Announcement

Collapse
No announcement yet.

R600 Gallium3D Loses To AMD Catalyst Legacy Driver On Windows 8

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

  • #31
    Originally posted by agd5f View Post
    The Linux catalyst driver is still supported and maintained as well as long as you are using a supported distro (kernel and xorg version). It's the same as on windows. The only difference is that windows doesn't change the kernel and display server as often as Linux distros do.
    The problem is that no major distro uses both the last supported Xorg and Kernel version.

    This model of Catalyst Legacy releases is not working for Linux.

    Comment


    • #32
      Originally posted by verde View Post
      The problem is that no major distro uses both the last supported Xorg and Kernel version. Not even the long term support Ubuntu 12.04 LTS.
      It does, though as of 12.04.2 they backported the Quantal X server/kernel and use them by default on new installations (at least on 64 bit). I guess it's related to UEFI. Kernel 3.2 and X server 1.11 are still in the repo, and the xorg-edgers PPA has X server 1.12.
      openSUSE 12.2 runs kernel 3.4 and X server 1.12. Still supported and poised to become the next Evergreen.
      Last edited by PsynoKhi0; 07-26-2013, 01:01 PM.

      Comment


      • #33
        Originally posted by droste View Post
        How did you do this? But the test itself is kind of weird here (juniper 5770).

        With compositing: ~32000
        Without compositing: ~7000 (no there's no zero missing at the end)
        Looks like Option "SwapbuffersWait" "off" in xorg.conf might help, without it I have ~7000 too.

        Comment


        • #34
          Originally posted by vadimg View Post
          Looks like Option "SwapbuffersWait" "off" in xorg.conf might help, without it I have ~7000 too.
          That might be the problem. But doesn't need catalyst do something like this too, to render tearfree?
          It's definitely not DPM, because it's not working here and the GPU runs always at highest clocks.

          Comment


          • #35
            Originally posted by droste View Post
            That might be the problem. But doesn't need catalyst do something like this too, to render tearfree?
            It's definitely not DPM, because it's not working here and the GPU runs always at highest clocks.
            What about CPU clocks? Is it getting clocked down in this configuration?

            Comment


            • #36
              Originally posted by droste View Post
              That might be the problem. But doesn't need catalyst do something like this too, to render tearfree?
              You can disable it in catalyst as well via the control panel.

              Comment


              • #37
                Originally posted by duby229 View Post
                What about CPU clocks? Is it getting clocked down in this configuration?
                Setting the CPU to the highest clock might bring a few points more, but SwapbuffersWait is probably the real reason for my results as vadimg has the same result with SwapbuffersWait.

                Comment


                • #38
                  Originally posted by plantroon View Post
                  radeon hd 3650 still gets lockups while or after running games/3D apps on free drivers (it's random but once u start 3D app it's 100% of getting a lockup within an hour)
                  Try R600_DEBUG=nohyperz

                  Comment


                  • #39
                    SwapbuffersWait basically stalls the GPU until buffers can be copied without tearing. It is always used for non-fullscreen windows, even if vsync is disabled, and as such it can have a huge effect on performance. It should always be disabled for benchmarking.

                    Personally, I just keep it disabled and live with a little bit of rare tearing, in exchange for much smoother overall performance.
                    Last edited by brent; 07-27-2013, 05:10 AM.

                    Comment


                    • #40
                      Something is really wrong with the testing, but the testing itself is very fit. :/
                      Something that affects both opensource drivers.

                      Comment

                      Working...
                      X