dummy-0.2.0 )
   video_cards_fbdev? ( >=x11-drivers/xf86-video-fbdev-0.2.0 )
   video_cards_glint? ( >=x11-drivers/xf86-video-glint-1.1.1 )
   video_cards_i128? ( >=x11-drivers/xf86-video-i128-1.2.0 )
   video_cards_i740? ( >=x11-drivers/xf86-video-i740-1.1.0 )
   video_cards_i810? ( >=x11-drivers/xf86-video-i810-1.6.0 )
   video_cards_impact? ( >=x11-drivers/xf86-video-impact-0.2.0 )
   video_cards_imstt? ( >=x11-drivers/xf86-video-imstt-1.1.0 )
   video_cards_mach64? ( >=x11-drivers/xf86-video-ati-6.6.0 )
   video_cards_mga? ( >=x11-drivers/xf86-video-mga-1.4.1 )
   video_cards_neomagic? ( >=x11-drivers/xf86-video-neomagic-1.1.1 )
   video_cards_newport? ( >=x11-drivers/xf86-video-newport-0.2.0 )
   video_cards_nsc? ( >=x11-drivers/xf86-video-nsc-2.8.1 )
   video_cards_nv? ( >=x11-drivers/xf86-video-nv-1.1.1 )
   video_cards_r128? ( >=x11-drivers/xf86-video-ati-6.6.0 )
   video_cards_radeon? ( >=x11-drivers/xf86-video-ati-6.6.0 )
   video_cards_rendition? ( >=x11-drivers/xf86-video-rendition-4.1.0 )
   video_cards_s3? ( >=x11-drivers/xf86-video-s3-0.4.1 )
   video_cards_s3virge? ( >=x11-drivers/xf86-video-s3virge-1.9.1 )
   video_cards_savage? ( >=x11-drivers/xf86-video-savage-2.1.1 )
   video_cards_siliconmotion? ( >=x11-drivers/xf86-video-siliconmotion-1.4.1 )
   video_cards_sis? ( >=x11-drivers/xf86-video-sis-0.9.1 )
   video_cards_sisusb? ( >=x11-drivers/xf86-video-sisusb-0.8.1 )
   video_cards_sunbw2? ( >=x11-drivers/xf86-video-sunbw2-1.1.0 )
   video_cards_suncg14? ( >=x11-drivers/xf86-video-suncg14-1.1.0 )
   video_cards_suncg3? ( >=x11-drivers/xf86-video-suncg3-1.1.0 )
   video_cards_suncg6? ( >=x11-drivers/xf86-video-suncg6-1.1.0 )
   video_cards_sunffb? ( >=x11-drivers/xf86-video-sunffb-1.1.0 )
   video_cards_sunleo? ( >=x11-drivers/xf86-video-sunleo-1.1.0 )
   video_cards_suntcx? ( >=x11-drivers/xf86-video-suntcx-1.1.0 )
   video_cards_tdfx? ( >=x11-drivers/xf86-video-tdfx-1.2.1 )
   video_cards_tga? ( >=x11-drivers/xf86-video-tga-1.1.0 )
   video_cards_trident? ( >=x11-drivers/xf86-video-trident-1.2.1 )
   video_cards_tseng? ( >=x11-drivers/xf86-video-tseng-1.1.0 )
   video_cards_v4l? ( >=x11-drivers/xf86-video-v4l-0.1.1 )
   video_cards_vesa? ( >=x11-drivers/xf86-video-vesa-1.1.0 )
   video_cards_vga? ( >=x11-drivers/xf86-video-vga-4.1.0 )
   video_cards_via? ( >=x11-drivers/xf86-video-via-0.2.1 )
   video_cards_vmware? ( >=x11-drivers/xf86-video-vmware-10.13.0 )
   video_cards_voodoo? ( >=x11-drivers/xf86-video-voodoo-1.1.0 )

   video_cards_tdfx? ( 3dfx? ( >=media-libs/glide-v3-3.10 ) )
   video_cards_fglrx? ( >=x11-drivers/ati-drivers-8.27.10 )
   video_cards_nvidia? ( || (
         >=x11-drivers/nvidia-drivers-1.0.8774
         >=x11-drivers/nvidia-legacy-drivers-1.0.7184
      )
   )
)"
LICENSE="${LICENSE} MIT"

