When inactive (logind session no longer active, e.g. VT switched to
another user, no longer DRM master) mutter stays dormant and doesn't
try to reconfigure outputs on hotplug events, among other things. That
means we also shouldn't keep leases active, or connectors leasible, when
not active, since dependent state are not kept up to date.
Address this by, treat all connectors as not leasable when inactive,
which effectively means leases are revoked.
This fixes a crash when monitors are hot plugged in a certain way when
the session is inactive, which an added test case tests.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4378>
When we receive a hotplug signal emission, we update the state of leased
and leasable resources. Some of this state depends on the current state
of any potential associated monitor (MetaOutput & MetaMonitor). In order
to have an up to date association of MetaOutput's and MetaMonitor's, we
must only update our own state after MetaMonitorManager has had a chance
to. To achieve this, make sure the MetaUdev::hotplug signal handler is
dispatched after MetaMonitorManager's handler by using
g_signal_connect_after().
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3943
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4378>
The original strategy to let the drag source client be in
charge of pointer cursor feedback during DnD has gotten
way too complex, with tablets and tools, toplevel drags,
and now the cursor shape protocol.
Instead, let the compositor be in charge of pointer cursor
feedback during DnD operation, it will know right away the
cursor renderer to update, and the appropriate feedback
through the current DnD action.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4368>
Specifically, only after checking crtc_state_impl->cursor_invalidated.
If that's false, we bail anyway, so no point getting the current cursor
position, which can get blocked by another thread holding the
seat_impl->state_lock in writer mode.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4377>
When the legacy KMS implementation running without a KMS thread receives
a cursor only frame, it doesn't do an actual page flip, but just posts a
"ready" callback without any timing information. In this case we failed
to discard the posted frame, meaning theh onscreen KMS throttling was
still thinking it was supposed to wait for the posted frame to be
presented.
Fixes: df7ac5b0a3 ("onscreen/native: Account for all posted frames")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/4021
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4375>
What it does now is handling the promotion of a "posted" frame to
"presented", or promoting it to the trash bin, if it was not carrying an
actual primary plane buffer. Promoting to presented effectively still
means "swapping the frm fb" though. Make the name reflect this.
Fixes: df7ac5b0a3 ("onscreen/native: Account for all posted frames")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4375>
meta_wayland_tablet_tool_update_cursor_surface was always creating a new
sprite for the cursor shape, which may result in allocating HW buffers
and uploading the shape to them.
Since meta_wayland_tablet_tool_update_cursor_surface gets called every
time the tablet tool moves, we don't want to repeat that work if the
cursor shape hasn't changed. We cache the sprite in the
MetaWaylandTabletTool struct for reuse next time, and clear the cached
sprite when the cursor shape changes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4371>
meta_wayland_pointer_update_cursor_surface was always creating a new
sprite for the cursor shape, which eventually results in allocating HW
buffers and uploading the shape to them.
Since meta_wayland_pointer_update_cursor_surface gets called every time
the pointer moves, we don't want to repeat that work if the cursor shape
hasn't changed. We cache the sprite in the MetaWaylandPointer struct for
reuse next time, and clear the cached sprite when the cursor shape
changes.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3993
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3996
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4371>
The cache lives on the MetaCursorTracker and thus is shared between all
CursorSpriteXcursors and avoids constantly having to hit the disk when
the cursor shape changes.
We still do a lot of work to get the sprite from the theme into a KMS
plane or drawn into the onscreen, but avoiding disk IO is a first, good
step.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4359>
This happens during shutdown if the cursor is over some clients.
It seems `kms` has lost its cursor manager because the
`prepare-shutdown` signal is emitted to the backend before clients
are closed. While we could just reverse that and close the clients
first, it's not strictly just a "clients" problem. So keeping
the existing shutdown order and just handling when it is NULL seems
like a more universal solution.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2993
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4370>
First instantiate the object, and set the instance pointer in
MetaSeatNative, then initialize it. This is needed due to initializing
libinput may create virtual input devices used for accessibility
features, such as mouse keys, and for this it needs to fetch the
MetaSeatImpl from MetaSeatNative.
Fixes: 5fc60eac ("seat-impl: Keep track of virtual input devices too")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3708
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4374>
The state should go from 'dispatched-one-and-scheduled-later' to
'scheduled-later', not 'scheduled-now' when being notified about a frame
being ready - otherwise we'll dispatch without proper pacing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4334>
The KMS thread handles updates posted asynchronously, but it expects to
only handle one such frame in flight from the compositor at a time. That
means that the triple buffering state tracking in MetaOncreen, that
keeps track of posted frames and when they become presented, must also
account for posted frames that doesn't contain an actual primary plane
pixel buffer.
This was not the case, causing MetaOnscreenNative to post multiple
frames to the KMS thread, which wasn't handled gracefully in certain
situations.
Before the KMS thread grows real support for it's own queue of separate
updates, make sure we keep the contract to the KMS thread in
MetaOnscreenNative, and only submit at most one KMS update for each CRTC
each cycle, even when there are no actual primary plane changes.
v2: Properly handle frame tracking when when KMS update empty
v3: In the page flip callback, only set the presented frame to frames
that has buffers. This is needed on older kernels which doesn't have
drmModeCloseFB() which would otherwise disable the CRTC when presented
frame with an actual buffer would be replaced with an "empty" frame,
causing the frame with the buffer to be released, with the buffer along
with it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4334>
The set of supported color modes of a monitor might change for the same
monitor, for example by the monitor providing different EDID blobs
depending on configuration done on the monitor itself.
When we have a color mode configured that is not actually supported by
the monitor at the moment, amend the configuration and fall back on the
default color mode.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3911
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4364>
MetaWaylandTabletTool is not a GObject thus we cannot use weak pointers
on them. The good news is that MetaWaylandTabletTool structs are perennial
and we do not need to do that.
Remove this unnecessary weak pointer handling, which can only cause
runtime warnings or crashes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4367>
Technically the impl device simple now also supports HDR, but there are
too many things that can go wrong, such as the colorspace prop getting
set, but not the HDR transfer function. For now, we will say that the
simple device impl does not have usable HDR support.
This gets picked up by the MetaOutputKms and propagtes the state to
everywhere.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4357>
While we do not intend to support HDR on the simple device impl, being
able to get to the default color mode requires support for those
properties. Otherwise, if the monitor is already in HDR mode when mutter
starts, we would be stuck there.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4357>
Changing to a cursor shape would set the cursor surface to NULL, so
trying to disable the cursor by setting the cursor surface to NULL was
detected as no-change. This commit fixes the check by taking into
account if the cursor shape is currently set.
Also adds a ref-test for it.
Fixes: 005b969227 ("wayland: Implement the cursor_shape_v1 protocol")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3997
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4358>
Make the prepare function a vfunc of MetaCursorSprite, moving the root
cursor prepare function into MetaCursorSpriteXcursor. This should solve
two issues:
- The root cursor prepare function was changed in f77d8e2a12
to solve an issue with fractional scales. The tool cursor prepare
function was missing this fix.
- The cursors created via the shape protocol had no prepare function at
all, so were not getting scaled.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3975
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4345>
In `should_constraint_be_enabled()` calling
`meta_wayland_surface_get_window()` could return a null window when the
toplevel is reset after locking pointer.
The null check was only done inside `HAVE_XWAYLAND` ifdef, and so if
mutter is built without xwayland it could lead to a crash.
Fixes: 6e818c8c38 ("build: Allow disabling xwayland")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4351>
The test is huge and on a very fast machine it takes 10 seconds. So in
Debian and Ubuntu builds we're finding a lot of machines/architectures
where it regularly requires more than a minute to complete.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4340>
Not only if the deadline timer is enabled. With the next commit, it'll
be semi-expected to happen even if the deadline timer is disabled.
Still leave the warning though, as a reminder that we'd rather prevent
this outside of the KMS thread.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4338>
If crtc->current_state.is_drm_mode_valid happens to be FALSE,
meta_kms_crtc_determine_deadline returns G_IO_ERROR_NOT_FOUND. This
previously resulted in silently and permanently disabling the deadline
timer, which is surprising and undesirable.
Bail early from ensure_deadline_timer_armed in this case instead. In
turn, G_IO_ERROR_NOT_FOUND from meta_kms_crtc_determine_deadline no
longer needs to be treated specially.
Fixes: 7493ed39ce ("kms/impl-device: Avoid retrying a failing deadline timer")
v2:
* Stop treating G_IO_ERROR_NOT_FOUND specially. (Sebastian Wick)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4321>
If ensure_deadline_timer_armed fails and crtc_frame->pending_update is
non-NULL, we need to process the pending update, or it may fall through
the cracks, and this stage view may freeze.
v2:
* Don't return immediately after calling
meta_kms_impl_device_do_process_update. (Sebastian Wick)
v3:
* Call meta_kms_impl_device_do_process_update only if
is_using_deadline_timer returned true, and log a warning if
crtc_frame->pending_update is non-NULL otherwise.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4321>
The KMS thread handles updates posted asynchronously, but it expects to
only handle one such frame in flight from the compositor at a time. That
means that the triple buffering state tracking in MetaOncreen, that
keeps track of posted frames and when they become presented, must also
account for posted frames that doesn't contain an actual primary plane
pixel buffer.
This was not the case, causing MetaOnscreenNative to post multiple
frames to the KMS thread, which wasn't handled gracefully in certain
situations.
Before the KMS thread grows real support for it's own queue of separate
updates, make sure we keep the contract to the KMS thread in
MetaOnscreenNative, and only submit at most one KMS update for each CRTC
each cycle, even when there are no actual primary plane changes.
v2: [Michel Dänzer]
* Drop Closes due to
https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4334#note_2378962.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4321>
No actual changes. We will use this version to hint GTK about
the availability of non broken wl_surface.offset support for
pointer cursors.
GTK can use this version to avoid cursor breakage when running
on older Mutter versions.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4327>
Any feedbacks for the same stage view but older frames than the one
being presented are discarded.
Since we now use per-frame lists of feedbacks, there can no longer be
any stale feedbacks where we previously discarded them, so remove that
dead code.
v2:
* Use ClutterFrameInfo::frame_count.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4306>
This test checks a particular issue, where a pointer moved in a way
resulting in the cursor position not changing, then moving again,
changing the cursor position, not damaging the area of the original
position, causing the cursor to get "stuck" until damaged due to some
other reason.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4331>
The previously painted rectangle of an overlay is not a global state,
but depends on what view it was painted on. There was also an issue
where an overlay being updated but not changing position, e.g. due to a
0,0 pointer movement or an absolute pointer movement with the position
not changing, not properly triggering damage of the old position when it
eventually actually moved.
Fix this by tracking damage per view, while also fixing the state
tracking to handle unchanged positions.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4331>
The thing before pretended to be a map but wasn't, so it required
iterating over it and checking the fields to find the requested one. The
code didn't do that and index into it anyway and because the shapes
start at a value of 1, everything was off.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4307>
The protocol changes for version 2 made it upstream but are not in a
release, yet. As soon as we have a release, we should replace this copy
with the wayland-protocols variant.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4307>
We run some ref-tests in the tty and kms test harnesses which does
support cursor planes. If we get "unlucky", the cursor could end up on
one of them and won't show up in the captured output.
Inhibit the hardware cursor when capturing the output for a ref-test.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4307>
To not depend on the system cursor theme for tests, always use
our own, even if the cursor theme appearance makes no difference.
We also need to add all of Adwaita's cursors in tree, to avoid any test
arbitrarily triggering missing cursor warnings or errors.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4307>
We will start using only the cursors we ship with the tests, which is a
snapshot of Adwaita cursors. left_ptr is not one of the supported ones,
so we use default instead.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4307>
Cogl and Clutter use the "current source dir" whereas the tests in
src/tests use the non-test source directory. The installed tests already
only use a single directory.
This commit changes everything to be src/test in the source and build
directory.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4307>
KMS can support only very specific cursor sizes and having a cursor that
doesn't match that size isn't something going wrong. Instead of
generating a warning, just log it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4307>
By setting cursor_stage_view->is_hw_cursor_valid and has_hw_cursor and
then skipping to the code below, we can reuse all the logic for when we
need to inhibit/uninhibit and update the sprite.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4307>
As in the commit mentioned below, we dropped the only place where
current_gl_draw_buffer is set. Which was still being used for glx and
ended up causing black screen with multi-monitors & nvidia.
Fixes: aa0600bf2 ("cogl: Remove Framebuffer:stereo-mode")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4323>
Use an enum and an array to enumerate the keyboard LEDs and
map xkb_led_index_t to libinput_led. There are future additions
coming and this will be extensible without becoming too spaghetti.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4301>
So that if somehow it does return early then we're not left with an
allocated `clutter_frame_clock_new_frame` that is never dispatched
(which then leads to the pool being exhausted a frame or two later).
It's not yet clear how it comes to this where the source is dispatched
and the state unscheduled, but at least the more detailed logging here
will help us to identify which state it came from.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3945
Fixes: 394bf5ab24 ("clutter/frame-clock: Add triple buffering support")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4316>
When a client binds to the a device, if obtaining a non-privileged fd
fails, instead of returning an error, put the client in a pending
queue.
Delaying the drm_fd is allowed by the protocol definition:
The compositor will send this event when the wp_drm_lease_device_v1
global is bound, although there are no guarantees as to how long
this takes [...]
A future commit will send the drm_fd events to the clients in the list
once Mutter can obtain a valid fd.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3921>
This might be possible if extended target volume is supported but
currently it isn't.
This is a requirement from the wayland color management:
"'reference_lum' may be higher than 'max_lum'. In that case reaching
the reference white output level in image content requires the
'extended_target_volume' feature support."
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4304>
This is a requirement from the wayland color management:
"With transfer_function.st2084_pq the given 'max_lum' value is
ignored, and 'max_lum' is taken as 'min_lum' + 10000 cd/m²."
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4304>
Ensure the window is placed in coordinates, before maximizing, tiling or
minimizing it. This should make the window maximized/tiled on the right
(or correctmost) monitor given the stored coordinates.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4311>
Ensure that windows get restored either in maximized and unmaximized
state. And ensure that monitors being removed result in windows
snapping back to reachable positions.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4311>
We don't want triple buffering when the estimated maximum update time is
known and no larger than a refresh interval. In that case, regular frame
clock dispatch is scheduled after the previous frame is presented, so no
third buffer is necessary.
Allowing triple buffering anyway was problematic when frames are skipped
for reasons other than the frame update taking too long (e.g.
https://gitlab.gnome.org/GNOME/mutter/-/issues/3884):
1. First frame dispatches, targets display refresh cycle (DRC) n, but
skips
2. Second frame dispatches, targets DRC n+1
3. First frame is presented at DRC n+1
4. Second frame is presented at DRC n+2
Without triple buffering:
2. First frame is presented at DRC n+1
3. Second frame dispatches, targets DRC n+2
4. Second frame is presented at DRC n+2
The second frame is presented at DRC n+2 in both cases, but with triple
buffering it targeted n+1, i.e. its contents might not be consistent
with when it's presented.
It gets worse with triple buffering if the second frame also skips:
4. Second frame skips, is presented at DRC n+3
That's a discrepancy of 2 refresh cycles between the target and
effective presentation time, which might be noticeable as more severe
stutter.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4282>
With triple buffering, frame_clock->next_presentation_time_us
corresponds to a different frame in
clutter_frame_clock_notify_presented.
Also remove superfluous has_next_presentation_time field, any
target_presentation_time value > 0 is valid.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4282>
The superseded frame is always the oldest frame so it's the head of
the queue, not the tail. Aside from anything else, the call to
`meta_onscreen_native_notify_frame_complete` that follows *does* use
the head and not the tail.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4310>
`gdctl show` now prints "monitor preferences", which currently consists
of only the luminance setting.
`gdctl prefs` is introduced, where one can run e.g. `gdctl prefs
--monitor DP-1 --luminance 80.0` to set the output luminance of the
monitor connected to DP-1 to 80%.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4271>
This aims to allow configuring the output luminance currently used via
the color state shaders. It will replace the output luminance property
of the debug control API.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4271>
compare_window_position calls into meta_window_stack_position_compare
with inverted inputs instead of duplicating the logic.
This comes at the cost of an extra meta_stack_ensure_sorted which should
be culled out.
To ensure this, need_resort is reset before the sorting process starts.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4183>
Since these will not be used from now on, we can finally clean up
meta_stack_windows_cmp, window_stack_cmp and meta_display_stack_cmp.
Now the only comparison function is meta_window_stack_position_compare.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4183>
When closing some wayland apps using text-input-v3, text_input_destructor
could remove resource in focus_resource_list without resetting focus.
This could be an issue if, the focus is on a wayland app not using text-
input-v3 after closing those. Text committing would fail, and the
candidate window would always show on previous entry.
This commit adds check in text_input_destructor, resetting focus if
necessary.
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/4776
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4164>
The steps for this test:
* For each testing color create an actor and assign the color to it.
* Create a src_color_state and set it to all actors.
* Create a target_color_state and set it to the stage_view.
* Wait for paint.
* Read the color for each actor which will be different after the
color-state transformation.
* Do the same transformation on CPU for each test color.
* Validate that the resulting colors match.
In the future this test will be extended to compare transformations
between color_state_params and color_state_icc.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4230>
This is a CPU-based transformation method that performs the
same transformation implemented on GPU shaders.
The transformation gets an array of pixels in RGB float format and
returns an array of pixels transformed.
This will be used in the next commit to compare the results
between CPU and GPU, validating the shader implementations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4230>
This commit adds another way how to disable the a11y manager ACL in development.
Setting the unsafe mode might not be as straightforward as setting an
environment variable for the session, and you can control it invidually too.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4217>
Track unique DBus senders and allow only thse which own a (at least so far)
pre-defined set of well known DBus names.
Carlos Garnacho: Renamed to a more generic helper, use g_bus_watch_name().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4217>
Listen to changes in MetaA11yManager configured modifiers, and propagate
these along to the MetaSeatNative. This lets the backend keep track of
configuration changes in a11y modifiers.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4217>
The ::a11y-modifiers-change method allows tracking for changes in the
configured modifiers, add this method to allow backends to get the
modifiers so that they can be passed along the lower layers.
Carlos Garnacho: Turn into a method instead of a signal argument, turn
into an array+length instead of a hashtable.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4217>
Implement the org.freedesktop.a11y.KeyboardMonitor interface,
allowing screen readers to interact with Mutter and grab
shortcuts or full keyboard interaction.
Carlos Garnacho: Move setup to ::constructed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4217>
This a11y manager will handle key event emission to screen
readers and other ATs. This initial commit only introduces
the object.
Carlos Garnacho: Make the object take a ::backend property
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4217>
These modifiers will be set by the backend from the main thread, and
need to be handled specially for them to be usable as both modifier
buttons, and their own regular action.
Carlos Garnacho: pass modifiers as array+lenght instead of hashtable.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4217>
The state handling about whether the a11y modifier is a first press (so
could be consumed for other actions), or results in the modifier action
(e.g. caps lock) is performed in the input thread. This information will
be propagated through the CLUTTER_EVENT_FLAG_A11Y_MODIFIER_FIRST_CLICK
flag in the related key events.
Carlos Garnacho: Drop synchronous wait for configuration changes
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4217>
Use a Clutter event flag to communicate the the fact that the event
is an a11y modifier first click. The accessibility modifiers will
require special handling in the input and main threads.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4217>
This object will for now only provide a way for assistive technologies
to receive keyboard events, however it is expected that it will be used for
the new a11y communication protocol in the future.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4217>
libinput 1.27 added a new "sticky" mode to the tap-and-drag drag lock -
the previous mode is now called the "timeout" mode.
In sticky mode the drag lock is explicitly terminated with an extra tap,
i.e. a full sequence is:
tap, down, [:move, up, wait, down:], up, tap
where the middle part can repeat and/or wait as long as required.
In the previous "timeout" mode the drag lock would automatically release
after a timeout - for users with low dexterity this timeout may be too
short.
The sticky mode is also how macos does drag lock.
Related: https://gitlab.gnome.org/GNOME/gnome-control-center/-/merge_requests/2798
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4292>
Since we now cascade when center-new-windows is true, we need to update the
always-above window positions to account for where the new windows will end
up being placed.
Assertions are also added for the two test windows' positions.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4202>
This commit introduces a new comparator, window_distance_cmp,
which uses either the top left or top right corner of a given window, based
on LTR, as reference for the squared distance.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4202>
Both northwest_cmp and northeast_cmp didn't account for the work_area's
relative x and y coordinates which would lead to improper sorting on
multi-monitor setups.
To fix this, we pass the work_area when necessary and use it to offset the
absolute coordinates.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4202>
Remove unnecessary sqrt() calls in northwest_cmp and northeast_cmp.
The square root would have been necessary if we needed the actual distance,
but we only care about the relative order, so it isn't.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4202>
This is a preparatory commit, it refactors away of the
done_check_denied_focus label, simplifying the control flow.
The only functional difference is that now the auto-maximization when
windows don't fit the work_area is always applied.
The next commits will adapt find_next_cascade to properly handle this.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4202>
This commit only reformats place.c, starting off uncrustify's output, no
functional change is intended.
It mostly aligns comments, as well as properly space out math,
follow current formatter enforced guidelines and fix alignment on a rogue
bracket.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4202>
By default, copy mode primary tries to use the GPU to perform the copy
and, in case of error, it fallbacks to CPU.
Add a flag (unused for the moment) to allow to use the CPU to perform
the copy.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4251>
_clutter_actor_get_animation_info() creates a new info when the actor
currently doesn't have one. That's unnecessary and wasteful in case
where we only need to check for transitions to remove, so switch to
_clutter_actor_get_animation_info_or_default() that falls back to
an empty static info.
Fixes: c250f602bd ("clutter/actor: Remove transitions when removing an effect")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4260>
More precisely, wait until no further udev hotplug events have arrived
in 2 seconds.
Keep a list of unique UpdateStatesData structs which have received
hotplug events, and call handle_hotplug_event for all of them once the
timeout expires.
This avoids problems due to some monitors generating hotplug events
when power saving is enabled on locking the session, and then
temporarily appearing disconnected.
v2:
* Make meta_test_disconnect_connect wait and run main context dispatch
before checking the number of logical monitors.
v3:
* Call g_source_unref immediately after g_source_attach, simplifies
source cleanup. (Sebastian Wick)
* Free hotplug devices list in meta_kms_finalize. (Sebastian Wick)
* Add KMS debug logging in on_udev_hotplug & hotplug_timeout.
* Bump timeout to 2 seconds. With my affected monitor, the second udev
hotplug event normally arrives almost a second after the first one,
occasionally more than 1.5 seconds though.
* Use UpdateStatesData with custom compare function for hotplug_devices
list data, since the GUdevDevice / device path string pointer values
are always different.
* Don't wait for timeout if meta_is_udev_test_device returns TRUE, to
hopefully fix VKMS CI tests. Drop meta_test_disconnect_connect changes
again.
v4:
* Use hash table instead of list for hotplug_events set. (Sebastian Wick)
* Add comment describing the keys & values in the hash table. (Sebastian Wick)
* Add KMS debug logging in handle_hotplug_event as well.
* Dropped Closes:, this might not suffice to fully address
https://gitlab.gnome.org/GNOME/mutter/-/issues/3831 after all.
v5:
* Simplify hash table annotation comment. (Sebastian Wick)
v6:
* Rename hotplug_timeout local string pointer to hotplug_event.
* Drop g_clear_pointer in favour of g_autofree in on_udev_hotplug.
(Sebastian Wick)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4209>
Handle NULL string pointer in meta_kms_update_states_in_impl.
Drop second parameter of meta_kms_update_states_sync, which wasn't used
by the external test caller anyway. Split out static update_states_sync
function which takes a hotplug event string pointer instead.
Preparation for next commit.
v2:
* Drop UpdteStatesData for encoded hotplug event strings.
v3:
* Put device path at the end of encoded string, allows simplifying
meta_kms_update_states_in_impl slightly further.
v4: (Sebastian Wick)
* Use g_autofree for hotplug_event string in on_udev_hotplug, fixes
leak.
* Store pointer to hotplug event device path string in local variable in
meta_kms_update_states_in_impl.
v5:
* Initialize `path` local variable to `NULL` and test it instead of the
`hotplug_event` parameter. Avoids (false-positive) compiler warning
about `path` possibly being used uninitialized.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4209>
It forced the power save mode to META_POWER_SAVE_ON when reading the
current KMS state.
This was problematic when a hotplug event is emitted while the session
is locked, which can be triggered by monitors polling all their inputs
for a signal: There's no mechanism to restore the previous power save
mode in this case, so the monitors would fail to actually enter power
saving mode but stayed on with a blank screen. This was at least one
cause of the symptoms described in
https://gitlab.freedesktop.org/drm/amd/-/issues/662 .
Moreover, I suspect it hasn't had any effect for the actual reading of
KMS state since 5f6aee3419 ("kms/update: Make power saving an update
wide change"), as changing the power save mode to META_POWER_SAVE_ON no
longer results in any immediate KMS state change, it's now only taken
into account for the next mode set.
It's not clear what the intended effect was in the first place, it was
originally added as part of 65db8efbe8 ("MonitorManager: add a KMS
backend") without rationale. It might have been cargo-culted from
somewhere else. It shouldn't be necessary from a KMS API PoV though.
