Looks like there is a new driver for the ATI cards, see URL. Their versioning scheme looks really strange, it seems that now 8.1 is more recent than 8.443 was. Maybe they want to stay in sync with the Windows version numbers. The file for this package is called ati-driver-installer-8-01-x86.x86_64.run and a permanent link to that package should be the following URL: http://ati.amd.com/support/drivers/linux/previous/linux-rf-cat81.html
Created attachment 141240 [details, diff] Patch from 8.443.1-r1 to 8.1 ebuild This patch against the latest ebuild currently in portage together with a rename to 8.1 made the package emerge for me.
Kindly review http://www.gentoo.org/doc/en/bugzilla-howto.xml#doc_chap6 (Zero-day bump requests)...
(In reply to comment #1) > Created an attachment (id=141240) [edit] > Patch from 8.443.1-r1 to 8.1 ebuild > > This patch against the latest ebuild currently in portage together with a > rename to 8.1 made the package emerge for me. > Your patch is wrong ... refer to dev overlay for je_fro I done pushed him the changes he will commit to tree most likely over the weekend.
Created attachment 141295 [details] full diff of ebuild changes for this release ... as I have added to je_fro's overlay http://overlays.gentoo.org/dev/je_fro/browser/x11-drivers/ati-drivers ebuild and patch is avaliable via the overlay for those wanting to test until je_fro gets around to adding it.
Created attachment 141297 [details] ebuild diff against the tree Will place patch up in just a min for those who do not want to pull it from je_fro's svn overlay.
Created attachment 141300 [details] ati-powermode-opt-path-2.patch description says it all I do believe :)
ebuild from je_fro's svn overlay works for me, ~amd64 multilib Thx
(In reply to comment #7) > ebuild from je_fro's svn overlay works for me, > ~amd64 multilib > > Thx > I haven't been using ati-drivers for too long, so don't count me as an expert. However, upgrading to 8.452 from je_fro's svn repo made me lose hardware acceleration. fglrxinfo reported OpenGL vendor as mesa. Reverting back to 8.443.1-r1 re-enabled hardware acceleration (fglrxinfo reported ATI Technologies Inc.). FWIW. Anybody else experience the same?
Works perfectly here on amd64! For the 1st time suspend and hibernate work!
(In reply to comment #9) > Works perfectly here on amd64! > For the 1st time suspend and hibernate work! > Nice, could you share your config? Here works nothing (IBM T60 x86 X1400 1400x1050 Xorg 7.3, 2.6.23.11 or so) especially suspend is now totally broken. Florian
(In reply to comment #10) > (In reply to comment #9) > > Works perfectly here on amd64! > > For the 1st time suspend and hibernate work! > > > > Nice, could you share your config? Here works nothing (IBM T60 x86 X1400 > 1400x1050 Xorg 7.3, 2.6.23.11 or so) especially suspend is now totally broken. > Florian > suspend and hibernate are touchy ... the wrong modules built into the kernel will show major trouble. As far as your resolutions not working email me directly you xorg log and kernel config.
(In reply to comment #8) > (In reply to comment #7) > > ebuild from je_fro's svn overlay works for me, > > ~amd64 multilib > > > > Thx > > > > I haven't been using ati-drivers for too long, so don't count me as an expert. > However, upgrading to 8.452 from je_fro's svn repo made me lose hardware > acceleration. fglrxinfo reported OpenGL vendor as mesa. Reverting back to > 8.443.1-r1 re-enabled hardware acceleration (fglrxinfo reported ATI > Technologies Inc.). FWIW. Anybody else experience the same? > This is most likely a broken eselect opengl set ati ... If you would update and retest and repost your result I will be more then happy to get you on the right track.
(In reply to comment #11) > (In reply to comment #10) > > (In reply to comment #9) > > > Works perfectly here on amd64! > > > For the 1st time suspend and hibernate work! > > > > > > > Nice, could you share your config? Here works nothing (IBM T60 x86 X1400 > > 1400x1050 Xorg 7.3, 2.6.23.11 or so) especially suspend is now totally broken. > > Florian > > > > suspend and hibernate are touchy ... the wrong modules built into the kernel > will show major trouble. As far as your resolutions not working email me > directly you xorg log and kernel config. > Oh, with this release my display is recognized correctly so the resoution is working. But hw 3d acceleration is broken, suspend is broken, video displaying is nearly broken. Florian
> Oh, with this release my display is recognized correctly so the resoution is > working. But hw 3d acceleration is broken, suspend is broken, video displaying > is nearly broken. Same for me. Resolution is OK, but no 3D and no video. Did not try suspend. Anyways, the ebuild from the overlay installs flawlessy. The problems are related to the drivers itself not to the ebuild.
(In reply to comment #14) > > Oh, with this release my display is recognized correctly so the resoution is > > working. But hw 3d acceleration is broken, suspend is broken, video displaying > > is nearly broken. > > Same for me. Resolution is OK, but no 3D and no video. Did not try suspend. > Anyways, the ebuild from the overlay installs flawlessy. The problems are > related to the drivers itself not to the ebuild. > Without knowing what card you have I am unable to check into why your card does not work please provide this info so we can get this resolved.
8.452 is in the tree :)
(In reply to comment #15) ... > Without knowing what card you have I am unable to check into why your card does > not work please provide this info so we can get this resolved. > Just look some posts above, there i posted a complete system describtion. Florian
I have a bug with a gentoo-sources-2.6.24 and this driver, fail to install
Created attachment 141798 [details] my builg.log
Created attachment 141800 [details] my emerge --info output
(In reply to comment #18) > I have a bug with a gentoo-sources-2.6.24 and this driver, fail to install > I know where the problem is and should have je_fro push the fix in next day or so.
