Commit graph

216 commits

Author SHA1 Message Date
Simon Ser
3d9f330ee0 scene: add scene_buffer_set_texture() 2024-03-14 22:06:10 +01:00
Alexander Orzechowski
86e1b24d84 wlr_scene: Track damage of null textures
If we hit this case, we effectively failed to render something, this might
be because a texture failed to upload or the texture is momentarily
unavailable after a GPU reset. If we fail to render, we have to continue
to track damage for the next frame in hopes that the texture becomes
available then.

An alternative approach would be to fail the commit completely if we
find this case, but in the case of gpu resets, clients may not commit
a new buffer for a while, and a frozen display does not help.

This fixes damage tracking issues after a gpu reset.
2024-03-13 07:53:33 +00:00
Alexander Orzechowski
2aebb8e444 wlr_scene: Fix crash when committing NULL buffer
If we commit a null buffer, we want to avoid the buffer damage path and
instead just update the node to avoid it.
2024-03-13 07:50:54 +00:00
Alexander Orzechowski
8da93e5936 wlr_scene: Add support for custom swapchains 2024-03-12 11:26:03 -04:00
Alexander Orzechowski
239354d644 wlr_scene: Decide when to commit output damage with buffers in damage ring
We will soon support custom swapchains. In order to track output damage
we should instead use the damage_ring which will hold all the buffers
we are currently tracking anyway across an arbitrary amount of swapchains.
2024-03-12 11:25:58 -04:00
Alexander Orzechowski
8220184a72 wlr_scene: Assert that buffer returned from swapchain is the right size 2024-03-12 11:16:30 -04:00
Simon Ser
c79ed0706d scene: simplify scene_node_update() condition when setting buffer 2024-03-12 12:11:35 +01:00
Alexander Orzechowski
2d97935e82 wlr_scene: Call surface enter scale events on surface create
We would fail to call scene_node_update() which would then call output
events for us. We need to make sure to update the node when we first map
a buffer, as the comment explained.
2024-03-12 01:57:07 -04:00
Simon Ser
431e8a7fd7 scene: ignore duplicate wlr_scene_output_layout_add_output() calls
An assert was added in [1] to avoid tracking the same output multiple
times. However, this is cumbersome for compositors [2]: they need to
add a special check for this. Additionally, this is inconsistent with
wlr_output_layout_add().

[1]: https://gitlab.freedesktop.org/wlroots/wlroots/-/merge_requests/4380
[2]: https://github.com/cage-kiosk/cage/pull/315
2024-02-20 11:04:46 +01:00
Simon Ser
842093bb84 Define _POSIX_C_SOURCE globally
Stop trying to maintain a per-file _POSIX_C_SOURCE. Instead,
require POSIX.1-2008 globally. A lot of core source files depend
on that already.

Some care must be taken on a few select files where we need a bit
more than POSIX. Some files need XSI extensions (_XOPEN_SOURCE) and
some files need BSD extensions (_DEFAULT_SOURCE). In both cases,
these feature test macros imply _POSIX_C_SOURCE. Make sure to not
define both these macros and _POSIX_C_SOURCE explicitly to avoid
POSIX requirement conflicts (e.g. _POSIX_C_SOURCE says POSIX.1-2001
but _XOPEN_SOURCE says POSIX.1-2008).

Additionally, there is one special case in render/vulkan/vulkan.c.
That file needs major()/minor(), and these are system-specific.
On FreeBSD, _POSIX_C_SOURCE hides system-specific symbols so we need
to make sure it's not defined for this file. On Linux, we can
explicitly include <sys/sysmacros.h> and ensure that apart from
symbols defined there the file only uses POSIX toys.
2024-02-15 15:41:12 +01:00
Paul Cercueil
220df2aa0f
Add more POSIX compliance macros to fix uClibc support
- Add POSIX 1993.09 compliance macro in source files that use
  "struct timespec";
- Add POSIX 2001.12 compliance macro in source files that use
  "struct sigaction" and the SA_SIGINFO macro, or the fchmod()
  function;
- Add POSIX 2008.09 compliance macro in source files that use the
  getline() function.

These compliance macros are enough for wlroots to compile with the
git-master version of uClibc-ng.