Also adjust the KMS hotplug test to assert that a hotplug event doesn't
implicitly change the power save mode.
v2: (Sebastian Wick)
* Fix shortlog of commit which added
meta_monitor_manager_native_read_current_state.
v3:
* Adjust KMS hotplug test for the change.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4209>
As a follow-up to 87cc6633a5, embed a partial copy of the Adwaita cursor theme
and point to it with XCURSOR_PATH to make the test completely independent
from the installed cursor themes.
Indeed, adwaita-icon-theme changed ever so slightly in version 48 so the
ref tests started failing when the new version was installed.
This is a copy of adwaita-icon-theme 47, which is what was used to generate
the ref images.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4252>
This requires reverting the previous changes intended for adapting to
triple buffering.
Fixes these tests randomly failing, also in CI pipelines of MRs not
directly related to any of this.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4290>
As per previous commit we're figuring out the bash location on shebang
using env, let's do it everywhere for coherency, even if in CI fedora
may not yet have issues about this.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4288>
Bash can be in both /usr/bin/bash or /bin/bash (or elsewhere!) depending on
the distro, so let's be generic using env to figure it out.
This comes from a packaging lintian error we get:
E: mutter-16-tests: wrong-path-for-interpreter /usr/bin/bash != /bin/bash
[usr/share/mutter-16/tests/socket-launch.sh]
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4288>
Avoids the following warning in the logs
mutter-x11-frame[4003]: Using
GtkSettings:gtk-application-prefer-dark-theme with libadwaita is
unsupported. Please use AdwStyleManager:color-scheme instead.
When initializing libadwaita, the AdwStyleManager default instance is
initialized taking care of loading the appropriate stylesheet.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4276>
It tests that the cursor renderer can successfully initialize with the
environment variable MUTTER_DEBUG_DISABLE_HW_CURSORS=1.
It would have caught the crash fixed by the previous commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4272>
Don't try to realize the cursor sprite for the HW cursor when it's set.
v2:
* Refactor is_hw_cursor_supported helper out of
realize_cursor_sprite_from_wl_buffer_for_crtc. (Jonas Ådahl)
v3:
* Keep meta_crtc_native_is_hw_cursor_supported check in
meta_cursor_renderer_native_update_cursor, to try and avoid
mysterious CI failure.
v4:
* Rename is_hw_cursor_supported → is_hw_cursor_available_for_gpu
and take a MetaGpuKms * parameter.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4272>
We want to ensure that also the stderr is matching our golden files
and at the same time we should be able to see it in logs (especially
when there's a python error to catch).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4277>
While having automatic completions is a very nice feature of gdctl (I
had suggested to use it too :)), it's not something that distros may
have by default and in particular it's not a package in main in ubuntu.
So, make the code less restrictive on completions, since completions is
not a core functionality of the tool and it can definitely work without
them.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4279>
While the `get_key_focus()` method returned the stage itself when
no explicit focus was set, it made sense for the corresponding
setter to accept the stage as synonym for NULL.
But now that the getter always returns the property value, it
makes more sense to expect NULL to unset the key focus.
Keep the current behavior of normalizing the key focus to NULL
in that case to minimize breakage, but print a warning to
use NULL instead.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4256>
Moving the key focus to the stage should be done by unsetting the
focus rather than setting it to the stage itself.
`clutter_stage_set_key_focus()` already "normalizes" the stage to
NULL internally, so this does not change the actual behavior of the
code.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4256>
The method currently returns the stage itself when the property
is NULL.
This has become particularly problematic as the method is detected
as getter by gobject introspection, and gjs now optimizes property
accesses by calling the getter method instead.
Address this by turning the method into a genuine getter without
falling back to the stage.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4256>
While the existing `get_key_focus()` methods looks like a getter of the
`key-focus` property and is detected as such by gobject introspection,
it behaves differently in that it returns the stage if no explicit
focus has been set.
This is about to change, so adjust the couple of cases that rely
on the fallback to the stage.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4256>
Previously if we had no measurements then `compute_max_render_time_us`
would pessimise its answer to ensure triple buffering could be reached:
if (frame_clock->state >= CLUTTER_FRAME_CLOCK_STATE_DISPATCHED_ONE)
ret += refresh_interval_us;
But that also meant entering triple buffering even when not required.
Now we make `compute_max_render_time_us` more honest and return failure
if the answer isn't known (or is disabled). This in turn allows us to
optimize `calculate_next_update_time_us` for this special case, ensuring
triple buffering can be used, but isn't blindly always used. So when
TIMESTAMP_QUERY support is missing we now take the same path as if
presentation timestamps are missing (introduced in 56fc09151d and
improved in 0555a5bbc1), which has the benefit of naturally switching
from double to triple buffering as required without actually having to
measure render times.
This makes a visible difference to the latency when dragging windows in
Xorg, but will also help Wayland sessions on platforms lacking
TIMESTAMP_QUERY such as Raspberry Pi.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1441>
This replaces the DISPATCHED state with new sub-states that are possible
with triple buffering:
DISPATCHED_ONE: Double buffering
DISPATCHED_ONE_AND_SCHEDULED: Scheduled switch to triple buffering
DISPATCHED_ONE_AND_SCHEDULED_NOW: Scheduled switch to triple buffering
DISPATCHED_ONE_AND_SCHEDULED_LATER: Scheduled switch to triple buffering
DISPATCHED_TWO: Triple buffering
Triple buffering is currently disabled until the test cases get updated to
handle it in the next commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1441>
Chronologically they already overlap in time as presentation may
complete in the middle of the dispatch function, otherwise they are
contiguous in time. And most switch statements treated the two states
the same already so they're easy to merge into a single `DISPATCHED`
state.
Having fewer states now will make life easier when we add more states
later.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1441>
Error diffusion was introduced in 0555a5bbc1 for Nvidia where last
presentation time is always unknown (zero). Dispatch times would drift
apart always being a fraction of a frame late, and accumulated to cause
periodic frame skips. So error diffusion corrected that precisely and
avoided the skips.
That works great with double buffering but less great with triple
buffering. It's certainly still needed with triple buffering but
correcting for a lateness of many milliseconds isn't a good idea. That's
because a dispatch being that late is not due to main loop jitter but due
to Nvidia's swap buffers blocking when the queue is full. So scheduling
the next frame even earlier using last_dispatch_lateness_us would just
perpetuate the problem of swap buffers blocking for too long.
So now we lower the threshold of when error diffusion gets disabled. It's
still high enough to fix the original smoothness problem it was for, but
now low enough to detect Nvidia's occasionally blocking swaps and backs
off in that case.
Since the average duration of a blocking swap is half a frame interval
and we want to distinguish between that and sub-millisecond jitter, the
logical threshold is halfway again: refresh_interval_us/4.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1441>
Although since all SCHEDULED* states basically have the same transitions
it's easier to show them as a single state. The mermaid diagram also won't
render correctly if there is a mix of combined and singular states. We
must use either always combined states or always singular states.
Fixes: 5b214dc2b7 ("clutter/frame-clock: Allow scheduling an update in the future")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1441>
It's analogous to discard_pending_page_flips but represents swaps that
might become flips after the next frame notification callbacks, thanks
to triple buffering. Since the views are being rebuilt and their onscreens
are about to be destroyed, turning those swaps into more flips/posts would
just lead to unexpected behaviour (like trying to flip on a half-destroyed
inactive CRTC).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1441>
All paths out of `meta_onscreen_native_swap_buffers_with_damage` from
here onward would set the same `CLUTTER_FRAME_RESULT_PENDING_PRESENTED`
(or terminate with `g_assert_not_reached`).
Even failed posts set this result because they will do a
`meta_onscreen_native_notify_frame_complete` in
`page_flip_feedback_discarded`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1441>
This is a case that triple buffering will encounter. We don't want it
to queue the same onscreen multiple times because that would represent
multiple flips occurring simultaneously.
It's a linear search but the list length is typically only 1 or 2 so
no need for anything fancier yet.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1441>
Introduces two new NamedTuples to deal with dimensions and positions.
The position is special in that x and y can be None. This was previously
wrongly declared to be only int. This commit fixes instances mypy found
where None positions were not handled.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4267>
It's not a single-element-tuple but something we can iterate and
produces strings.
Also ignore typing for the requests module because we'd have to install
more things for it to be available.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4267>
Now with the new parsing mechanism we are able to set `Super` as
default for overlay-key. The parsing of the `Super` string will
fail at the new mechanism will kick in by appending `_L` and `_R`.
The parsing of the new two strings then will succeed and we will
be able to use both Super_L and Super_R to activate the overview.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4050>
The overlay-key and locate-pointer-key are special keys. They
can be used together with other keys to create a combo or they
can be used as a single key. This means that we are treating a
modifier key as a regular key while still allowing to use it as
part of a combo. This requires a special treatment that we can't
extend to an arbitrary list of keys.
However, we would like to use both Super_L and Super_R to
activate the overview. In order to allow this, introduce a new
parsing mechanism. With the new mechanism, if we fail to parse
the configured string, we will try to parse again by appending
_L first and _R later. If both succeed then we will use their
combos for handling the special key.
With this in place, we can configure Super as overlay-key. The
parsing of Super will fail, but Super_L and Super_R will succeed.
Allowing us to use both.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1277
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4050>
The second combo is still disabled. This is in preparation for a
subsequent commit that is going to parse the preference in a
different way and might end up with two combos defined.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4050>
Special keys are going to be represent by up to two combos. This
functions is able to handle them by adding all the keycodes that
these combos resolve to. Special keys don't have modifiers, so there
is no need to devirtualize them.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4050>
This function allows, given an array of combos, to get all the
keycodes associated with them. Right now it is used for just
one combo, but will be used for two combos in order to handle
special keys like the overlay-key and the locate-pointer-key.
As it is now, this function is still useful for aggregate the
GArray creation/destruction in a single place.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4050>
We are going to need to call this function more than once to add
all the keycodes to the same GArray. Switch to handle GArrays and
rename the function in order to give a hint that keycodes are going
to be added to the passed GArray.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4050>
We use them in a single place, there is no need to internally keep
track of them. The MetaKeyBindingManager has their resolved key
combo (overlay_resolved_key_combo and locate_pointer_resolved_key_combo)
which is enough. Get the special binding combos from the preference
right away when needed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4050>
While we don't actually require the newer version, there was some
fallout from its change to optimize GObject property accesses by
calling the getter function.
Including that change in the CI image (and therefore derived toolbox
images) should give it more test exposure, so we can hopefully
find remaining issues (if any) before the release.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4268>
By definition, headless means no HW display output, so initializing HW
cursor support makes no sense.
Fixes hitting the g_warning in tests when there's a GPU device
available, breaking them.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4259>
Version 1 of the presentation protocol requires that 0 be sent for the
refresh rate for variable refresh rate. Fix this by checking the mode
during the presentation event.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4227>
Probe all RGBA 8888 variants.
If there's a GBM device, prefer a format also supported by GBM, if any.
Fall back to dumb BOs otherwise.
This allows the HW cursor to work correctly using BGRA8888 on s390x.
v2:
* Rename get_cursor_format_info → find_cursor_format_info.
* Log device path if we can't find any suitable cursor plane format.
v3:
* Split cursor_planes_support_format helper out of
find_cursor_format_info to make logic clearer. (Sebastian Wick)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4255>
Without doing this, a non-linear color state transformation could result
in premultiplied colour values larger than the alpha value, which
manifested with artifacts such as parts of the cursor shining brighter
than SDR white (with HDR enabled).
This was reported in the GNOME Shell Matrix room on August 8th 2024, and
I later hit it myself.
v2:
* Fix add_pipeline_snippet function formatting. (Sebastian Wick)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4224>
Only a tiny number of tablet pads that have more than one mode
group, everyone else has one mode toggle button and one group.
Let's not display "Group 0" for everyone if (almost) no-one has a Group
1 anyway.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4248>
The "seat" property is already used by the parent
ClutterVirtualInputDevice - re-using this property means the parent's
propery never gets set so clutter_virtual_input_device_get_seat()
returns NULL.
This causes mutter to crash if a tablet pad sends a key event via a pad
ring or strip.
Since this type is only ever constructed with a MetaSeatNative as seat
we can extract that object through the parent and drop our property.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4247>
If there are any in-progress transitions on any properties of the
effect, these will cause a crash next time they tick and update, as they
will try to access a `@effects.${effect_name}.${property_name}` property
on the `ClutterActor` which no longer resolves to an effect. In some
cases this will be because `priv->effects` itself is now `NULL` on the
`ClutterActor`.
This can be triggered by rapidly toggling screen time limits on and off
in gnome-shell with a low screen time limit which has already been
reached for the day. It will alternately add a desaturation effect and
fade-in transition, then remove the effect, then the transition will
update and crash.
Avoid this by removing relevant transitions when removing an effect.
Do the same for the other two groups of metas: constraints and actions,
as they will be subject to the same bug (under different reproducer
conditions). And the same for when any of these three meta groups are
cleared, as that could also trigger the same bug.
Signed-off-by: Philip Withnall <pwithnall@gnome.org>
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/8168
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4222>
HDR being enabled was controlled by toggling a property on
org.gnome.Mutter.DebugControl, which affected how the color space and
HDR metadata of the output was configured. Replace this with a higher
level MetaMonitor / MetaOutput level "color mode" enum, that is also
reflected in the monitor configuration API.
This enum is then used to derive the color space and HDR metadata at the
lower level where it matters. The ForceHDR debug control property is
still left there, as it only affects the color space and transfer
function of the view related to a monitor.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4192>
Finding configurations relevant for inspiration when creating a new one
can be useful for finding more things to inherit from previous
configurations than the scale, so put the configuration gathering code
in a helper.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4192>
Available modes are 'default', which is always added, and BT.2100,
which is added if the BT.2020 color space, and the PQ transfer function,
is supported by the output.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4192>
Commit 48d070dae changed the logic to compare the modifiers mask as a
whole.
Unfortunately, that does not work with all combinations of modifiers, as
some may not be reported when the ISO_Next_Group key is notified.
Revert to the original logic which is to compare against each modifier
mask individually.
Fixes: commit 48d070dae - keybindings: Check for ISO_Next_Group keysym
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4237>
In process_iso_next_group(), we would use try to match mask and keycode
of ISO_Next_Group to tell whether the key combo has been activated.
That works on X11, but not on Wayland with the native backend, because
the keycode does not match.
But we do not need to go all through that burden to match the key combo
we could just use the keysym instead, which would work even when there
is no physical key for ISO_Next_Group.
All we need to do is check whether the symbol is ISO_Next_Group and the
modifier mask matches, which simplifies the code as well.
(Note that we still need to keep the resolved iso_next_group_combo
key combo around because the X11 backend grabs that key combo.)
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3883
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4232>
It is a dependency of the new gdctl tool, and therefore no longer
confined to tests. Install it via the common-dependencies script,
so that the dependency is also satisfied for gnome-os and systemd
system extensions.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4233>
The named enum (NamedEnum) is used to describe a mapping for enum values
to strings. Enums using this define a function that defines the mapping,
and the named enum handles converting from to strings. This replaces
existing manually coded translations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4190>
It's based on `get-state.py`, but with the intention to expand its
functionality into not only listing information, but setting and
changing monitor configurations. It's meant to complement monitor
configurations from Settings with something that has more level of
control.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4190>
The gravity argument is no longer effective since the previous
commit. Drop the argument, and rename it to "meta_window_resize_frame",
all callers have been updated.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3972>
Ignore the gravity argument in the move_resize_internal() vmethod,
and use the window getter to get it. It should be functionally
equivalent, and will eventually replace the argument entirely.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3972>
We really only need the logind API which can be provided either by
libsystemd or libelogind.
This also pulls out finding a logind provider from the native backend
only code because we make use of it in the generic backend code as well.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4223>
This page does not seem to have been archived properly when the wiki got
shut down so the link does not work anymore. There is no replacement
page in the handbook either.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4229>
The reference images were generated using the Adwaita theme, so make sure that
the test client also draws a cursrom from the Adwaita theme rather than reading the
default theme from the environment.
Ideally the test suite would embed the cursor theme and point XCURSOR_PATH to
it, so that even if the Adwaita icon theme changes in the future the test
wouldn't suddently break; I did not implement this yet.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4225>
Currently, this has been living in StWidget, moving that to Clutter
allows us to properly track the accessibility state changes in the
actors provided inside Clutter as well as simplifying things for a
future move from Atk.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4089>
When the cursor is captured via metadata, schedule updates when the
sprite changes or position moves, so we can gather the cursor change
during dispatch.
Also use the new skipped-paint stage watch to avoid having to guess
whether something will be painted by peeking at the pending damage.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4067>
This is intended to allow being notified about a stage update happening,
but painting didn't happen. This is possible today by using other
signals and keeping track of painting happened, but it saves us some
state tracking by just being told so.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4067>
Some are replaced with g_warning(), some are not replaced yet, as they
will be handled differently.
One warning was turned into a META_DEBUG_DISPLAY, due to it being easily
hit in CI, while being extremely unlikely otherwise.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4005>
These warnings were altered due to the vast majority of cases they are
in practice harmless and just cause bug reports or red herrings due to
people seeing warnings and treating them as something that might be
relevant, while they rarely are.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4005>
Move the wrapper functionality where it is actually used, use a single
mechanism to determine if we're already in a dbus-runner environment and
make the mocked launching simpler.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4022>
The fallbacks were neccesarry for tests but we added some more advanced
logind mocking which will get us the right values from dbus. There is no
point in those fallbacks anymore.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4022>
In this scenario:
1. Only an "empty" content update (which results in no visible output
change) from client A arrives before the frame deadline, so a frame
event is sent for that at the deadline.
2. Another "empty" content update from client A results in a frame event
being scheduled for the next frame deadline.
3. A non-"empty" content update from client B arrives before start of
vblank, and the resulting output frame manages to hit the next
display refresh cycle.
The result was that the frame events from steps 1+2 were sent during the
same display refresh cycle, so the frame rate reported by client A was
higher than the display refresh rate.
This change fixes that, at the cost of frame events being sent out
later in the display refresh cycle if there's no new frame for the
next cycle.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3559
Fixes: 8f27ebf87e ("clutter/frame-clock: Start next update ASAP after idle period")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3878>
The stage view color state is not the output color state but usually a
linear version of it. For direct scanout, we need the view color state
to match the output color state instead.
Fixes: 20c7653d49 ("compositor-view/native: Don't scan out surface with color state mismatch")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4196>
In order to ensure consistent behavior with the composited path as well as across
different KMS drivers. In the future we'll want to use other values as well,
requested by client via the upcoming color representation protocol.
Note that right now KMS drivers default to different values. Most use BT709 and
narrow range, notably Intel and AMD, but some others do not.
BT709/narrow is arguably a much better default than BT601/narrow as the former is
used for most contemporary video content and the later more for still images.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4204>
Allow a transaction to have a timing constraint.
Any transaction with a timing constraint will be deferred at its initial
commit, without testing the target time. The timing constraints are cleared
later in an on_before handler immediately before repaint.
The new frame clock api to schedule later ticks is use to ensure we get an
appropraitely timed tick to clear the constraint.
Signed-off-by: Derek Foreman <derek.foreman@collabora.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3355>
Allow creating a queue of future times to tick at. This adds a new clock
state where we let the clock go idle, but will wake at a time in the
future to tick.
We can still schedule ticks while in this new state, but we're assured a
tick at or shortly after the scheduled time.
This will be used later by wayland code that schedules future presentation.
Signed-off-by: Derek Foreman <derek.foreman@collabora.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3355>
By introducing a BufferImpl that handles the buffer
allocation/de-allocation bits and making the driver responsible for
creating the correct impl.
This allow moving various Buffer specific vfuncs from Driver as well as
getting rid of the gl_handle field from Buffer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4181>
This consolidates duplicate code in meta_drm_buffer_gbm_blit_to_framebuffer
to use the newly added meta_drm_buffer_gbm_create_native_blit_image, which
also has the side-effect of caching creation of the EGLImage per GBM BO.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4027>
Creating an EGLImage is rather expensive and is taking the bulk of the
time the secondary GPU copy path is using for each frame. By caching
these per GBM BO we avoid this expensive recreation, which seems to
significantly improve FPS throughput in these scenarios, e.g. an
AMD or Intel iGPU with an NVIDIA dGPU.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4027>
Allowing to disable font rendering integration, making it possible to
build Mutter without pango/harfbuzz/fribidi dependencies.
This commit also adds a new clutter-pango header that is used to include
pango specific bits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4106>
There is no real need to re-create a new cairo_font_options_t now that
the API is internal. Instead, create the font_options once and just
update it attributes.
Actors already register for the emitted font-changed signal to re-create
a new PangoContext.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4106>
When the `org.gnome.desktop.interface` schema is not found, currently
we were not initializing the font_options which means we needed to
handle that on the backend side. Instead, generate the font_options at
that moment.
As the settings are loaded the moment we assign a backend to the
settings `_clutter_settings_set_backend` which is called just after the
backend is constructed which is too early for any actor to use it for
creating a PangoContext, so the change is safe.
Also, as the font_options getter is only used in ClutterActor when
creating the PangoContext, drop the getter. As we might just store that
info somewhere else in the future.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4106>
Currently, we were first reading the settings, creating a FontSettings
struct and then mapping the string associated on that struct back to
their corresponding cairo type. A lot of dancing for not much benefits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4106>
The cairo_font_options is only meant to be consumed by ClutterActor when
creating a PangoContext and as those APIs are never used externally,
mark them private to not expose more cairo APIs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4106>
As those properties are never set externally and just end up mapping the
gsettings values, in order to create a cairo_font_options_t.
Instead, simplify the whole thing and just create the
cairo_font_options_t from the resulting FontSettings.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4106>
Not with VRR though, because the deadline timer fires at most every 33ms
then.
Previously, the deadline timer was used only for cursor-only updates.
Using it for other updates means they pick up the latest cursor position
available at the deadline, resulting in the lowest possible input→output
latency for cursor movement.
TTBOMK this unlocks the full potential of the KMS thread given the
current atomic KMS API.
v2:
* Don't call meta_kms_update_merge_from with twice the same update
pointer.
v3:
* Don't arm deadline timer if crtc_frame->pending_page_flip is true.
v4:
* Tweak want_deadline_timer indentation per check-code-style CI job.
v5:
* Also check crtc_frame->await_flush for want_deadline_timer.
v6:
* Tweak coding style to keep lines shorter. (Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3964>
Previously it was just dropped on the floor with no feedback, which
could result in the main thread waiting for something which never
happens.
Fixes: 0d9fd1ead7 ("kms/impl-device: Destroy submitted update in disarm_all_deadline_timers")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3964>
To swap_buffer_result_feedback from page_flip_feedback_discarded. The
former is where META_KMS_ERROR_DISCARDED from disarm_all_frame_sources
gets handled here.
Fixes: af250506fb ("kms/impl-device: Queue result when discarding submitted update")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3964>
Currently, we cast the CoglXlibRenderer pointer into a CoglX11Renderer
which is something that we would get rid of later on, along with the
usage of a data field once the Renderers become a subtypes of
CoglRenderer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4187>
We closed all the DMA buffer fds after creating the DMA buffer backed
CoglFramebuffer, which meant that the fds we passed to CoglDmaBufHandle
were already closed. This broke screen casting with DMA buffers.
Fixes: 8509b74532 ("Make DMA buffer allocation format and modifier aware")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4177>
This makes it possible to use udev even for backends which are not
native. Specifically we want to start controling backlights in mutter
which has to happen for both native and X11.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4130>
The generic backend now provides the launcher which is the last native
specific dependency. This makes it possible to move the udev component
to the generic backend.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4130>
For now we only pause/resume the rendering in the base class but in the
next few commits, we'll move udev from native to generic and need to
pause/resume it as well.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4130>
By connecting to changes of the new property in backend native itself.
This removes the dependency on the native backend and allows us to move
it to the generic backend.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4130>
As they are no longer used separately by the GL3/GLES2 drivers.
The remaining Buffer/Texture/Attribute/Clip stack helpers are kept in
separate files as they are complex and are probably better kept split.
Further cleanups might be done in future commits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4132>
This uses the ref test infrastructure inside the cursor tests screen
cast client, and verifies the content on the screen cast matches the
content of the compositor, both when using the 'embedded' cursor mode,
and the 'metadata' cursor mode.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3859>
It doesn't contain anything screen cast specific, except the
XDG_RUNTIME_DIR, so make it usable from elsewhere. While at it, prepare
it for being able to run the glib test functionality.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3859>
This also avoids loosing precision since the float won't be casted to an
int before being used to configure the test output.
Currently there is only one test that expects the scale to be calculated
dynamically, all the other always assume scale ending up being 1.0
unless explicitly configured in some way.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3859>
Only actively test MemFd, and don't ever ask for or expect MemPtr, since
we don't support those anyway. Also, for convenience, ask the buffers to
be automatically mmap:ed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3859>
We create the cursor view data when we update the cursor, but allow the
texture to be realized earlier than that. The texture change listener
invalidates the hw cursor validity state, but since it's initialized to
invalid, the hardware cursor will still be correctly handled.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3859>
This makes the overlay code work which is notably used for screen
sharing and recording if the client doesn't support meta-data cursors.
