Commit Graph

2115 Commits

Author SHA1 Message Date
Robert Mader
52c4b85161 cogl: Turn CoglScanout into an object
We need an object to hold additional scanout related information, such
as scaling and positioning data. Turn CoglScanout into such an object,
moving the interface into CoglScanoutBuffer.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3177>
2024-02-10 17:41:12 +00:00
Carlos Garnacho
0eb9c43908 wayland: Trigger seat focus change on popup changes
Popups were missing the "input focus" unification in the pointer
seat, triggering MetaWaylandKeyboard focus changes underneath. On
one hand this missed moving all associated focus with it, on the
other hand this made keyboard and global input focus get out of
sync, and bring funky behavior like keyboard focus loss after
dismissing popups.

Fixes: 7b232d9f65 ("wayland: Keep track of the "input focus" on MetaWaylandSeat")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3256
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3568>
2024-02-08 11:32:31 +00:00
Robert Mader
a422ae022b wayland/buffer: Move scanout lifetime handling
There doesn't seem to be a good reason to keep this code in
`MetaWaylandSurface`. Moving it to `MetaWaylandBuffer` cleans things
up and will allow us to tread buffers differently depending on their
type.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3559>
2024-02-02 17:52:30 +01:00
Robert Mader
03a98343d2 wayland/dma-buf: Take MetaWaylandBuffer in try_acquire_scanout
And do some minor variable order cleanup while on it.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3559>
2024-02-02 17:40:37 +01:00
Sebastian Wick
fbf68bf88c crtc-kms: Get rid of meta_crtc_kms_supports_format
Instead get the assigned primary plane and use
meta_kms_plane_is_format_supported.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3512>
2024-01-31 00:58:29 +00:00
Sebastian Wick
60fd80b616 crtc-kms: Get rid of meta_crtc_kms_get_modifiers
Instead get the assigned primary plane and use
meta_kms_plane_get_modifiers_for_format.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3512>
2024-01-31 00:58:29 +00:00
Carlos Garnacho
381048cf90 wayland: Restore IM state flushing before handling key events
The code that maybe flushed IM state before processing a key event
became ineffective at commit 7716b62fa2, since the handle_event()
method on MetaWaylandTextInput won't handle key events, only IM
events and touch/button press events causing IM state to be
committed. Basically, the events that directly change the IM state.

Move this ineffective code to the the filter_event() method handling
the key presses in order to let the IM maybe filter them, and handle
them so that any key event that is let through (both key events
previously injected by the IM, and key events that the IM chooses to
ignore) will ensure that the pending IM state is flushed before the
key event is handled and emitted to the client.

This brings back lost guarantees of orderly event emission when IMs
alternate key events and IM actions.

Fixes: 7716b62fa2 ("clutter: Separate ClutterInputFocus event processing and filtering")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3090
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3536>
2024-01-27 10:18:40 +00:00
Dor Askayo
b8deb4caa0 wayland: Emit frame callbacks when the frame is pending presentation
When Wayland clients send commits without a buffer attached ("empty"
commits), they may lead to stage updates that do not result in any
frame being submitted for presentation ("empty" updates).

Due to how frame scheduling is handled, there can be many such
"empty" updates in a single refresh cycle. If frame callbacks were
emitted after each of these "empty" updates, and if the client
sending "empty" commits was using frame callbacks to throttle the
same logic that results in these "empty" commits being sent, it would
result in a feedback loop between Mutter and the client where the
client would send "empty" commits and Mutter would reply almost
immediately with a frame callback causing the client to send "empty"
commits continuously.

As such, when an "empty" update is detected, frame callbacks are
scheduled to be emitted only once in every refresh cycle, avoiding the
feedback loop.

When a "non-empty" update is detected, frame callbacks are instead
emitted immediately to allow clients to draw their next frame as soon
as possible. It is safe to emit frame callbacks in this case because
the frame for the current refresh cycle is already "finalized" and
that any commit sent by the client at that point would only be handled
in a future refresh cycle.

To implement this, the previous logic had used
meta_frame_native_had_kms_update() to detect "non-empty" updates,
assuming that those would always result in a KMS presentation with the
native backend.

However, this approach misses the fact that virtual monitors do not
use KMS, and as such do not result in KMS presentation even for
"non-empty" updates. As a result, frame callbacks would not be emitted
immediately, resulting in unintended throttling of client rendering.

Instead, assume that it is safe to emit frame callbacks immediately
whenever an update results in the frame clock waiting to be notified
of presentation, since this is also when commits sent by clients are
scheduled to be handled in a future refresh cycle.