pkg_setup() {
use minimal || ensure_a_server_is_building

# Patches required for compiz to work with AIGLX,
# but they slow EXA down (bug #147841).
if use aiglx; then
   einfo "AIGLX patches will be applied."
   ewarn "These patches are known to cause problems with EXA enabled."
   PATCHES="${FILESDIR}/aiglx-01-no-move-damage.patch
      ${FILESDIR}/aiglx-02-dont-backfill-bg-none.patch
      ${FILESDIR}/aiglx-05-offscreen-pixmaps.patch
      ${PATCHES}"
fi

# EXA fixes included: 6773, 6881 (?), 7671 (1st patch, 2nd=?), 7639
# EXA (7671 2nd?) and Beryl fixes ('white cube' - UNTESTED!!!)
if use beryl; then
PATCHES="${FILESDIR}/xorg-server-9999-beryl-fixes-mesa-git.patch
${PATCHES}"
# Chwilowo nie... sprawdzic czy naprawia crashe
#       ${FILESDIR}/xorg-server-9999-mesa-glapi.patch
fi
if use exa; then
PATCHES="${FILESDIR}/xorg-server-9999-exa-fixes.patch
${PATCHES}"
fi

# SDL only available in kdrive build
if use kdrive && use sdl; then
   conf_opts="${conf_opts} --enable-xsdl"
else
   conf_opts="${conf_opts} --disable-xsdl"
fi

# Only Xorg and Xgl support this, and we won't build Xgl
# until it merges to trunk
if use xorg; then
   conf_opts="${conf_opts} --with-mesa-source=${WORKDIR}/${MESA_P}"
fi

# localstatedir is used for the log location; we need to override the default
# from ebuild.sh
# sysconfdir is used for the xorg.conf location; same applies
# --enable-install-setuid needed because sparcs default off
CONFIGURE_OPTIONS="
   $(use_enable ipv6)
   $(use_enable dmx)
   $(use_enable kdrive)
   $(use_enable !minimal xvfb)
   $(use_enable !minimal xnest)
   $(use_enable !minimal xfree86-utils)
   $(use_enable !minimal install-libxf86config)
   $(use_enable dri)
   $(use_enable xorg)
   $(use_enable xprint)
   $(use_enable nptl glx-tls)
   --sysconfdir=/etc/X11
   --localstatedir=/var
   --enable-install-setuid
   --with-fontdir=/usr/share/fonts
   ${conf_opts}"

local diemsg="You must build xorg-server and mesa with the same nptl USE setting."
if built_with_use media-libs/mesa nptl; then
   use nptl || die "${diemsg}"
else
   use nptl && die "${diemsg}"
fi

# (#121394) Causes window corruption
filter-flags -fweb

# Nothing else provides new enough glxtokens.h
ewarn "Forcing on xorg-x11 for new enough glxtokens.h..."
OLD_IMPLEM="$(eselect opengl show)"
eselect opengl set --impl-headers ${OPENGL_DIR}
}

src_unpack() {
x-modular_specs_check
x-modular_dri_check

# Don't patch before everything's ready
#PATCHES="" mesa_cvs_src_unpack
mesa_git_src_unpack
xserver_git_src_unpack

x-modular_patch_source

# Set up kdrive servers to build
# Bug #150052 - anything that uses vm86.h is broken on non-x86 arches.
# That translates into the following set:
vm86_devices="chips epson glint i810 mach64 mga neomagic
   nv r128 radeon siliconmotion vesa via"

if use kdrive; then
   einfo "Removing unused kdrive drivers ..."
   for card in ${IUSE_VIDEO_CARDS}; do
      # Skip binary drivers
      if [[ ${card} = video_cards_nvidia ]] \
         || [[ ${card} = video_cards_fglrx ]]; then
         continue
      fi