[jadahl] Changed the approach the way coordinates are scaled.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3859>
It took a MetaTestClient, which wasn't used. To allow it be used for windows
that doesn't come from a MetaTestClient, remove the MetaTestClient
namespace, and update the callers.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3859>
This adds wl_seat and wl_pointer boilerplate, and a pointer-entered
signal on the surface helper struct. It's intended to be used by a
future cursor test case client.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3859>
Following previous commit, rename _clutter_paint_volume_init_static()
to clutter_paint_volume_init_from_actor(), and also
_clutter_paint_volume_copy_static() to
clutter_paint_volume_init_from_paint_volume().
Make clutter_paint_volume_init_from_paint_volume() follow the dst/src
semantic in its arguments, which also allows removing
_clutter_paint_volume_set_from_volume() which is exactly the same now.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4175>
And change clutter_paint_volume_free() to always free the paint volume.
Remove all calls to clutter_paint_volume_free() on static variables.
Having to call a free function on a static variable always seemed a bit
odd, and this genuinely confuses Coverity (and me).
Coverity CID: #1505838
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4175>
As far as I can see this doesn't happen in practice, but theoretically
the uninitialized values can be used if:
1. `did_placement` is TRUE
2. `flags` contains both `META_MOVE_RESIZE_RESIZE_ACTION` and
also `META_MOVE_RESIZE_MOVE_ACTION`
3. `!meta_window_is_tied_to_drag (window)` is FALSE
In that case, the `frame_rect` variable (with uninitialized values) is
passed to `unconstrained_rect`, then passed to `constrained_rect`,
then finally the (uninitialized) X and Y values are read in the
`if (did_placement)` branch.
This is probably a regression from 3047b2ce26. I don't know if this is
the appropriate fix.
Coverity CID: #1511378
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4175>
As they are now shared between GL3 and GLES3 texture driver, there is no
need anymore for having them in a separate header and included.
Cleanup the names of the functions while doing so.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4163>
Abstract away the common bits between GL3 and GLES2 TextureDriver
implementations by sharing the common bits in a parent class.
Ideally, we would move the various vfuncs that are GL specific from the
abstract TextureDriver type but that can be done at a later stage once
there is actual work on adding a Vulkan driver.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4163>
As they make more sense there. Maybe we could have a common
OpenGLTextureDriver
that would share the common texture_2d_* vfuncs but that can be done
later.
By moving those vfuncs to the TextureDriver, we can get rid of the
texutre_2d
nop driver implementations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4163>
Trying to update the opacity before the window actor get assigned a
surface will be ignored.
As result, the initial opacity set by X11 clients using the EWMH
property _NET_WM_WINDOW_OPACITY is not applied at first.
To avoid the issue, make sure to set the opacity once the window actor
has a surface assigned.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3513
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4170>
As those tests are not really useful, see previous commit. Removing them
allows us
to get rid of gdk-pixbuf in this code path and so, allowing to
completely remove it
once the MetaBackground API is refactored.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4133>
As is, ClutterImage is not really useful, it only serves for rendering a
CoglTexture as an actor. Shell, has a subclass that adds more features
that unfortunately cannot be upstreamed without bringing more gdk-pixbuf
usage inside libmutter, eg implementing GIcon/GLoadableIcon.
It also has requirements based on whether the image is symbolic or not.
Things that Clutter so far doesn't care about.
So just remove ClutterImage & let shells re-implement it themselves if
needed based on their needs.
Note, that once we have ClutterSnapshot, it should be straightforward to
write a custom actor that renders a CoglTexture or so.
This "un"fortunately means getting rid of various interactive tests that
either didn't compile at all or are not useful as is, like all the
remaining interactive tests.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4133>
g_hash_table_steal leaked the MetaMonitorSpec key in the old
compositor->outputs hash table:
==1059254== 62 (32 direct, 30 indirect) bytes in 1 blocks are definitely lost in loss record 8,232 of 13,059
==1059254== at 0x48489F3: calloc (vg_replace_malloc.c:1340)
==1059254== by 0x4C65C19: g_malloc0 (gmem.c:133)
==1059254== by 0x4956FE8: meta_monitor_spec_clone (meta-monitor.c:108)
==1059254== by 0x4A7B9E8: meta_wayland_compositor_update_outputs (meta-wayland-outputs.c:555)
==1059254== by 0x4A7C09E: meta_wayland_outputs_init (meta-wayland-outputs.c:796)
==1059254== by 0x4A63B60: meta_wayland_compositor_new (meta-wayland.c:874)
==1059254== by 0x49B58D0: meta_context_start (meta-context.c:523)
==1059254== by 0x10A8D7: main (mutter.c:148)
==1059254==
==1059254== 62 (32 direct, 30 indirect) bytes in 1 blocks are definitely lost in loss record 8,233 of 13,059
==1059254== at 0x48489F3: calloc (vg_replace_malloc.c:1340)
==1059254== by 0x4C65C19: g_malloc0 (gmem.c:133)
==1059254== by 0x4956FE8: meta_monitor_spec_clone (meta-monitor.c:108)
==1059254== by 0x4A7B9E8: meta_wayland_compositor_update_outputs (meta-wayland-outputs.c:555)
==1059254== by 0x4A7BA8C: on_monitors_changed (meta-wayland-outputs.c:572)
==1059254== by 0x4F5E9BF: g_closure_invoke (gclosure.c:833)
==1059254== by 0x4F72D82: signal_emit_unlocked_R.isra.0 (gsignal.c:3887)
==1059254== by 0x4F747A8: signal_emit_valist_unlocked (gsignal.c:3519)
==1059254== by 0x4F7A665: g_signal_emit_valist (gsignal.c:3262)
==1059254== by 0x4F7A722: g_signal_emit (gsignal.c:3582)
==1059254== by 0x49691BD: meta_monitor_manager_notify_monitors_changed (meta-monitor-manager.c:1241)
==1059254== by 0x496EA8D: meta_monitor_manager_rebuild (meta-monitor-manager.c:3968)
v2:
* Use g_autoptr. (Sebastian Wick, Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4149>
Fixes leak:
==9634== 60 (16 direct, 44 indirect) bytes in 1 blocks are definitely lost in loss record 8,198 of 13,049
==9634== at 0x48489F3: calloc (vg_replace_malloc.c:1340)
==9634== by 0x5DDCD29: drmModeGetPlaneResources (xf86drmMode.c:1243)
==9634== by 0x4AD94C9: init_planes (meta-kms-impl-device.c:1010)
==9634== by 0x4ADC7B5: meta_kms_impl_device_init_mode_setting (meta-kms-impl-device.c:2350)
==9634== by 0x4AD27BC: meta_kms_impl_device_atomic_initable_init (meta-kms-impl-device-atomic.c:1416)
==9634== by 0x4DD3B32: g_initable_new_valist (ginitable.c:249)
==9634== by 0x4DD3C19: g_initable_new (ginitable.c:163)
==9634== by 0x4ACEAE8: meta_create_kms_impl_device (meta-kms-device.c:662)
==9634== by 0x4ACECD1: create_impl_device_in_impl (meta-kms-device.c:722)
==9634== by 0x4B01837: meta_thread_impl_dispatch (meta-thread-impl.c:542)
==9634== by 0x4B00D06: impl_source_dispatch (meta-thread-impl.c:175)
==9634== by 0x4C5C81E: g_main_dispatch (gmain.c:3357)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4149>
Fixes leak:
==6470== 64 bytes in 1 blocks are definitely lost in loss record 8,677 of 13,062
==6470== at 0x48489F3: calloc (vg_replace_malloc.c:1340)
==6470== by 0x4C65C19: g_malloc0 (gmem.c:133)
==6470== by 0x4C746C5: g_rc_box_alloc_full (grcbox.c:106)
==6470== by 0x527B63B: mtk_region_create (mtk-region.c:75)
==6470== by 0x527B69C: mtk_region_copy (mtk-region.c:96)
==6470== by 0x4A54F56: meta_screen_cast_stream_src_maybe_record_frame_with_timestamp (meta-screen-cast-stream-src.c:1023)
==6470== by 0x4A54A9F: meta_screen_cast_stream_src_maybe_record_frame (meta-screen-cast-stream-src.c:871)
==6470== by 0x4A49431: on_after_paint (meta-screen-cast-virtual-stream-src.c:190)
==6470== by 0x49787F6: notify_watchers_for_mode (meta-stage.c:195)
==6470== by 0x4978A2F: meta_stage_paint_view (meta-stage.c:271)
==6470== by 0x533D152: clutter_stage_paint_view (clutter-stage.c:490)
==6470== by 0x497A476: paint_stage (meta-stage-impl.c:410)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4149>
Fixes leaks:
==1060013== 96 (32 direct, 64 indirect) bytes in 1 blocks are definitely lost in loss record 10,897 of 13,064
==1060013== at 0x4F81D57: g_type_create_instance (gtype.c:1929)
==1060013== by 0x4F64ABF: g_object_new_internal.part.0 (gobject.c:2606)
==1060013== by 0x4F66ADD: g_object_new_internal (gobject.c:2603)
==1060013== by 0x4F66ADD: g_object_new_with_properties (gobject.c:2769)
==1060013== by 0x4F67A30: g_object_new (gobject.c:2415)
==1060013== by 0x52F7C7B: clutter_color_state_new_full (clutter-color-state.c:339)
==1060013== by 0x4939CD0: update_color_state (meta-color-device.c:725)
==1060013== by 0x4939DDE: meta_color_device_new (meta-color-device.c:759)
==1060013== by 0x493CB7B: update_devices (meta-color-manager.c:205)
==1060013== by 0x493CE65: meta_color_manager_monitors_changed (meta-color-manager.c:264)
==1060013== by 0x49341CB: meta_backend_monitors_changed (meta-backend.c:371)
==1060013== by 0x4969150: meta_monitor_manager_notify_monitors_changed (meta-monitor-manager.c:1235)
==1060013== by 0x496928F: meta_monitor_manager_setup (meta-monitor-manager.c:1273)
==1060013==
==1060013== 96 (32 direct, 64 indirect) bytes in 1 blocks are definitely lost in loss record 10,898 of 13,064
==1060013== at 0x4F81D57: g_type_create_instance (gtype.c:1929)
==1060013== by 0x4F64ABF: g_object_new_internal.part.0 (gobject.c:2606)
==1060013== by 0x4F66ADD: g_object_new_internal (gobject.c:2603)
==1060013== by 0x4F66ADD: g_object_new_with_properties (gobject.c:2769)
==1060013== by 0x4F67A30: g_object_new (gobject.c:2415)
==1060013== by 0x52F7C7B: clutter_color_state_new_full (clutter-color-state.c:339)
==1060013== by 0x4939CD0: update_color_state (meta-color-device.c:725)
==1060013== by 0x4939DDE: meta_color_device_new (meta-color-device.c:759)
==1060013== by 0x493CB7B: update_devices (meta-color-manager.c:205)
==1060013== by 0x493CE65: meta_color_manager_monitors_changed (meta-color-manager.c:264)
==1060013== by 0x49341CB: meta_backend_monitors_changed (meta-backend.c:371)
==1060013== by 0x4969150: meta_monitor_manager_notify_monitors_changed (meta-monitor-manager.c:1235)
==1060013== by 0x496EA7D: meta_monitor_manager_rebuild (meta-monitor-manager.c:3968)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4149>
Fixes leak:
==5763== 96 (32 direct, 64 indirect) bytes in 1 blocks are definitely lost in loss record 10,901 of 13,065
==5763== at 0x4F81D57: g_type_create_instance (gtype.c:1929)
==5763== by 0x4F64ABF: g_object_new_internal.part.0 (gobject.c:2606)
==5763== by 0x4F66ADD: g_object_new_internal (gobject.c:2603)
==5763== by 0x4F66ADD: g_object_new_with_properties (gobject.c:2769)
==5763== by 0x4F67A30: g_object_new (gobject.c:2415)
==5763== by 0x52F7C46: clutter_color_state_new_full (clutter-color-state.c:339)
==5763== by 0x52F7C03: clutter_color_state_new (clutter-color-state.c:312)
==5763== by 0x52F7110: clutter_color_manager_get_default_color_state (clutter-color-manager.c:150)
==5763== by 0x52E7543: get_default_color_state (clutter-actor.c:17836)
==5763== by 0x52E765D: clutter_actor_unset_color_state (clutter-actor.c:17864)
==5763== by 0x52CF056: clutter_actor_constructor (clutter-actor.c:5646)
==5763== by 0x4F64DD3: g_object_new_with_custom_constructor (gobject.c:2524)
==5763== by 0x4F67275: g_object_new_internal (gobject.c:2604)
==5763== by 0x4F67275: g_object_new_valist (gobject.c:2945)
Fixes: 2693cac83a ("clutter/color-manager: Add a method to get the default color state")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4149>
The same MetaCrtcMode objects can be reached via the
MetaOutputInfo::modes array & MetaMonitorMode::crtc_modes arrays and via
the MetaGpuPrivate::modes list, so all of them need to hold their own
references.
In turn, those references need to dropped in meta_output_info_unref and
meta_monitor_mode_free. (meta_gpu_finalize is already dropping the
MetaGpuPrivate::modes list references)
Fixes leak:
==4092== 123 (32 direct, 91 indirect) bytes in 1 blocks are definitely lost in loss record 11,396 of 13,056
==4092== at 0x4F81D57: g_type_create_instance (gtype.c:1929)
==4092== by 0x4F64ABF: g_object_new_internal.part.0 (gobject.c:2606)
==4092== by 0x4F676BA: g_object_new_internal (gobject.c:2923)
==4092== by 0x4F676BA: g_object_new_valist (gobject.c:2945)
==4092== by 0x4F67A0C: g_object_new (gobject.c:2418)
==4092== by 0x4AB578B: meta_crtc_mode_virtual_new (meta-crtc-mode-virtual.c:51)
==4092== by 0x4B081BB: meta_virtual_monitor_native_new (meta-virtual-monitor-native.c:83)
==4092== by 0x4AC58CC: meta_monitor_manager_native_create_virtual_monitor (meta-monitor-manager-native.c:603)
==4092== by 0x496803B: meta_monitor_manager_create_virtual_monitor (meta-monitor-manager.c:621)
==4092== by 0x4A4A204: create_virtual_monitor (meta-screen-cast-virtual-stream-src.c:625)
==4092== by 0x4A4A39D: ensure_virtual_monitor (meta-screen-cast-virtual-stream-src.c:663)
==4092== by 0x4A4A461: meta_screen_cast_virtual_stream_src_notify_params_updated (meta-screen-cast-virtual-stream-src.c:684)
==4092== by 0x4A568C5: on_stream_param_changed (meta-screen-cast-stream-src.c:1565)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4149>
SwapBuffers will implicitly flush the query along with the other GPU
work for the frame.
The comment was about this function being called for direct scanout,
which hasn't been the case since 56580ea7c9 ("backends/native: Assume
zero rendering time for direct scanout buffers").
This eliminates one GPU work flush to the kernel per frame with the
Mesa radeonsi driver, leaving only a single flush per frame.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4158>
Indirectly via cogl_framebuffer_flush, since we also need to call
glFlush now. Need to do it in cogl_onscreen_egl_swap_* because
meta_onscreen_native_swap_buffers_with_damage uses
cogl_context_get_latest_sync_fd.
Doing it before the swap was problematic because the swap may do GPU
work of its own, which wasn't covered by the EGL sync object created in
_cogl_context_update_sync. This could result in visual artifacts. See
the discussion starting at
https://gitlab.freedesktop.org/mesa/mesa/-/issues/11996#note_2678104 for
details.
For similar reasons, move the cogl_framebuffer_finish calls after the
swap as well.
As a bonus, this eliminates one of 3 GPU work flushes to the kernel per
frame with the Mesa radeonsi driver, because the glFlush/glFinish call
in cogl_framebuffer_flush/finish doesn't have any GPU work to flush
after SwapBuffers.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4158>
The buffer resource doesn't need to exist for a timeline to exist so we
still have to send the timeline release. Only guard the
wl_buffer_send_relase call.
Fixes: e8b890ab53 ("wayland: Implement linux-drm-syncobj-v1")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4157>
Even though the touch moved to being set up in CLUTTER_ENTER, it
is still meant to lock onto the pressed surface, we however will
receive crossing events if the touch moves between actors/surfaces,
triggering warnings when the touch info registration is attempted
on the already existing touch sequence.
The reasons to handle this in CLUTTER_ENTER still apply, so ensure
the touch info for the sequence does not previously exist to avoid
the warning.
Fixes: 2e82a2049f ("wayland: Register touchpoint info on CLUTTER_ENTER")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4166>
We try to emit wl_touch.cancel in the situations where grabbing input
would take input away from the wl_touch interface. This however does
not play fine with grabs induced by wayland interfaces (e.g. xdg popups).
A more natural place to handle this is the MetaWaylandSeats' default
input handler, specifically the focus() vmethod. Here, we may know if the
focus surface matches the current surface as picked by MetaWaylandTouch,
and if it does not, either a grab or another input handler stole input
away from the default event handler.
A case where this will notably not happen anymore is in the transition
to a xdg_popup grab, since the grabbing client will still be handling
input, and touch input will not be unfocused away despite the transition
to a grab.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3752
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4166>
The logic to allow src/wayland to update the pointer cursor changed
to bypassing grabs in commit e69e4fa6, since it is nowadays the
responsibility of the DnD source to update the pointer cursor in reaction
to the negotiated action.
This is not entirely correct, and was done at the expense of regressions.
Change the logic so we explicitly check for a DnD grab existing, or use
the active focus instead (i.e. grab-aware). This fixes the regression, and
keep DnD cursor icons working.
This is a partial revert of commit e69e4fa6db.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3460
Fixes: e69e4fa6db ("Revert "wayland: Check focus surface to set a pointer cursor"")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4166>
Make create_transform_snippet method more consistent.
Abstract TransferFunction struct: convert it to ColorOpSnippet.
This transform snippet will be defined by these ColorOpSnippets.
These ColorOpSnippets are similar to the prescriptive DRM API for
color transformation.
A standard transform snippet would have as ColorOpSnippets:
1. eotf
2. luminance_mapping
3. color_space_mapping
4. inv_eotf
Update uniforms the same way the transform snippet is defined.
Update color transform key to consider how the transform snippet is
generated.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4144>
When creating a new color state from the primitives Colorimetry, EOTF
and Luminance; it is needed to previously check their tags to properly get
their values and avoid UB.
This check is duplicated and is a bit unreadable.
Using this new function helps keeping readability.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4144>
Most of the implementation at color state was specific to a color state
generated from parameters so move it to a new class Params.
In the next commits a new color state ICC class will be added.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4144>
The EOTFs snippets were only added when color states were not equal.
It makes more sense to only add the whole color transformation pipeline
in that case.
This makes the assumption to always append EOTFs snippets.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4144>
This makes DMA buffer allocation in all layers take an array of possible
modifiers to allocate using, or zero modifiers if implicit modifiers
should be used.
The format hard coding previously used is moved to the screen cast code,
or in case of the (unused by default) shadow buffer buffers use the same
format as the the CoglOnscreen.
This also means the CoglDmaBufHandle and MetaDrmBuffer got taught how to
distinguish between planes. It's mostly unused in practice, so rather
untested.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3900>
It's only implemented by the GBM render device backend, and uses EGL,
thus does not distinguish scanout capable modifiers.
A filter enum is added to Cogl, since it'll be used via the Cogl API,
but the actual Cogl API isn't added yet.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3900>
clutter_stage_schedule_update() sets the field `update_scheduled` to
`TRUE` as an optimization to make redundant updates a no-op. This failed
if there was a pending event and if the stage was not yet mapped.
What happened is:
* clutter_stage_schedule_update() is called
- ClutterStage::update_scheduled is set to TRUE
- frame clock scheduled
* frame clock dispatches
- frame is discarded early, no actual stage update happens
* device is created (e.g. virtual device from remote desktop session)
- `device-added` event reaches ClutterStage::event_queue
* stage is shown
- clutter_stage_schedule_update() is called
- ClutterStage::update_scheduled is TRUE
- ClutterStage::event_queue has events in it
- These two conditions means clutter_schedule_update() becomes a
no-op
At this point, no more updates will happen from
clutter_stage_schedule_update().
Fix this by resetting `ClutterStage::update_scheduled` to `FALSE` even
if the frame was discarded due to the stage not yet being mapped.
A test case is added that replicates the above descibed events.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3804
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4152>
To allow tests to emulate GNOME Shell behavior, where the stage showing
is delayed. Do so by adding the `"options"` property to the plugin
object type. The property expects a vararg GVariant, and supports
the `"show-stage"` (boolean) vararg entry.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4152>
This is needed to allow tests to manipulate the behavior of the test
shell plugin during startup. Since the plugin is created and started
when the MetaDisplay is created, it needs to be handled via MetaContext,
by setting the options after creating the context, but before starting.
For simplicity reasons, make the options an opaque GVariant, passed via
a an `"options"` property when the plugin object is created, if the
passed options is non-NULL. Only passing the options when non-NULL
allows for backward compatibility.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4152>
This disables the visibility change effects for the dragged window,
which makes it feel a bit smoother.
TODO: Double-check whether this is indeed desirable. Note: ChromeOS imp
does this and Chromium-X11 used to do this at client side.
Status:
- [x] Basic window drag triggering
- [x] Exclude the dragged window from event targets
- [x] Event forwarding (window drag vs wayland grabs)
- [x] Offset calc relative to toplevel geometry
- [x] Attach already mapped windows
- [x] Properly support not-yet-mapped windows
- [x] Disable visibility change animations
- [ ] Dnd events stream adaptations
Signed-off-by: Nick Diego Yamane <nickdiego@igalia.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4107>
- Event stream adaptations. When there is a toplevel-drag in place, do:
- Send wl_data_source.dnd_finished and end the session successfully
- Send wl_data_source.cancelled and end the MetaWindowDrag when ESC key
is pressed.
- Modify MetaWaylandDataDevice such that, when a toplevel-drag is
running, it does:
- Propagate motion events, so that they can be processed further by
MetaWindowDrag.
- Ends the associated MetaWindowDrag upon release event.
- Hook up the window mapping process in MetaWindowWayand, such that:
- the initial position of the window attached to the ongoing
toplevel-drag instance can be calculated and set.
- the appropriate gravity and flags can be set when calling MetaWindow's
meta_window_move_resize_internal, which allows it for example to be
moved freely (unconstrained) as per current dragging cursor.
Status:
- [x] Basic window drag triggering
- [x] Exclude the dragged window from event targets
- [x] Event forwarding (window drag vs wayland grabs)
- [x] Offset calc relative to toplevel geometry
- [x] Attach already mapped windows
- [x] Properly support not-yet-mapped windows
- [x] Disable visibility change animations
- [x] Dnd events stream adaptations
Signed-off-by: Nick Diego Yamane <nickdiego@igalia.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4107>
- For already mapped windows, the window drag session is started
straight away;
- For about-to-be-mapped window (ie: undocking window use case):
- The "shown" signal for the dragged window triggers the actual
MetaWindowDrag once it's mapped.
- MetaWindowWayland now handles the case of toplevel-drag and position
the window about to be mapped according to the toplevel-drag
parameters.
- While attached to a toplevel-drag, the window state is updated to:
- Actor's "reactive" state is set to false, which in practice excludes
it from the possible drag target list;
- WindowActor's "tied to drag" state is set to true, which results in
initial placement constraints to be skipped, so newly created
(detached) windows can be freely dragged around.
- Toplevel drag session ends upon:
- dnd drop and cancellation.
- xdg_toplevel_drag_v1 object destruction (client-side).
- data source destruction.
Status:
- [x] Basic window drag triggering
- [x] Exclude the dragged window from event targets
- [x] Event forwarding (window drag vs wayland grabs)
- [x] Offset calc relative to toplevel geometry
- [x] Attach already mapped windows
- [x] Properly support not-yet-mapped windows
- [ ] Disable visibility change animations
- [ ] Dnd events stream adaptations
Signed-off-by: Nick Diego Yamane <nickdiego@igalia.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4107>
- This adds a new ClutterActor parameter to meta_compositor_drag_window
and meta_window_drag_begin such that it's possible to inject an
external grab actor, which is then used to implement a "non-grabbing"
mode that allow a MetaWindowDrag to be used in conjunction, for
example, with a Wayland drag-and-drop session;
- This new API is going to be used to implement the toplevel drag
protocol in a subsequent commit.
Signed-off-by: Nick Diego Yamane <nickdiego@igalia.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4107>
The cogl_scanout_get_dst_rect() fell back on the buffer dimensions as
the destination rectangle when nothing was explicitly set. This,
however, is not necessarily correct. For example, if a buffer is larger
the CRTC resolution, but the surface is scaled to exactly match the CRTC view,
the expected destination size should match the CRTC resolution, not the
buffer dimension, which would be the case if no explicit destination was
set.
In meta_wayland_try_aquire_scanout() we're in a good position to
determine the destination rect in the CRTC primary plane, since we have
all the prerequisits, i.e. that the surface effectively covers the whole
CRTC, the actor allocation box (the non-black border part), the scale
and transform of the view.