This issue was mostly hidden because frame callbacks would be sent
immediately when the target presentation time for the frame had
changed compared to the previous frame. However, this behavior was
removed in 26d8b9c69 ("wayland: Remove unnecessary dispatch of frame
callback source"), exposing the issue.

Fixes: a7a7933e0 ("wayland: Emit frame events in GSource after "empty" updates")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3263
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3549>
2024-01-26 22:22:52 +00:00
Dor Askayo
3e4a330ae7 clutter/frame-clock,wayland: Calculate frame deadline during scheduling
Calculate the frame deadline in ClutterFrameClock's
calculate_next_update_time_us() rather than in MetaWaylandCompositor's
on_after_update().

The specifics of the deadline calculation for a given frame should be
implementation detail of the frame clock and and remain internal to
allow extensibility.

This extensibility is specifically useful for scenarios where a
different deadline calculation is needed due to alternative frame
scheduling logic, such as for VRR.

No change in behavior.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3521>
2024-01-22 15:39:48 +00:00
Dor Askayo
d5c2f20d55 wayland: Flatten logic in on_after_update()
Slightly improves the cognitive complexity of the function.

No change in behavior.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3521>
2024-01-22 15:39:48 +00:00
Dor Askayo
26d8b9c69b wayland: Remove unnecessary dispatch of frame callback source
To avoid communicating lower frame rate to clients through frame
callbacks, it is important to avoid delaying the source dispatch when
a dispatch is already scheduled.

To that end, the previous logic would emit pending frame callbacks
immediately in case a source dispatch was still scheduled for the
previous refresh cycle and then (potentially) schedule another source
dispatch for the current refresh cycle.

However, emitting pending frame callbacks immediately would send
frame events for every pending frame callback, including for the
current "empty" update. Scheduling another source dispatch for the
current cycle was then unnecessary and potentially undesirable
because there may not even be another "empty" update during the cycle.

Instead, let the already-scheduled source dispatch handle emitting any
pending frame callbacks, and do not schedule an additional source
dispatch for the current cycle as it may not be needed.

This approach is useful because it removes an implicit assumption
that the refresh rate is fixed and that target presentation time
remains constant within a refresh cycle. This assumption does not
apply for VRR.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3521>
2024-01-22 15:39:48 +00:00
Dor Askayo
8c88dbfbe8 wayland: Rename source_ready_time_us to frame_deadline_us
The value of this variable represents the last point in time in
which an update would be allowed to scheduled for the given frame.

Rename it for clarity and in preparation for the next commits.

No change in behavior.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3521>
2024-01-22 15:39:48 +00:00
Dor Askayo
83713a38bb wayland: Remove unreachable condition
The value returned from clutter_frame_get_target_presentation_time()
is always same as the value returned from
clutter_frame_get_min_render_time_allowed() when they are called
consecutively because both functions effectively return the value of
frame->has_target_presentation_time. This is with the assumption
that this variable is only ever modified by the same thread that
also executes on_after_update().

As such, a case where the former returns FALSE after the latter
returned TRUE is not possible, which means the line that sets
"target_presentation_time_us = 0;" is effectively unreachable.

Acknowledging this fact allows the call to
clutter_frame_get_target_presentation_time() to be moved outside the
"else" case and into the "if" condition itself. This is done in
preparation for the next commits.

No change in behavior.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3521>
2024-01-22 15:39:48 +00:00
Robert Mader
c19eef3f5e wayland/dmabuf: Bump protocol version to 5
It adds the following clarification:
```
Starting from version 5, the invalid_format protocol error is sent if
all planes don't use the same modifier.
```

We already send an error, just the wrong one.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3450>
2024-01-22 14:20:30 +00:00
Robert Mader
8b1015a5c7 wayland: Switch to stable linux-dmabuf protocol
It is the first protocol to keep the `-vN-` despite being declared
stable.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3450>
2024-01-22 14:20:30 +00:00
Carlos Garnacho
7b232d9f65 wayland: Keep track of the "input focus" on MetaWaylandSeat
This is the unified focus (key, IM, pads, ...) for the focus window.
Just like MetaWaylandPointer and others keep track of the "current"
surface, this is the "current" surface for those (not necessarily
the focused surface, e.g. in the case of compositor grabs).

