Announcement

Collapse
No announcement yet.

Xserver 1.14 support will arrive with Catalyst 13.6

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

  • Xserver 1.14 support will arrive with Catalyst 13.6

    Since "some" are waiting for this to come, there is a statement
    of an AMD engineer on the semi-official bug tracker:

    http://ati.cchtml.com/show_bug.cgi?id=807#c1

  • #2
    Nice. Now it would be good if we can get Xserver 1.13 support and support for kernels >3.4 in the Legacy driver.

    Comment


    • #3
      Originally posted by Vim_User View Post
      Now it would be good if we can get Xserver 1.13 support and support for kernels >3.4 in the Legacy driver.
      Don't hold your breath for new Xserver support in the legacy driver.
      I'd be very surprised if it sees any (minor) update at all.

      The message by AMD has been clear (more or less):
      If you're using a HD 4K and below, go use the FOSS driver instead.

      Comment


      • #4
        Originally posted by entropy View Post
        Don't hold your breath for new Xserver support in the legacy driver.
        I'd be very surprised if it sees any (minor) update at all.

        The message by AMD has been clear (more or less):
        If you're using a HD 4K and below, go use the FOSS driver instead.
        I would be happy if I could do that, I don't play games on that hardware anyways, but my laptop is unusable with the FOSS driver (overheating, no UVD support) and many people have the same problem. So that is not an option at all and AMD should know this, there are bug reports and at least bridgman was made aware of the issue. If that is really the kind of support AMD believes their customers deserve than it is no wonder that they have financial problems, why should we buy their hardware again?

        Comment


        • #5
          I can understand your frustration. In fact we are still waiting for proper power management. You'd think that would have been a higher priority than the last major feature especially since the lack thereof makes a lot of systems unusable. If if the hardware does work, in the end it doesnt matter if it burns up your lap. I'da thought that is common sense. So in that regard I certainly agree with you.

          However AMD making the OSS driver their official supported driver is not a bad thing. In fact it is a very good thing, and the driver simply would not have come even this far if they hadnt done that. I personally would -LOVE- to see catalyst dropped from Linux altogether. At least the OSS driver works. Thats more than catalyst can do in most setups.

          Comment


          • #6
            Originally posted by duby229 View Post
            However AMD making the OSS driver their official supported driver is not a bad thing.
            Of course not, but they have made it at a time where the OSS driver was (and still is) not in the state to replace the non-free driver for many of their customers and simply left them alone with their problems: "You have a HD2/3/4000? Sorry, but this three years old hardware (also, remember that the HD4000 series is still the top of the line integrated product for the FX series CPUs and still sold) is so obsolete we will support it only with a legacy driver that maybe or maybe not will be updated once in a while. If this is not what you want then use the free driver. Oh, the free driver overheats your hardware and lowers its lifetime with that? Well, we don't care!" That is how it sounds to me.

            I personally would -LOVE- to see catalyst dropped from Linux altogether. At least the OSS driver works. Thats more than catalyst can do in most setups.
            The Catalyst driver works for me, as long as I am able to install it. I could do that recently again on my main machine, but only because the Slackware team decided to move back to Xserver 1.13 (from 1.14), just because the lack of support for that from AMD. Now we expect that Slackware 14.1 is coming soon and I can't use it on my laptop because AMD thinks that Xserver 1.13 and kernel 3.8 are to new for me.

            For me AMD's driver policy comes down to: "Eat shit or buy new hardware! Oh, you have already bought new hardware, well, then you can wait until we are in the mood to support somewhat newer software, but only if we want to or maybe if it is released on Ubuntu."
            They can bet that if I buy new hardware it will not be from AMD anymore.

            Comment


            • #7
              Just pointing out that if the only 4k hw still actively sold are the chipset integrated ones, overheating is no problem: the TDP of those is <5W, the heatsinks + fans can handle that at full load.

              Comment


              • #8
                Should clarify: 5w number pulled from quick googling, may be higher in the most powerful chipsets. Still, well below the mobos cooling ability.

                Comment


                • #9
                  Originally posted by curaga View Post
                  Just pointing out that if the only 4k hw still actively sold are the chipset integrated ones, overheating is no problem: the TDP of those is <5W, the heatsinks + fans can handle that at full load.
                  But missing UVD support for these may be an issue.

                  Comment


                  • #10
                    Originally posted by Vim_User View Post
                    But missing UVD support for these may be an issue.
                    UVD support will be in available out of the box in the next few months, it needs kernel 3.10 and Mesa 9.2. We don't have it now, but it's not far off.

                    Power management is certainly lacking, which certainly is annoying, but the open source team are doing what they can: the code is still undergoing the technical review.

                    Comment


                    • #11
                      Originally posted by archibald View Post
                      UVD support will be in available out of the box in the next few months, it needs kernel 3.10 and Mesa 9.2. We don't have it now, but it's not far off.
                      UVD support will not come for the HD3200 and HD4200. Most integrated chipsets sold use the HD4200.

                      Comment


                      • #12
                        Originally posted by Vim_User View Post
                        UVD support will not come for the HD3200 and HD4200. Most integrated chipsets sold use the HD4200.
                        Hey thanks, this should save us a bunch of time. If you know that it's not going to come out then I guess we can stop working on it...

                        Comment


                        • #13
                          Originally posted by bridgman View Post
                          Hey thanks, this should save us a bunch of time. If you know that it's not going to come out then I guess we can stop working on it...
                          I better should have said: Will not come soon.
                          http://phoronix.com/forums/showthrea...095#post328095
                          Originally posted by agd5f
                          rs780/880 chips like the 4250 do not yet have UVD support.
                          If you have more information when we can expect UVD support for those chips just post it.
                          Also, if you have information when (or better if) we can expect support for kernels >3.4 and Xserver >1.12 in the legacy Catalyst drivers it would also be worth to know that for us users of "Legacy" hardware.

                          Comment


                          • #14
                            It's good news that Catalyst 13.6 will support Xserver 1.14.


                            I hope that Ubuntu 13.04 will also be better supported (I did not upgrade because hybrid graphics users (intel/amd) said they had to install an older intel driver version to get things work).

                            For my laptop, I stick with the catalyst driver because it has good power management and support the last opengl version (just in case I need it)
                            For my desktop, I have an hd4670, and I must say that I experienced too much issues with fglrx and I'm much more happy with the opensource driver (though I set the performance to low to make the card not getting too hot). My experience with the open source driver is relatively problem free.

                            I've read in an old post that the lenght of the code for power management in fglrx was more than the mesa code. Is it true? Does it mean that power management for the open source driver, when it comes, won't be as complete as that fglrx's one? (because it would require too much work)

                            Comment


                            • #15
                              a quick question about UVD. I found Review AMD A6-1450 APU "Temash" on the web which mentions UVD 4.2
                              What would that version update include? I am not able to find anything about it

                              Comment

                              Working...
                              X