This tweaks the CoglScanout API a bit to make it explicit that the
dst_rect must be unconditionally provided, and removes the fallback to
the buffer dimension as the destination rectangle, which sometimes
resulted in a destination rectangle being larger than the primary plane
itself, resulting in clipping and incorrect scaling.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3773
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4147>
The "dst_rect" calculated is in "CRTC space", meaning the bounding box
used for calculating it should be the view layout dimension, scaled by
the view scale and transform by the view transform. Previously it was
only transformed, not scaled. While fixing this, rename the variables to
make it a bit more clear what coordinate space they are expected to be
in.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4147>
Correct dependency analysis for required DT_NEEDED entries is
made difficult due to the lack of DT_RPATH, despite all other
installed libraries in the build setting an RPATH.
Whilst it is true dlopen() still works, this is only because
libmutter is already linked to the main executable, which in turn
links to the private libraries via DT_RPATH.
Signed-off-by: Ikey Doherty <ikey@serpentos.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4150>
This allows drag-and-drop within the same surface with touch input.
Test case: drag-and-drop a file into a directory in Nautilus.
Prior to this change that would only work if the finger first leaves
then re-enters the surface.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4091>
For zwp_text_input_v3_set_surrounding_text(), the protocol specifies:
"If the client is unaware of the text around the cursor, it should
not issue this request, to signify lack of support to the compositor."
Mutter currently doesn't clear its stored surrounding text when the
input focus changes, re-using the existing text for the newly focused
client. This already seems problematic, but since clients aren't
supposed to set surrounding text to NULL or an empty string if they
don't have one (and instead should simply avoid calling
`set_surrounding_text()`), this is clearly a bug in Mutter.
Fix it by unsetting the stored surrounding text when removing input
focus from a client.
Fixes: 33088d59db ("wayland/text-input: Pass char based offset to ClutterInputFocus")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3604>
In add_drm_device, calling g_hash_table_insert with NULL for the hash
table (which happens for any device added after mutter finished starting
up) would presumably just drop the key/value pointers on the floor,
leaking the string & MetaRenderDevice object referenced by them. Fixing
this should help for the referenced issue.
In meta_backend_native_take_render_device, g_hash_table_steal_extended
would presumably return FALSE (and log a warning), so the behaviour would
be "correct" by chance.
Issue: https://gitlab.gnome.org/GNOME/mutter/-/issues/3785
Fixes: 58c3734d78 ("backend/native: Prepare render devices earlier than KMS")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4136>
When using a tablet tool in relative mode motion compression may
apply. Doing so drops the axes from the event, leading to a segfault
later when we're trying to broadcast_axis() an event without axes.
Fix this by making sure we copy the axes over during motion compression.
All but the wheel are absolute so we can just take them from the new
event but if we do have wheel data add them together and where the wheel
changes direction skip motion compression.
We can take a few shortcuts here because the clutter implementation
guarantees exactly CLUTTER_INPUT_AXIS_LAST axes so we only need to
put in warning checks in case that ever changes.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3766
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4117>
The cursor and anchor position of the surrounding text *must* be within
(or right before/after) the string that is submitted as surrounding text.
Everything else is a client error that we shouldn't accept and log as such.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3650>
meta_wayland_text_input_focus_delete_surrounding() is accessible from JS and
used by gnome-shell via ClutterInputMethod.delete_surrounding(). Mutter
should never crash on invalid function calls from JS, so g_return_if_fail()
instead of asserting in case the parameters are wrong here.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3650>
This allows to capture the next update of all stage views using
meta_backend_renderdoc_capture. Finer control over what stage view is
captured when can be added in the future.
Currently, renderdoc does not support the GL_OES_EGL_image extension
that we depend on for importing dma-bufs. This means that dma-buf
support is broken when mutter in launched with renderdoc.
One can hack up renderdoc to pretend to support GL_OES_EGL_image and
specifically EGLImageTargetTexture2DOES to restore the full mutter
functionality but renderdoc captures replay with a black rectangle where
those dma-buf buffers are being painted.
See:
https://registry.khronos.org/OpenGL/extensions/OES/OES_EGL_image.txthttps://registry.khronos.org/OpenGL/extensions/OES/OES_EGL_image_external.txt
It looks like the renderdoc maintainer does not want to merge support
for the extension: https://github.com/baldurk/renderdoc/pull/2845
We would only need support for dma-buf EGLImages which so it might be
possible to convince the maintainer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4061>
In contrast to realtime scheduling, this doesn't risk us getting
SIGKILL:ed when the kernel is doing busy looping in
drmModeAtomicCommit() for some reason, but will according to testing,
right now, give us more or less the same benefit when it comes to
dispatch lateness and commit lateness.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4124>
Either prefer "normal" or "realtime", via an enum, instead of a boolean.
Also make it configurable with an env var
`MUTTER_DEBUG_KMS_SCHEDULING_PRIORITY`, which can be set to either
`normal` or `realtime`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4124>
Overlays are always cursors, and "redraw overlay" doesn't communicate
that. Add "cursor" or "cursor overlay" to some functions to make it a
bit more obvious in a couple of places.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4066>
There were things hooked up (connecting signals, adding stage watches),
but it was a bit disorganized, with "adding watches" doing more, and
"init callbacks" being a bit vague. Split things up to
* setup view - hook up things that need to listen on a stage view
* setup cursor tracking - track cursor positions, sprite changes
* the rest - monitor changes etc
This also properly handles a race condition when we'll enable before the
idle callback creating the view from the virtual monitors is run.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4066>
The hardware cursor support in the cursor renderer now acts on a per
stage view basis, so no need to inhibit the hardware cursor, it isn't
going to try to put a hardware cursor on a virtual CRTC anyway.
Instead use overlay inhibitation to decide whether an cursor overlay
should be painted or not. Inhibit the overlay when hidden/metadata, but
not when embedded.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4066>
There's a workaround (ec18a559c6 ("monitor-manager: Ignore
tablet-mode-switch on initial config of native portrait mode panels")) for
panels that default to portrait mode, where touch mode is disabled, but
a touchscreen is present. This workaround gets applied when we receive
the first reading from the orientation manager.
Make sure this workaround is applied as expected by adding a test for it.
The test needs to be first in the list of orientation tests, because mutter
only applies the workaround on the first orientation event it receives.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4126>
The MetaWindowConfig can indicate whether the configuration reflects on
the initial configuration, before the window is first mapped.
Add a smaller helper (private) window function to create the appropriate
"type" of MetaWindowConfig depending whether the window was already
showed or not.
This is preparation work for the following commits where this function
will be used.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4076>
Use the meta_window_is_fullscreen() API instead of accessing the
fullscreen field of the MetaWindow structure directly.
This is both a (small) cleanup and preparation work for the next commit.
No function change (intended).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4076>
This is intended to be used in place of the window rect and fullscreen
flags.
That will also allow for a pre-configuration signal to be added, passing
the configuration so that a plugin can tweak the configuration before it
gets applied first.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4076>
The client may not wait for devices to be created before enabling,
meaning we might activate the input capture session before the device
emulation is started. This would result in events silently being
dropped.
This helps with the flaky input-capture test case.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4128>
Since 1eba07f6 the size of a boxed value is read when calling
_cogl_boxed_value_set_x(). This results in an uninitialized read when
setting uniforms using cogl_pipeline_set_uniform_*() since the size was
not initialized by _cogl_pipeline_override_uniform() when these
functions call _cogl_boxed_value_set_*().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4125>
Use the 'seq' field in the spa_meta_header struct to track buffer
sequence numbers. This can be useful for predictable buffer tracking in
consumers, or for debugging purposes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4065>
This will become important when dispatching happens from a frame clock
driver, where the driver itself triggers from a GSource. If the driver
source has a higher priority, we'll risk dispatching before the frame
clock has returned to a state expecting to be dispatched.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4065>
We can't unfortunately use the macros for the sub-types as the whole GL
variant of the texture wants access to the struct fields even when using
a macro like COGL_DECLARE_INTERNAL_TYPE. So we would have to re-do the
whole driver integration first.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4100>
Touch devices are handled very similarly to how absolute pointer
devices, by creating either shared or standalone devices depending on
what kind of monitor it's associated with.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4064>
A EIS client might want to unbind a device capability; doing so should
effectively remove the device, which we didn't. Instead we always
created devices that a seat bind event had capabilities set for.
Fix this by explicitly keeping track of what is our "keyboard", our
"pointer", and whether we have a set of abs pointers, and don't create
duplicates if we already have devices created. For absolute pointer
devices, just keep track if we should have them, because we might have
many, or none, if we happen to be headless at the time being.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4064>
We don't need two separate configuration helpers, they do the same
thing. This will simplify making the viewport dependent device creation
more generic and reusable.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4064>
For some reason the variable for the *Native subclass of
MetaVirtualInputDevice was suffixed with _evdev. Long long ago the input
backend in clutter was called "evdev", so might be because of that.
Anyway, lets rename to something more closely related, i.e. _native.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4064>
The underlying data structure of MtkRegion is pixman_region32, which
gives us boxes, not rectangles. Use the new get_box() API to bypass
going via rectangles to get the boxes directly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4103>
Go one step further and pass regions. Sometimes the rectangles were
already a region, e.g. the swap-buffer case, and sometimes it still
potentially needs to pass through a rectangle array, e.g. damage with a
view transform.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4103>
Only before passing to EGL does it need to become a int tuple array.
It's used in non-EGL places which now become more easy to read. While at
it, make use of the new (and tweaked) helper function for flipping
rectangles from "cogl space" to "GL space".
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4103>
The other similar API all operate with the assumptions that (0,0) is at
the top left, so lets make damage regions behave this way too. Add a
helper to flip the rectangles, to make it a bit more convenient. It'll
be used in more places in a follow up.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4103>
Now that connectors can be configured as for lease, rename the function
to meta_kms_connector_is_non_desktop() to make clear that it returns
the hardware configuration rather than the user configuration.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4112>
Add a flag to MetaMonitor indicating if the monitor is available for
lease and store/update it from the monitor configuration.
Also, add unit tests validating that the configuration is applied and
that invalid configurations fail.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4112>
Move components without dependencies to the top and try to move
components with dependencies close to the component they depend on.
While this is an improvement, we really should start tracking and
documenting the actual dependencies between our components so that the
order here becomes comprehensible.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4018>
The constructor already copies the pipeline which means we can add a
snippet safely without affecting the parent pipeline, without creating
another copy.
Fixes: 6b07141f1a ("clutter/paint-nodes: Make paint nodes handle color transformations")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4037>
Instead of just adding the cogl header boilerplate, we'll redirect the
main function with #define and #undef to cogl_main(). The real main
calls the hooks-generated cogl_hooks() which chains up to the users
main() which is now called cogl_main().
<boilerplate>
#define main cogl_main
<user shader>
<hook functions>
void
cogl_hooks () {
cogl_main ();
<hook code>
}
#undef main
void main () { cogl_hooks(); }
This allows the user shader to continue using shaders which seem like
they define the main function and output to the framebuffer but also
gives the CoglPipeline a chance to add hooks.
This in particular makes our ClutterColorState transform hooks work on
user programs which are used by ClutterShaderEffects.
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/7804
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/7805
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3662
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4037>
To avoid mostly going through struct fields in macros as we might soon
move those first to a WindowConfiguration & maybe even make the window
struct private with time.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4111>
As it ends up using the global default context. So just call that
directly to easily spot the remaining usages of get_default_context.
Note that the helper will probably be removed later this cycle once the
remaining usages in meta & libst have been replaced with passing around
the context
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4077>
As they are not used anywhere, in the next commit
we will just remove the whole thing and use glib helpers directly as
there is nothing specific about ClutterThread anymore
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4077>
cogl_framebuffer_finish can result in a CPU-side stall because it waits for
the primary GPU to flush and execute all commands that were queued before
that. By using a GPU-side EGLSync we can let the primary GPU inform us when
it is done with the queued commands instead. We then create another EGLSync
on the secondary GPU using the same fd so the primary GPU effectively
signals the secondary GPU when it is done rendering, causing the latter
to wait for the former before copying part of the frames it needs for
monitors attached to it directly.
This solves the corruption that cogl_framebuffer_finish also solved, but
without needing a CPU-side stall.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4015>
This adds meta_egl_create_sync and meta_egl_destroy_sync to be able to
create and dispose EGLSync objects, respectively, as well as
meta_egl_wait_sync to be able to wait for an EGLSync on the GPU.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4015>
The hook would be optional if anything called `meta_plugin_get_info()`,
except that since commit 4fdbb466e1 ten years ago, nothing has done so.
It seems a bit pointless to export information that is completely
unused, so stop doing that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4104>
The hook would be optional if anything called `meta_plugin_get_info()`,
except that since commit 4fdbb466e1 ten years ago, nothing has done so.
It seems a bit pointless to export information that is completely
unused, so stop doing that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4104>
While in double buffering we only care about one previous presentation,
triple buffering will sometimes need to refer to the presentation from
two dispatches prior. So it might help to separate those frame stats
more clearly. This way each frame's dispatch and presentation times are
stored more cohesively such as to not be overwritten during overlapping
frame lifetimes.
Having two types of frame reference (dispatch and presentation) moving
at difference speeds meant that they could not be stored in a ring. Not
all dispatches become presentations and so storing them in a ring would
necessitate very complex conflict avoidance. Instead, a simple reference
counting model was used.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3961>
So that we maintain a perfectly balanced number of callbacks:
dispatch == notify_ready + notify_presented
Otherwise you can't put any useful logic inside notify_ready and be sure
you're handling all the empty frames.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3961>
e994fbf02 moved warping the pointer to before the destruction of the
resource to prevent dereferencing the constraint after destruction.
This however meant that the constraint was still active when the motion
event caused by the warp is handled, which would constrain the pointer
back again to its original position.
This moves the warping of the pointer back to after the destruction of
the resource and instead just retrieves the seat earlier while the
constraint is still valid.
Fixes: e994fbf02 ("wayland/pointer-constraints: Warp pointer before destroying resource")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3696
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4098>
This reverts commit bad48ea815.
Commit 666e5f1f9 (keybindings: Port to Clutter, 2013-10-03) changed the
Meta.KeyEventHandler callback so that it would take a Clutter.KeyEvent
instead of an XIDeviceEvent. This removed the need for the "(type
gpointer)" annotation, since Clutter works with introspection.
The gpointer annotation caused GJS to simply skip over the callback
argument. Since this was only the second-last argument, reintroducing it
required updating gnome-shell JS code that used the last 'binding'
argument. This was likely inconvenient at the time, since work was
happening in a separate 'wayland' branch, and compatibility with the
X11-only branch had to be kept. So shortly after, the gpointer
annotation was reintroduced with commit bad48ea81 (introspection: Make
MetaKeyHandlerFunc compatible with master, 2013-12-09), unbreaking the
gnome-shell JS code.
But now that the 'wayland' branch has long been merged, we can fix this
properly by removing the gpointer annotation and updating the JS
handlers instead.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4079>
The script is used to install dependencies that are needed both
for the CI image and the gnome os system extension.
To avoid installing dependencies in an environment where they are
already satisfied, they should be guarded by an appropriate check,
most commonly pkgconf.
Right now these checks are straight-forward:
- if a dependency is not in the base image, it will
be built when building the CI image
- if a dependency is not in GNOME OS, it will be
built when building the GNOME OS system extension
However we will soon add another caller of the script, when
allowing to export a system extension from our development
toolboxes to the host.
That case is trickier, as the dependency *is* satisfied by
the build environment (because it is based on the CI image),
but it will still be needed by the exported system extension.
To account for that case, add a pkgconf wrapper that only searches
in PKG_CONFIG_DIRS in the specified destdirs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4075>
When building a system extension for GNOME OS, any extra
dependencies are needed both in the container (for building)
and the extension's destdir (for running).
Because of that, the --destdir flag that was added in commit
8aeb6dc86 does not only install to the specified directory
as expected, but also runs a second install step without
destdir (i.e. to the system).
However that behavior is not a good fit when we extend our
existing toolbox tooling to build system extensions for
Fedora instead of GNOME OS.
To account for that, make the --destdir option cumulative
and install the project to all provided destdirs (or /
if omitted).
This gives us the flexibility to install to the system, a
different destdir, or both:
```
$ ./install-meson-project.sh # install to system
$ ./install-meson-project.sh --destdir /new/dest # install to destdir
$ ./install-meson-project.sh --destdir /new/dest --destdir / # both
```
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4075>
It will be used to force the color state of all monitors to be as if
they managed to turn on HDR mode. This is useful for debugging shaders
in an environment where an HDR mode capable display is not available,
such as the nested mode.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4035>
During an onscreen swap, the cogl_onscreen_swap_buffers_with_damage()
function ensures that the Cogl renderer creates a sync object every
frame. This sync object is later shared with the Wayland clients that
utilize the explicit sync protocol.
However, in headless mode, the function mentioned above is not called.
As a result, the sync object the Cogl renderer stores seems to be not
created. This causes cogl_context_get_latest_sync_fd() function to
return an invalid sync fd, causing Mutter to not be able to materialize
the sync timeline point that the clients wait for when the explicit sync
protocol is in use.
This change simply adds a call to the cogl_framebuffer_flush() function
to the offscreen swap path to make sure that there is a sync object that
can be shared with the clients, which will be signalled when all the
queued operations before the swap are completed.
Signed-off-by: Doğukan Korkmaztürk <dkorkmazturk@nvidia.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4056>
While the events when processing keybindings are quite obviously
key events, the individual structs are all opaque and all public
API is provided by the generic ClutterEvent type. That means that
any handler that uses the passed in event must cast it to
ClutterEvent anyway, so drop the cast to ClutterKeyEvent when
processing key events.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4080>
The option is advertised as alternative to --help, but doesn't
actually work. While not super important for a script that isn't
user facing, it's a trivial fix and doesn't hurt.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4074>
As that is where the whole text rendering integration happens
And would allow us to get rid of some over-abstraction in cogl-pango,
simplify
ClutterSnapshot integration as well
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4004>
As we will be merging CoglPango inside Clutter, so prepare for that
Note this removes the Cogl shared atlas debug usage, as it is private to Cogl
The usage of cogl_c_args was replaced with cogl_debug_c_args to avoid COGL_COMPILATION
pre-moving to clutter. The whole meson file will be removed later anyways.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4004>
Instead of duplicating a string we own already we can just steal it from
the array that we're using.
This is safe since we're sure about the tokens GStrv length and we are
always stealing the last element, so there is no risk that g_strfreev
would eventually leak something.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4071>
Previously ClutterStageWindow was an interface with only one base
implementation (MetaStageImpl) which others inherit from.
This just makes it a class so that we can use _GET_CLASS() API instead of
the costly (by comparison) _GET_IFACE() vtable lookups.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4048>
The meta_input_device_get_wacom_device function may return NULL, so
we should check it before use. I'm not 100% sure that the code as
currently written could reach here with a NULL wacom_device, but
that could change in the future.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4070>
clutter_primaries_to_wayland made sense when there only existed
ClutterColorspace. Now that ClutterPrimaries also exist, it makes more
sense to change that func to clutter_colorspace_to_wayland.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4062>
Quoting Ebassi https://www.bassi.io/articles/2023/02/20/bindable-api-2023/:
Whenever you’re describing a function that takes a callback, you
should always annotate the callback argument with the argument that
contains the user data using the (closure argument) annotation
You should not annotate the data argument with a unary (closure).
The unary (closure) is meant to be used when annotating the callback
type
Recently gobject-introspection became a bit more strict with this and
that generated some warnings:
Warning: Cogl: invalid "closure" annotation: only valid on callback
parameters
This commit fix all the closure annotations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4058>
In some use cases there is a need to dynamically change the preferred
primary GPU, or get rid of the preference altogether. This is currently
not possible due to a change in udev introduced by systemd v247. This
version made the tags "sticky", meaning there is no way to remove them
once attached. When a tag gets removed, only the CURRENT_TAGS property
reflects that change, the removed tag will remain in the TAGS property.
This also bumps libgudev version to 238, since that version introduces
a function, which we need to get the current tags.
Related: https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1562
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4055>
This TF can't be defined as a TransferFunction enum because it needs a
gamma_exp value too.
Add to EOTFType enum a new type: EOTF_TYPE_GAMMA.
With this new type, now EOTFs are unions that can have either
a TransferFunction enum or a gamma_exp.
Set gamma_exp as uniform.
Add the support of it in the color management protocol.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4020>
These properties now are tagged unions:
- ClutterColorimetry:
Can be from colorspace or primaries;
- ClutterEOTF:
Can be from known tf or custom gamma exp (next commit);
- ClutterLuminance:
Can be defined explicitly or derived;
Make the color management protocol use them too.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4020>
ColorState is inmutable so the GObject properties aren't necessary.
Also move ClutterColorstate and ClutterTransferFunction enums to
color-state.h now that they are not used as GObject properties.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4020>
For non-debug and non-plain cases - i.e. mainly release builds.
This ensures we use the same options in all places and draws a cleaner
distinction between g_assert() and g_return_if_fail() - the later will
still be done in release build while the former are meant for debug
only.
One advantage of doing this is that it allows us to use non-trivial
asserts more generously, such as calling `g_list_length()`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3424>
Add a test that:
- Creates 2 clients
- Leases a connector using the first client
- Tries to lease the same connector using the second client
- Checks that the first lease succeeded
- Checks that the second lease failed
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4031>
Add a test that:
- Creates 2 clients
- Leases a connector using the first client
- Checks that the first client receives a lease_fd event
- Checks that both clients receive a connector withdrawn event
- Finishes the lease
- Check that both clients have access to the connector
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4031>
Add a test that:
- Creates 2 clients
- Releases a device for each client
- Checks that releasing a device for one client doesn't affect the other
- Checks that an error is raised if a released client is used
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4031>
Add an event queue to the DRM lease client that allows to check that the
compositor sent the expected events in the expected order.
For the moment, the event queue is used to check that the initial
connection to the compositor works as expected.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4031>
Add the most basic possible test that checks that:
- The DRM lease protocol is available
- At least one device is advertised
- At least one connector is available for lease
Organize the code in a way that it is possible to created multiple DRM
lease clients at the same time as future tests will need this.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4031>
Add the required files to run DRM lease using the KVM/tty backends.
DRM lease tests need at least one connector available for lease. Instead
of adding the test to "wayland-unit-tests.c", add it to its own file and
set the environment variable MUTTER_DEBUG_LEASE_CONNECTORS=Virtual-1 to
make the only connector created by VKMS by default available for lease.
This commit doesn't add any tests, just the required files to start
adding them.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4031>
MetaWaylandDrmLeaseDevice and MetaWaylandDrmLeaseConnector hold a
reference to each other.
In both cases, the reference count was increased. Do not increase the
reference count when lease_connector->lease_device is stored to break
the reference count cycle.
Fixes: fb08a597e1 ("wayland/drm-lease: Advertize initial connectors")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4031>
drmModeAddFB() doesn't take a format, but depth and bits per pixel.
These can be used to determine whether there should be an alpha channel
or not, and is roughly assumed to result in either XR24 or AR24 if one
passes 24 or 32 as depth, with 32 as bpp.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3926>
In multi-GPU setups, when one of the GPUs is used to render the desktop
and the other one has all of their devices available for lease,
meta_kms_impl_device_list_lessees() can be called after the device fd
has been unhold.
Make sure to hold and unhold it in this function.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4030>
When running with COGL_DEBUG=show-uniforms, display the values of the
uploaded uniforms.
This helps because it's not possible to know the uniform values only from
the shader source.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3965>
This function returns a string of the boxed value considering type, size
and count; and adds a name.
e.g:
vec2 scale = vec2(1.0000, 1.0000)
Also handle NULL cases on _cogl_boxed_value_equal.
This will be used by the next commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3965>
Use _cogl_boxed_value_destroy to free the correspondent array instead of
free, which was only freeing float_array.
Add a new func _cogl_boxed_value_array_alloc to allocate the
corresponding array considering the type.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3965>
_cogl_boxed_value_transpose is only used for the matrix type.
Now is named _cogl_boxed_value_copy_transposed_value.
Gets the transposed value considering count (to store in array type or not).
The caller doesn't have to worry about count anymore.
Instead of calling memcpy, add function _cogl_boxed_value_copy_value
which copies the value to the corresponding boxed value property
considering type and count.
The caller doesn't have to worry about type and count anymore.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3965>
There are three array properties, considering type:
1. float *float_array
2. int *int_array
3. void *array
Drop using array. This was inteded to be used when type is matrix and
count > 1, however some functions use it and others use
float_array. Fix this using always float_array, this starts to make all funcs
consistent handling the properties.
There's a mem leak when destroying the boxed value, only freeing the array
prop. Free the corresponding array considering the boxed value type to fix it.
_cogl_boxed_value_set_x func is still inconsistent, the next commits
will solve it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3965>
Colorspace transformation matrices were hardcoded considering only known
colorspaces like bt709 or bt2020.
Now that a colorspace can be defined from its primaries, allow getting the
colorspace transformation matrix from them and drop the hardcoding.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3948>
This is because next commits will generate the colorspace
mapping matrix from the colorspace primaries and won't be required
to define any color space mapping matrix.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3948>
The single pixel buffer allows setting colors with 32 bpc.
Attempt to use pixel formats that allow higher precision than BGRA_8888.
