Announcement

Collapse
No announcement yet.

Updated and Optimized Ubuntu Free Graphics Drivers

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

  • Oibaf, would it be possible to enable ARM builds on this PPA? I don't know a whole lot about the PPA process but if it's just flipping a build switch it would be nice. I've got Debian working on several ARM systems and having to build the Freedreno driver (xf86-video-freedreno/xserver-xorg-video-freedreno) from source is rather annoying as well as the new mesa it requires. There is a Freedreno xserver-xorg-video-freedreno source repository out there but it doesn't pull from git or stay up to date it seems. I would definitely switch to Ubuntu for ARM if git mesa with Freedreno was available. For reference my target hardware at the moment is my Galaxy Note phone, which I currently have Debian jessie + Freedreno working on as well as my HP TouchPad which I have mostly the same setup on.

    Comment


    • Cannot get hardware decoding going

      Hello everyone. Thank you for participating in the forum.

      I have tried everything I can think of to get hardware decoding going with the oibaf ppa (short of using Ubuntu 14.10) but no luck.

      System is:
      core2 duo 2.1GHz (E6400)
      intel dg965wh ACH8
      8G dual channel ram
      SATA RAID5 (software)
      XFX HD 6570 512K
      Ubuntu 14.04 desktop 64 livecd install

      fresh install with no updates or proprietary software
      add the oibaf ppa
      update
      dist-upgrade
      install mpv and mdadm
      mpv --hwdec auto bluray.mkv (ripped with makemkv)
      mpv --hwdec vdpau bluray.mkv
      mplayer -vo vdpau bluray.mkv

      The performance is the same with or without the ppa. On scenes with lots of dynamic detail the graphics stutters and pixelates as if there are dropped frames.

      I tried adding mesa-vdpau-drivers as suggested by the ppa and that did not help either.

      After installing the ppa it looks as if mpv and mplayer are trying to use hardware decoding but it seems to always pick an ffmpeg or lavc codec. I am not sure if I am correct about that because I have never seen the system working properly so I do not know what the stdout/err is supposed to look like...

      Mplayer looks like it also tries to use hardware decoding but says the codec needs to be 'enabled at compilation'. It seems to be looking for a Windows VC1 codec. If I specify a hardware decoder that does not exist mplayer complains about the codec, throws a 'system too slow' error, and the video lags the audio, freezes, then skips ahead to try and catch up, as if frame dropping is disabled.

      VLC performs similarly even with hardware decoding enabled under 'codecs'.

      Similar result with the proprietary AMD driver. Similar results with the xorg-edgers ppa. Similar results with the latest AMD driver from their website.

      I also tried using a Geforce 210 with proprietary Nvidia driver with similar results.

      Then the Geforce card died completely before I could try the latest driver from Nvidia web site. Both graphics cards were used, from ebay. The XFX appears to be fine. The Geforce 210 might not have much/any hardware decoding if the Phoronix testing on other 2xx cards is any guide (they basically did not reduce peak CPU load at all) so I am not sure that was a fair test, while it was still functioning anyway.

      I have a GT 730 being delivered tomorrow to try again. This one is new but also from ebay at nearly 50% discount auction. I prefer to use the OSS driver if possible and the least expensive graphics required for the job since at higher costs upgrading the processor is an alternate route that has additional benefits.

      I have four systems total that I would like to get it working on and will need at least 2 more graphics cards to do it (two systems are Pentium 4 and upgrading the processor is not a viable option, the third is another Core2 Duo E6400 in an ABIT), but I need to get at least one working first so I know I have a functional plan... maybe I will buy two Core2 Quad processors and use the graphics cards for the Pentiums? Feels like a bad compromise... I should be able to get this working at very low cost with used graphics?

      What am I doing wrong? I spent weeks on this trying to debug and getting lost in the complexities with no positive result. There is a long thread at AVS Forum where I have posted dozens of experiments, trying to get it working. I have a hard time believing I am that dumb... it must be something I overlooked?

      The thing that is most confusing is the way the applications seem to be aware of hardware decoding, and trying to use it, but always picking software decoding anyway.

      I have another fresh install ready to go, no updates or proprietary software applied yet. What command lines will accomplish hardware decode?

      Thank you very much for your time and effort.

      Comment


      • I have to disable this PPA now because I couldn't login to my Ubuntu 14.04 session (Unity):
        Code:
        [  1987.501] (EE) 
        [  1987.501] (EE) Backtrace:
        [  1987.502] (EE) 0: /usr/bin/X (xorg_backtrace+0x48) [0x7f9594875d28]
        [  1987.502] (EE) 1: /usr/bin/X (0x7f95946cd000+0x1aca19) [0x7f9594879a19]
        [  1987.502] (EE) 2: /lib/x86_64-linux-gnu/libpthread.so.0 (0x7f95937ca000+0x10340) [0x7f95937da340]
        [  1987.502] (EE) 3: /usr/lib/x86_64-linux-gnu/libdrm_nouveau.so.2 (nouveau_bo_set_prime+0x1) [0x7f958df0b301]
        [  1987.502] (EE) 4: /usr/lib/xorg/modules/drivers/nouveau_drv.so (0x7f958e110000+0xc430) [0x7f958e11c430]
        [  1987.502] (EE) 5: /usr/bin/X (0x7f95946cd000+0x17d2a9) [0x7f959484a2a9]
        [  1987.502] (EE) 6: /usr/bin/X (0x7f95946cd000+0x17cc1e) [0x7f9594849c1e]
        [  1987.502] (EE) 7: /usr/bin/X (0x7f95946cd000+0x55a1e) [0x7f9594722a1e]
        [  1987.502] (EE) 8: /usr/bin/X (0x7f95946cd000+0x598aa) [0x7f95947268aa]
        [  1987.502] (EE) 9: /lib/x86_64-linux-gnu/libc.so.6 (__libc_start_main+0xf5) [0x7f9592209ec5]
        [  1987.502] (EE) 10: /usr/bin/X (0x7f95946cd000+0x44dde) [0x7f9594711dde]
        [  1987.502] (EE) 
        [  1987.502] (EE) Segmentation fault at address 0x0