Until I find out why the check is failing simple enable "PCI_LEGACY" in 2.6.24.
Those who understand how to patch can try .... Index: ati-drivers-8.452.ebuild =================================================================== --- ati-drivers-8.452.ebuild (revision 164) +++ ati-drivers-8.452.ebuild (working copy) @@ -74,7 +74,7 @@ fi if kernel_is ge 2 6 24; then - CONFIG_CHECK="PCI_LEGACY" + CONFIG_CHECK="!PCI_LEGACY" ERROR_PCI_LEGACY="${P} requires support for pci_find_slot (CONFIG_PCI_LEGACY)." fi That should resolve the problem and throw an error if not enabled.
According to the Unofficial Ati Linux Diver Wiki, the new naming is Catalyst and versioning goes by the following key: "The new driver release carry a new name (the same used for the Windows counterpart) and version numbering (7.11 means year 2007, month November). It is still a proprietary closed-source driver."
(In reply to comment #24) > According to the Unofficial Ati Linux Diver Wiki, the new naming is Catalyst > and versioning goes by the following key: > > "The new driver release carry a new name (the same used for the Windows > counterpart) and version numbering (7.11 means year 2007, month November). It > is still a proprietary closed-source driver." > We are well aware of this and have decided against it. Reasonings include many users are familiar with current naming etc..
(In reply to comment #18) > I have a bug with a gentoo-sources-2.6.24 and this driver, fail to install > We have redone the check for PCI_LEGACY it is fixed in overlay je_fro will push it to the tree soon as he has time.
Small comment, the dependency =virtual/libstdc++-3.3* is unneeded with this driver version, as far as I can tell. Won't hurt though.
(In reply to comment #12) > > This is most likely a broken eselect opengl set ati ... If you would update and > retest and repost your result I will be more then happy to get you on the right > track. > Second time around works like a charm. Hardware acceleration works well now.
I just emerged at-drivers 8.452, and I have encountered problems with DRI. Although, xorg.0.log reports that DRI and Composite are both enabled, they are not. When I run strace glxinfo or strace glxgears I get the following error: ioctl(5, 0x80146454, 0xbfc5b800) = 0 mmap2(NULL, 65536, PROT_READ|PROT_WRITE, MAP_SHARED, 5, 0x5) = 0xb7fa3000 ioctl(5, 0xc0086420, 0xbfc5b800) = 0 ioctl(5, 0xc01c64a6, 0xbfc5b750) = 0 ioctl(5, 0xc01c64a6, 0xbfc5b750) = 0 ioctl(5, 0xc01c64a6, 0xbfc5b750) = 0 ioctl(5, 0xc01c64a6, 0xbfc5b750) = 0 ioctl(5, 0xc01c64a6, 0xbfc5b750) = 0 ioctl(5, 0xc01c64a6, 0xbfc5b750) = 0 ioctl(5, 0xc01c64a6, 0xbfc5b740) = 0 ioctl(5, 0xc01c64a6, 0xbfc5b740) = 0 ioctl(5, 0xc01c64a6, 0xbfc5b200) = 0 ioctl(5, 0xc01c64a6, 0xbfc5b200) = 0 ioctl(5, 0x4008642a Also in dmesg I find the following errors: [fglrx] Reserve Block - 0 offset = 0X0 length = 0X40000 [fglrx] Reserve Block - 1 offset = 0X7ff5000 length = 0Xb000 [fglrx] Reserve Block - 2 offset = 0X7fb5000 length = 0X40000 [fglrx] interrupt source 20008000 successfully enabled [fglrx] enable ID = 0x00000004 [fglrx] Receive enable interrupt message with irqEnableMask: 20008000 [fglrx] interrupt source 10000000 successfully enabled [fglrx] enable ID = 0x00000005 [fglrx] Receive enable interrupt message with irqEnableMask: 10000000 [fglrx:firegl_lock_free] *ERROR* lock was not held by 2! (*lock=0x80000001) [fglrx:firegl_unlock] *ERROR* firegl_lock_free failed! [fglrx:firegl_lock_free] *ERROR* lock was not held by 2! (*lock=0xc0000001) [fglrx:firegl_unlock] *ERROR* firegl_lock_free failed! [fglrx] interrupt source 10000000 successfully disabled! [fglrx] enable ID = 0x00000000 [fglrx] Receive disable interrupt message with irqEnableMask: 10000000; dwIRQEnableId: 00000005 [fglrx] interrupt source 20008000 successfully disabled! [fglrx] enable ID = 0x00000000 [fglrx] Receive disable interrupt message with irqEnableMask: 20008000; dwIRQEnableId: 00000004 [fglrx] PCIe has already been initialized. Reinitializing ... [fglrx] Reserve Block - 0 offset = 0X0 length = 0X40000 [fglrx] Reserve Block - 1 offset = 0X7ff5000 length = 0Xb000 [fglrx] Reserve Block - 2 offset = 0X7fb5000 length = 0X40000 [fglrx] interrupt source 20008000 successfully enabled [fglrx] enable ID = 0x00000004 [fglrx] Receive enable interrupt message with irqEnableMask: 20008000 [fglrx] interrupt source 10000000 successfully enabled [fglrx] enable ID = 0x00000005 [fglrx] Receive enable interrupt message with irqEnableMask: 10000000 [fglrx:firegl_lock_free] *ERROR* lock was not held by 2! (*lock=0x80000001) [fglrx:firegl_unlock] *ERROR* firegl_lock_free failed! Does anybody know what is causing this? Thank you
i think this bug could be closed now, this version is in portage next one is released so open a bumb request for this one.
Closing, as it was bumped. Report your errors on other bugs.
