p.s: I've got an i810 - so perhaps this problem was affecting more than just i915 people?
]]>//cut
Well, the kernels I compiled fix the problems with i915 cards. I am not sure if it will work with any other card, so you are probably better off with offical packages.
Yes, you upgraded the kernel package correcly. And yes, tiling error does goes away for me when I apply the custom kernel (reboot is required!). Try compiling 2.6.30-rc7, it's told that it fixes many problems.
P.S. For i915 owners - I'll put patched kernels on this website as they come along, latest version is 2.6.29.4-1: http://archlinux.pondi.hr
]]>The fact is I followed karabaja4's links to download the patched kernels. I downloaded the latest two (the .3-1 and the .4-1) and tried installing them. I am not an expert so I suppose I should do:
pacman -U <filename>
I actually did it and the manager seems to have installed something. I have actually no way of understanding if it actually has now patched my kernel (any tips would be great). I am having all these doubts because the xorg log file continues to display stuff about the tiling being broken... Maybe I will now try to find the even older kernels. However, can someone please tell me in they are also still experiencing this problem. And why does my system keep giving my the tiling error. All those people that had their problems solved by this patch: did the tiling error go away? And those who still have the problem: are they still recieving the tiling errro.
EDIT: I actually tried the 2.6.29.1-2 kernel and even that one gives me the tiling error!!!!! Is it possible. People are saying their problems began after the upgrade from version 2.6.29.1-2!!! What is happening here?
Here is some output:
glxgears gives similar output with all 4 kernel versions i tried:
1074 frames in 5.0 seconds = 214.759 FPS
1126 frames in 5.0 seconds = 225.097 FPS
1049 frames in 5.0 seconds = 209.717 FPS
lscpi gives:
00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03)
and my X log file (Xorg.0.log) contains the infamous lines:
(II) intel(0): [drm] installed DRM signal handler
(**) intel(0): Framebuffer compression enabled
(**) intel(0): Tiling enabled
(==) intel(0): VideoRam: 262144 KB
(II) intel(0): Attempting memory allocation with tiled buffers.
(EE) intel(0): Failed to set tiling on front buffer: rejected by kernel
(EE) intel(0): Failed to set tiling on back buffer: rejected by kernel
(EE) intel(0): Failed to set tiling on depth buffer: rejected by kernel
(II) intel(0): Tiled allocation successful.
(II) intel(0): [drm] Registers = 0xdc100000
(II) intel(0): [dri] visual configs initialized
(II) intel(0): vgaHWGetIOBase: hwp->IOBase is 0x03d0, hwp->PIOOffset is 0x0000
(II) EXA(0): Offscreen pixmap area of 31457280 bytes
(II) EXA(0): Driver registered support for the following operations:
(II) Solid
(II) Copy
(II) Composite (RENDER acceleration)
I am having difficulty understanding the output of pacman. Did it actually patch the kernel?:
[2009-06-02 17:17] >>> Updating module dependencies. Please wait ...
[2009-06-02 17:17] >>> MKINITCPIO SETUP
[2009-06-02 17:17] >>> ----------------
[2009-06-02 17:17] >>> If you use LVM2, Encrypted root or software RAID,
[2009-06-02 17:17] >>> Ensure you enable support in /etc/mkinitcpio.conf .
[2009-06-02 17:17] >>> More information about mkinitcpio setup can be found here:
[2009-06-02 17:17] >>> http://wiki.archlinux.org/index.php/Mkinitcpio
[2009-06-02 17:17]
[2009-06-02 17:17] >>> Generating initial ramdisk, using mkinitcpio. Please wait...
