Announcement

Collapse
No announcement yet.

AMD Catalyst 8.8 Linux Driver

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

  • #21
    Originally posted by wojak View Post
    Hi,
    Since catalyst 8.3 i have this problem with new catalyst.
    I have connect PC to TV (dvi-hdmi) and there is a black border aroud. It is in all resolution. When i change ref. rate from 60Hz to 50Hz i doesn't help.Only in res. 1920x1080_30Hz all is correct but this ref. rate is to small.

    http://img378.imageshack.us/my.php?image=picturecu6.jpg
    The ghetto fix posted here works for me with the

    aticonfig --set-dispattrib=tdms2i,positionX:0

    etc. and all. The same thing happens in windows but the CCC has a slider to decrease overscan to 0 which fixes it there.

    Technical support and discussion of the open-source AMD Radeon graphics drivers.

    Comment


    • #22
      Originally posted by kiwi_kid_aka_bod View Post
      But have they fixed that damned checkerboard corruption in MythTV / Wine / lots of other programs?

      Really, if they haven't fixed that after the amount of wailing and gnashing of teeth around here, then someone at AMD ought to be waterboarded.
      I would also like to know if the ati/linux tearing and checkerboarding is fixed. If it isn't I'll probably install the fanless nvidia card I've got lying around into my mythfe and use that until I get reports that fglrx or substitute works well for mythfe purposes.

      I'd rather not because
      1) It will probably increase the powerdraw & heat inside the box by 50% (its a 780G box with a CPU, 4GB ram, WD GP drive and not much else).
      2) The HDMI out is nice for the times I have to boot to windows (ie to play the blu-rays with some level of WAF)

      The Mythdora 5.0 distribution with the livna april fglrx drivers (4.876 IIRC) work ... sort of, but the tearing is driving me nuts. The two most recent drivers checkeboard and are not useable at all.

      Comment


      • #23
        Originally posted by blabub View Post
        If there still no solution for wine? Or a workaround?
        if your still having issues with the checkerboard of doom another user pointed
        out that its kinda a regression bug and if you put "virtual X Y" with X being the next mutliple of 64 of your screen res in your screens section, wine will work and not totally fubar your screen. IE i use 3360x1050 my virtual is 3392x1050 soo it goes

        ...
        Subsection Display
        Depth 24
        Virtual 3392 1050
        ....
        Last edited by kusuriya; 20 August 2008, 03:38 PM.

        Comment


        • #24
          Originally posted by energyman View Post
          there is something fishy with xv.

          tvtime: a small jerk every couple of seconds.
          mplayer -vo xv a big jerk every couple of seconds
          I get a jerk every second or so when I'm scrolling in firefox or a terminal, but only with compiz activated (which makes it completely inusable). Without compiz I only get the jerks in mplayer with -vo xv.

          Besides that, the hardlocks when shutting down X or switching to a vt with Ctrl-Alt-Fx (that I get since I upgraded 8.5 to 8.7) are still there.

          Great release, ATI . I think I'll go back to 8.5, because 8.6, 8.7 and 8.8 only added bugs.

          Comment


          • #25
            Originally posted by Fran View Post
            Great release, ATI . I think I'll go back to 8.5, because 8.6, 8.7 and 8.8 only added bugs.
            What's with the naming of the drivers, they seem to be intentionally confusing. The press release says 8.5 or 8.8 but the driver calls itself 4.87 or 5.12.

            Comment


            • #26
              Originally posted by liels View Post
              What's with the naming of the drivers, they seem to be intentionally confusing. The press release says 8.5 or 8.8 but the driver calls itself 4.87 or 5.12.
              as I understand it, it is a throwback from the days when the linux drivers had different version numbering than the windows ones. Soo while the catalyst package its self did a version jump to match windows... the drivers themselves have not done that for whatever reason (theres probly a great internal reason but I am not an ATI dever soo i dont know.)

              Comment


              • #27
                Originally posted by Michael View Post
                For all fixes you don't have to wait three months. It's a matter of where they are in the development cycle for a given month, how invasive the fix is, how much interest there would be in backporting the fix, and whether their real customers are impacted by the bug.
                Well, I would think that the number of annoyed users on this forum would demonstrate the interest in some kind of fix.

                Also, last time I checked I wasn't a figment of someone's imagination (how exactly could someone check that...) so I at least think I'm a real customer...

                Not having a go at you Michael, just taking cheap pot-shots

                Right now I don't even know if they've added this to their internal bug tracker, never mind reproduction, analysis, or potential fixes, so we may as well treat it as 3 years to fix time, and even then we should thank ourselves lucky.

                Whilst bridgeman's presence and communication is appreciated, this closed mentality when it comes to some very nasty issues is very disheartening. AMD (along with Intel) have really done a great deal to engage with and support the community. What seems to be missing is the desire to let the community help them. Yes the "community" includes annoying PITA's, but it's also a large and varied pool of diverse platforms performing a plethora of roles. Perhaps there is some inner circle of trusted beta testers that I'm not aware off, but if so, then it is not working well enough. Issues are getting through that should be seen and solved before ever getting past beta, let alone into release candidate or actually released to public.

                Comment


                • #28
                  Please, please... anybody can give a decent speculation on the XV libs ?

                  Comment


                  • #29
                    sundown
                    libAMDXvBA looks like supposed to control UVD block on R600s (including stuff like denoise, color vibrance, flesh tone, etc., h.264, VC1).

                    libXvBAW.so looks like modified version of libXvMCW.

                    Well, probably we'll see UVD support on Linux some day...

                    Comment


                    • #30
                      Originally posted by taipan View Post
                      Has the same poor 2d performance.

                      I think I have to wait until amd release the docs for m86 and r600 cards.

                      KDE 4 is far from been usable.

                      If I turn on effects on kwin (Xrender or Opengl) then if I resizing a window it will takes ages, I need a calendar to measure that.

                      Their drivers it's like their support, sucks big time.

                      I've not tested 8.8 yet, but with 8.7 KDE4 with enabled composite (opengl or xrender) is unusable. Maximizing window takes more than one second on radeon 4850 and core2quad@3,2GHz.

                      So why all people telling only about nvidia poor 2d performance? Nvidia is at least usable on geforce8 and above cards. ATI - not.

                      Comment

                      Working...
                      X