Commit graph

12 commits

Author SHA1 Message Date
Pierre-Loup A. Griffais
4ca73da9f6 vblankmanager: initial system to let steamcompmgr know when to repaint
For now just mark one repaint per refresh interval without any particular
alignment, as that can be hard to get in nested. We'll use the display
timing extensions for that later.

We'll have it align properly in embedded mode soon.
2020-01-26 14:59:51 -08:00
Pierre-Loup A. Griffais
4079d0c785 steamcompmgr: rework a couple of things to plan for decoupling paints
- Add a commit wait thread that just waits for the work associated with
   an in-bound dma-buff is done. We'd like to be getting a fence directly
   from Wayland, but in the absence of that we'll make our own by tapping
   into implicit-sync, DX11-style. gpuvis logging of the outcome of the
   waits in the thread are consistent with the end of GPU work.

 - Rework wayland commit handoff between wlserver and steamcompmgr to let
   steamcompmgr have full access to the commit queue, letting it remove
   elements in the middle and still keep going, compared to the previous
   interface. This lets steamcompmgr get one commit per window per loop,
   for now. We should be able to kick off waits and ref/unref resources
   for all pending commits soon, with some work on the accounting first.

This should also fix an issue introduced recently where the overlay
could get really starved in embedded mode, as we were just getting
the first commit in the queue for a single frame.
2020-01-23 00:52:17 -08:00
Pierre-Loup A. Griffais
7b2220003c Let it take a command to run in its DISPLAY after "--". 2020-01-18 17:14:49 -08:00
Pierre-Loup A. Griffais
90fa0770e8 nested: plumb some more output options.
Output window size, scaling filter, border.

TODO: make output dynamically resizeable, and some settings reloadable.
2020-01-18 16:05:19 -08:00
Pierre-Loup A. Griffais
c70e71d1e2 I think I've come to terms with not building wlr stuff as C++.
Might as well clean up some of the initial madness.
2020-01-01 15:30:38 +09:00
Pierre-Loup A. Griffais
bfd02982ab Kill EGL with fire, rest of GL will follow shortly.
Working nested Vulkan path now using surface from SDL2 window.

For some definition of working, tile corruption there too.
2020-01-01 15:30:37 +09:00
Pierre-Loup A. Griffais
2be0bdbcbb Start of Vulkan compute rendering code. Getting FBIDs (glxgears still OK). 2020-01-01 15:30:37 +09:00
Pierre-Loup A. Griffais
bc7207c69d More DRM flipping/importing, getting half of glxgears on screen.
Lower half is getting eaten by something, presumably underflow as it's black.

Make EGL and DRM coexist; make ensure_win_resources safe to call regardless
of what resources we want, so outer code doesn't have to care.
2020-01-01 15:30:37 +09:00
Pierre-Loup A. Griffais
77d0e72c69 Make commandline coexist across both sides, plumb nested size control. 2020-01-01 15:30:37 +09:00
Pierre-Loup A. Griffais
95e132179d Chop off some more concepts that were higher-level than I thought initially. 2020-01-01 15:30:37 +09:00
Pierre-Loup A. Griffais
b39c5a67d4 Hook up steamcompmgr output pipeline.
Get steamcompmgr's GL context from EGL and rip out GLX, as we need EGL to
import dma-bufs.

When new buffers are committed to surfaces, push their information to
steamcompmgr (without any feedback for now, fire-and-forget).

Can render games now.
2020-01-01 15:30:37 +09:00
Pierre-Loup A. Griffais
86efc186c5 Merge with rootston.
Enable C++ for bits of the code, but nothing that includes wlroots as that
doesn't want to build as C++.

Enable SubstructureRedirectMask mode and handle MapRequest for now; this lets
us get the WL_SURFACE_ID ClientMessages.

This uses a build of wlroots that has its internal XWM ripped out.
2020-01-01 15:30:37 +09:00