[2009-06-02 17:17] ==> Building image "default"
[2009-06-02 17:17] ==> Running command: /sbin/mkinitcpio -k 2.6.29-ARCH -c /etc/mkinitcpio.conf -g /boot/kernel26.img
[2009-06-02 17:17] :: Begin build
[2009-06-02 17:17] :: Parsing hook [base]
[2009-06-02 17:18] :: Parsing hook [udev]
[2009-06-02 17:18] :: Parsing hook [autodetect]
[2009-06-02 17:18] :: Parsing hook [pata]
[2009-06-02 17:18] :: Parsing hook [scsi]
[2009-06-02 17:18] :: Parsing hook [sata]
[2009-06-02 17:18] :: Parsing hook [filesystems]
[2009-06-02 17:18] :: Generating module dependencies
[2009-06-02 17:18] :: Generating image '/boot/kernel26.img'...SUCCESS
[2009-06-02 17:18] ==> SUCCESS
[2009-06-02 17:18] ==> Building image "fallback"
[2009-06-02 17:18] ==> Running command: /sbin/mkinitcpio -k 2.6.29-ARCH -c /etc/mkinitcpio.conf -g /boot/kernel26-fallback.img -S autodetect
[2009-06-02 17:18] :: Begin build
[2009-06-02 17:18] :: Parsing hook [base]
[2009-06-02 17:18] :: Parsing hook [udev]
[2009-06-02 17:18] :: Parsing hook [pata]
[2009-06-02 17:18] :: Parsing hook [scsi]
[2009-06-02 17:19] :: Parsing hook [sata]
[2009-06-02 17:19] :: Parsing hook [filesystems]
[2009-06-02 17:19] :: Generating module dependencies
[2009-06-02 17:19] :: Generating image '/boot/kernel26-fallback.img'...SUCCESS
[2009-06-02 17:19] ==> SUCCESS
[2009-06-02 17:19] upgraded kernel26 (2.6.29.4-1 -> 2.6.29.4-1)
Here's the compiled version of newest kernel 2.6.29.3-1 for i915 cards - download
]]>UXA/DRI2:
* Virtual screen size limited to 2048x2048 for pre-965 platforms: bug#21190.
* front-buffer rendering broken: bug#19174.
* XvMC disabled: bug#20790.
* VT switch broken on some machines: bug#19578.
- KMS:
* This kernel patch is required to fix tiling issue on 965+ with KMS enabled: bug#21246.
* only native resolution supported: bug#20801.
* video overlay not supported: bug#20901.
* no user-level control for backlight: bug#20963.
- glReadPixels fail with tiling on some platforms with specific memory configurations: bug#18420,bug#18586,bug#19317.
- GEM kernel not working with PAE (CONFIG_HIGHMEM64G), so now GEM is disabled with PAE in 2.6.28+ kernel as a temporary workaround: bug#17993.
- 3D broken with tiling enabled on 8xx platform: bug#18878,bug#20473.
- G41 DRM support: bug#21095, with patch available.
- No support for DRI if screen wider/higher than 2048 px on pre-i965: bug#10479, to be fixed by shatter.
They already fixed the tiling issue, the rest needs to be patched... As u can see only native resolution is supported using kms...
]]>Does anyone have this problem with the newest intel drivers? >>> http://bugs.archlinux.org/task/14594
Endless story these intel drivers.
]]>It is the same crappy performance and bugs that were found in 2.6.29 with 2.6.3 xf86-video-intel except for the tiling issue. I've downloaded 2.6.30-rc4 from www.kernel.org and compiled it myself, installed xf86-vide-intel-newest from aur and no improvements what so ever... I'm begining to think that the situation with intel drivers will never get better. Lower resolutions are not switching in full screen (u get native resolution with framed window the size of the resolution you selected in game), redrawing message popups is still slow, glxgears 480fps, urban terror 20-22fps...
I have intel X3100 FYI...
I'm afraid I'll have to agree--I've been mixing and matching and tweaking since January, and the best performance I've gotten so far would've been with xf86-video-intel-2.7 + kernel-2.6.30rc3, which yielded about 65% of Q2 2008 package performance.
]]>I have intel X3100 FYI...
]]>I think this is i915 cards issue, sorry. You might get better kernel performance with kernel 2.6.30-rc4 which solves many other bugs: http://aur.archlinux.org/packages.php?ID=25304
Thanks I will try to compile tonight.
]]>Looking at the FPS's you guys get I guess 375 FPS in the default glxgears window for a "00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 0c)" is really bad.
I think this is i915 cards issue, sorry. You might get better kernel performance with kernel 2.6.30-rc4 which solves many other bugs: http://aur.archlinux.org/packages.php?ID=25304
]]>I'm using the kernel supplied in this thread, do I need to build mesa-git, intel-dri-git etc. from AUR to get better performance?
Actually etracer seems to be running alright.
I know glxgears is no benchmark, but the difference between 375 FPS and above 1000 FPS is quite big.
Could it be because I'm using "AutoAddDevices" "False"?
]]>BTW, I have a question for you (since you are the one of the few people with i915). Does your glxgears show more FPS when you kill xorg and start it again (more than a first time)? For example, I start my computer, my xorg and glxgears shows 1000fps. I kill my xorg with ctrl-alt-backspace, and start it again, now glxgears shows 1200fps.
The difference here is barely noticable.
When running glxgears for the first time I got around 1150-1160, after restarting Xorg it showed around 1190 fps
]]>