      real_card=${card#video_cards_}

      # Differences between VIDEO_CARDS name and kdrive server name
      real_card=${real_card/glint/pm2}
      real_card=${real_card/radeon/ati}
      real_card=${real_card/%nv/nvidia}
      real_card=${real_card/siliconmotion/smi}
      real_card=${real_card/%sis/sis300}

      disable_card=0
      if ! use ${card}; then
         # (bug #136370) Radeon needs fbdev and vesa
         if ! use x86 \
            && use video_cards_radeon; then
            if [[ ${card} = fbdev ]] \
               || [[ ${card} = vesa ]]; then
               continue
            fi
         fi
         disable_card=1
      # Bug #150052
      elif ! use x86 &&
         [[ ${vm86_devices/${card#video_cards_}/} != ${vm86_devices} ]]; then
         ewarn " $real_card does not work on your architecture; disabling."
         disable_card=1
      fi

      if [[ $disable_card = 1 ]]; then
         ebegin " ${real_card}"
         sed -i \
            -e "s:\b${real_card}\b::g" \
            ${S}/hw/kdrive/Makefile.am \
            || die "sed of ${real_card} failed"
         eend
      fi

   done

   # smi and via are the only things on line 2. If line 2 ends up blank,
   # we need to get rid of the backslash at the end of line 1.
   if ! use video_cards_siliconmotion && ! use video_cards_via; then
      sed -i \
         -e "s:^$VESA_SUBDIRS.*$\\\:\1:g" \
         ${S}/hw/kdrive/Makefile.am
   fi
fi

# Make sure eautoreconf gets run if we need the autoconf/make
# changes.
if [[ ${SNAPSHOT} != "yes" ]]; then
   if use kdrive || use xprint; then
      eautoreconf
   fi
fi
x-modular_reconf_source
}

src_install() {
x-modular_src_install

dynamic_libgl_install

server_based_install

# Install video mode files for system-config-display
insinto /usr/share/xorg
doins hw/xfree86/common/{extra,vesa}modes \
   || die "couldn't install extra modes"

# Bug #151421 - this file is not built with USE="minimal"
if ! use minimal; then
   # Install xorg.conf.example
   insinto /etc/X11
   doins hw/xfree86/xorg.conf.example \
      || die "couldn't install xorg.conf.example"
fi
}

pkg_postinst() {
switch_opengl_implem

# Bug #135544
ewarn "Users of reduced blanking now need:"
ewarn " Option \"ReducedBlanking\""
ewarn "In the relevant Monitor section(s)."
ewarn "Make sure your reduced blanking modelines are safe!"
}

pkg_postrm() {
# Get rid of module dir to ensure opengl-update works properly
if ! has_version x11-base/xorg-server; then
   if [ -e ${ROOT}/usr/$(get_libdir)/xorg/modules ]; then
      rm -rf ${ROOT}/usr/$(get_libdir)/xorg/modules
   fi
fi
}

dynamic_libgl_install() {
# next section is to setup the dynamic libGL stuff
ebegin "Moving GL files for dynamic switching"
   dodir /usr/$(get_libdir)/opengl/${OPENGL_DIR}/extensions
   local x=""
   for x in ${D}/usr/$(get_libdir)/xorg/modules/extensions/libglx*; do
      if [ -f ${x} -o -L ${x} ]; then
         mv -f ${x} ${D}/usr/$(get_libdir)/opengl/${OPENGL_DIR}/extensions
      fi
   done
eend 0
}

server_based_install() {
use xprint && xprint_src_install

if ! use xorg; then
   rm ${D}/usr/share/man/man1/Xserver.1x \
      ${D}/usr/$(get_libdir)/xserver/SecurityPolicy \
      ${D}/usr/$(get_libdir)/pkgconfig/xorg-server.pc \
      ${D}/usr/share/man/man1/Xserver.1x
fi
}

switch_opengl_implem() {
   # Switch to the xorg implementation.
   # Use new opengl-update that will not reset user selected
   # OpenGL interface ...
   echo
#    eselect opengl set --use-old ${OPENGL_DIR}
   eselect opengl set ${OLD_IMPLEM}
}