First attempt to use half float format, fallback to ABGR_2101010 and
finally fallback to BGRA_8888.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3913>
lease_disappeared() was calling meta_drm_lease_disappeared() and
set_connectors_as_available().
meta_drm_lease_disappeared() emits the "revoked" signal, connected to
on_lease_revoked(), which marks the leased connectors as available by
calling set_connectors_as_available().
Therefore, calling set_connectors_as_available() again from
lease_disappeared() is redundant.
Remove the redundant call to set_connectors_as_available() and drop two
unused functions.
Fixes: fc44437876 ("wayland/drm-lease: Withdrawn leased connectors")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3675
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4029>
Because for `COGL_DRIVER_GL3`, `_cogl_context_get_gl_extensions` needs to
know that the GL version *really* is >= 3.0 before it calls `glGetStringi`
which didn't exist prior to GL 3.0 or ES 3.0.
This was causing crashes on Xilinx Mali implementations that only support
ES 2.0 (hence `glGetStringi` == NULL), but were being forced to call
that function before the GL version check which should tell you the
function isn't supported.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4019>
The NULL check was inverted, meaning we'd grab with no leader device.
That meant updates coming from the what-should-have-been leader device
getting lost because incorrectly being classified as non-leader of
the grab.
Fix this by only allowing to grab if we have a device, and always mark
the current tool device as the grab leader.
Fixes: e4004a7c4f ("wayland: Use the tool's current_tablet device instead of caching it")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4033>
GLES 2.0 does not have RGB8 and RGBA8 as sized internal formats. There
is OES_rgb8_rgba8 which adds RGB8 and RGBA8 but only for
RenderbufferStorageOES and not for TexImage2D which I wrongly assumed.
It seems like there is currently no GLES2 extension which adds RGB8 and
RGBA8 to TexImage2D so we have no choice but to fall back to unsized
internal formats in those cases as long as we don't want to drop GLES2
support.
This should be fine in practice and we should get our 8bpc textures.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3680
Fixes: 7f943613a8 ("cogl: Use sized internal renderable formats")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4036>
meta_wayland_color_management_dispose func is only called when the compositor
is shutting down, in that case the wl_globals are already automatically removed.
meta_wayland_color_management_dispose calls wl_global_remove again,
this makes a SIGSEGV when color_management is enabled and mutter is being
shut down.
Stop calling wl_global_remove to fix it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4032>
XPending() will do a recvmsg() syscall if there are no items in the queue.
In most cases, this is unnecessary because we know that there is data to be
read of the connection or there are items already read which simply need
to be processed.
Discovering both of those conditions can be done without recvmsg() in the
hot paths.
Before this path, every iteration of the main loop had the potential to
submit a recvmsg() syscall. This reduces that overhead drastically.
XFlush() on the other-hand knows if it needs to write data or not and will
do no IO in the case the buffer is empty.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3653
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4006>
We only need to wake up the other side of the GAsyncQueue if we transition
from 0 to 1 item in the queue. Otherwise, we can be certain that the other
side has received a wakeup and will eventually flush the queue.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4007>
The wrong type resulted in a crash when returning from the function,
because g_slist_free was called instead of g_free for the old_struts
list data pointers.
Fixes: e005d035c0 ("boxes: Define cleanup function for MetaStrut and use auto-pointers")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4028>
Track toplevels being saved, and save state some time after. This
will make session state somewhat remembered on shell crashes, as
long as there was time to snapshot the data in disk.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3825>
Allow saving the session gvdb file in the background, with as little
overhead in the main thread as possible. We still need to serialize
all created/deserialized MetaSessionState to a GVDB hashtable there,
in order to avoid these being poked from the async task thread.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3825>
The xdg_session_manager_v1 global interface is the generator
of xdg_session_v1 objects for clients. These will notify of an
unique ID that can be used for future instantiations.
Once a xdg_session_v1 object is obtained, toplevels can be added
to be managed by it, and clients may get a hint about whether the
toplevel was restored to a saved state.
Changes by Carlos Garnacho: Integrate with MetaSessionManager core
object. Flesh out event emission of xdg_session_v1 and
xdg_toplevel_session_v1 objects, handle sessions being
replaced/deleted.
Changes by Sebastian Wick:
* make lifetimes of xdg_sessions entirely determined by the wayland and
handle its destruction via the signal
* fix session destruction vs deletion
* do not drop refcount of replaced session state temporarily to make
sure the replacing session keeps the state
* disconnect signals of destroyed and replaced sessions
* disconnect window-unmanaging signal handler for
MetaWaylandXdgToplevelSession
* call wl_resource_destroy in xdg_toplevel_session_remove to make it a
destructor
* handle session being destroyed before topevel-sessions
* handle the toplevel going away before the topevel-sessions
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3825>
This object is a windowing-specific implementation of MetaSessionState,
allowing to save window state for toplevel surfaces of a Wayland client
using the xdg_session_management_v1 protocol.
This object is detached from windowing logic itself, and will be
integrated in later commits.
Changes from Carlos Garnacho: Integrate state serialization with
MetaSessionState and MetaSessionManager.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3825>
Make this core object own the MetaSessionManager, for the window management
code to access.
At this level, we will be able to integrate with systemd notification
system, and use systemd fdstore to keep the mapped memory warm for
us for the case of soft reboot. This is at the moment not implemented
here.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3825>
This core object will be the manager of "client sessions", allowing
the windowing-specific paths to generate MetaSessionState objects to
track their clients.
This object is unused at the moment, and will be integrated in later
commits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3825>
This is an abstract base class to implement a "client session",
carrying the accounting of the windows, and allowing to serialize/read
their state into a Gvdb table.
Since different windowing backends may require slightly different
data to be saved for each window, this is meant to have windowing-specific
implementations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3825>
A nick property is a bit similar to the nick of a GObject property, in
that it's a shorter version of the name. It's intended to be used to
store state on the file system, where the state depends on the desktop
environment being used. E.g. gnome-shell sets the name "GNOME Shell",
which is, if no nick is explicitly set, transformed into the nick
"gnome-shell", which will be used for file paths.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3825>
Preparation for next commit, which may merge multiple KMS updates with
sync_fds for modesets. Waiting for all sync_fds to signal before
processing the merged KMS update would be rather involved, for now just
leave implicit sync enabled for it. We're still relying on implicit sync
for modesets in general anyway.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3999>
Previously these tests tried to write to /usr when run as an
installed-test, which happens to work on Gitlab-CI because we're running
as root inside a container, but will not work when running in a more
realistic scenario as an unprivileged user (which is how Debian's
autopkgtest framework runs this test suite).
This also avoids leaving non-package-manager-managed detritus in /usr.
In color-management-tests, we can just delete the code that sets
XDG_DATA_HOME.
In color-management-profile-conflict-test, we also need to copy
the conflicting vx239-calibrated.icc into the temporary XDG_DATA_HOME
to get onto the code path that this test is intended to exercise.
Resolves: https://gitlab.gnome.org/GNOME/mutter/-/issues/3658
Signed-off-by: Simon McVittie <smcv@debian.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4012>
g_array_sized_new() creates a new GArray with a preallocated size, but,
after creation, the array length is still zero [1].
Store the modifiers in a EGLuint64KHR array and use g_array_new_take()
to create a new GArray with the correct size.
Because no modifiers were returned, gbm_surface_create() was used
instead gbm_surface_create_with_modifiers() on multi-GPU setups.
[1] https://docs.gtk.org/glib/type_func.Array.sized_new.html
Fixes: aec85281ba ("native/renderer: Retrieve the right modifiers set for each GPU")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3998>
When a surface is destroyed, the existing feedback surfaces are marked
as inert by setting the wl_resource user_data to NULL. This wasn't
handled in the feedback surface destructor.
Fixes: 2341346c90 ("wayland: Implement the color management protocol v4")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/4000>
The tablet tool is initialized with a device but if that device is later
removed we never update tool->device. This eventually causes a crash
when we're passing that device into
meta_wayland_input_invalidate_focus().
The tool keeps track of the current tablet anyway so instead of caching
this pointer in the tool, use the current tablet's device.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3642
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3959>
`(int) (1.0f * (float) INT_MAX)` doesn't necessarily result in INT_MAX
due to how floating point arithmetics. Handle this better by setting
INT_MIN/MAX explicitly, when the floating point value post scaling
exceeds the corresponding limit.
This fixes resizing of electron windows.
Fixes: 6e8c7c5f84 ("Add experimental mode to use native scaling of Xwayland clients")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3997>
Meson variables don't have a scope, so using `prefix` in a loop
for D-Bus interface prefixes actually overshadows the global
variable that holds the build prefix.
As far as I can tell, the only place where the "wrong" value is
read is the build summary, but let's not tempt fate and fix it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3995>
The value returned by meta_frame_get_frame_bounds() is owned by the
caller so we can assign it directly without copying to avoid a leak and
a bit of unnecessary work. Also it can't ever return NULL, so there is
no need to check for that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3994>
As with Xsettings, we want to use the X11 UI scaling factor to set the
cursor size, so that clients use a larger theme, both when using
`native-xwayland-scaling` and a physical layout mode.
Fixes: 6e8c7c5f84 ("Add experimental mode to use native scaling of Xwayland clients")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3992>
This is different from MetaSetting's UI scaling factor, and different
from the effective Xwayland scale.
The MetaSetting's UI scaling factor is the scaling factor used by
gnome-shell chrome itself.
The effective Xwayland scale is, with `native-xwayland-scaling` enabled,
the scale everything X11 is scaled with.
The X11 UI scaling factor is intended to be the scaling factor X11
clients are told to use, and how to derive that differs depending on the
layout mode and the effective Xwayland scale.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3992>
When !3952 introduced the new tracing macros, they were only defined in
the HAVE_PROFILER case, causing builds without profiler support to fail.
Also it introduced an unconditional call to cogl_is_tracing_enabled()
which is not available without HAVE_PROFILER.
Fixes: 777c63507 ("cogl/trace: Allow defining and setting sysprof trace counters")
Fixes: 322ac42a6 ("stage-impl: Trace the damage region")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3991>
It is intended to replace using GetResources() and ChangeBacklight().
It moves from a normalized 1-100 numbers, to directly exposing the
hardware. This more closely maps to how gsd-backlight.c in
gnome-settings-daemon normally works, and simplifies the API a bit to
not have to deal with rounding issues.
There is still no KMS uAPI for this, so it still only hooks up to
XRANDR. Being private API, it doesn't try very hard to predict how the
KMS uAPI will look. When that day comes, it will likely need some
adaptations.
Part of the motivation here is to get something for gsd-backlight.c to
use where it can work more similarly to how the current common case
(sysefs) works, while attempting to migrate away from libgnome-rr.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3861>
It's now a signal emitted when `meta_output_set_backlight()` is called.
It is also no longer normalized between 1 and 100; that is now handled
done via the D-Bus API layer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3861>
Add new param to request SyncTimeline metadata on buffer that contains
the sync points.
Add new Buffers param depending on the SyncTimeline metadata and that
adds extra space for the syncobj fd.
When we receive a buffer from PipeWire with the SyncTimeline, also set
up the spa_data with the syncobj fd and the sync type. We use the
GenericFd type.
When we render a buffer with SyncTimeline, set the acquire & release
points for the syncobj.
[ Michel Dänzer:
* Flesh out Wim's skeleton and make it compile
* Dequeue explicit-sync buffers only if the release point has signaled
* Tweak commit log
v2:
* Always check spa_buffer->n_datas as well when getting the
SPA_META_SyncTimeline metadata (Doğukan Korkmaztürk)
* Always put space before opening parens.
v3:
* Keep a per-stream list of dequeued buffers with unsignaled release
point. (Wim Taymans)
* Fix indentation of meta_drm_timeline_is_signaled parameters in
dequeue_pw_buffer.
v4:
* priv->dequeued_buffers handling fixes in dequeue_pw_buffer.
v5:
* Fix bumping sync points in maybe_set_sync_points.
v6:
* Guard new code by #ifdef HAVE_NATIVE_BACKEND as needed.
* Remove unused sync_fd in dequeue_pw_buffer.
v7: (Sebastian Wick)
* Refactor can_reuse_pw_buffer helper out of dequeue_pw_buffer.
* Separate loops for priv->dequeued_buffers and pw_stream_dequeue_buffer
in dequeue_pw_buffer.
* Add g_assert after g_hash_table_lookup.
* Prevent calling meta_drm_timeline_is_signaled with non-NULL *error.
* Use g_autoptr & g_steal_pointer for timeline in maybe_create_syncobj.
* Log meta_drm_timeline_create_syncobj / meta_drm_timeline_import_syncobj
errors in maybe_create_syncobj.
* Use g_warning_once instead of g_warning.
* Do not close (spa_buffer->datas[1].fd) in on_stream_remove_buffer,
g_hash_table_remove closes it.
v8:
* Fix pw_stream_dequeue_buffer loop condition in dequeue_pw_buffer.
(Sebastian Wick)
* Use g_clear_error in can_reuse_pw_buffer. (Sebastian Wick)
* Add comment above dequeued_buffers list declaration in
MetaScreenCastStreamSrcPrivate.
v9:
* Call g_clear_error from dequeue_pw_buffer instead of
can_reuse_pw_buffer. (Sebastian Wick)
v10: (Jonas Ådahl)
* Move added #includes to separate HAVE_NATIVE_BACKEND block.
* Add and use enum entries for spa_buffer::datas array access.
* Clean up the end of can_reuse_pw_buffer.
* Change first argument of dequeue_pw_buffer to MetaScreenCastStreamSrc.
* Rename GSList iterator to "l".
* Clean up error handling in dequeue_pw_buffer.
* Fix second parameter to second can_reuse_pw_buffer call in dequeue_pw_buffer.
* Fix coding style of code added in on_stream_param_changed.
v11:
* Move local_error declaration inside HAVE_NATIVE_BACKEND in
dequeue_pw_buffer.
v12: (Jonas Ådahl)
* Use new syncobj_data_from_buffer helper to retrieve acquire/release
spa_data.
* Generally assume spa_buffer->n_datas is large enough, just spot-check
it in strategic places.
* Use GList for dequeued_buffers list.
* Warn in meta_screen_cast_stream_src_dispose if dequeued_buffers list
isn't empty.
v13:
* Simplify empty list test in meta_screen_cast_stream_src_dispose.
(Jonas Ådahl)
v14:
* Refactor maybe_remove_syncobj helper out of on_stream_remove_buffer.
This allows using g_return_if_fail in the former. (Sebastian Wick,
Jonas Ådahl)
v15:
* Guard syncobj_data_from_buffer & enum definitions by
HAVE_NATIVE_BACKEND, since they're unused outside of that guard.
]
[ Jonas Ådahl
v16:
* Destroy the timelines hash table after destroying the stream.
* Don't warn when getting timeline data.
* Fix n_datas sanity check condition.
]
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3876>
Build upstream version 1.2.0 for the CI image if needed.
v2:
* Build upstream 1.2.0 tag instead of 1.2.2.
v3:
* Build upstream 1.2.3 tag for CI, it has a needed fix.
v4:
* Enable systemd support in pipewire build. (Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3876>
Will be used to look up a MetaDrmTimeline object for a syncobj file
descriptor.
v2:
* Add comment above timelines hash table declaration in
MetaScreenCastStreamSrcPrivate. (Bilal Elmoussaoui)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3876>
v2: (Doğukan Korkmaztürk)
* Use drmSyncobjQuery instead of drmSyncobjTimelineWait.
* Fix libdrm function name in error message.
v3:
* Add separate boolean is_signaled out parameter. (Sebastian Wick)
* Use g_strerror instead of strerror.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3876>
Creates a kernel syncobj and returns a file descriptor representing it.
v2:
* Call drmSyncobjDestroy also after drmSyncobjHandleToFD returns 0, or
we leak the original syncobj reference. (Sebastian Wick)
* Add errno based error messages.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3876>
Allow scale-aware Xwayland clients to scale by an integer scale
themselves, instead of letting them render them at 1x scale and then
scaling up the texture, making it look blurry.
When monitor framebuffers are scaled, this special cases Xwayland and
sends output regions in a way that Xwayland think everything is N times
as large as the logical region, where N is the ceil of the max monitor
scale.
This is done by introducing a "stage" vs "protocol" coordinate space for
X11, where the "protocol" coordinate space is "stage" multiplied by a
scaling factor.
Xwayland thus will have its own "effective scale", sent via
wl_output.scale. The effective Xwayland scale is also used for the
internal MetaWaylandSurface scale internally, unless there is a viewport
dst size set on the same surface, in which case the scale is still set
to 1, to not interfere with wp_viewport semantics.
We're guarding this behind a new experimental feature
"xwayland-native-scaling", which can only come into effect when enabled
together with "scale-monitor-framebuffer".
[v2]:
Move stage_to_protocol/protocol_to_stage to generic window class.
This means parts that aren't aware of any windowing system specific
logic, only that coordinates originate from there for a given window,
can still get their coordinates properly handled.
In particular, this means coordinates from IBus that originates from the
client, can be scaled properly when that client is using X11.
Also make them properly introspected.
[v3]:
Split up coordinate transform API.
Make it one that takes a MtkRectangle, and another that takes a point.
This means the rounding strategy becames explicit when transforming a
point, while when it's a rectangle, it's still always "grow".
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3567>
This scale is currently a lie, it doesn't do anything. What it
represents is the current highest monitor scale, and will eventually be
used to, when configured to do so, scale X11 coordinates as well as
coordinates given to Xwayland.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3567>
We know let Xwayland set the RANDR names from the connectors. To stop
relying on layouts and coordinates to match the primary logical monitor,
instead use the connector name of the first monitor.
Also make the X11 client sanity checking check that the right X11 output
is primary as part of the monitor tests.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3567>
With the next commits we'll introduce a new mode for scaling of Xwayland apps,
we'll want to put this mode behind an experimental setting though, so add
that setting.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3567>
This replaces the `legacy-ui-scaling-factor` entry in
`org.gnome.Mutter.DisplayConfig`, with the motivation being to no longer
expose X11 specific state via the monitor configuration API.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3567>
Hooks up the wayland protocol to the color state luminances. The color
state handles the default levels so we can just pass everything through
after we checked for all the error conditions.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3953>
This uses the luminance levels of the color states to anchor the white
of content instead of hard-coding the levels.
This also starts using uniforms for parts of the mapping which means we
don't have to generate and compile a shader when the luminance levels
change.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3953>
They represent the minimum and maximum luminance levels of the primary
color volume and the reference luminance level (reference white, SDR
white, ...) in the reference viewing environment.
They help anchoring the white level, optionally help with preserving the
dynamic range and help with adjusting from a "dark" to a "bright"
viewing environment.
The values have defaults which depend on the transfer characteristics.
This reflects the wayland color management protocol.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3953>
Actions might get detached sometime during event processing,
at a time that the stage did already prepare an emission chain
holding references to the actions and actors that need to handle
events. This means actions might become detached, but still handle
the incoming event, or possible crossing events generated in-place
when the actor becomes unparented.
Avoid this situation, by skipping event handling on actions that
went detached, we will just instruct to continue event processing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3988>
Hiding it in debug logging was a little too hidden. Someone might want
to know why performance has degraded without having to restart in debug
mode hoping they can reproduce the issue.
Also remove an assertion that would issue spurious warnings. We should not
always expect IMPORT_STATUS_NONE (implying the first failure must be on
the first frame). Instead we might start with IMPORT_STATUS_OK for a number
of frames and then have a sporadic failure some time later.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3928>
Instead, get it from the context. See next commit
For ClutterText, we had to switch to using constructed
as the ClutterContext will be set for the ClutterActor in the
constructor phase
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3977>
gnome-shell switched to rst2man to generate its man pages, so
we can drop asciidoc from the image.
It's unclear (at least to me) what's currently pulling in rst2man,
so it seems safer to explicitly pull that in ourselves.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3983>
Or, optionally, when the MUTTER_FRAMES_PLATFORM_LIBRARY env var is set
to "adwaita". For debugging purposes, it is possible to disable loading
any platform library with MUTTER_FRAMES_PLATFORM_LIBRARY set to "none".
Add the CSS class "ssd-frame" to the MetaFrame (i.e. GtkWindow)
instances
as that lets libadwaita pick the right style, and won't do anything for
non-Adwaita styles.
This patch is specially careful not to link against libadwaita.
Closes https://gitlab.gnome.org/GNOME/mutter/-/issues/2830
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3981>
Launching pipewire and wireplumber is racy, as there is an arbitrary
amount of time between pipewire is launched, and that the socket is
bound.
In order to eliminate this race, bind the pipewire sockets ourselves,
and launch pipewire (and wireplumber) when there is activity on the
socket. This is using the systemd method of doing socket activation,
which consists of passing the number of passed file descriptors via
$LISTEN_FDS, and the pid of the launchee via $LISTEN_PID.
The former is easy, just pass the file descriptors, but the former is
more tricky when using python, as executing code before exec() is poorly
supported and likely to be deprecated. To address this, socket
activation services are wrapped in a socket-launch.sh helper which sets
the $LISTEN_PID to itself before calling exec().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3973>
list = ['one']
list.append('two,three'.split(','))
results in ['one', ['two', 'three']]
while
list += 'two,three'.split(',')
results in ['one', 'two', 'three']
which is what is expected.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3973>
While when running with meson, we want the output to be in meson-logs
for convenience, such a path isn't feasible when running as an installed
test. To address this, make the destination path configurable via an
environment variable.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3973>
This aims to replace a reoccuring arbitrary sleep followed by a focus
assertion. The problem with these is that the sleep does not reliably
wait for long enough, due to arbitrary system resource usage in CI and
elsewhere. By instead waiting, and instead asserting we don't have any
intermediate wrong focus, we remove the race, and test for more invalid
situations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3973>
See previous commit log on the effects of this.
This means the deadline evasion needs to be added in both cases in
clutter_frame_clock_notify_presented.
v2:
* Use meta_kms_update_set_sync_fd. (Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3958>
If the KMS thread is using the deadline timer, and a valid sync_file
descriptor is passed in:
1. The update is deferred, and the deadline timer is left armed, until
the sync_fd signals (becomes readable).
2. Implicit synchronization is disabled for the KMS update.
This means cursor updates should no longer miss a display refresh
cycle due to mutter's compositing GPU work finishing too late.
v2:
* Use g_autoptr for GSource in meta_kms_impl_device_handle_update.
(Sebastian Wick)
v3:
* Use meta_kms_update_get_sync_fd, don't track sync_fd in
CrtcFrame::submitted_update. (Jonas Ådahl)
v4:
* Clean up CrtcFrame::submitted_update members in crtc_frame_free.
v5:
* Coding style cleanup in meta_kms_impl_device_handle_update.
(Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3958>
If the window_drag->handle ClutterActor:visible property is FALSE,
then we avoid a full-framebuffer damage on the monitor when beginning
and ending a drag.
Testing with `mutter --wayland --display-server` still shows a full-
framebuffer damage on the first drag, but that appears to be unrelated
to this. Subsequent full-framebuffer damage which would occur on
drag-begin and drag-end have been elided.
Related: #3630
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3975>
Toplevels get the main monitor from their MetaWindow and have no main
monitor when the toplevel is not mapped.
Subsurfaces get the main monitor from their parent surface.
DnD and cursors get the main monitor from the current cursor position no
matter if the cursor is actually being shown or not.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3941>
The main monitor is role dependent. For a toplevel this comes from the
MetaWindow main monitor, for a subsurface from the parent surface, for
pointer and dnd from the cursor position.
The next commit will use meta_wayland_surface_set_main_monitor in the
different roles to keep the property up-to-date.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3941>
As a writable property, nothing sets it in both Mutter/GNOME Shell
making it a dead-code. As we will be moving pango related bits to
gnome-shell, remove this one already
Removes pango-ft2 dependency
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3969>
As GNOME Shell always sets that to FALSE anyways
This does not drop the separate caching whether the mipmapping is
disabled or not, in case shell wants to make use of that in the future
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3962>
So far, the tests relied on the host system to provide pipewire and
wireplumber. This seperates the tests from the host system which is
especially useful if the tests are run in a toolbox which has a
different pipewire installed than the host. It also should make it
harder to have a mismatch between the pipewire library and the pipewire
daemon.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3966>
We're already inhibiting real time scheduling when reading new KMS state
after hot plugs, as well as when during mode sets, due to the kernel not
being able to reliably handle these within the 250 ms limit. However, we
didn't do this during initial probing, which meant that occasionally
we'd run into these kind of issues during startup.
Handle this by always inhibiting real time scheduling up front, and
don't uninhibit until all initially discovered device have finished
processing their initial mode set.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3628
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3960>
If an application is maximised, clutter_stage_pick_and_update_device()
goes into the
if ((flags & CLUTTER_DEVICE_UPDATE_IGNORE_CACHE) == 0)
condition and returns the current actor for the device. This means no
CLUTTER_LEAVE/ENTER events are generated and in turn means the focus is
never invalidated and updated.
This leads to tool->focus_surface always being NULL and all events are
discarded.
Notably, tool->current is set to the right surface but
that one never changes either so meta_wayland_tablet_tool_set_current_surface()
exits early too because surface == tool->current and we thus never call
meta_wayland_input_invalidate_focus().