Since this unified focus will exist regardless of keyboard
capabilities (e.g. even if just for "logical" focus like IM/clipboard
that does not depend on input devices), it does not make sense
to trigger a focus sync on keyboard capability changes, the focus
is staying the same, we however need to focus the keyboard interface
to the already existing focus when the capability is enabled.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3511>
2024-01-12 14:39:17 +00:00
Carlos Garnacho
962eb9e054 wayland: Hook focus synchronization to MetaDisplay signals
Instead of letting the MetaDisplay be aware of the Wayland compositor,
and take care of updating its focus. This makes the MetaWaylandCompositor
able to track focus changes by itself, using MetaDisplay as the source
of truth.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3511>
2024-01-12 14:39:17 +00:00
Carlos Garnacho
a2d2e04d80 wayland: Use MetaWaylandCompositor API to drive focus synchronization
Keep this within the wayland code itself, and avoid poking MetaDisplay
API that will go away.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3511>
2024-01-12 14:39:17 +00:00
Carlos Garnacho
9383171958 wayland: Move Wayland focus synchronization code out of core
Handle focus synchronization in MetaWaylandCompositor itself. This
is so far plumbed so that MetaDisplay still drives focus synchronization
directly.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3511>
2024-01-12 14:39:17 +00:00
Bilal Elmoussaoui
94f9d88371 x11: Drop error trap helpers
They are no longer that useful as they end up calling
mtk functions nowadays

Followup of https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3230

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3483>
2024-01-10 13:58:18 +00:00
Bilal Elmoussaoui
7d6e7773bf window: Make Window.set_transient_for a vfunc
So we can move the xgroup_leader to WindowX11. See next commit

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3211>
2024-01-09 18:59:36 +00:00
Bilal Elmoussaoui
bc9cd123e9 window: Move xwindow to WindowX11
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3211>
2024-01-09 18:59:36 +00:00
Bilal Elmoussaoui
d98b0eb71e window: Move xvisual to WindowX11
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3211>
2024-01-09 18:59:36 +00:00
Bilal Elmoussaoui
2a75661883 region: Move RegionBuilder to Mtk
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3501>
2024-01-09 18:47:34 +00:00
Bilal Elmoussaoui
fced59b33d region: Make Region.transform private
It is only used once in MetaWaylandSurface

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3501>
2024-01-09 18:47:34 +00:00
Bilal Elmoussaoui
4d53e4d156 region: Move Region.crop_and_scale to Mtk
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3501>
2024-01-09 18:47:34 +00:00
Bilal Elmoussaoui
6e7d314e75 region: Move Region.scale to Mtk
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3501>
2024-01-09 18:47:34 +00:00
Bilal Elmoussaoui
6f9e75b6f2 boxes: Move Rectangle.is_adjacent_to to Mtk
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3501>
2024-01-09 18:47:34 +00:00
Bilal Elmoussaoui
346e9a7f2c docs: Remove remaining SECTIONs
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3489>
2024-01-09 15:23:04 +00:00
Bilal Elmoussaoui
5964ccb385 wayland: Fix building without native backend
Closes https://gitlab.gnome.org/GNOME/mutter/-/issues/1026

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3480>
2024-01-03 14:46:41 +01:00
Bilal Elmoussaoui
cebde5c2d1 backend/native: Move DrmFormatBuf to cogl-drm-formats
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3480>
2024-01-03 14:46:41 +01:00
Sebastian Wick
59d4343057 wayland-buffer: Enable YCbCr support for wayland shm
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3371>
2023-12-20 23:51:14 +00:00
Sebastian Wick
c8f4c85bfb wayland-buffer: Support MultiTexture formats via wayland shm
Finding the shm offset and shm stride for each plane is the main issue.
The rest is just creating multiple textures for each plane.

One assumption is that shm planes are always contiguous in memory so the
next plane comes directly after the size of the current plane.

The size of a plane is determined by the height and stride. There is
only a single stride parameter for shm buffers but we assume that the
first plane is always non-subsampled which gives us a number of "logical
elements" on one line (stride / bpp of the first plane). The stride of
the other planes is then the number of logical elements devided by the
subsampling factor and multiplied by the bpp of the plane.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3371>
2023-12-20 23:51:14 +00:00
Alynx Zhou
33088d59db wayland/text-input: Pass char based offset to ClutterInputFocus
Wayland's text-input-v3 uses byte based offset for cursor and anchor of
surrounding text, but Clutter needs char based offset here. This commit
converts byte based offset to char based offset before passing them to
ClutterInputFocus.

Fixes <https://gitlab.gnome.org/GNOME/mutter/-/issues/3102>.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2712>
2023-12-20 19:31:05 +00:00
Alynx Zhou
27bdf0c577 wayland/text-input: Use byte based offset in delete_surrounding_text
ClutterInputFocus/GtkIMContext uses char based offset for
delete_surrounding, however, text_input_v3 uses byte based offset for
it. Currently only GTK with mutter can work correctly via text_input_v3
because they both forget to convert between char based offset and byte
based offset.

