Announcement

Collapse
No announcement yet.

AMD Releases Cayman Documentation, Open Driver Is Close

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

  • #51
    Originally posted by yesterday View Post
    What I'm trying to say is that the patent will stop the drivers from having these features, because no company or well-known open source organization is going to work on implementing patented software.
    Heh... Really??

    Linux.
    VP3/Theora.
    VP8/WebM.

    Patented stuff in the mix. FOSS. RedHat, SuSE, Debian, Ubuntu, FSF, etc. use the stuff and develop for it ALL THE TIME.

    It depends on whether you have an effective license in a pledge that forms one or a full-on license grant- if you do, you work with it. If you don't...well, you don't.

    Comment


    • #52
      Originally posted by glisse View Post
      GL3 or 4 are not a no go, we will have 99% (% top of my head where 1% is patented stuffed) of it advertised through appropriate extension. Question will be do we advertise GL version 3 or 4 without the patented stuff and throw glerror to program trying to use those, or do we go another way.
      How are the Mesa team members handling the attempt to use S3TC? Don't the drivers throw an error when you try that right now? I'd say unless it's more than 20% of the feature on the ARB edge, you'd do the same thing. If it's more than that, you don't advertise it unless even though it's a larger percentage, it's not as harmful as the first.

      Comment


      • #53
        Originally posted by glisse View Post
        GL3 or 4 are not a no go, we will have 99% (% top of my head where 1% is patented stuffed) of it advertised through appropriate extension. Question will be do we advertise GL version 3 or 4 without the patented stuff and throw glerror to program trying to use those, or do we go another way.
        Are there even any apps out there that try to use OpenGL 3 or 4 and DON'T use something like floating point rendering targets? I was under the impression that it was a pretty fundamental feature of GL3. On the other hand, i guess apps that use that aren't likely to provide an older path that would work on the drivers anyway, so maybe it wouldn't cause any new problems.

        Comment


        • #54
          Originally posted by Svartalf View Post
          How are the Mesa team members handling the attempt to use S3TC? Don't the drivers throw an error when you try that right now?
          No, where did you get that from?

          I have perfect s3tc support on my Intel Arrandale GPU and AFAIK r300g also has full support. I'm currently fiddling around with r600g, where mipmap aren't correctly handled yet (and the CS checker needs some update, but for now it works to just disable the texture checks). Trying to find out how r300g handles these, maybe this gives me a clue on how to do it for r600g.

          Comment


          • #55
            Originally posted by LiquidAcid View Post
            No, where did you get that from?

            I have perfect s3tc support on my Intel Arrandale GPU and AFAIK r300g also has full support. I'm currently fiddling around with r600g, where mipmap aren't correctly handled yet (and the CS checker needs some update, but for now it works to just disable the texture checks). Trying to find out how r300g handles these, maybe this gives me a clue on how to do it for r600g.
            mipmap have been working for a long time unless somethings recent broke them, if you have issue please open a bug and provide test case. We don't see issue with cs checker, which kernel ? Again it's best to open bug for this.

            Comment


            • #56
              I'm talking about r600g here:

              Comment


              • #57
                Originally posted by V!NCENT View Post
                I mean it would be great to have OpenGL>2.1, but when you look at it, it is actually not a big deal.
                All of the games I mentioned are OpenGL 2.1, they don't require any OpenGL3 stuff.

                Comment


                • #58
                  Originally posted by LiquidAcid View Post
                  I'm pretty sure that glisse *only* talks about r600g
                  Test signature

                  Comment


                  • #59
                    Well, then he's definitely wrong, because texture compression never really worked for r600g. How could this even work when the CS checker rejects these texture formats completly?

                    Comment


                    • #60
                      Originally posted by LiquidAcid View Post
                      Well, then he's definitely wrong, because texture compression never really worked for r600g. How could this even work when the CS checker rejects these texture formats completly?
                      It seems you are confusing things, mipmap & st3c compressed texture are too different orthogonal things. mipmap works and all gl test using mipmap that we have prove it works. We don't support st3c or any kind of texture compression. In your bugs it's unclear wether you tried to force st3c adevertising or not, doing so might broke the driver as we don't support st3c texture yet. Also quake4 might be doing something uncommon/out of normal GL spec with mipmap.

                      Yes i only talk about r600g

                      Comment

                      Working...
                      X