Signed-off-by: Paul Cercueil <paul@crapouillou.net>
2024-02-12 17:41:45 +01:00
Simon Ser
4ec901bb33 scene: release shm buffers once uploaded
Once a shm buffer is uploaded there's no reason for wlr_scene_buffer
to keep it locked any further. Allow the compositor to re-use it.
2024-01-27 12:21:54 +00:00
Simon Ser
ab924064f2 scene: drop wlr_scene_set_presentation() 2023-12-27 13:41:20 +01:00
Simon Ser
d68ba9d6c2 presentation-time: drop wlr_presentation arg
There can only be a single presentation-time global advertised to
clients, this it's unnecessary to pass around the wlr_presentation
pointer.
2023-12-27 13:41:01 +01:00
Alexander Orzechowski
62169bf5ac wlr_scene: fix transformed outputs backend damage clearing
When we cleared the pending backend damage when the output committed,
we would not take into account the output transform. It's easiest to fix
this by just changing pending_commit_damage to always have transformed
coordinates.
2023-12-04 07:06:35 -05:00
Alexander Orzechowski
364d7cba42 wlr_scene: Apply output damage once
We would apply it twice if we failed direct scanout. Once when we attempt
direct scanout and again when we composite.
2023-12-04 06:58:35 -05:00
Alexander Orzechowski
fe8916fef0 wlr_scene: Don't damage whole damage ring when exiting direct scanout
Direct scanout damage will just accumulate on the damage ring while in
direct scanout and properly damage when we exit anyway. On the flip side
since we now manage backend damage submission ourselves, this won't break
that either.
2023-12-03 05:29:05 +00:00
Alexander Orzechowski
0619c99dc7 wlr_scene: Use wlr_damage_ring_rotate_buffer()
Compositors who use _build_state are no longer required to call
damage_ring_rotate themselves. This is a minor breaking change.
2023-12-03 05:29:05 +00:00
Alexander Orzechowski
06d301bf1a wlr_scene: Track output damage separately
The damage ring is only supposed to keep track of damage within
the swapchain
2023-12-03 05:29:05 +00:00
Simon Ser
56ec13596a Cleanup wlr_matrix.h includes
Many files used to require wlr_matrix but no longer do.
2023-11-25 08:37:43 +01:00
Simon Ser
2eb225236e util/transform: add wlr_output_transform_coords()
We hand-roll this in multiple places.
2023-11-23 11:07:34 +01:00
Simon Ser
9e702e9cfe util/transform: move over wl_output_transform helpers
These aren't really tied to wlr_output.
2023-11-23 11:03:57 +01:00
Alexander Orzechowski
303f23d7dd wlr_scene: Update outputs if subpixel status changes
This is important for buffers that are meant to render any sort of
text.
2023-11-15 18:42:12 -05:00
Alexander Orzechowski
50446fe75e wlr_scene: Update outputs when primary output is changed
Helpers care about the primary output. They need to know when it changes
even if intersection status has not changed necessarily.
2023-11-15 18:42:12 -05:00
Alexander Orzechowski
889c5ed5ff wlr_scene: Update outputs when output scale/transform changes
We want to call the outputs updated signal when an output scale or transform
changes. Otherwise helpers like the scene surface helpers will not be
notified of scale changes and not pass them to clients.
2023-11-15 18:38:12 -05:00
Alexander Orzechowski
3d3cca9164 wlr_scene: Introduce wlr_scene_subsurface_tree_set_clip 2023-11-14 17:27:08 +00:00
Alexander Orzechowski
d1ddd4ca3a wlr_scene_surface: Introduce scene_surface_set_clip 2023-11-14 17:27:08 +00:00
Alexander Orzechowski
c6d356648c set_buffer_with_surface_state: Take whole surface struct 2023-11-14 17:27:08 +00:00
Alexander Orzechowski
1c0b10b61a wlr_scene_subsurface_tree: Addon to scene
We will need this later so that we can determine the subsurface tree
that owns any given scene node.
2023-11-14 17:27:08 +00:00
Alexander Orzechowski
33b437d574 wlr_scene: Amend scene_buffer.point_accepts_input to take coordinate pointers
The pointers mean that we can mutate them. This will be useful later
so we can hide details from the compositor when we clip subsurface trees.
2023-11-14 17:27:08 +00:00
Alexander Orzechowski
291431c14f scene/output_layout: Add assert for duplicate output insertion
Ensure that the output was not added multiple times to the scene
output layout.
2023-10-08 11:18:11 -04:00
Kirill Primak
b06c2f3d1f scene-output-layout: assert lo->output == so->output 2023-10-08 13:11:29 +03:00
Alexander Orzechowski
e1e911d425 scene_output_layout: Don't destroy output when output layout is destroyed
There were a couple of problems with this:
1. The behavior is unexpected. Typically objects in wlroots won't
also destroy objects that they depend on. For instance, wlr_scene_output
will not destroy the wlr_output when it's destroyed. It shouldn't be any
different for scene layouts.
2. This fixes a crash where because wlr_output_layout and wlr_scene_output
are both addons to wlr_output, we might get into a situation where
wl_list_for_each_safe might malfunction. See [1]

This means that the compositor needs to manually destroy the output
when they destroy the layout, hence ~breaking. Compositors can just
call `wlr_scene_output_destroy()` if they want to destroy both at the
same time.