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3616
Fixes: fb8ac5dff7 ("wayland: Track current tablet tool focus surface")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3956>
Clutter's "device-removed" signal is sent in
clutter_seat_handle_event_post(). Our tablet code is set up to handle
that signal to then notify wayland clients of removed tablet devices.
However, returning CLUTTER_EVENT_STOP for a DEVICE_REMOVED event means
we never get to the point where we send out the signals and thus never
remove the tablets.
Fixes: a37fd34bbb ("wayland: Make MetaWaylandSeat in charge of its own tablet seat")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3615
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3955>
This leads to possibly missed key release events being propagated down to
clients, and in the case of X11 clients, to stuck keys e.g. after switching
workspace with Ctrl+Alt+Left/Right and ending up with a X11 client focused.
Fixes: 11a4d56185 ("keybindings: Send trigger when a key accelerator is deactivated")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3935>
And take it into account in meta_kms_crtc_get_deadline_evasion.
This uses the same fundamental approach as clutter frame clock scheduling:
Measure the deadline timer dispatch duration, keep track of the longest
duration, and set the timer to fire such that the longest measured
dispatch duration would result in it completing shortly before start of
vblank.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3612
v2:
* Move DEADLINE_EVASION_CONSTANT_US addition from
meta_kms_crtc_determine_deadline to meta_kms_crtc_get_deadline_evasion.
* Calculate how long before start of vblank dispatch completed for
debug output in crtc_frame_deadline_dispatch.
* Shorten over-long lines in crtc_frame_deadline_dispatch.
v3:
* Take VRR into account in crtc_frame_deadline_dispatch &
meta_kms_crtc_update_shortterm_max_dispatch_duration. (Robert Mader)
v4:
* Check if deadline has already passed in meta_kms_crtc_determine_deadline,
set the deadline for one refresh interval later if so.
* Fix indentation in crtc_frame_deadline_dispatch.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3934>
Using GL_BGRA8_EXT as internalformat for TexSubImage2D was not allowed
in the EXT_texture_format_BGRA8888 extension. This changed recently:
https://registry.khronos.org/OpenGL/extensions/EXT/EXT_texture_format_BGRA8888.txt
1.4, 23/06/2024 Erik Faye-Lund: Add GL_BGRA8_EXT for ES 2.0 and later.
Mesa already supports this which is why 7f943613a8 ("cogl: Use sized
internal renderable formats") worked as intended. Technically spec
compliant and our CI had an up-to-date driver.
So while this is no bug, it's still not great because older drivers will
generate GL errors. This commit changes this specific format back to an
unsized internal format which means we could, in theory, get less than
8bpc framebuffers.
We can try to revert this commit when newer driver versions have
propagated far enough.
Fixes: 7f943613a8 ("cogl: Use sized internal renderable formats")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3954>
Let's add a bunch of tests for the newly introduced monitor config
migration code. We're verifying that:
- monitor configs get migrated and monitors.xml is updated on startup
- existing monitor configs in LOGICAL layout mode (ie.
"scale-monitor-framebuffers" was enabled before updating mutter) get
migrated
- the migration of various more complex monitor layouts from PHYSICAL
to LOGICAL layout mode works
- for monitor layouts where conversion to LOGICAL fails, the PHYSICAL
layout mode still is migrated, while for LOGICAL at least primary
monitor and disabled monitors are preserved
- simple monitor configurations (with no scaled monitors, or with only
"irrelevant" scaled monitors at the end of the layout) do not go through
the conversion code paths
- monitor configs in PHYSICAL layout mode with integer scale factors
which will result in non-integer logical monitor sizes get converted
to the closest fractional scale factor
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3596>
We don't want to do the work of the layout mode detection and conversion
every time we read the monitors.xml file.
Instead, when the detection logic is used, set a flag to automatically
update the config files after the parsing is finished.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3596>
Introduce some "best effort" conversion code to migrate monitor configurations
from PHYSICAL (the old default) to LOGICAL (the new default on wayland)
layout mode.
This conversion will only be used when the old PHYSICAL layout-mode
configuration is not compatible with the new LOGICAL layout-mode one.
This only applies if 1) there's a monitor that needs scaling in the
layout, and 2) the scaled monitor comes before other monitors in the
coordinate system (ie. it's not the rightmost or bottommost monitor).
There are two algorithms added here to convert monitor layouts:
- One for "simple" 1-dimensional monitor layouts, where all monitors are
aligned on a vertical or horizontal strip.
Here's a few (inaccurate) examples of how this would look with different
layouts (left side is PHYSICAL, right side is LOGICAL, x is the origin of
the coordinate system, the numbers are scales of the monitors):
```
x──────┬──────┬──────┐ x ┌──────┐
│ 2 │ 1 │ 2 │ 2┌──┤ 1 ├──┐2
│ │ │ │ ──► └──┤ ├──┘
└──────┴──────┴──────┘ └──────┘
x ┌──────┐ x ┌──────┐
┌────┤ 1 │ ┌──┤ 1 │
│ 2 │ │ ──► └──┤ │
└────┴──────┘ 2 └──────┘
x ┌────┐
┌──────┤ │ x──────┐
│ │ │ │ ├─┬────┐
│ 1 │ 3 │ ──► │ 1 │3│ 1 │
│ │ │ │ ├─┴────┘
└──────┤ │ └──────┘
│ ├────┐
│ │ 1 │
└────┴────┘
```
- A second more complex algorithm for 2-dimensional monitor layouts with
a common baseline that all monitors are aligned to.
And examples for this one:
```
x ┌──────┐
┌──────┤ │
│ 1 │ 2 │ x──────┐
│ │ │ │ 1 ├────┐
└──┬───┴───┬──┘ ──► │ │ 2 │
│ 3 │ ├──┬───┴────┘
│ │ └──┘3
└───────┘
x ┌──────┬──────┐
│ 1 │ │
│ │ │ x──────┬──────┐
┌─────┴──────┤ 1 │ │ 1 │ │
│ │ │ │ │ │
│ │ │ ──► └──┬───┤ 1 │
│ 3 │ │ │ 3 │ │
│ ├──────┘ └───┤ │
│ │ │ │
│ │ └──────┘
└────────────┘
x ┌───────┐
┌──────┐ │ │ x ┌───────┐
│ 2 │ ┌──────┤ 1 │ │ │
│ │ │ 1 │ │ 2 ┌──────┤ 1 │
└────┬─┴────┴─┬────┴───────┘ ──► ┌──┤ 1 │ │
│ │ ├──┴┬─────┴───────┘
│ 2 │ │ 2 │
│ │ └───┘
└────────┘
```
These algorithms will fail for any more complex 2d monitor layout, eg.
```
x ┌───┬────┐
│ 2 │ 1 │
│ ├────┘
┌───┴┬──┘
│ 1 │
└────┘
x───┬───┬───┐
│ 1 │ 2 │ 1 │
├───┼───┼───┤
│ 1 │ 1 │ 1 │
├───┼───┼───┤
│ 1 │ 1 │ 1 │
└───┴───┴───┘
```
In those cases where the conversion failed, we fall back to aligning
the monitors on a horizontal line, preserving the scale, the primary
monitor and the disabled monitors.
Note that we also need to convert the scale factor in some cases,
because LOGICAL layout mode also behaves different here:
When the scale results in a fractional logical monitor size (eg.
the native monitor width is 2560px, and a scale of 3 is set =>
2560px / 3 = 853.333px), in LOGICAL mode we won't use that scale.
Instead we have an algorithm (see
meta_monitor_get_closest_scale_factor_for_resolution()) to find
the nearest fractional scale factor which doesn't result in
fractional logical monitor size. We reuse this algorithm here for
the conversion.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3596>
We'll reuse meta_monitor_get_closest_scale_factor_for_resolution() for the
conversion of monitor configs, and during those conversions, we probably don't
want to impose the same limits to fractional scales that we usually impose.
This means that we can even convert physical layout configs where the user
manually changed to a value higher than what our fractional scale calculations
would allow.
Move this check into the calling function so that it's not imposed by
meta_monitor_get_closest_scale_factor_for_resolution() directly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3596>
When there is no layout mode set in a logical monitor config, we currently just
assume the configuration matches the mode that the system expects. This blows
up when the layout mode expected by the system changes (eg. by turning on
"scale-monitor-framebuffers" in mutter): Suddenly configs fail the validation
check and get thrown away.
Since we now can add one configuration for each layout mode to the config store,
we can do better here: Let's only add configurations to the store where we
verified beforehand that the monitor layout is compatible with that mode, either
because we set it ourselves using the <layout_mode> key, or by detecting which
modes the layout is compatible with.
Also update monitor config ifiles to adjust for the new layout_mode, as
they all are assumed to be "logical".
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3596>
Verify even more assumptions we make about logical monitor configs:
- Have a more explicit check that the monitor modes in the logical monitor are
all equal
- Complain if scale factor with physical layout mode is fractional
- Make sure that scale factor with logical layout mode actually scales to a
non-fractional width and height
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3596>
We'll need a few of those things from the monitor config store soon, also it's
generally useful to have a prefix which makes it clear where functions are
defined.
So factor some things out into a new monitor-config-utils.c file.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3596>
Store and load the layout mode for each logical monitor configuration in
monitors.xml by introducing a new <layoutmode> element. The value of the
element can be either "logical" or "physical". The layout mode is also
made part of the monitor configuration key.
Right now this isn't doing a lot:
When no <layoutmode> is found on the config (this is the case with all
existing configs), we'll keep using the layout mode expected by the system,
without updating the config file.
When changing an existing, or introducing a new configuration, we'll now
store the current layout mode with the config though, and load it again
on the next start of mutter. This is still not problematic as long as
mutters expected layout mode doesn't change (eg. by turning on/off
"scale-monitor-framebuffers").
When the expected layout mode of mutter switches between
restarts, the monitor config is now still loaded but remains unused,
and mutter will create (and store) a new one with the other layout mode.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3596>
We'll introduce some new migration code with the next few commits to introduce
a layout_mode property in monitors.xml. This will be significantly easier
without keeping around the old monitor migration code, so drop it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3596>
We have meta_verify_logical_monitor_config() already, and it does a few checks that
meta_verify_monitors_config() doesn't do yet, so let's also call
meta_verify_logical_monitor_config() when verifying the whole config.
We'll rely on this being part of meta_verify_monitors_config() soon, because we'll
stop calling meta_verify_logical_monitor_config() from the config parser.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3596>
We forgot to check whether multiple groups of monitors are actually
all connected with each other, so fix that.
[jadahl: Rewrote algorithm to detect split groups]
[jadahl: Added test case]
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3596>
There is one define for the format (GL_BGRA_EXT) and one for the
internal format (GL_BGRA8_EXT). Use them appropriately.
This also adds defines to consistenly not use the _EXT postfix.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3914>
The opaque fp16 Cogl format variants need a required format that is
already premultiplied whereas the fp16 formats with an alpha channel can
be either straight or premult.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3914>
The first event happening for a new touch will be the CLUTTER_ENTER
event generated when picking it. Use this event for registration of
the touch info, so that MetaWaylandEventHandler.get_focus_surface may
get the right focus surface for the device/sequence on the first try.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3889>
And notably, don't cancel touch when an event handler is being
removed. Touch events are largely unaffected by most Wayland
grabs (pointer constraints, popups), so we might be cancelling
input too early if one of these wayland grabs was effective when
touch interaction began, but stopped sometime between touch updates.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3889>
Touch events are implicitly grabbed to the surface they began in,
and are not affected by the focus handler. However these events
will appear to come from the core pointer device, which might lead
to the wrong device being updated.
Ignore events with a sequence, since the default focus handler
does not intend to do anything with them.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3889>
On a hybrid machine with i915 primary and nvidia-drm (470) secondary,
`meta_render_device_egl_stream_initable_init` calls
`meta_kms_inhibit_kernel_thread` to change from the default 'kernel'
thread type to 'user'. And soon after that it would
`meta_render_device_egl_stream_finalize` because I'm not actually
using that GPU, and calls `meta_kms_uninhibit_kernel_thread`.
So during startup Mutter would default to a realtime kernel thread,
switch to a user thread (which doesn't support realtime), and then
switch back to a realtime kernel thread.
In the middle of all that, while the thread type was 'user' and
realtime disabled, something was invoking `ensure_crtc_frame` which
created a `CrtcFrame` without a deadline timer. Soon after that the
thread type changed back to 'kernel' with deadline timers expected, but
our existing `CrtcFrame` has no deadline timer associated with it. And
so it would never fire, causing the cursor to freeze whenever the primary
plane isn't changing. And the problem was permanent, not just the first
frame because each `CrtcFrame` gets repeatedly reused (maybe shouldn't
be called a "Frame"?).
Now we adapt to switching between kernel and user thread types by adding
and removing the deadline timer as required.
Close: https://gitlab.gnome.org/GNOME/mutter/-/issues/3464
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3950>
Colord is a system service which will result in a polkit dialog showing
up when connecting a remote session.
We want to get rid of colord eventually anyway, so disconnecting virtual
monitors from colord isn't an issue.
Fixes: f5ce2ddf3c ("color-manager: Create color devices also for virtual monitors")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3942>
If we finish compositing in time, the composited result will be
submitted prior to the deadline timer is triggered, and we'll be fine,
and if not, at least the cursor updates will be smooth, which makes it
appear smoother than not.
There is a risk that this can negatively impact composited updates when
moving the cursor, so make it possible to toggle a paint-debug flag for
now until this has been more tested.
This also mean we need to disarm the deadline timer after handling
update, as there might be a scheduled cursor update pending, but we
already handled it, so disarm the timer.
Here is an illustration of the difference.
In the following scenario, with disarming, the composited frame E, and
the cursor movement C gets presented. With this branch, only the cursor
movement C gets presented.
```
* A: beginning of composited frame
* B: begin notification reaches KMS thread
* C: cursor moved
* D: calculated deadline dispatch time (disabled with the branch)
* E: KMS update posted
* F: KMS update reaches KMS thread
* G: actual deadline (and with branch and gets committed)
Compositor thread: --------A---------------E---------
\ \
\ \
KMS thread: -----------B------C----D---F-G----
```
In the following scenario, by not disarming, the cursor update C will be
presented, and the would-be-delayed composited frame E would be delayed
anyway, i.e. fixing cursor stutter.
```
* A: beginning of composited frame
* B: begin notification reaches KMS thread
* C: cursor moved
* D: calculated deadline dispatch time (and with branch will be dispatched)
* E: KMS update posted
* F: actual deadline
* G: KMS update reaches KMS thread (and with branch gets postponed)
Compositor thread: --------A---------------E---------
\ \
\ \
KMS thread: -----------B------C----D-F-G------
```
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3184>
Although we track updates for EGL_DEVICE, they are often empty because
the primary plane has a custom page flip method. That means there's
no CRTC latched yet, but we do know exactly which CRTC is associated
with the flip. Set it so the update can still be processed.
Fixes: 27ed069766 ("kms/impl-device: Add deadline based KMS commit scheduling")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3939>
If other handlers (e.g. DnD) are on top of the popup grab focus, we
may want it to move outside same-client surfaces as the popup grab
specifies.
Check that it is the current handler before making same-client checks,
so that these handlers on top have an opportunity to find other
surfaces, e.g. during DnD from a popup.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1681
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3937>
Move away from tracking presses/releases directly, since there might be
other event handlers on top that might prevent the popup event handler
to fully track all events. The replacement is using event state modifiers,
which will use information set from the backend, and is enough to determine
there's no more pressed buttons without tracking prior event history.
This makes the popup event handler able to interact with other event
handlers that might be on top, and consume button release events for
themselves (e.g. DnD), no longer resulting in a stuck popup grab.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3937>
The relationship between MetaKmsConnector and MetaDrmLease is already
stored in MetaDrmLeaseManager::leased_connectors.
Change the type of MetaDrmLeaseManager::connectors to a GList.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3922>
As in the protocol definition for wp_drm_lease_connector_v1::withdrawn:
Sent to indicate that the compositor will no longer honor requests
for DRM leases which include this connector. [...] Compositors are
encouraged to send this event when [...] the connector gets leased
to a client.
Withdrawn the leased connectors and, if they are available once the
lease finishes, advertise them again.
Related to: https://gitlab.freedesktop.org/wayland/wayland-protocols/-/merge_requests/322/
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3922>
And stop passing in the color states from the RendererNative. We also
keep the color states updated by listening for changes in the color
device.
The RendererX11Cm has a single view and no mapping to a specific color
device, so we handle the absense of a color device as well and rely on
ClutterStageView to have the default color states.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3930>
This allows us to destroy and create a new offscreen dynamically, when
the rotation or color state changes.
An idle gsource with priority higher than CLUTTER_PRIORITY_REDRAW is
used to ensure the an offscreen exists when required without having to
allocate in the redraw process.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3930>
This is usually an indication of a failed drop operation if the
drop site didn't request any target. Check for this specifically
on XDnD button release, so that we can cancel the DnD operation
right away.
Inspired on a fix from Jonas Ådahl.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3821>
We were relying on the drag source surface to keep receiving events
thanks to it being implicitly grabbed by the button press. This
broke at some point, making the Xdnd drag source unable to keep
directing the DnD operation as it is expected by X11 clients.
To fix this, make the Xdnd MetaWaylandEventInterface stick itself
to the drag source surface keeping the focus of the device driving
the DnD operation, so that the X11 client can still handle events
from it.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3511
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3821>
Currently, we would only disable a11y if a certain flag is passed
but the function is always called with NONE flag. Instead
drop the flag, use a new environment variable for that
That value is then used by actors to short-circuit get_accessible
implementation and return NULL if the accessibility is not enabled
Also clean the other accessibility functions
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3917>
As GNOME Shell has a more complete solution with StScrollable /
StViewport / StScrollbar and does not make use of the Clutter actor.
We might want to "upstream" GNOME Shell infrastructure later at some
point but the current "solution" is too poor, so just drop it
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3917>
As they are better fit in ClutterActor
The accessible_role is intentionally put in the public fields
as ClutterActorAccessible needs access to that without going
in recursion
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3917>
Instead of doing that in both MetaStage & CallyStage.
This allows ClutterStage to also emits the relavant acessibility
bits directly without having a roundtrip through Cally
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3917>
The generic term updated can mean anything. This is specifically about
calibration related updates like changing the sink colorimetry
(Colorspace, HDR metadata) and changes to the white point for night
light etc.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3904>
Previously the color device was destroyed when it was attached to a
monitor that was going away. However, the MetaMonitor objects are
ref-counted and can stay around for longer, even if the underlying
resources went away. We need color devices for as long as the
MetaMonitors are alive.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3904>
Every monitor should eventually have a corresponding color device. To
make sure this can work, we must handle situations where the color
manager didn't connect to colord yet, and thus isn't ready.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3904>
When triple buffering, `meta_onscreen_native_prepare_frame` for the next
frame is called before `notify_view_crtc_presented` for the previous frame.
So our booleans were unfortunately still TRUE in the second prepare_frame,
resulting in two frames with the same property updates.
When double buffering, having roughly one frame interval between
`meta_onscreen_native_prepare_frame` and `notify_view_crtc_presented`
meant that property updates signalled between the swap and presentation
wouldn't get attached to a KMS update, and would be forgotten when
`notify_view_crtc_presented` resets the flags to FALSE.
To solve these we now keep a separate flag and counter per property,
tracking invalidation and pending updates respectively. The latter is a
counter rather than a boolean in support of triple buffering where two
updates may be pending concurrently (next and posted).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3912>
Add a new service client type for a filechooser portal client, and
expose the x11_interop protocol to it.
This will be used to make Nautilus a file chooser portal implementation.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3694>
Because if the current theme has exceeded the dimensions of
`DRM_CAP_CURSOR_WIDTH/HEIGHT` then the warning is just going to repeat
every time the cursor changes. We still fall back to software cursors
just fine so it's not important to repeat the warning.
In Mutter 46 the warning was "Invalid theme cursor size". Same problem.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3597
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3924>
We still use GLSL 100 syntax which means `varying`, `attribute` and
`gl_FragColor` but GLSL 140 wants us to use `in` and `out`. This
provides some simple `#define`s to make it still look like the GLSL 100
syntax is supported.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3907>
This adds a check which makes sure that the required GLSL/GLSL ES
versions are supported.
It also splits out the GLES version check into its own function, just
like GL does.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3907>
As it was originally the function to be used before
making it private and providing safer wrappers around it for x11/win32.
Nowadays, it is only used in x11 and only internally in mutter, exposing
a 'safer' variant costs us exposing more of x11 renderer APIs without
much benefits.
With this change, the only internal xlib renderer we need from meta is
set_foreign_display which can't be easily worked around
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3910>
material is almost no longer used in the code base and the
few remaining references makes it confusing when looking at parts
of the codebase. So rename the rest as well.
Note that this renames a DeformEffect property and the only extension
making use of it doesn't use the property so i think it is okay to do
so without deprecating the old property for a few releases
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3910>
Avoids exposing the former in the next commit even if MtkRectangle
uses signed integers compared to RectangleMapEntry. But as
those rectangles are defined inside of Cogl, it should be okay
I guess?
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3909>
The OpenGL specs say rounding is preferred, but not required. Let's
avoid that uncertainty by choosing a test value that rounds and truncates
to the same integer either way. Only green needs fixing since our red,
blue and alpha values already follow this rule.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3903>
Simply to make it clear that the renamed function is specific to a
particular X11 initialization mode (mandatory Xwayland), put that in the
name, so that it's easier to understand when this function is relevant.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3329>
This means that for X11 sessions we'll do it before any windows are
mapped, and before any plugin implementation is started. Doing it before
a plugin is started is important, because things that the plugin does
during startup can have consequences on how compositing on Xorg works.
For the Xwayland case, we'll do it relatively in the setup phase. It
appears to have been harmless to do it later in the post-opened signal,
but there is no harm in doing it as one of the earlier steps.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3089
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3329>
by executing `global.context.get_debug_control().exported = true`.
This makes it possible to get access to the debug service without having
to start with `--enable-debug`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3902>
The ID is required for the wayland protocol. In the future we might want
to spend a bit more effort to re-use existing color states when a new
one is requested and also try to re-use IDs instead of just counting up.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3897>
This means the pipeline can be manipulated after retrieving. This also
fixes a leak when adding pipelines to the cache, as we the pipeline
would take a ref, but when adding, we wouldn't clean up our own ref.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3897>
As the only info we can pass now is whether we have any idle
closures. Removes the timeout handling which is now is always set to
either 0 or -1 based on whether we have idle closures
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3887>
The drm subsystem has been moving over to gitlab for some time now and
the old anongit.freedesktop.org remote is becoming unusable. Contains
the same repo with the same tags, so this shouldn't result in any
differences.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3896>
According to gmacros.h, G_STATIC_ASSERT is undefined for __GI_SCANNER__
because it causes confusion. Which is understandable, and so is the
complaint about the lingering G_STATIC_ASSERT after preprocessing.
This problem seems to be unique to just clutter-deform-effect.h because
it's the only header file containing a top-level G_STATIC_ASSERT. The
rest are all in C files.
Fixes: 138e5d4f54
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3591
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3898>
Having an always-on-top window affects focus granting logic if the
to be showing window overlaps with any of them. Instead of triggering
the focus denying logic if a new window ever so slightly touches an
always-on-top window to only triggering if it's covered more than 60% by
always-on-top windows.
This is intended to make using always-on-top windows a bit less annoying
and not cause as many unintended focus-on-map denials.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3879>
When we show a window, we'll check if it overlaps with an existing
always-on-top window with the intention to deny focus. However, we did
this potentially before having placed the window, meaning we effectively
checked as if it was placed at (0, 0), which created unexpected results.
Instead check the overlap state after placing. A window placement test
case is added to verify this works as expected.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3879>
Replace a boolean argument and a temporary MetaWindow struct field with
a `MetaPlaceFlag` passed where relevant. This includes
`meta_window_move_resize_internal()` and `meta_window_constrain()`, as
placement may happen during constraining, and also
`meta_window_force_placement()`.
The struct field (denied_focus_and_not_transient) was only ever set in
meta_window_show(), before meta_window_force_placement(), and
immediately unset as a side effect of that. In .._show() we'll always
force placement if the window wasn't already placed, and in
meta_window_constrain(), we'd only ever call meta_window_place() if the
window wasn't already placed, meaning the variable would only ever be
relevant during `meta_window_show()`. Having it as a flag makes that
relationship and temporary state clearer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3879>
The function checking whether a 'always-on-top' window covers the
showing window now has that in the name, to make it more obvious. That
function was also changed to use the more common way of iterating a
list, and now uses auto cleanup pointers for the list.
The condition itself was updated to follow the current coding style.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3879>
Not sure it makes sense to translate only parts of the debugging output
and on top of that, those files were not part of POTFILES, so they were not
translated anyways..
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3895>
It won't be used until later when we flip, and in fact assigning
it early could have led to its own assertion failing on the next frame
in the unlikely event that we return with "Failed to ensure KMS FB ID...
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3891>
This avoids the following critical warning happening sometimes when a
Wayland client exits taking all its window with it in an arbitrary
order:
CRITICAL: meta_window_set_stack_position_no_sync: assertion 'window->stack_position >= 0' failed
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3870>
This, in contrast to 'assert_stacking' only checks showing windows. This
is useful when doing workspace tests, where one want to check what
windows are currently visible.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3870>
When a transient window becomes transient, check if the parent is
sticky, and if it is, make the transient sticky as well. This handles
situations where e.g. a utility dialog (such as search and replace) is
opened on a sticky window, also making the utility dialog sharing the
same stickyness state.
