Announcement

Collapse
No announcement yet.

Canonical Working On Mesa Code Again For Mir

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

  • Canonical Working On Mesa Code Again For Mir

    Phoronix: Canonical Working On Mesa Code Again For Mir

    With Canonical's small X.Org team back to publishing patches on the Mesa mailing list, it looks like they might be trying again soon for pushing forward their Mir EGL back-end...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    we'll see in the coming days if Canonical tries to mainline their Mir back-end itself for Mesa or just simply is trying to reduce the maintenance workload in supporting the out-of-tree code by pushing forward some code (i.e. today's patches)
    I don't think it's only that.

    The Mir backend lives in a part of Mesa that doesn't change much, whereas the DRIimage v7 code (the patches here) lives in a part that can change more. It makes more sense to merge it (and avoid other programers to write again this code, which is partially needed for DRI3 support).

    We need this for Prime on Wayland, and for writing glamor dri3 helpers. And RAOF also published again more complete patches, because I asked him, telling him the needs of glamor (the __DRiimage v6 bits were missing from previous patches, and glamor would need them, so he added them) and Prime on Wayland.

    Comment


    • #3
      Originally posted by mannerov View Post
      I don't think it's only that.

      The Mir backend lives in a part of Mesa that doesn't change much, whereas the DRIimage v7 code (the patches here) lives in a part that can change more. It makes more sense to merge it (and avoid other programers to write again this code, which is partially needed for DRI3 support).

      We need this for Prime on Wayland, and for writing glamor dri3 helpers. And RAOF also published again more complete patches, because I asked him, telling him the needs of glamor (the __DRiimage v6 bits were missing from previous patches, and glamor would need them, so he added them) and Prime on Wayland.
      Yes, but why merging ONE distro specific patches?

      Comment


      • #4
        Originally posted by Drago View Post
        Yes, but why merging ONE distro specific patches?
        I think you failed at reading there.

        Comment


        • #5
          Originally posted by Drago View Post
          Yes, but why merging ONE distro specific patches?
          If I read the post your quoting correctly, its saying that these patches [B]do[B] contain work that is useful for more than just MIR, and parts of Wayland in this case as well.

          Comment


          • #6
            Originally posted by zeealpal View Post
            If I read the post your quoting correctly, its saying that these patches [B]do[B] contain work that is useful for more than just MIR, and parts of Wayland in this case as well.
            Not to mention that there is no Mir dependant code in these patches, ie no mir egl platform.

            Comment


            • #7
              Originally posted by pdffs View Post
              Originally posted by Drago
              Yes, but why merging ONE distro specific patches?
              I think you failed at reading there.
              But succeeded in trolling! Yey!

              Comment


              • #8
                Originally posted by pdffs View Post
                I think you failed at reading there.
                I think blind hate clouded his neurons to the point he can't even read.

                Comment


                • #9
                  The Mesa crew should reject the patches.

                  There is nothing stopping Canonical from producing Mir. They want it, they can have at it.

                  But Canonical is notorious for producing -~NOTHING~- for the community in return. So the community should not be charged with supporting Mir.

                  Mir is Canonical's baby. Canonical and Canonical alone should support it.

                  Comment


                  • #10
                    Originally posted by halfmanhalfamazing View Post
                    There is nothing stopping Canonical from producing Mir. They want it, they can have at it.

                    But Canonical is notorious for producing -~NOTHING~- for the community in return. So the community should not be charged with supporting Mir.

                    Mir is Canonical's baby. Canonical and Canonical alone should support it.
                    Have you even read the article? I will quote it for you:
                    The patch-set is basically about supporting DRI Image 6/7 in the Gallium3D drivers.

                    The drivers enable DRI Image 7 support for the Gallium3D drivers currently supporting DMA-BUF's import/export capabilities: Nouveau, LunarG's ILO Intel Gallium3D, and Radeon (R300, R600, and RadeonSI). There's also the necessary prep work to Gallium3D's driver-independent DRI and DRM winsys code.
                    Additionally, the code can be beneficial to Wayland's PRIME support with DMA-BUF import/export handling and also DRI3.
                    So these patches are actually beneficial for the community, which blows your whole argument right out of the water.

                    Comment

                    Working...
                    X