xprint_src_install() {
# RH-style init script, we provide a wrapper
exeinto /usr/$(get_libdir)/misc
doexe ${S}/Xprint/etc/init.d/xprint
# Patch init script for fonts location
sed -e 's:/lib/X11/fonts/:/share/fonts/:g' \
   -i ${D}/usr/$(get_libdir)/misc/xprint
# Install the wrapper
newinitd ${FILESDIR}/xprint.init xprint
# Install profile scripts
insinto /etc/profile.d
doins ${S}/Xprint/etc/profile.d/xprint*
insinto /etc/X11/xinit/xinitrc.d
newins ${S}/Xprint/etc/Xsession.d/cde_xsessiond_xprint.sh \
   92xprint-xpserverlist.sh
# Patch profile scripts
sed -e "s:/bin/sh.*get_xpserverlist:/usr/$(get_libdir)/misc/xprint \
   get_xpserverlist:g" -i ${D}/etc/profile.d/xprint* \
   ${D}/etc/X11/xinit/xinitrc.d/92xprint-xpserverlist.sh
# Move profile scripts, we can't touch /etc/profile.d/ in Gentoo
dodoc ${D}/etc/profile.d/xprint*
rm -f ${D}/etc/profile.d/xprint*
}

ensure_a_server_is_building() {
for server in ${IUSE_SERVERS}; do
   use ${server} && return;
done
eerror "You need to specify at least one server to build."
eerror "Valid servers are: ${IUSE_SERVERS}."
die "No servers were specified to build."
}

xserver_git_src_unpack() {
local EGIT_REPO_URI="git://anongit.freedesktop.org/git/xorg/xserver"
local EGIT_PROJECT="xorg-server"
git_src_unpack
}

mesa_git_src_unpack() {
mkdir ${WORKDIR}/${MESA_P}
local EGIT_REPO_URI="git://anongit.freedesktop.org/git/mesa/mesa"
local EGIT_PROJECT="mesa"
local S=${WORKDIR}/${MESA_P}
git_src_unpack
}

files/xorg-server-9999-exa-fixes.patch

Code:

diff -Naur exa.old/exa_migration.c exa/exa_migration.c
--- exa.old/exa_migration.c 2006-12-28 01:11:31.000000000 +0100
+++ exa/exa_migration.c 2006-12-28 22:53:13.000000000 +0100
@@ -518,6 +518,7 @@
ScreenPtr pScreen = pixmaps[0].pPix->drawable.pScreen;
ExaScreenPriv(pScreen);
int i, j;
+ Bool op_can_accel = can_accel;

/* If this debugging flag is set, check each pixmap for whether it is marked
* as clean, and if so, actually check if that's the case. This should help
@@ -593,6 +594,31 @@
   * If not, then we migrate toward FB so that hopefully acceleration can
   * happen.
   */
+
+ if (op_can_accel && !can_accel &&
+    pExaScr->info->UploadToScratch != NULL ) {
+
+    /*
+    * Rely on uploadToScratch to make one system pinned
+    * pixmap available for the accelerator.
+    */
+
+    int pinnedDst = 0;
+    int pinnedSrc = 0;
+    for (i=0; i<npixmaps; ++i) {
+    if (pixmaps[i].as_dst && exaPixmapIsPinned(pixmaps[i].pPix)) {
+       pinnedDst = 1;
+       break;
+    }
+    if (!pixmaps[i].as_dst && exaPixmapIsPinned(pixmaps[i].pPix)) {
+       pinnedSrc++;
+    }
+    }
+    if (!pinnedDst && (pinnedSrc < 2)) {
+    can_accel=TRUE;
+    }
+ }
+
   if (!can_accel) {
      for (i = 0; i < npixmaps; i++)
      exaMigrateTowardSys (pixmaps[i].pPix);
diff -Naur exa.old/exa_offscreen.c exa/exa_offscreen.c
--- exa.old/exa_offscreen.c 2006-12-28 01:11:31.000000000 +0100
+++ exa/exa_offscreen.c 2006-12-28 22:36:43.000000000 +0100
@@ -251,42 +251,44 @@

ExaOffscreenValidate (pScreen);