This is also more in line with the semantics of making a window sticky,
where transient would implicitly become sticky as a side effect.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3870>
If a transient window is sticky (visible on all workspaces) and it gets
activated, we'd call move_worskpace() which would effectively unstick
it, which is rather unexpected. It'd also effectively unstick its parent
as well, due to moving a transient window also moves its descendants and
ascendants.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3870>
Introduced in libinput 1.26 this feature allows restricting the
tablet tool pressure range to a subset of its physical range. The
use-case is either to require some higher-than-usual minimum pressure
before the pen reacts, or lower-than-usual pressure to reach the maximum
logical pressure.
libinput takes a [0.0, 1.0] normalized range which we expose as percent
in the gsettings. The wacom driver doesn't have an exact equivalent but
it has a Threshold setting (range [1, 2048]) that defines when a button
is generated for tip down.
See gsettings-desktop-schemas!84
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3794>
This mapper hooks into CLUTTER_BUTTON_PRESS/RELEASE events with a
clutter button of zero but a nonzero evcode (e.g. BTN_STYLUS).
It then looks up the available button actions and implements
switch-monitor and keybindings using the MetaTabletActionMapper parent
class.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3649>
Unlike most other schemas the path for a tool requires a bit of
processing (serial number or tablet vid/pid if there's no serial number).
Let's make the tool settings available through the MetaInputSettings
instead of having to duplicate that path composition in the caller.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3649>
Stylus actions that don't map into LMR or back/forward are now created
as a clutter button event with a button number of zero. Nothing is
actually done with those events for now, they're just discarded later.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3649>
For stylus buttons we apply a button mapping, e.g. secondary button ->
right. This mapping was previously applied to the clutter event's evdev
code only, not the actual clutter button. As a result, gnome-shell would
always treat the BTN_STYLUS as middle and BTN_STYLUS2 as right,
regardless of the mapping.
Move the mapping up so we first adjust our evcode, then proceed with
the usual mappings.
Note that this temporary breaks the stylus mapping to Back/Forward which
will be fixed in a follow-up commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3649>
Use the helper function for mapping a stylus tool evdev code to a
clutter button code. This fixes a (theoretical) issue - if a tool were
to send any button other than the one we handled those would likely be
BTN_SIDE and friends and we'd likely end up with negative button
numbers. The BTN_TOOL_PEN range is not predicable enough to do any sort
of calculation conversion because things like BTN_TOOL_DOUBLETAP have
specific meanings that aren't actually buttons.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3649>
This complements the existing clutter->evdev and evdev->clutter helpers,
but this time for buttons we expect from a stylus tool. We also need to
convert left/middle/right for the Wacom puck/lens cursor tools but that
particular conversion is lossy.
Note that these are more restrictive than the normal codes - if we
get "other" buttons from a stylus we don't really know what they could
possibly map to. So we safely map what looks like buttons from a mouse
but otherwise complain and return zero.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3649>
This is prep work for using the same functionality for tablet tools as
well. The new MetaTabletActionMapper takes care of the event bubbling
via the device-added/removed and input-event signals and provides the
helper functions to cycle outputs and/or emulate keybindings.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3649>
The pointer returned by cogl_color_copy() was not stored anywhere and
the allocation it was pointing to got leaked. However we also don't need
to use that function here and can just copy the struct values directly.
Fixes: dc52ccc75 ("cleanup: Port from ClutterColor to CoglColor")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3884>
Instead of using cairo for scaling and rotating cursors before putting
them on a plane, use Cogl. For now still download them back to the CPU
so we can place them on a dumb buffer, but can explore rendering to a
DMA buffer directly as a future improvement.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
A ClutterPipelineNode took a reference to the passed pipeline and used
it during the lifetime of the paint node. In theory this meant that a
paint node pipeline could be changed, and reused, affecting the previous
paint node drawing.
Referencing instead of copying also meant the ClutterPipelineNode isn't
able to itself manipulate the pipeline any further.
Fix both of these issues by copying, instead of adding a reference to,
the passed pipeline. While resulting in some additional allocation
over head, it means we can now eventually handle color transformations
automatically for ClutterPipelineNode's.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
When running with COGL_DEBUG=show-source, log what pipeline and shader
program is used when painting, and what shader source code corresponds
to each shader program.
This helps identify what shader is used when painting to what
framebuffer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
So far 'color-state' is added, intended to tag pipelines with color
state transformation capabilities. Color state transformation snippets
are tagged with it. Eventually handlers of pipelines will use this
information to on-demand decorate pipelines with color transformation
snippets.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
A snippet is assigned a capability, which is then transferred to the
pipeline it gets added to. Managers of pipelines can then check whether
a pipeline it got handed whether it has a certain capability, and if
not, handle the situation where it is not.
The intention is to allow Clutter to tag color state transformation
snippets with a capability, allowing generic paint nodes to handle color
transformations when needed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
This looks at the color state it got from the actor on construction, and
the target color state from the paint context and generates (and caches)
color aware pipelines used for painting. One of the purposes here is for
mixing SDR and HDR content and painting to a HDR monitor. If HDR (or
optical blending) isn't activated, the produced shaders will be
equivalent to what we had before.
Also add some names to the piplines, as this helps identifying what
pipeline source is associated with what pipeline.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
In some of the places we need more context than just the CoglContext, so
prepare for that by passing around the paint context, which carries
this, everywhere instead. It won't be needed everywhere, but lets stay
consistent.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
Color aware rendering needs shaders / pipelines that adapt to what
output they render to. For example if we want to render to a linear
BT.2020 intermediate framebuffer on one monitor, and a non-linear sRGB
direct target buffer on another, the shader for the same paint node or
content will depend on where they are going to be presented.
In order to help keeping track of what shader should target what
monitor, without having to regenerate them each time, introduce a
pipeline cache that knows how to handle differentiating between
transforming between different color state.
The cache is meant to handle caches for multiple pipeline users, where
each user might potentially want to keep track of multiple pipelines
itself. Lookup should be O(1), and in order to achieve this, separate
the cache into 3 levels.
The first level is the "pipeline group", where e.g. a ClutterContent
type allocates a group where it can store its pipelines. Each group has
a fixed number of "slots" where it can store a pipeline. Each slot has a
hash table where the key is derived from a color state transform, and
where the value is a CoglPipeline where the thame color state
transformation is expected to be handled.
A content will when painting know about its own color state, and the
target state it should render into, retrieve a cached pipeline for the
correct transform, or if the cache didn't have it, generate it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
Allow compositing in a linear color space and do so either when forcing
it via the debug controls D-Bus API, or when the experimental HDR mode
is enabled.
This relies on paint nodes etc to actually transform everything into the
linear target color space, which isn't done yet, so enabling it right
now will cause a broken result. Yet, introduce this now, so that
painting can be fixed piece by piece.
Linear blending is automatically enabled on monitors where HDR is
enabled, as this makes it possible to use an linear color space when
blending content from different color spaces with different transfer
functions.
Linear blending requires extra precision, i.e. 16 bit per channel
in the intermediate buffer due how the values are distributed,
so only enable the experimental HDR mode if the Cogl context supports
half float formats.
By default, no intermadiate linear offscreen framebuffer is used.
To test, do e.g.
./tools/debug-control.py --toggle ForceLinearBlending
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
The initial target color state is the color state of the stage view
being painted to. If we're painting to an arbitrary framebuffer, it's
currently hard coded to sRGB/electrical.
The content color state is not set on construction, but when starting to
paint, it's set to the color state of the stage itself. Whenever an
actor is painted, it'll set the color state to the color state of
itself. The intention is that offscreen rendering pushes a target color
state that causes painting to it to not necessarily be in the stage view
color state.
Pass color state with offscreen framebuffer, as this avoids hard coding
sRGB in the lower level bits of paint contexts. It's still practically
hard coded, only that it's derived from somewhere else (e.g. the stage
or window actor).
Nothing is actually using this yet, but will eventually.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
This aims to provide the color state used for compositing for a specific
view. It's currently hard coded to default, but will eventually be
configured depending on the monitor and configuration. It's intended to
be used as a target color state for rendering with color awareness.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
The cogl shader cache uses snippet pointers as cache keys, meaning even
if we generate two identical snippets; if they have different pointer
addresses, they'll generate separate cache entries. Handle this by
caching our snippets on the context.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
This helper generates shader snippets that converts pixels from one
color state to another. For example if there is content with sRGB color
state that should be converted to linear BT.2020 before being written to
the framebuffer, a shader that makes the pixel go through the following
steps will be generated:
1. sRGB EOTF
2. Luminance gain (hard coded for now)
3. Color space mapping
The intention is that it should be possible to composite in a linear
color space, into an intermediate framebuffer, which is then passed
through an inverse EOTF to produce linear content in the output color
state.
When transforming from BT.2020/PQ to sRGB/sRGB, clamping to the sRGB
max luminance is done.
Cases where direct transform is also handled, i.e. where one doesn't go
via an linear intermediate buffer, in which case there might be both an
EOTF and an inverted EOTF in the same shader snippet to still do color
space mapping using optical color encoding. This will be used for e.g.
transforming cursor sprites.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
An unknown color space isn't very useful to have, as there is not very
actionable what to do with it. Rename it to 'default'. Later it'll be
used to an implicit color space, which in practice will be treated as
sRGB.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
The context can be set via a property (currently unused) or via the
current global singleton as a fallback. It means API that acts on an
actor can avoid going via any globals.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3433>
g_unix_fd_list_append() duplicates the provided fd. If that fails, it
returns -1 as fd index and sets - if provided - the passed GError
accordingly.
However, currently, mutter does not check the return value (the fd index
of the appended fd) and thus passes an invalid fd list via dbus to the
remote desktop session user.
Fix this error by also checking the fd index. If the fd index is invalid
(< 0), simply pass the error message of the g_unix_fd_list_append() call
to the caller.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3871>
This reverts commit a3082b8eb3.
We don't find the VKMS device with this commit because it is on seat0
and not on META_BACKEND_TEST_INPUT_SEAT.
The other way around, i.e. returning seat0 in all cases also doesn't
work because *something* hangs if the default seat referrs to the real
seat0 instead of the nonesense META_BACKEND_TEST_INPUT_SEAT.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3869>
mktemp can create a tempfile relative to a directory passed in via -p.
It also uses the $TMPDIR variable for the same purpose. When the
template is specified via -t, $TMPDIR takes precedence over -p. When the
template is specified via a positional argument, -p takes precedence.
Since fec38819ac $TMPDIR is set via the
dbus runner which took precedence.
virtme-ng doesn't seem to share /tmp with the host system which results
in the exit status from the test in the VM not propagating back to the
test harness.
Fix that by making sure we always create the tempfile for the result in
the build directory.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3867>
Fixes error building against libdrm >= 2.4.122:
../src/backends/native/meta-kms-plane.c:67:8: error: redefinition of ‘struct drm_plane_size_hint’
67 | struct drm_plane_size_hint {
| ^~~~~~~~~~~~~~~~~~~
In file included from /usr/local/include/libdrm/drm.h:1025,
from /usr/local/include/xf86drm.h:40,
from ../src/backends/native/meta-kms-plane-private.h:20,
from ../src/backends/native/meta-kms-plane.c:21:
/usr/local/include/libdrm/drm_mode.h:866:8: note: originally defined here
866 | struct drm_plane_size_hint {
| ^~~~~~~~~~~~~~~~~~~
Suggested by Jonas Ådahl.
v2:
* Use has_type. (Sebastian Wick)
v3: (jadahl)
* Bump meson requirement to 1.3.0 for compiler.has_type()
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3556
Fixes: 0ca933baec ("backend/native: Adds support for SIZE_HINTS Cursor Plane Property")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3848>
This code is called from handlers connected to signals of a
MetaWindow. It cannot happen that the window will end up NULL
in these, so exchange with a g_assert() as we in fact expect
it to be non-NULL.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3860>
We cannot use a function with the same signature for signals with different
arguments, if we want to rely on the user data parameter. Separate into
two signal handlers calling the same function inside.
Fixes: b9ba34ac6f ("wayland/activation: Fix signal callback signature")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3860>
The two dialog creation virtual functions returned by these functions have to
be unreferenced by the caller (and are actually unreferenced in other places in
the code).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3790>
This gives greater control to the callers on the place where a grab is being
activated, this may make a difference in the handling of crossing events
triggered through it, e.g. by having callers rely on having already obtained
a ClutterGrab prior to handling the resulting effects.
The "input only" grab has also been turned inactive by default, in order to
to have the ClutterGrab pointer available for checks at the MetaWaylandEventHandler
focus changing methods triggered through grab activation.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3463
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3747>
If the titlebar of a window has been moved above the screen by a user
via an unconstrained move, then any constrained user resize following
this move will cause the window to jump below the top of the screen or
cause other glitchy behavior.
This commit removes the constraint that the titlebar of a window must be
below the top of the screen for any resize that is both (1) triggered by
a user and (2) is a resize that affects only the left, right, or bottom
edges of the window. This allows users to move a window partially above
the screen and then resize the window to be wider or resize the bottom
edge of the window to make it taller or shorter.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1206
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3764>
This unbreaks building here. The compilation error was due to
MetaX11Display having an incorrect typedef in a now preprocessed out
part, which will be fixed in a later commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3851>
Allows gnome-shell or other compositors to detect which features
were built when building libmutter automatically without having to
expose various build options themselves
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3776>
Moving a window is a compositor action and happens immediately. Waiting
here is pointless. Make sure instead that the action happens immediately
by asserting the position.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3842>
wait and wait_configure after a single resize is useless and a race.
Resize is a client side action which doesn't result in a configure and a
wait doesn't sync for a resize as well.
Sometimes the resize is paired with another action, such as maximize,
fullscreen or show. In those cases a configure will be generated and a
previous resize is accounted for.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3842>
Commit e775052429 changed the code such that resetting the actor is done
when a surface role is assigned. The dnd surface assigned vfunc doesn't
chain up which means the code to reset the actor is never hit and the
dnd surface never shows up.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3540
Fixes: e775052429 ("wayland/actor-surface: Reset the actor on role-assignment")
In these checks, it's an important detail to preserve subpixel information
in order to correctly determine whether the coordinates are inside or
outside a view.
Otherwise, small enough motion towards the left/top might get rounded
to 0, be seen as "inside the view", and the pointer coordinates be allowed
to escape the viewport constraints.
This was figured out by Pascal Nowack before me, with a difference of
minutes. Credit where credit is due.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3530
Fixes: 6c972546f1 ("mtk: Add Rectangle.contains_point")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3820>
In meta_seat_impl_remove_virtual_input_device(), the 'device'
variable is first removed from MetaSeatImpl, then a "device
removed" event is generated with it.
The problem here is that, if this is the last reference of
'device', the removal from MetaSeatImpl will destroy it. Then
the freed variable will be used to create the "device removed"
event, which is a use-after-free situation.
Fix that by owning an extra ref to 'device' as long as the
function is executing. Do this by declaring a g_autoptr
variable with the extra ref. This g_autoptr variable is cleaned
up by the end of the function, which achieves the desired effect.
Spotted by Coverity.
CID: #1594046
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3835>
Right now the unmapped signal doesn't always fire which means we didn't
see a surface that's being unmapped in these code paths before. In
particular the resource, window and role can be gone. Handle those
cases.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3783>
This allows us one less level of indentation of all the tests. It is not
entirely true, the X11 backend test case script can run without it, but
it isn't valuable enough to run without the native backend being enabled
to complicate building.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
Instead of having get_seat_id() handle most cases, and then special
casing another case outside of it, let it handle them all, making all
users just able to call get_seat_id().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
Except the X11 backend test, it still needs an X server. Eventually we
can replace it with the equivalent that uses Xwayland, but that needs a
"scaled down" mutter that runs as the host compositor for Xwayland, that
doesn't expose anything on the session bus.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
It's constructed as a native backend headless backend, but uses a custom
monitor manager (based on MetaMonitorManagerNative) that creates a fake
monitor. Rendering is unconditionally done with the surfaceless
renderer.
The test devices used now use virtual devices, meaning some changes to
the tests to e.g. not set names, and not dealing with input devices
directly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
This is where the viewports are updated, and so lets move connecting the
signal together with it. It also helps a future change to the test
backend where it creates a custom monitor manager.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
Test touch devices are autocleaned up, but if a test case explicitly
needs it to be removed at a certain point, make sure to also clear the
pointer so that it doesn't get removed twice.
Right now it's harmless, but in a later commit it'll expected to only
remove a device once.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
`update_secondary_gpu_state_post_swap_buffers` decides what our front
buffer object will be. There is only one answer. So return it as the
function result instead of making the caller figure it out.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3830>
It's always equal to `onscreen_native->next_frame` and we can't eliminate
that copy so easily. Removing the parameter removes all ambiguity about
where the next frame will come from.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3829>
Inside the "if (clutter_actor_has_accessible (actor))" condition,
the 'atk_child' variable is set and a signal is emitted on it.
There is a classic ref/unref dance around the signal to guarantee
that 'atk_child' won't be destroyed.
However, this ref/unref dance doesn't work, because the unref is
done *before* the 'atk_child' variable is used again. So if this
was the last reference to it, it would have been destroyed in the
unref call, then used for another signal emission a few lines down.
That's a use-after-free.
Fix that by declaring the 'atk_child' variable with g_autoptr. This
delays the unref until the very end of the function, and is NULL safe.
Also add a sneaky assertion, just for extra safety.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3828>
Starting in mesa 24.1 (good in 24.0.9, bad in 24.1.1), the current
kernel somehow manages to hang our VKMS based tests. Let's update the
kernel so we can bump the CI image without running into those hangs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3826>
When a client wants to lease DRM resources, it will attach the list of
connectors is wants to lease and then submit the request.
Once the request is submitted, destroy the DRM lease request object.
Co-authored-by: José Expósito <jexposit@redhat.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3746>
When a new device is connected, register a new global for it.
When a device is gone, remove the global. Upon receiving this event,
the client should destroy any matching wp_drm_lease_device_v1 object.
To destroy a wp_drm_lease_device_v1 object, the client must first
issue a release request. Upon receiving this request, send a released
event and destroy the object.
Co-authored-by: José Expósito <jexposit@redhat.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3746>
When the list of connectors available for lease changes send
wp_drm_lease_device_v1.connector events for added connectors and
wp_drm_lease_connector_v1.withdrawn events for removed connectors,
followed by a wp_drm_lease_device_v1.done event.
Co-authored-by: José Expósito <jexposit@redhat.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3746>
Add meta_kms_impl_device_open_non_privileged_fd() that returns a
non-master file descriptor for a MetaKmsImplDevice.
It'll be required to implement wp_drm_lease_device_v1_send_drm_fd() in a
future commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3746>
The manager keeps track of which connectors are leasable in general,
which connectors and resources are already part of a lease, and keeps
track of when leases get revoked.
When leasing out connectors, the required drm resources to drive the
connectors are included in the lease as well.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3746>
When a plane is leased, it is assigned to a CRTC which is leased. When
trying to find a primary plane for a modeset, skip the assigned planes
on leased CRTCs to avoid sharing the resources with the leased process.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3746>
This allows us to keep track when primary and cursor plane assignments
on a CRTC are unassigned. With this commit, all planes which are
assigned are actually in use and can't be assigned to anything else.
We'll make use of that fact when we search for a leasable primary plane.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3746>
The lease_objects function takes connectors, CRTCs and planes which are
turned into a drm lease. The resulting lease can be revoked with
revoke_lease. With list_lessees the currently active leases can be
queried.
Co-authored-by: Sebastian Wick <sebastian.wick@redhat.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3746>
Add an environment variable (MUTTER_DEBUG_LEASE_CONNECTORS) that allows
set a ":" separated list of connector names as available for lease.
The names of the connectors can be found in "/sys/class/drm".
To illustrate it with an example, the names of the connectors and its
status can be fetched with this command:
$ for p in /sys/class/drm/*/status; do con=${p%/status}; echo -n "${con#*/card?-}: "; cat $p; done
DP-1: disconnected
DP-2: disconnected
DP-3: disconnected
DP-4: disconnected
DP-5: connected
DP-6: connected
DP-7: disconnected
eDP-1: connected
HDMI-A-1: disconnected
And, to set "DP-5" and "DP-6" available for lease, the environment
variable can be set like:
MUTTER_DEBUG_LEASE_CONNECTORS=DP-5:DP-6
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3746>
When mutter creates a dma-buf buffer for screencasting, the buffers
stride will, among other attributes, also be defined.
However, mutter currently only sets the buffer stride, when actually
recording a frame, but not when adding it.
This behaviour disallows screencast consumers (clients) to already
import the respective buffer (i.e. for Vulkan creating a VkImage for the
dma-buf image), as the stride is not yet communicated to the client.
Since the stride won't change after adding the respective buffer,
directly set the buffer stride, when adding the PipeWire buffer. This
allows screencast consumers (clients) to do optimizations in their
encoding paths.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3827>
meta_flush_input() creates and runs a main loop in order to block until the
input thread has flushed all its events. This main loop is created on the
default main context though (NULL is passed to g_main_loop_new()), which
means that while the main loop runs, the default main context is active (aka
stage updates will happen).
This causes an issue with tests, specifically when there already is a stage
update queued before running meta_flush_input(): meta_flush_input() will
(as expected) block until the input thread flushed all its events to the
main thread. But while that is happening, the main thread will be doing the
stage update that was already queued, without the new events (the input
thread is just starting to flush those). Then meta_flush_input() returns,
and in our test we see that the stage has been updated, except it wasn't
updated with the latest events. The test now continues and fails.
To fix the issue, make meta_flush_input() truly blocking, so that it only
flushes the input thread, but doesn't drive the global main context while
waiting for that. After the flushing is finished, tests must now manually
iterate the main context themselves to ensure that a stage update happens.
This breaks a few stacking tests, because "move_cursor_to" and "click" use
meta_flush_input() internally. For those commands we now need to dispatch
a stage update afterwards.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3807>
Currently AtkActor iface is implemented by CallyActor as well
as a ton of infrastructure just for one action in ClutterText, the
activate action.
Instead, simplify the whole thing and move the AtkAction implementation
where it supposed to be.
This refactoring would help with merging Cally within Clutter in
the future
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3814>
Let the ClutterFrame (or rather MetaFrameNative) own both the scanout
object and the framebuffer object, and let the frame itself live for as
long as it's needed. This allows to place fields that is related to a
single frame together, aiming to help reasoning about the lifetime of
the fields that were previously directly stored in MetaOnscreenNative.
Also take the opportunity to rename "current" to "presenting", to make
it clearer that frame's buffer is what is currently presenting to the
user.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3799>
It's possible to destory the xdg_toplevel and xdg_surface which removes
the surface role and then re-use the surface on a new toplevel.
Mutter used to not create new SurfaceActors for the new toplevel which
were then shared between the old and new toplevel. This happens because
the actors can stay around after the wl resource is destroyed (e.g. for
closing effects).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3667>
Instead of resetting in the xdg_surface.get_toplevel call.
It also changes XDG popus handling to deal with cases where the surface
was part of a window before, the same way the toplevel does.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3667>
Wayland tests also get kvm and tty test variants, but running tty tests
on your main session makes them fail. The intention for tty tests is to
skip when not run from a tty, so fix that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3811>
Enable the docs in the main build, then inherit the artifacts. It should
result in less CPU resources being used.
GitLab doesn't support multiple artifacts per job, so we still have to
download the whole build, as the build job can't hand out just the
references, and it seems to be planned as a premium-only feature sadly:
https://gitlab.com/gitlab-org/gitlab/-/issues/18744
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3816>
If view initialization fails then don't add the view, rather than
adding a dummy offscreen view. This avoids flooding the log with
offscreen frame clock confusion:
Before:
```
libmutter-WARNING **: 15:47:27.763: Failed to allocate onscreen framebuffer for /dev/dri/card0: Failed to allocate surface: Function not implemented
Clutter-WARNING **: 15:47:28.557: (../clutter/clutter/clutter-frame-clock.c:419):clutter_frame_clock_notify_presented: code should not be reached
Clutter-WARNING **: 15:47:28.563: (../clutter/clutter/clutter-frame-clock.c:419):clutter_frame_clock_notify_presented: code should not be reached
Clutter-WARNING **: 15:47:28.567: (../clutter/clutter/clutter-frame-clock.c:419):clutter_frame_clock_notify_presented: code should not be reached
(repeats forever)
```
After:
```
libmutter-WARNING **: 16:09:04.945: Failed to create view for Unknown 46" on None-1: Failed to allocate onscreen framebuffer for /dev/dri/card0: Failed to allocate surface: Function not implemented
```
Relates to:
https://launchpad.net/bugs/1967707https://gitlab.gnome.org/GNOME/mutter/-/issues/2489https://gitlab.gnome.org/GNOME/mutter/-/issues/2295
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3115>
AMD driver is buggy when it comes to generate textures of 1px height [1]
while this is not a mutter problem, we should not block on that either.