This commit fixes it in mutter by saving committed surrounding text in
MetaWaylandTextInput and converting char based offset to byte based
offset with the UTF-8 encoded surrounding text.

Fixes <https://gitlab.gnome.org/GNOME/mutter/-/issues/2146>.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2712>
2023-12-20 19:31:05 +00:00
Michel Dänzer
14d84707fc wayland/surface: Move committed buffer size check out of if (buffer)
A surface commit may change the buffer scale but not attach a new
buffer. In that case, the size of the previously attached buffer needs
to be consistent with the new buffer scale.

Fixes: 7649e2f3ab ("wayland/surface: Move buffer size check to meta_wayland_surface_commit")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3466>
2023-12-19 11:53:26 +01:00
Michel Dänzer
17640d9c98 wayland/surface: Use committed texture width/height in _surface_commit
meta_wayland_surface_get_buffer_width/height uses the currently applied
buffer, which may have a different size.

Fixes: 7649e2f3ab ("wayland/surface: Move buffer size check to meta_wayland_surface_commit")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3466>
2023-12-19 11:23:22 +01:00
Michel Dänzer
98c8c03729 wayland/surface: Check buffer size also for role-less surfaces
The Wayland protocol spec doesn't say anything about this being required
only for surfaces with a role.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3463>
2023-12-18 21:56:59 +00:00
Michel Dänzer
7649e2f3ab wayland/surface: Move buffer size check to meta_wayland_surface_commit
Multiple reasons:

* More consistent with the protocol spec language.
* Ensures the size is checked and the protocol error sent from a
  protocol processing context, instead of whatever context
  meta_wayland_surface_commit might get called from.
* The latter implies that surface->resource is guaranteed to be valid.

Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3211
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3463>
2023-12-18 21:56:59 +00:00
Michel Dänzer
0a0d1e2066 wayland/surface: Track committed and applied scale separately
Preparation for next commit, no functional change intended.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3463>
2023-12-18 21:56:59 +00:00
Carlos Garnacho
a36616f81d core: Drop focus_frame argument from meta_display_set_input_focus()
Sort that out in the X11 display, where it matters, without the need
of this argument in generic API.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3269>
2023-12-18 17:55:27 +00:00
Florian Müllner
5fc6ff19fa wayland/xwayland: Include x11-display-private
The include is currently satisfied by

window-x11-private → iconcache → x11-display-private

The icon cache is about to be removed, so add the missing include
directly.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3452>
2023-12-15 14:14:33 +00:00
Michel Dänzer
50980f995b tree-wide: Fix spelling of "inhibition"
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3444>
2023-12-06 16:01:36 +00:00
Michel Dänzer
200fc76f14 wayland/idle-inhibit: Pass actor instance in on_actor_changed
This matches the other g_signal_connect / g_clear_signal_handler calls
for these signal handlers.

Fixes: a3c62bf8aa ("wayland/idle-inhibit: Add state tracking to fix races")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3444>
2023-12-06 16:01:36 +00:00
Michel Dänzer
212283c13b wayland/idle-inhibit: Handle NULL inhibitor->actor
Fixes: a3c62bf8aa ("wayland/idle-inhibit: Add state tracking to fix races")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3444>
2023-12-06 16:01:36 +00:00
Michel Dänzer
c47730894e wayland/dma-buf: Prefer sync_file for dma-buf readiness source
Similarly to https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3389,
this avoids over-synchronization due to the client reading from the
buffer, e.g. if drawing a frame re-uses the previous frame contents.

v2:
* Rename owned_fd → owned_sync_fd (Sebastian Wick)
* Put curly braces around do/while loop

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3408>
2023-12-05 14:53:21 +00:00
Michel Dänzer
0e471bcea7 wayland/dma-buf: Refactor create_source helper
No functional change intended.

v2:
* Make local source assignment separate from its declaration
  (Ivan Molodetskikh)

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3408>
2023-12-05 14:53:21 +00:00
Sebastian Wick
97c1d19017 wayland-buffer: Use MetaFormatInfo to find supported formats
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3280>
2023-11-22 12:11:38 +00:00
Sebastian Wick
dbdc821ca7 multi-texture-format: Expose immutable fields directly
Split the struct into mutable and immutable parts. Access the mutable
parts via getters and the immutable parts via a single struct. This
avoids copying around the immutable parts.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3280>
2023-11-22 12:11:38 +00:00
Sebastian Wick
6f1edfc776 common/drm-formats: Rename to MetaFormatInfo and simplify API
Every format kind has it's own function to search for a MetaFormatInfo
from which contains all the information.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3280>
2023-11-22 12:11:38 +00:00