- DBG_OFFSCREEN (("Alloc 0x%x -> 0x%x (0x%x)\n", size,
-       area->base_offset, area->offset));
+ DBG_OFFSCREEN (("Alloc 0x%x -> 0x%x (0x%x) locked %d\n", size,
+       area->base_offset, area->offset, area->state == ExaOffscreenLocked));
return area;
}

void
ExaOffscreenSwapOut (ScreenPtr pScreen)
{
+ ExaOffscreenArea *nextArea;
ExaScreenPriv (pScreen);
+ ExaOffscreenArea *area = pExaScr->info->offScreenAreas;

ExaOffscreenValidate (pScreen);
- /* loop until a single free area spans the space */
+
for (;;)
{
- ExaOffscreenArea *area = pExaScr->info->offScreenAreas;
-
   if (!area)
      break;
- if (area->state == ExaOffscreenAvail)
+ if (area->state == ExaOffscreenAvail ||
+    area->state == ExaOffscreenLocked)
   {
      area = area->next;
-    if (!area)
-    break;
+    continue;
   }
- assert (area->state != ExaOffscreenAvail);
+
+ nextArea = area->next;
   (void) ExaOffscreenKickOut (pScreen, area);
   ExaOffscreenValidate (pScreen);
+ area = nextArea;
}
ExaOffscreenValidate (pScreen);
- ExaOffscreenFini (pScreen);
+ ;
}

void
ExaOffscreenSwapIn (ScreenPtr pScreen)
{
- exaOffscreenInit (pScreen);
+ ;
}

void

files/xorg-server-9999-beryl-fixes-mesa-git.patch

Code:

diff -Naur glx.old/glxdri.c glx/glxdri.c
--- GL/glx/glxdri.c 2006-12-28 01:11:31.000000000 +0100
+++ GL/glx/glxdri.c 2006-12-28 23:01:05.000000000 +0100
@@ -371,13 +371,7 @@
   type = GL_UNSIGNED_SHORT_5_6_5;
}

- if (!(glxCountBits(pixmap->drawable.width) == 1 &&
-    glxCountBits(pixmap->drawable.height) == 1)
- /* || strstr(CALL_GetString(GL_EXTENSIONS,
-    "GL_ARB_texture_non_power_of_two")) */)
- target = GL_TEXTURE_RECTANGLE_ARB;
- else
- target = GL_TEXTURE_2D;
+ target = GL_TEXTURE_RECTANGLE_ARB;

CALL_PixelStorei( GET_DISPATCH(), (GL_UNPACK_ROW_LENGTH,
               pixmap->devKind / bpp) );
diff -Naur glx.old/indirect_table.c glx/indirect_table.c
--- GL/glx/indirect_table.c 2006-12-28 01:11:31.000000000 +0100
+++ GL/glx/indirect_table.c 2006-12-28 23:00:16.000000000 +0100
@@ -1232,7 +1232,7 @@

/*****************************************************************/
/* tree depth = 13 */
-static const int_fast16_t VendorPriv_dispatch_tree[155] = {
+static const int_fast16_t VendorPriv_dispatch_tree[159] = {
/* [0] -> opcode range [0, 131072], node depth 1 */
2,
5,
@@ -1480,11 +1480,16 @@
/* [152] -> opcode range [65536, 65552], node depth 13 */
1,
LEAF(88),
+ 156,
EMPTY_LEAF,

+ /* [156] -> opcode range [65544, 65552], node depth 13 */
+ 1,
+ LEAF(96),
+ EMPTY_LEAF,
};

