Announcement

Collapse
No announcement yet.

AMD Catalyst 12.6 For Linux Disappoints

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #31
    Originally posted by Kano View Post
    @xgt001

    i dont use arch.
    I use Debian SID amd64 and it segfaults also...

    Comment


    • #32
      this release fixes A LOT of bugs.

      Oh but this is soooo disappointing...

      it is? why? what do you expect from one release to the next? World shattering performance improvements?

      Stuff like this is only disappointing if someone artificially increased expectations beforehand.

      Comment


      • #33
        compare it to 12-6 beta and you find none...

        Comment


        • #34
          Originally posted by Kano View Post
          compare it to 12-6 beta and you find none...
          Isn't that the point of a beta, though? To test the release before it comes out and make sure there are no regressions?

          I'm not sure why anyone would expect the final release to be different than the beta.

          Comment


          • #35
            Originally posted by smitty3268 View Post
            I'm not sure why anyone would expect the final release to be different than the beta.
            Not anyone, just fglrx-haters.

            Comment


            • #36
              Originally posted by devguy View Post
              I was unable to get the Catalyst 12.6 beta work with neither kernel 3.4.x nor XServer 1.12 on Ubuntu 12.04. Worked acceptably with 3.3.x and 1.11, though.
              are you on a 32 bit system?? there are known issues for 3.4 kernel on 32bit architecture while installing catalyst

              Comment


              • #37
                On the aticonfig / amdconfig issue:

                Replace /etc/ati/control with the version from 12.6b will fix the Non supported hardware detected when using aticonfig / amdconfig.

                Comment


                • #38
                  so debian fucked up their xorg-server and that is amd's fault?


                  btw, kernel 3.0.X here... why the race to the latest kernels if there is ZERO difference for all practical reasons....

                  Comment


                  • #39
                    Just posting to say that I'M NOT HAPPY AMD!!!!

                    Just fresh installed Debian Wheezy on a PC with a fairly new 6550 graphics card.. The open source drivers were unusable because the display was so corrupted.. Debian Wheezy recently got the new 12.1 xserver-xorg-core package and now the AMD Catalyst 12.6 x86_64 drivers are broken on 64-bit Linux!! Something about the Catalyst 12.6 driver needing to pass a 64-bit pointer but is instead passing a 32-bit pointer + some corruption to xserver-xorg which causes newer 64-bit compiled versions of xserver-xorg-core (12.1) to crash instantly before loading the desktop!!

                    I'm not happy AMD, not happy at all.. At this point, both your open source and Catalyst drivers are broken for me.. I had to pin the old xserver* from Debian stable (Squeeze) to get the Catalyst 12.6 drivers to stop causing xserver-xorg to crash because it's sending a partially corrupted pointer on a 64-bit OS! Apparently this has been a known bug for over a month against the 12.6 Catalyst drivers and AMD doesn't seem to have done anything about it!!!

                    FIX IT PLEASE!!!
                    Last edited by Sidicas; 07-02-2012, 12:00 AM.

                    Comment


                    • #40
                      Originally posted by Sidicas View Post
                      Just posting to say that I'M NOT HAPPY AMD!!!!

                      Just fresh installed Debian Wheezy on a PC with a fairly new 6550 graphics card.. The open source drivers were unusable because the display was so corrupted.. Debian Wheezy recently got the new 12.1 xserver-xorg-core package and now the AMD Catalyst 12.6 x86_64 drivers are broken on 64-bit Linux!! Something about the Catalyst 12.6 driver needing to pass a 64-bit pointer but is instead passing a 32-bit pointer + some corruption to xserver-xorg which causes newer 64-bit compiled versions of xserver-xorg-core (12.1) to crash instantly before loading the desktop!!

                      I'm not happy AMD, not happy at all.. At this point, both your open source and Catalyst drivers are broken for me.. I had to pin the old xserver* from Debian stable (Squeeze) to get the Catalyst 12.6 drivers to stop causing xserver-xorg to crash because it's sending a partially corrupted pointer on a 64-bit OS! Apparently this has been a known bug for over a month against the 12.6 Catalyst drivers and AMD doesn't seem to have done anything about it!!!

                      FIX IT PLEASE!!!
                      looks like more of a debian issue not sure though.. even fedora 64 bit runs catalyst 12.6 with xserver 1.12!!!

                      Comment


                      • #41
                        Originally posted by energyman View Post
                        so debian fucked up their xorg-server and that is amd's fault?.
                        Actually AMD fucked up when they linked the driver to libpciaccess (using 32 bit pointers on a 64 bit distro is not such a good idea)
                        Dont know why its only an issue on Debian

                        EDIT: There is an ugly hack to work arround the issue at http://ati.cchtml.com/show_bug.cgi?id=522#c5
                        Last edited by rechapit; 07-02-2012, 09:34 AM. Reason: Added link to hackfix

                        Comment


                        • #42
                          Originally posted by rechapit View Post
                          Actually AMD fucked up when they linked the driver to libpciaccess (using 32 bit pointers on a 64 bit distro is not such a good idea)
                          Dont know why its only an issue on Debian

                          EDIT: There is an ugly hack to work arround the issue at http://ati.cchtml.com/show_bug.cgi?id=522#c5
                          But if you use Debian 'unstable', why do you expect everything to work?
                          I don't complain that the open source radeon driver has issues with Linux 3.4.4 on Gentoo but rather file a bug report for it.

                          (the issue is, that ColorTiling2D stopped working with the newer kernels and causes problems with xrandr expanding to external displays)

                          Comment


                          • #43
                            The same error is of course in debian wheezy, which will be frozen soon. i just verified that that hack works, but that shows how stupid fglrx devs are - the reason was posted when the beta was out and they did NOTHING. why call it beta first when the final is the same?

                            Comment


                            • #44
                              Originally posted by Kano View Post
                              why call it beta first when the final is the same?
                              Not the same. Look at this thread: http://phoronix.com/forums/showthrea...-on-HP-Envy-15

                              Comment


                              • #45
                                Well a differnt control file does not mark a different driver, it was just recompiled with the wrong options.

                                Comment

                                Working...
                                X