Announcement

Collapse
No announcement yet.

Bringing The R600 Gallium3D Driver Up To Speed

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

  • #11
    Originally posted by d2kx View Post
    Marek has been doing amazing work over the last months. If any of the volounteers deserved to get hired/payed, then it would be him.
    Very much agreed!

    Comment


    • #12
      Current Status

      Originally posted by V!NCENT View Post
      What is up with the overal status of Gallium3D and the likes?
      -Mesa is already a State Tracker, right? Is it fully converted to Gallium3D compatibility? Or is it lacking compared to standard/classic Mesa?
      -What is up with all the other State Trackers and where do they 'live'? OpenGL lives in Mesa, but where is OpenCL, OpenVG and all the rest?
      -Is the only thing that is lacking in r600 and up the Gallium driver, or is there more to it?
      Hi,

      I was wondering that myself as well.

      The only other thing I found is a preliminary openCL frontend for Gallium3D, named Clove

      Comment


      • #13
        Originally posted by whizse View Post
        Marek has been doing amazing work over the last months. If any of the volounteers deserved to get hired/payed, then it would be him.
        Very much agreed!
        Yeah it would be nice if he got hired by Red Hat to work on R600+.

        [cynical mode] But can you imagine if ATI hired him? They'd lock him up in the office and not let him commit a single line of code.
        [/cynical mode]

        Comment


        • #14
          ... and of course let's not forget to cheer for Jerome Glisse, who actually writes r600g

          Comment


          • #15
            Originally posted by netkas View Post
            anything i did wrong ?
            Dunno, the port is likely no operational yet. Note that it only works on r7xx, r6xx is missing the bytecode stuff.

            Comment


            • #16
              Originally posted by glisse View Post
              Dunno, the port is likely no operational yet. Note that it only works on r7xx, r6xx is missing the bytecode stuff.
              Will you merge to Mesa master before or after the r6xx bit?

              Comment


              • #17
                Originally posted by nanonyme View Post
                Will you merge to Mesa master before or after the r6xx bit?
                Dunno, i think i will merge once i am happy with the instruction packing & scheduling and i don't know if i will add r6xx before,after,while ...

                Comment


                • #18
                  Originally posted by glisse View Post
                  Dunno, i think i will merge once i am happy with the instruction packing & scheduling and i don't know if i will add r6xx before,after,while ...
                  Fair enough. I guess we'll just have to keep eyes open.

                  Comment


                  • #19
                    Originally posted by monraaf View Post
                    [cynical mode] But can you imagine if ATI hired him? They'd lock him up in the office and not let him commit a single line of code.
                    [/cynical mode]
                    Sure, just like we did with Alex Deucher

                    Comment


                    • #20
                      Originally posted by bridgman View Post
                      Sure, just like we did with Alex Deucher
                      There's a good side and a bad side of that though of people ending up directly working for you. Good side is they sign the NDA and get access to full AMD resources in developing the drivers. The bad side is that they sign the NDA so they're bound by it when it comes to what exactly they can push out. The good side is bigger than the bad side though, considering no one seems to be that anxious of reverse-engineering the parts of the chips AMD/ATi doesn't want getting public.

                      Comment

                      Working...
                      X