        Comment


        • Originally posted by simonbcn View Post
          I have to disable this PPA now because I couldn't login to my Ubuntu 14.04 session (Unity):
          Code:
          [  1987.501] (EE) 
          [  1987.501] (EE) Backtrace:
          [  1987.502] (EE) 0: /usr/bin/X (xorg_backtrace+0x48) [0x7f9594875d28]
          [  1987.502] (EE) 1: /usr/bin/X (0x7f95946cd000+0x1aca19) [0x7f9594879a19]
          [  1987.502] (EE) 2: /lib/x86_64-linux-gnu/libpthread.so.0 (0x7f95937ca000+0x10340) [0x7f95937da340]
          [  1987.502] (EE) 3: /usr/lib/x86_64-linux-gnu/libdrm_nouveau.so.2 (nouveau_bo_set_prime+0x1) [0x7f958df0b301]
          [  1987.502] (EE) 4: /usr/lib/xorg/modules/drivers/nouveau_drv.so (0x7f958e110000+0xc430) [0x7f958e11c430]
          [  1987.502] (EE) 5: /usr/bin/X (0x7f95946cd000+0x17d2a9) [0x7f959484a2a9]
          [  1987.502] (EE) 6: /usr/bin/X (0x7f95946cd000+0x17cc1e) [0x7f9594849c1e]
          [  1987.502] (EE) 7: /usr/bin/X (0x7f95946cd000+0x55a1e) [0x7f9594722a1e]
          [  1987.502] (EE) 8: /usr/bin/X (0x7f95946cd000+0x598aa) [0x7f95947268aa]
          [  1987.502] (EE) 9: /lib/x86_64-linux-gnu/libc.so.6 (__libc_start_main+0xf5) [0x7f9592209ec5]
          [  1987.502] (EE) 10: /usr/bin/X (0x7f95946cd000+0x44dde) [0x7f9594711dde]
          [  1987.502] (EE) 
          [  1987.502] (EE) Segmentation fault at address 0x0
          Definitely it uses this PPA is a pain in the ass. Today I have updated again and now:

          Comment


          • Originally posted by simonbcn View Post
            Definitely it uses this PPA is a pain in the ass. Today I have updated again and now:

            PITA? If something is PITA it is you. The guy is doing incredible work and you mean it is PITA. If running ppa-purge is too bothersome for you when something breaks (because you are running very recent untested code.) than
            just stick to stock Ubuntu or Windows and enjoy.

            Comment


            • Recent builds seem to break something with Intel drivers. I am not able at the moment to provide any info, since I am quite busy, and had to ppa-purge right away. Is there anyone who can confirm issues with intel cards, or someone who could provide some additional data about this?

              Comment


              • Just to break the chain a bit; two computers using AMD hardware are fine with the PPA currently.

                One uses a 2400 PRO; the other is a dual-graphics 7660G + 7670M laptop (haven't tested the dGPU).

                Comment


                • Originally posted by simonbcn View Post
                  Definitely it uses this PPA is a pain in the ass. Today I have updated again and now:
                  You you aware that you are running drivers straight from mesa git, right? This is suppose to break every now and then, that is how devs fix stuff, they test it out and if you are using Oibaf PPA you are simply testing latest Mesa development code which is not stable at all (not suited for everyday use, though most of us use it anyway :P ).

                  Oibaf is maintaining this PPA for your convienence in his own spare time and if you don't like it, there is always ppa-purge, you ungrateful fuck.

                  Comment


                  • Did some tests today with my AMD Radeon R7 260X (GCN 1.1, BONAIRE).

                    Counter-Strike: Global Offensive is 25-30% faster with stock Ubuntu 14.10 (Mesa 10.3, but HyperZ enabled manually) than it is with Ubuntu 14.10 + Oibaf PPA (Mesa 10.4-git, also with HyperZ enabled), so it regressed a little.

                    Independent of the Mesa version used, using Linux 3.18rc1 gives a small 8-10% performance boost over the stock Linux 3.16 kernel used in Ubuntu 14.10 (haven't tried Linux 3.17).

                    So Mesa 10.3 + Linux 3.18rc1 is the fastest combination for me with the opensource stack, but AMD Catalyst is still quite a bit faster.

                    Comment


                    • Originally posted by reCAPTCHA View Post
                      Recent builds seem to break something with Intel drivers. I am not able at the moment to provide any info, since I am quite busy, and had to ppa-purge right away. Is there anyone who can confirm issues with intel cards, or someone who could provide some additional data about this?
                      I can confirm some issues with Intel hardware. A lot of flashing and blinking occurs in the terminal window, have not noticed issues with other applications.

                      I'm on a fresh 14.10 install. Terminal was working properly before I added the Oibaf ppa.

                      Intel i3 530 CPU: http://ark.intel.com/products/46472

                      [ 19.240] (II) intel(0): Using Kernel Mode Setting driver: i915, version 1.6.0 20080730
                      [ 19.240] (II) intel(0): SNA compiled: xserver-xorg-video-intel 2:2.99.916+git1410241932.d08a5f~gd~u (Oibaf <[email protected]>)

                      Comment

                      Working...
                      X