Announcement

Collapse
No announcement yet.

The First Wayland Benchmarks From Fedora 20 Show Great Promise

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

  • The First Wayland Benchmarks From Fedora 20 Show Great Promise

    Phoronix: The First Wayland Benchmarks From Fedora 20 Show Great Promise

    Since last week it's been possible to run the GNOME Shell on Wayland with Fedora 20. The user-experience isn't yet refined and easy, but Linux enthusiasts can easily get a GNOME 3.10 session running on Wayland for testing purposes using F20 packages. In this article are the first graphics benchmarks from Fedora 20 when running GNOME 3.10 on Wayland with XWayland support and then from running a clean X.Org Server.

    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
    Thats the (part of) Linux community looking at things.

    For Mir it is "regression"
    For Wayland it is "promise"

    Comment


    • #3
      Performance is doing much worse than XMir, which isn't very promising at all. Performance hit is over 10% in Xonotic in comparison to XMir from a month ago which only had a 5% hit.

      Comment


      • #4
        Originally posted by verde View Post
        Thats the (part of) Linux community looking at things.

        For Mir it is "regression"
        For Wayland it is "promise"
        Yes we got few fps less than pure Xorg, but the overall desktop is much smoother because they are not as idiot to run a whole Xorg running a DE inside a Mir that does practically nothing but enabling security issues.

        Comment


        • #5
          I don't know for Gnome's XWayland window manager, but for Weston XWayland window manager, we don't bypass compositing at all yet.

          That explains the overhead due to useless copies (X compositing), and that should be solved in the near future.

          Comment


          • #6
            So, Are the steam linux games going to be able to run in wayland? How's that going to be handled?

            Comment


            • #7
              XWayland should allow to run any games that ran under X.

              Currently a few issues (slightly slower performance, tearings, popup windows when fullscreen not showing, a few input region issues) would prevent this to be a good experience.

              Comment


              • #8
                And just like XMir, almost all of the performance regressions here can be solved via Composite Bypassing, since they are all full-screen games.

                I'm more interested in the small things: application start up times (gedit and other applications were shown to have a large amount of time spent waiting on the X Server), smoothness when dragging windows around, etc.

                Also, this is just GNOME's implementation of a Wayland Compositor. By the time KDE's compositor comes out, GNOME might have fixed theirs. And if not, maybe KDE's will be better. Who knows...

                Comment


                • #9
                  Originally posted by mannerov View Post
                  XWayland should allow to run any games that ran under X.

                  Currently a few issues (slightly slower performance, tearings, popup windows when fullscreen not showing, a few input region issues) would prevent this to be a good experience.
                  I just hope than wayland support will be as easy as a simple new feature in an update of the game, and not to port the games almost from 0 again to be able to play it in wayland, in that case, valve would not be happy, because all their work is in the currently state of linux (and the same for the other people than had ported games to linux).
                  But if the "slightly slower performance" problem could be fixed, they would not even be bothered to port to wayland the games already released.
                  Last edited by edoantonioco; 07 October 2013, 03:11 PM.

                  Comment


                  • #10
                    Originally posted by Daktyl198 View Post
                    And just like XMir, almost all of the performance regressions here can be solved via Composite Bypassing, since they are all full-screen games.
                    Non-full-screen applications will probably run slower.

                    Comment

                    Working...
                    X