-static const void *VendorPriv_function_table[96][2] = {
+static const void *VendorPriv_function_table[99][2] = {
/* [ 0] = 0 */ {NULL, NULL},
/* [ 1] = 1 */ {__glXDisp_GetConvolutionFilterEXT, __glXDispSwap_GetConvolutionFilterEXT},
/* [ 2] = 2 */ {__glXDisp_GetConvolutionParameterfvEXT, __glXDispSwap_GetConvolutionParameterfvEXT},
@@ -1581,6 +1586,9 @@
/* [ 93] = 65541 */ {__glXDisp_CreateContextWithConfigSGIX, __glXDispSwap_CreateContextWithConfigSGIX},
/* [ 94] = 65542 */ {__glXDisp_CreateGLXPixmapWithConfigSGIX, __glXDispSwap_CreateGLXPixmapWithConfigSGIX},
/* [ 95] = 65543 */ {NULL, NULL},
+ /* [ 96] = 65544 */ {NULL, NULL},
+ /* [ 97] = 65545 */ {NULL, NULL},
+ /* [ 98] = 65546 */ {__glXDisp_GetDrawableAttributesSGIX, __glXDispSwap_GetDrawableAttributesSGIX},
};

const struct __glXDispatchInfo VendorPriv_dispatch_info = {

files/xorg-server-9999-mesa-glapi.patch

Code:

--- ../Mesa/src/mesa/glapi/glX_API.xml 2006-12-27 13:53:26.000000000 +0100
+++ ../Mesa/src/mesa/glapi/glX_API.xml 2006-12-28 06:46:52.000000000 +0100
@@ -183,7 +183,6 @@
</function>
</category>

-

<category name="GLX_MESA_copy_sub_buffer" number="215">
<function name="CopySubBufferMESA">

files/aiglx-01-no-move-damage.patch

Code:

Disable damage notifications on move for manually redirected windows.
The automatic compositor needs damage notification on move, but a
compositing manager doesn't.

--- ./composite/compwindow.c.spiffiffity 2006-03-29 12:51:53.000000000 -0500
+++ ./composite/compwindow.c 2006-04-13 23:18:42.000000000 -0400
@@ -571,12 +571,15 @@
}
else
{
+ CompWindowPtr cw = GetCompWindow (pWin);
+
   ptOldOrg.x -= dx;
   ptOldOrg.y -= dy;
   REGION_TRANSLATE (prgnSrc, prgnSrc,
            pWin->drawable.x - ptOldOrg.x,
            pWin->drawable.y - ptOldOrg.y);
- DamageDamageRegion (&pWin->drawable, prgnSrc);
+ if (pWin->redirectDraw && cw->update == CompositeRedirectAutomatic)
+    DamageDamageRegion (&pWin->drawable, prgnSrc);
}
cs->CopyWindow = pScreen->CopyWindow;
pScreen->CopyWindow = compCopyWindow;
@@ -655,7 +658,8 @@
/*
* Report that as damaged so it will be redrawn
*/
- DamageDamageRegion (&pWin->drawable, &damage);
+ if (cw->update == CompositeRedirectAutomatic)
+ DamageDamageRegion (&pWin->drawable, &damage);
REGION_UNINIT (pScreen, &damage);
/*
* Save the new border clip region

files/aiglx-02-dont-backfill-bg-none.patch

Code:

Disable backfilling of windows created with bg=none, which otherwise
would force a framebuffer readback.

--- ./composite/compalloc.c.spiffiffity 2006-03-13 16:59:55.000000000 -0500
+++ ./composite/compalloc.c 2006-04-12 16:37:50.000000000 -0400
@@ -478,6 +478,7 @@
* Copy bits from the parent into the new pixmap so that it will
* have "reasonable" contents in case for background None areas.
*/
+#if 0
if (pGC)
{
   XID val = IncludeInferiors;
@@ -492,6 +493,7 @@
            w, h, 0, 0);
   FreeScratchGC (pGC);
}
+#endif
return pPixmap;
}

files/aiglx-05-offscreen-pixmaps.patch

Code:

--- ./hw/xfree86/xaa/xaaInit.c.offscreen-pixmaps 2006-07-05 14:31:42.000000000 -0400
+++ ./hw/xfree86/xaa/xaaInit.c 2006-08-30 16:47:29.000000000 -0400
@@ -98,6 +98,30 @@
xfree(infoRec);
}