[1] https://gitlab.freedesktop.org/wlroots/wlroots/-/merge_requests/4358#note_2106260

This reverts commit 1a731596c5.

Co-authored-by: Kirill Primak <vyivel@eclair.cafe>
2023-10-07 13:31:20 -04:00
Alexander Orzechowski
5fb0007e02 output_event_commit: Remove committed and buffer
The newly introduced state struct can be used to retrieve this.
2023-10-06 10:06:01 +00:00
Alexander Orzechowski
1b0694b794 treewide: Migrate from sizeof(struct) to sizeof(*pointer) where practical 2023-10-03 01:51:07 -04:00
Leo Li
b18a849fc3 wlr_scene: Use cached node coordinates where possible
Avoids recalculating node coordinates.
2023-09-08 21:11:19 +00:00
Rose Hudson
f5917f0247 scene_output_layout: make output adding explicit
Closes: https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3709
2023-08-22 11:53:28 +02:00
Kirill Primak
10ba8ebc70 Don't assume xdg_surface.{toplevel,popup} is non-NULL
This assumption will become incorrect with future commits.
2023-08-21 16:30:29 +00:00
Alexander Orzechowski
fd540f6d07 wlr_scene: Don't damage when setting opaque region
Opaque region is a optimization hint, (bugs outstanding) it will not
change the output contents, therefore damage does not need to be submitted.

However, we still need to update the visibility state of the other
nodes in the tree. To do this call scene_update_region() by ourselves
but not `scene_node_update()` which will damage the outputs.
2023-07-22 19:23:36 -04:00
Alexander Orzechowski
a32180afa7 wlr_scene: Fix damage tracking with non atomic opaque region configuration
We need to intersect the opaque region with the node size or else we'll
get damage tracking effects with compositors attempting to use
wlr_scene_buffer_set_opaque_region() along with resizing the buffer
at the same time in a certain order.

Consider this: I have a new buffer that I want to commit to my scene buffer
that is smaller than the old one. However, I still have the old opaque
region that is the size of the old larger buffer, so that means that
for the small moment between when we reconfigure the opaque region for the
new buffer the opaque region will be oversized. Scene logic will then
try to apply occluding optimizations outside of the node boundaries
causing damage artifacts.
2023-07-22 19:23:36 -04:00
Ronan Pigott
b18c76642a scene: apply rounding when translating buffer damage
Without a round in this case the damage region is translated to truncated coordinates, potentially
misplacing it relative to the actual position of that region in the output buffer.
2023-07-16 02:55:09 -07:00
Leonardo Hernández Hernández
4de9c9c47f scene: send wl_surface.preferred_buffer_scale 2023-07-15 10:28:25 +00:00
Simon Ser
67447d6cb4 presentation-time: add separate helper for zero-copy
The backend is not able to tell whether a surface is being
presented via direct scan-out or not. The backend will set
ZERO_COPY if the buffer submitted via the output commit was
presented in a zero-copy fashion, but will no know whether the
buffer comes from the compositor or the client.
2023-07-14 15:20:35 +02:00
Simon Ser
fe84bfc8d9 scene: add wlr_scene_output_sample_event
This contains the output and a bool indicating direct scan-out.
2023-07-14 15:16:12 +02:00
Simon Ser
88942d43fd scene: rename output_present event to output_sample
Using "present" is confusing here: the event is emitted when the
buffer is being sampled to be displayed on an output, not when it's
being presented on-screen.

Rename to match the presentation-time terminology.
2023-07-14 15:06:13 +02:00
Alexander Orzechowski
d8585d661c wlr_scene: Fix buffer damage for output state
We would transform the damage based on output state which might be
out of date.
2023-07-12 17:20:12 +00:00
Simon Ser
fe06e5f49a Use wl_container_of() instead of casts
This slightly improves type safety.

The culprits were found with:

    git grep -E '\([a-z0-9_ ]+ \*\)\W?[a-z]'
2023-07-11 20:16:17 +02:00
Rose Hudson
214df8eda0 scene_output: optionally record and report timings 2023-07-10 09:57:02 +00:00
Simon Ser
7a9f8d8d6b Use struct initializers instead of memset()
This is a bit more type-safe.
2023-07-07 17:31:11 +02:00
Alexander Orzechowski
57a0313779 wlr_scene: Don't recompute node location during rendering
We already compute it during the render list creation, we should cache
it and re-use the value.
2023-07-01 09:48:25 +00:00