ATI Catalyst 8.7 for Linux has been released. ATI claims it's driver version 8.512. Catalyst itself after installing claims it's 8.513. I used 8.512 since that's the official claim. Release notes: https://a248.e.akamai.net/f/674/9206/0/www2.ati.com/drivers/linux/catalyst_87_linux.html New features/bugfixes overview: - The display no longer fails to display an image after the installation of the Linux graphics driver. - Corruption is no longer noticed when switching from the laptop's display to the external CRT display device. - SECAM, NTSC or PAL signal changes when using aticonfig or Catalyst Control Center and now gets applied to the TV display output. - An X segmentation fault no longer occurs when launching SUZI/textured video. - Creating two successive connections no longer results in direct GL context to fail. As always with ATI, there are more bugs than features and most of the time, playing video, switching to a console or using 3D shows off ATI's new "Blue Screen of Death for Linux (TM)" technology. Enjoy the new version of the most brain-damaged and broken binary blob ever. I attached an updated ebuild (based on ati-drivers-8.501.ebuild). Only the download URL needed changing (along with wiping of $Header and resetting arches, as per ebuild writing guidelines). Therefore, the diff is very short: - # $Header: /var/cvsroot/gentoo-x86/x11-drivers/ati-drivers/ati-drivers-8.501.ebuild,v 1.3 2008/07/09 22:33:19 opfer Exp $ + # $Header: $ - SRC_URI="${ATI_URL}/ati-driver-installer-8-6-x86.x86_64.run" + SRC_URI="${ATI_URL}/ati-driver-installer-8-7-x86.x86_64.run" - KEYWORDS="~amd64 x86" + KEYWORDS="~amd64 ~x86"
Created attachment 161060 [details] ati-drivers-8.512.ebuild
this driver is known to be buggy
(In reply to comment #2) > this driver is known to be buggy As far as I can tell, all fglrx drivers are known to be buggy. See sarcasm in #0 :P Anyway, this driver introduces official support for the X2000 and X3000 Mobility Radeon series as well for the Radeon HD 4800 series. I failed to mention that in #0. So for this hardware, this driver is actually recommended.
Created attachment 161333 [details] ati-drivers-8.512-r1.ebuild Changed ebuild to include patch from http://phorogit.com/index.php?p=fglrx-packaging.git&dl=plain&h=a94205d6bb86dbf154c3e519b6da869c529cafe4 to make fglrx work with 2.6.26 kernels.
Created attachment 161335 [details, diff] fglrx-8.7-2.6.26.patch Patch for fglrx 8.7 to work with kernel 2.6.26 from http://phorogit.com/index.php?p=fglrx-packaging.git&dl=plain&h=a94205d6bb86dbf154c3e519b6da869c529cafe4
Created attachment 161338 [details, diff] Diff between ati-drivers 8.501 and 8.512
I am using the patched fglrx drivers with 2.6.26 and so far no problems. The module loads as expected and I can use xv, suspend and gl which are the most important features of the fglrx driver for me ;)
(In reply to comment #4) > Created an attachment (id=161333) [edit] > ati-drivers-8.512-r1.ebuild > > Changed ebuild to include patch from > http://phorogit.com/index.php?p=fglrx-packaging.git&dl=plain&h=a94205d6bb86dbf154c3e519b6da869c529cafe4 > to make fglrx work with 2.6.26 kernels. > Fine, but why does the ebuild patch the driver with the 2.6.25 kernel, too? Works for me without patching the kernel 2.6.25! I don't had a look at the patch and cannot say if it is recommended to patch the 2.6.25. Besides this the ati-drivers 8.512 (don't worry about 8.513 in the catalyst, this difference have the older drivers, too) works for me with the 2.6.26 (tested some games with wine)
There are some symbols removed in 2.6.26 which might allready not be exported in 2.6.25. So some users might stumble upon this. We could make it optional in 2.6.25 if the the symbols get exported but I am not sure what the corresponding setting is in the kernel options
thanks for the ebuilds!
*** Bug 234354 has been marked as a duplicate of this bug. ***
It is not working on MSI MEGA BOOK with amd64 and ATI Radeon XPRESS 1150 with second monitor 1600x1200 (LCD ViewSonic VP201b). Picture on second monitor is destroyed. Last working driver is ati-drivers-8.476.
It seems to me this is obsolete because 8.522 is in portage...
Actually, it's not just obsolete, it's fixed. 8.512 went in at the same time as 8.522.
I'm closing this then.