+void
+XAAEvictPixmaps(void)
+{
+ XAAScreenPtr pScreenPriv;
+ XAAInfoRecPtr infoRec;
+ ScreenPtr pScreen;
+ int i;
+
+ xf86MsgVerb(X_INFO, 3, "XAA: Evicting pixmaps\n");
+
+ for (i = 0; i < screenInfo.numScreens; i++) {
+ pScreen = screenInfo.screens[i];
+ infoRec = GET_XAAINFORECPTR_FROM_SCREEN(pScreen);
+
+ pScreenPriv = pScreen->devPrivates[XAAScreenIndex].ptr;
+ infoRec = pScreenPriv->AccelInfoRec;
+
+ infoRec->offscreenDepths = 0;
+ infoRec->Flags &= ~OFFSCREEN_PIXMAPS;
+
+ XAAMoveOutOffscreenPixmaps(pScreen);
+ XAAInvalidatePixmapCache(pScreen);
+ }
+}

Bool
XAAInit(ScreenPtr pScreen, XAAInfoRecPtr infoRec)
--- ./GL/glx/glxdri.c.offscreen-pixmaps 2006-08-30 16:46:45.000000000 -0400
+++ ./GL/glx/glxdri.c 2006-08-30 16:45:52.000000000 -0400
@@ -367,6 +367,19 @@
int bpp;
GLenum target, format, type;

+ /* When the GLX_EXT_texture_from_pixmap is used, as it's
+ * implemented here, we want to pull pixmap out of video memory
+ * and into host memory. */
+ {
+ extern void XAAEvictPixmaps(void);
+ static int evictedPixmaps;
+
+ if (!evictedPixmaps) {
+ XAAEvictPixmaps();
+ evictedPixmaps = TRUE;
+ }
+ }
+
pixmap = (PixmapPtr) glxPixmap->pDraw;
if (!glxPixmap->pDamage) {
glxPixmap->pDamage = DamageCreate(NULL, NULL, DamageReportNone,

Last, but not least: my ebuild was made just for fun, I'm not going to compete with anyone; however, if someone can provide some fixes/testing for them, I would be grateful.

Post last updated on: 20061228 23:30:29 CET 2006
Builds with xorg-server up to commit: Thu Dec 28 13:21:54 PDT 2006
_________________
RLU #137109 | BYKOM STOP! | Chcesz do mnie napisać? | jog
"Mam plany gotowe i wszystko, no i wiesz, idę do tego, no, mebloroba..." (Raaf)

Last edited by argasek on Thu Dec 28, 2006 10:33 pm; edited 1 time in total

LAj
Apprentice

Joined: 22 May 2004
Posts: 294
Location: Avellino[Italy]

Posted: Thu Dec 28, 2006 7:43 pm Post subject:

Phlogiston wrote:

Primozic wrote:

Here you go:
updated tarball

Thank you guys,

but where are the x11-drivers/xf86-input-* ebuilds from CVS ?
why we have that keyword?

emerging xorg-server
they are not calculated as deps
_________________
They produce notes of any size as their own properties askng interests also.This is seignorage and goes to privates banks(FED, BCE).You don't have to pay to use $$,you can simply use your notes recovering monetary sovereignty.Public debt is not right!

argasek
Bodhisattva

Joined: 06 May 2004
Posts: 1120
Location: Sol [0,0], Poland, Kraków

Posted: Thu Dec 28, 2006 8:58 pm Post subject:

Umm, I just found my files/xorg-server-9999-beryl-fixes-mesa-git.patch being borked, I'll post correct version soon. Edit: fixed ebuild and patches provided within previous post (fixed means: patches apply correctly , not the bugs are fixed). _________________ RLU #137109 \| BYKOM STOP! \| Chcesz do mnie napisać? \| jog "Mam plany gotowe i wszystko, no i wiesz, idę do tego, no, mebloroba..." (Raaf)

Display posts from previous:

	Gentoo Forums Forum Index Unsupported Software	All times are GMT Goto page Previous 1, 2, 3 ... , 38, 39, 40 Next
Page 39 of 40

Jump to:

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

Copyright 2001-2013 Gentoo Foundation, Inc. Designed by Kyle Manna © 2003; Style derived from original subSilver theme. | Hosting by Gossamer Threads Inc. © | Powered by phpBB 2.0.23-gentoo-p9 © 2001, 2002 phpBB Group