Created attachment 17290 [details] xorg.conf Chipset: X3100 (should be 965GM) Platform: Acer Aspire 5315 Arch: i686 xf86-video-intel: >=2.2 xserver: 1.3.0 / 1.4.2 mesa: 6.5.2 / 7.0.3 kernel: 2.6.24 Distro: Gentoo All intel drivers with version 2.2 or greater (tested up to 2.3.2) do not display (normal) text. The only thing which is shown, are the underscores in menus for instance. I tried xorg-1.3 and xorg-1.4 - but none seems to work. It does not happen with driver version 2.1 ... but here other issues exist (hanging X when using SDL/OpenGL).
Created attachment 17291 [details] xorg.log - with xorg-1.3 The xorg.log for the xorg-1.3
Created attachment 17292 [details] xorg.log - with xorg-1.4 xorg.log for xorg-server-1.4.2
Created attachment 17293 [details] lspci -v output Just to be sure: the lspci -v output
I suspect this might be caused by XAA->EXA transition in 2.2.0. Could you try XAA (by setting AccelMethod XAA in xorg.conf) with 2.2 or 2.3.2?
Ok... changing this, it works in the xorg-1.3 / i810-2.3.2 combination :) Thanks.
sounds "fixed". how about combination of xserver-1.4.2/xf86-video-intel-2.3.2?
This works too... (now even DRI is working :D) But - I would call it fixed, if it worked with EXA ;)
(In reply to comment #7) > This works too... (now even DRI is working :D) > > But - I would call it fixed, if it worked with EXA ;) > oh, I see the point. :) reassign to zhenyu to see if he has some quick fix or not...
Created attachment 17382 [details] [review] try allocating exa 965 state in non-stolen mem Please try this patch against 2.3.2. And test with current git master is also helpful to see if this problem is fixed upstream.
(In reply to comment #9) > Please try this patch against 2.3.2. René, if you want to try this yourself, you can create an overlay with a copy of the 2.3.2 ebuild and add a PATCHES variable somewhere. It'll automatically patch the sources. Contact me privately if you need any help with the Gentoo-side of testing :)
re-assign to Carl and ping bug reporter as well...
Sorry for the delay - but I had no (direct) access to the device. Anyways - I've tried the attached patch - this does not make a difference. But I've also tried the current git master: And this seems to work with both XAA and EXA :) So I would declare this as solved/fixed, though I don't know which patch fixed it exactly :) (and I'm too lazy to do a git bisect)
(In reply to comment #12) > Anyways - I've tried the attached patch - this does not make a difference. > But I've also tried the current git master: And this seems to work with both > XAA and EXA :) Cool! > So I would declare this as solved/fixed, though I don't know which patch fixed > it exactly :) (and I'm too lazy to do a git bisect) Judging by the number of commits between 2.3.2 and master, you're looking at maybe 10 rebuilds maximum. But of course, that's up to you ;) Closing?
(In reply to comment #13) > > So I would declare this as solved/fixed, though I don't know which patch fixed > > it exactly :) (and I'm too lazy to do a git bisect) > Judging by the number of commits between 2.3.2 and master, you're looking at > maybe 10 rebuilds maximum. But of course, that's up to you ;) Ok - I tried it, but git got me ;). Bisect did not show a reasonable result =| > Closing? Yes, except someone objects :)
ok. sounds interesting... I'll mark the bug as fixed. Also, I didn't see others report out similar issue, so it might be related with some config issue on your machine... anyway, who knows. It's glad to know it works for you. We have a stable driver release soon. you can get it at http://www.intellinuxgraphics.org/download.html could you also have a try with the package there to see if you will run into the same issue? please reopen then if it does... thanks.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.