Already tried it out. It's another simple bump. The new version number is 8.471.3. Change 8-2 to 8-3 in the SRC_URI. The ati-powermode-opt-path-2 patch still applies cleanly. Very pleased because this version fixes all the bugs that have been making me tear my hair out since I got this card a few weeks ago.
Sigh; was this 0day bug really needed?!? http://www.gentoo.org/doc/en/bugzilla-howto.xml#doc_chap6 (Zero-day bump requests)
I thought I was being helpful by not just saying "bump please!" but actually saying that this is what little you have to do, I've tried it and it works. Won't bother next time.
Well, first: I'm sorry if this also counts as 0 day annoyance.. This ebuild as well as the older ebuilds have one problem in common, which i discovered only recently (actually with the last version of the drivers). They failed emerging with the following error: FATAL: modpost: GPL-incompatible module fglrx.ko uses GPL-only symbol 'lock_acquire' Well, last time i ignored this, as i thought it's a driver problem, but upon the same error with the new drivers, i got suspicious and investigated: The driver won't compile if the kernel has CONFIG_LOCKDEP enabled. Which is the result of enabling e.g CONFIG_DEBUG_LOCK_ALLOC or CONFIG_PROVE_LOCKING. see: http://article.gmane.org/gmane.linux.debian.devel.bugs.general/375953 So, i think it's important to tell the user to disable the last two settings in the kernel hacking section, in case this error arises. With CONFIG_LOCKDEP disabled the drivers compile perfectly.
What about adding the fix proposed for this atieventsd bug at: http://ati.cchtml.com/show_bug.cgi?id=992#c26 It's simple enough to be in a gentoo ebuild and it fixes an important bug for KDE users.
Created attachment 145810 [details, diff] Patch for ati-drivers-8.471.3 ebuild This patch applies to ati-drivers-8.455.2-r1.ebuild and bumps it to Catalyst 8.3 version, fixing the location of the Xauth files for xdm and kdm, thus preventing lockups and freezes when logging out from KDE.
Created attachment 145826 [details, diff] Better patch for ati-drivers-8.471.3 ebuild This is a better one.
Created attachment 145827 [details, diff] Better patch for ati-drivers-8.471.3 ebuild Sorry for the bugspam, the previous patch was wrong, this is the correct one.
Okay folks, I've got the new ebuild in my overlay that appears to work on my macbook pro. Any other testers would be appreciated :) Thanks, je_fro
Jeffrey, the version number in your overlay is wrong. It should be 8.471.3.
done!
(In reply to comment #9) > Jeffrey, the version number in your overlay is wrong. It should be 8.471.3. > The version string does not matter as long as it reflects a bump. Come 8.5 which is May release the version string will be moved to the normal catalyst release. We just do the release version way it is now due to problems with our current PV.
Updated from 8.452 to je_fro overlay 8.471.3 today. Seems to work fine :-) I have tested video playback and opengl. Video playback have some tearing, but that also existed on 8.452. Opengl works. I have played 4-5 hours WoW with no crash or clitches (besides known problems). (--) fglrx(0): Chipset: "ATI Radeon HD 3600 Series" (Chipset = 0x9598) (II) fglrx(0): VESA VBE OEM Product: RV635 Portage 2.1.4.4 (default-linux/amd64/2006.1/desktop, gcc-4.1.2, glibc-2.6.1-r0, 2.6.23-gentoo-r9-sax-r2 x86_64) ================================================================= System uname: 2.6.23-gentoo-r9-sax-r2 x86_64 AMD Athlon(tm) X2 Dual Core Processor BE-2400 ACCEPT_KEYWORDS="amd64" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-march=athlon64 -msse3 -O2 -pipe" CHOST="x86_64-pc-linux-gnu" CXXFLAGS="-march=athlon64 -msse3 -O2 -pipe" MAKEOPTS="-j3"
(In reply to comment #12) Another success here. AMD64, 2.6.24. Verified glxgears, googleearth and video playback (xv). All works fine. Oddly enough I couldn't apply the patch from this bug report and had to edit the ebuild manually.
Kernel 2.6.24-gentoo-r3 on x86 here. (--) fglrx(0): Chipset: "ATI MOBILITY RADEON X700" (Chipset = 0x5653) (--) fglrx(0): (PciSubVendor = 0x1025, PciSubDevice = 0x007e) Works great, and *FINALLY* the vsync works again ! This version is well worth adding to the tree IMHO.
> > Works great, and *FINALLY* the vsync works again ! This version is well worth > adding to the tree IMHO. > And so it shall be. Thanks guys and gals! je_fro