So, given that the test purpose is to check the journal flushing during
unref more than its behavior, we can still check this in all the drivers
keeping a warning in the AMD case.
See: https://gitlab.freedesktop.org/mesa/mesa/-/issues/11269
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3793>
Some tests sadly behaves differently depending on the driver in use.
While this shouldn't happen we can't block on these driver issues, so
add a test utils private function to get the driver information so that
we can adapt test behavior depending on this.
This will allow to disable / enable tests at runtime instead of failing
in all the implementations, which is still better for catching
regressions in the parts we may be ignoring otherwise.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3793>
When running the tests in non-undefined mode we should skip the known
failing tests instead of the whole suite, so mark them as such so that
we can still run the tests partially checking the things that still
work, avoiding to regress on them too.
We also run the tests in normal mode so that we can ensure that the
failures are still happening in that way.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3793>
We need to use a different $XDG_RUNTIME_DIR path to be able to start most
tests in parallel, and we can use a temporary directory for that which python
cleans up when done.
Also, given that most of settings are stored in HOME use temporary one
for that too, to prevent mutter to fail because it may load some local
configuration (e.g. monitors.xml) that don't meet the expectations or
that may change the test behavior in an unexpected way.
As per this, CI needs to be adapted for new args handling
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3793>
When testing mutter using `META_DBUS_RUNNER_WRAPPER=rr` we may get a
not found-device error, given that it's not a case we support, we can
ignore it as we do with permission denied one, limiting this to the RR
case.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3793>
Showing the stage always results in an initial update being queued. If the
virtual input events from the test ends up being dispatched after that, the
`wait_for_update()` checks gets out of sync, as the first update ended up
waiting for the initial update, not the one from the input event.
Fix the gesture test by adding another call to `wait_stage_updated()` right
after `clutter_actor_show()`.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3521
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3795>
Allow a screen cast stream source to say that nothing changed in terms
of cursor metadata, and treat this together with a cursor-only frame as
we not recording anything.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3803>
If the presentation time isn't known, e.g. if the monitor is virtual and
the actual presentation happens far away, the presentation time we
actually received tends to be the time a frame was presented to the next
layer, meaning practically immediately after painting.
When scheduling another update after that, don't assume that if the next
calculated update is not the immediate next update, schedule an update
sooner, as that will in such cases always be true, meaning we ended up
busy looping with constant frame updates being scheduled.
Fix this by only triggering that logic if the last presentation time was
actually vsync:ed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3803>
We're not doing anything significant in the KMS thread anyway, so don't
make it a kernel thread, and don't ask to be real time scheduled (which
we wouldn't be anyway, but for clarity).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3805>
Don't try to find the card, and then the render node from it, just ask
udev to list the render nodes directly. This avoids running into
permission errors when the user cannot open /dev/dri/card* even without
mode setting capabilities.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3805>
By turning a macro that exists in the codebase to a proper
function so that gnome-shell could make use of it as well
instead of using a region for it contains_point api...
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3801>
And also "completion" time to measure when the commit returned.
This is structured so as to measure all timestamps first before logging
anything. That way our results shouldn't be (don't seem to be) affected
by the logging itself.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3265>
While it should not be expected that we pick the pointer into a
MetaSurfaceActor that is disembodied of its MetaWaylandSurface/MetaWindow,
the paths where this should be enforced are somewhat scattered.
So account for the situation in picking code, and prefer a NULL surface
over a crash. This operates on the assumption that this inconsistent state
where Mutter didn't know better to pick a correct surface actor will be fixed
by later crossing events resolving the intermediate state, and that no
other input events will be received meanwhile.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3393
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3729>
When unmapping a subsurface, it does lose early its connection to the
parent surface. This is however a deciding factor in determining whether
the surface (role) has a window.
Make the subsurface actor unreactive if its connection to the parent
MetaWindow was severed, since it should not be eligible for picking anymore.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3729>
These actors are expected to be destroyed along with their surface, this
however happens later in the process, so there is a moment where actors
are eligible for picking, but do not have a surface anymore.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3729>
Prior to the grabs/focus rework in !3420, Wayland grabs were handled
separately from ClutterGrabs. This required explicitly checking for
ClutterGrabs as those were expected to prevent events from reaching
Wayland clients.
Now after !3420, Wayland client grabs also result in ClutterGrabs, which
means that this check causes input events for popups with grabs to not
get sent to ibus anymore. Instead the events are getting sent to the
client directly, which results no ibus support in popups (unless the
client handles that itself by using a different GTK_IM_MODULE).
However due to the changes from !3420 checking for ClutterGrabs is also
no longer necessary and the meta_wayland_text_input_update() focus check
is now sufficient to only forward events to ibus, when the focus is
actually on a Wayland client. So to fix this we can simply remove the
check.
Fixes: 2a584a8f0 ("wayland: Make use of Wayland event grabbing mechanism")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3502
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3787>
Reads exposed size hints for the given cursor plane. Chooses nearest
minimum cursor size out of the hints with respect to the user chosen
cursor size from the UI. Allocates optimized Hardare cursor size,
hence drm buffer
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3165>
Instead use an abstract "logical monitor id" that is generated from the
logical monitor. Instead of using low level numbers from the mode
setting devices, use either data from the EDID, or the connector, if the
EDID is not useful.
This should help with windows remembering monitor positions when the
same monitor reappears but with another mode setting device ID.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3753>
Replace the sync_focus() calls with a set_focus() do-it function taking
a surface. This is in line with the rest of the things that happen at
the default MetaWaylandEventInterface.focus implementation, and will
make these correctly observe the presence of grabs, since
meta_wayland_seat_get_input_focus() will return the would-be focus
in these cases.
This change makes the "focused" client selection truly
in sync with the keyboard focus.
Fixes: 5ca10c31d1 ("wayland: Follow seat's input focus client for clipboard selections")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3490
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3789>
Replace the sync_focus() calls with a set_focus() do-it function taking
a surface. This is in line with the rest of the things that happen at
the default MetaWaylandEventInterface.focus implementation, and will
make these correctly observe the presence of grabs, since
meta_wayland_seat_get_input_focus() will return the would-be focus
in these cases.
This change makes the "focused" client selection truly
in sync with the keyboard focus.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3498
Fixes: 9bdb00c459 ("wayland: Follow seat's input focus client for primary selections")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3789>
Don't expose its headers, remove all getters, and move the rest to
cogl-output-private.h. Next commits will clean it up even more, but
for now, just make it private.
This will help next commits remove CoglOutput.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3792>
When cloning an actor we were applying a global scale to it, based on
the size of the clone itself and of the cloned actor.
This implied that the transformed size of the clone was not the one that
was set, but it was taking taking in account the actor scaling too.
So in practice we were scaling it twice.
While this had no visual implications it indeed was causing troubles
when a ClutterClone was reactive because in such case only the scaled
area of the scaled clone was considered reactive.
Assume you had an actor of 100x100 pixels, scaling it to a clone of
50x50 pixels:
- The scale applied to the clone was 0.5
- The transformed size of the clone was: 25x25 pixels
- The clone was reactive only in that sub-area
To avoid this, never touch the clone transformation matrix, but only
transform it at paint time.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2959>
We were leaking one instance because _cogl_winsys_egl_context_created()
could be potentially be called twice (via cogl_display_setup() one as
part of cogl_renderer_check_onscreen_template() and the other when
called from clutter_backend_do_real_create_context()), and so we'd ended
up overwriting the reference we had.
However, we really didn't use it anywhere and once used to call the
relevant functions it's just useless.
So let's just keep it as local variable
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3788>
We were leaking the color profile path keys but also it wasn't clear how
the ownership was passed to the new hash-table, so let's just remove it
from the pending hash table and add it to the new one including the
expected reference.
This is safe because we were still adding a temporary extra ref to the
profile
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3788>
The cursor was following the "current" surface (i.e. the logical
focus, unaffected by e.g. grabs), so MetaWaylandPointer was always
providing window cursors, just so the MetaCursorRenderer would maybe
discard them and show the compositor cursor instead. E.g. in the
presence of grabs.
This clear barrier between grabs being compositor business deserving
a compositor cursor, and non-grabs being client business turned a bit
blurrier in grabs-pt5 where "client" things like popups and DnD would
also involve grabs. The fixes in that regard in the branch went on
the lenient side, introducing situations where grabs do exist but we
are preferring client-side cursors anyways.
Fix this by making MetaWaylandPointer aware of grabs at the time of
updating the client-side cursor, by following the "focus" surface
(i.e. the effective focus, affected by grabs outside of MetaWaylandInput).
The focus surface and cursor will be updated on focus changes, also
induced by grab changes.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3460
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3736>
In some cases, reparenting a window with its frame may fail; this seems
to happen especially during initialization of a window that may be
unmapped and re-mapped quickly and multiple times.
If this happens, we're never going to receive a remove event on the
stack tracker and so we may end up adding it twice to the list of the
windows to synchronize with the compositor, breaking its assumption that
the stack list is unique, and eventually leading to a crash because we
do not end up removing all the instances of a window on its destruction.
In particular we may end up in this situation:
Syncing Window 10485927: 0x555558863540 (actual xid is 10485927),
user time is 10485928 frame is 0x5555588715c0, frame xid 6291591
Syncing Window 14680081: 0x5555588664b0 (actual xid is 14680081),
user time is 14680082 frame is 0x555558871d80, frame xid 6291595
Syncing Window 6291460: 0x55555796dc80 (actual xid is 10485763),
user time is 10485764 frame is 0x555557a6f630, frame xid 6291460
Syncing Window 6291465: 0x555557a68af0 (actual xid is 14680067),
user time is 14680068 frame is 0x555557a73e80, frame xid 6291465
Syncing Window 6291509: 0x555557f9d830 (actual xid is 8388623),
user time is 0 frame is 0x555557fac780, frame xid 6291509
Syncing Window 6291586: 0x5555586e1690 (actual xid is 4194363),
user time is 0 frame is 0x55555886e550, frame xid 6291586
Syncing Window 6291591: 0x555558863540 (actual xid is 10485927),
user time is 10485928 frame is 0x5555588715c0, frame xid 6291591
Where the same meta window 0x555558863540 is added twice because that's
both mapped by the window itself (10485927) and by its frame (6291591).
This happens because for historical reasons the xids hash table managed
by the x11-display maps both the X11 windows, their frames and their
user time windows as the meta-window, and so if we don't filter out them
properly we end up duplicating the entries in the compositor list.
Such duplicates finally end up making mutter to crash in
meta_compositor_sync_stack() because we could end up trying to access
to an invalid window, given its actor has been destroyed but not all the
instances have been removed from the compositor windows list:
0x00007ffff71059 in meta_compositor_sync_stack (compositor=0x555555b8,
stack=0x555558701b80) at ../../mutter/src/compositor/compositor.c:773
773 if ((old_window->hidden || old_window->unmanaging) &&
(gdb) print old_window
$1 = (MetaWindow *) 0x0
So, in order to prevent this, check that XReparentWindow does not fail,
and in case of failure, reset the window state to the one it had before
we failed and more importantly, remove the association between the frame
X11 window and the MetaWindow, since this is not true anymore and so
that at the next stack synchronization there won't be any meta window
associated to that frame XID (unless there aren't further stack changes
impacting on that). Without this we would have instead waited for the
remove event that we predicted, but that could never happen because no
ReparentNotify is emitted in such case.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3732>
Since we listen the X events in the same thread, and they are delivered
through the main loop, there's not any need to set the frame details on
windows before the reparent operation, because such action could fail.
So move the code order, as preparation for handling the error.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3732>
That was only there to force updating MetaFrame's bounds
but that is something that is already updated in
MetaWindowX11.move_resize_internal. So just drop all of that
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3254>
Since the `switch` didn’t have a default case, the `cull_front` and
`cull_back` variables could technically be used uninitialised if the
`cull_mode` was unrecognised.
That seems unlikely to happen as presumably other code makes sure the
`cull_mode` is valid, but it doesn’t hurt to add a `default:` case to
squash the compiler warning.
Signed-off-by: Philip Withnall <pwithnall@gnome.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3770>
In the future, the module will automate uploading the release
tarball. We already use the CI pipeline to generate the tarball,
so it's easy to hook up the module and provide some testing
before the module goes into production.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3775>
We currently assume that the `CI_COMMIT_TAG` variable matches the
version component of the generated dist tarball.
That is usually correct, but sometimes errors happen and a wrong
tag is pushed, and the real release uses something like "46.0-real".
Account for that by building the artifact path from `meson introspect`
and exporting it as environment variable.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3775>
For a Wayland only build, we would like to avoid linking against
libXcursor which on it turn, links back to some of the X11 deps.
In order to achieve that, we include a small subset of xcursor.
In case Mutter is built with X11 or with both Wayland & X11, we link
against libXcursor and don't make use of the in-tree implementation.
This patch mimics what GTK 4 do by shipping an in-tree copy of xcursor.
Especially that libwayland-cursor does not provide an alternative to
xcursor itself.
Helps #2272
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3607>
When meta_wayland_pointer_constraint_remove() is called, it can trigger
a meta_wayland_event_handler_invalidate_focus() via:
meta_wayland_pointer_constraint_destroy()
meta_wayland_pointer_constraint_disable()
meta_wayland_input_detach_event_handler()
meta_wayland_input_invalidate_all_focus()
meta_wayland_event_handler_invalidate_focus()
Which then would result in a "focus-surface-changed" signal which would
call meta_wayland_pointer_constraint_remove() a second time. This
happens after surface_remove_pointer_constraints() has already been
called in the first meta_wayland_pointer_constraint_remove() call,
leading to "data" being NULL.
To prevent this issue disconnect the signal handler before calling
meta_wayland_pointer_constraint_disable() when destroying a constraint.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3476
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3749>
In case of empty regions (e.g. when locking the pointer) the pointer
was only forced to stay within the boundaries of its current pixel
(i.e. culling subpixel position), instead of the position where the
pointer lock did start.
Fixes: 07d24fe50 ("backends/native: Allow infinitely small pointer constraint regions")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3749>
Since 07d24fe50 regions are not translated to their on-screen
coordinates anymore, but are relative to the origin stored in the
constraint. This origin however was not considered when checking whether
the pointer was within the constraint region. This meant that the
constraint region would appear to always be placed at 0,0 instead of on
the surface.
Fix this by using the cursor position relative to the origin.
Fixes: 07d24fe50 ("backends/native: Allow infinitely small pointer constraint regions")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3409
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3749>
We close wayland popups when a button or touch release happens outside
of the grab, except we don't want to close them when that button release is
actually the release of the press that was opening the grab in the first
place.
We never see the press event that opened the grab, so the first event we
see is actually always a release. Make sure to not close the popup on that
event, and instead only close the popup if we see the press count drop from
1 to 0.
This fixes a bug where popup would close right after they open. To
reproduce, click to open a popup, hold pressed and move the cursor over
shell chrome, then release. Or alternatively test with a popup that gets
opened with a long-press gesture (eg. long touch long press on libadwaita
tabs), just doing the touch long-press and then release.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3631>
Until this surface or its parent is finalized.
This makes sure that any `MetaWaylandSubsurfacePlacementOp` referencing
this surface for sibling will be applied as intended.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3737>
This reverts commit 35d92e0fac.
This turned out to cause trouble, because it can prevent
MetaWaylandSurface::applied_state.subsurface_branch_node from ever
getting linked up for a sub-surface.
It shouldn't be necessary anyway, since permanently_unmap_subsurface /
wl_subcompositor_get_subsurface reset the sub-surface state as defined
by the protocol.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3737>
GSettings overrides can be active and set the default value depending on
the XDG_CURRENT_DESKTOP environement variable. For the tests we run, we
rely on the default settings by using the GSettings memory backend but
we also need to make sure not overrides are in place.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3730>
because we want to switch between two workspaces. In some configurations
there is only a single workspace at this point so trying to get current
workspace + 1 gets us a NULL pointer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3730>
The internal representation of the min/max width of windows include what
is outside of the window geometry, so when the window geometry changes,
but the min/max size did not in the same commit, we'd be left with an
out of date min/max size, potentially causing windows to shrink when
configured.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3755>
local_error in meta_egl_query_device_string() is using g_autoptr,
meaning that it was getting freed after g_propagate_error(). This then
would result in error->message becoming invalid, causing crashes when
logging the error message later on.
Fixes: 8234f5bc7 ("egl: Return success status from meta_egl_query_device_string")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3758>
Just shovel the data through our own stdin/stdout, which will end up at
the right place (e.g. /dev/null).
This should hopefully solve `mutter-dist` failing due to a D-Bus method
call timeout in CI.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3757>
An action can be performed both on a signal emitted by a window or by
a well-known signal name, so that one can do:
[window-id]::signal => command
And so these would work:
w/1::position-changed => resize w/2 20 30
::monitors-changed => move w/1 30 40
We only support "monitors-changed" global signal name or connecting to
window signals for now, but this can be easily expanded adding support
for more global signal values or parsing other kinds of object instances
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3701>
Allows to use:
set_strut 0 0 MONITOR_WIDTH*0.5 MONITOR_HEIGHT/4 top [default]
We use the currently focused window as the reference monitor if any or
the primary one if no monitor is explicitly provided.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3701>
We may need to compare monitor size to fractions and while it can be
done using inverse divisions, they were not supported as we didn't parse
floating point values.
So, just add support for both multiplications and divisions, so that we
can easily do stuff like `MONITOR_HEIGHT*3/4`.
Additions would be easy to support too if we don't care about operator
priorities, but that's out of scope for now.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3701>
We just made an half assed attempt to wait, but e.g. when clients were
waiting for a frame event, the attempt was not enough, as the clients
would not ack any configure until the time they were scheduled to paint
again.
Fix this by actually tracking newly pending configurations, and waiting
for them to return.
On X11, still be lazy and hope for the best, as there is no similar
configuration tracking there.
Some tests were updated to use just 'wait' when there was no actual need
for any extra waiting.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3701>
When a window with an input shape on its decoration window becomes
undecorated and meta_window_x11_update_input_region() gets called via
notify::decorated, the buffer_rect of the window has not been updated
yet while the decorated property has. This would lead to us comparing
the input shape of the client window to the buffer_rect which still
includes the decoration window. This would fail to detect the common
case when the client window has no input shape set, leading to the input
region being set to the size of the client window rather than NULL. If
the window is then resized later, the input shape would remain at the
previous size.
This was not a problem before 6bd920b35, because then we were (wrongly)
always comparing to the client_rect.
Fix this by choosing the correct rect for comparison depending on
whether the window is decorated.
Fixes: 6bd920b35 ("x11/window: Use correct bounding rect to determine NULL input region")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3451
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3720>
There is nothing to allocate for a 0-sized files, and indeed
posix_fallocate() will error out if the passed len isn't greater
than 0.
Now that anonymous files are used to back the memory selection
source, this fixes unsetting the selection when the screen is
locked.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3752>
When a key binding is removed, and a trigger key sequence is dispatched
before the idle callback that resolves and updates the actual binding,
we should handle that gracefully.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3711>
Two new fields: ref_count and removed, are added to MetaKeyHandler, and
it would be freed only if the ref count has reached 0. When handler is
removed from key_handlers GHashTable, key_handler_destroy() would mark
removed as TRUE, and do an unref. handler->removed is checked in
get_keybinding, and binding with handler removed would not be used.
Also in MetaKeyBinding, it now has the ownership of the name field, to
avoid it being freed before logging. Create or copy a binding would
do a ref inc for handler, and free one would unref handler.
Fixes https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/1870.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3711>
When a client resizes on its own, make sure the new size is passed
through the window constraints machinery directly, to trigger any
potential window management rule that might apply.
Fix a couple of tests to make use of this behavior by introducing a new
'wait_size' command that waits until a window has been resized to a
expected size.
This replaces the fix introduced in 0e736af301 ("window: Ensure
constraints after a Wayland client resize").
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3700>
Starting the timeout to move from hidden to suspended before the window
is mapped means we don't have a previous window configufration which we
need to get the new window configuration with the suspended state.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3731>
Sometimes the test runners are saturated with other work. Bump the test
timeouts by a multiplier of 5 with the hope that they now will be much
more likely to have time to finish in time.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3735>
These functions are specific for X11 windows only so we should check
if the passed window is an X11 one, not just a MetaWindow since we're
casting to the actual type at later point.
Fixes changes part of commit e1e6534eb
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3727>
If a window sends a configure stacking request, we were comparing the
active window with the event window even though they were different
client types (e.g. wayland and x11).
This was leading to a critical error, so let's handle this by ensuring
that the active window is of the same kind of the event window before
doing x11-specific checks. Behaving as different applications in case.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3727>
Because `meta_kms_impl_device_simple_initable_init` is called in the
middle of `meta_kms_device_new`, the crtcs list for `MetaKmsDevice`
has not been populated yet. And thus the loop to detect missing
cursor planes and create fake ones never iterated. But the crtcs list
does already exist in `MetaKmsImplDevice` so iterate over that instead.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3264
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3676>
Prior to commit 5dfed8a431, the MetaWaylandKeyboard would always remember
the last key press serial, and consider it valid after the key was released,
as long as no other key presses/releases happened in between.
That commit improved things so that MetaWaylandKeyboard can track multiple
keys being pressed simultaneously, but also changed so that the serial for
a key press is immediately forgotten after the key press event was received.
This may break in situations like testing or keyboard macros where key
press and release is handled in a quick sucession, so the client reaction
to the key press (e.g. popping up a menu) might arrive too late.
Add a sort of spiritual successor to this handling, and make keyboard
press serials corresponding to the last key up forgotten at the next
key press/release received.
Fixes: 5dfed8a431 ("wayland: Preserve serial for all pressed keys")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3458
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3721>
We did not track the current surface (i.e. the logical focus) too
thoroughly, so there might be chances that a stale surface pointer
here becomes the focus. Track its destruction (like it's done at e.g.
MetaWaylandPointer) and unset the current surface early, in order
to avoid possible invalid memory access.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3372
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3715>
It seems only the iconcache used to use it, but this is gone since
commit d16ddc42ce.
Even before that, the Xrender usage was removed in commit 556e7694de,
albeit leaving a redundant include <X11/extensions/Xrender.h> in its
place then, which comes from libXrender.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3716>
Fix an obvious copy paste error that slipped through the cracks.
Fortunately it doesn't have a visual impact for well behaving clients
but only makes us not hit direct-scanout paths, assuming no other bugs
in the stack.
Fixes: f21762ea6e (wayland: Add support for preferred_buffer_scale/transform)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3717>
This call is meant to replace meta_wayland_keyboard_get_focus_client(),
since we will always have a MetaWaylandSeat with an input focus (or not),
but we may or may not have a MetaWaylandKeyboard.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3707>
When there is a new owner but there is no matching mime type we clear
the saved mimetype and the saved clipboard but an outstanding async
meta_selection_transfer_async can set the saved clipboard.
Abort the async transfer when we have a new owner.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3678>
The precondition checks in meta_selection_source_memory_new can return
NULL if the mimetype is NULL but callers expect the error to be set when
NULL is returned.
Let's just make sure we never call it with a NULL mimetype.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3678>
The current code checking keyboard serials for popup/grab
validation is a bit simple, tracking one key press exclusively.
This may break expectations if a client uses a serial
corresponding to a previous key that is still pressed.
Keep track of the serials corresponding to all pressed keys,
and ensure these are reset across focus changes, since the
validity of those serials is already outdated. The code does
still keep track of a single (last) key release serial, since
the validity lifetime is somewhat underdefined with those if
we keep track of multiple keys simultaneously.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3267
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3644>
Instead of initializing to 'suspended', which will send the `SUSPENDED`
xdg_toplevel state, set it to hidden at first. If the window is placed
on an inactive workspace, it'll eventually enter the 'suspended' state,
but will have had some time in non-suspended state to get map, even if
not visibly.
This fixes inital suspended state when mapping a window maximized.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3229
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3475>
This, in contrast to meta_window_should_be_showing() reports whether a
window should be showing despite not being showable. This is useful to
know the intended visibility state that should happen in the immediate
future.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3475>
d991961ae changed this code from client rect to buffer rect to account
for the fact that for SSD windows the decorations are now included in
the bounding region, but it kept the variable name as client_area and
the comment was also still referring to the client rect.
Fixes: d991961ae ("x11: Use input region from frame window for decorated windows")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3697>
When updating the input region we check whether the input shape reported
by XShape matches the bounding region of a window to determine when it
was not set by the client. We would then use a NULL input region instead
which always matches the full size of the window. The code however was
using the client rect for this comparison, which does not include the
window frame. Since d991961a the frame is considered part of the input
region.
This meant that for SSD windows where the input region would match the
bounding region, we would not detect that and fail to set the input
region to NULL, but instead set it to the reported input shape.
Usually this would not be the case due to the GTK frame window having
shadows and a resize region, but in the presence of an issue that causes
GTK to wrongly detect _GTK_FRAME_EXTENTS as not supported, GTK would not
draw shadows or set an input shape. And due to GTK not updating its
input shape, there would be no further calls to
meta_window_x11_update_input_region() after the initial one.
The input region would therefore remain at the fixed size from the
initial call. This was causing windows to become click-through outside
of the region corresponding to their initial size after being resized.
Fixes: d991961ae ("x11: Use input region from frame window for decorated windows")
Related: https://gitlab.gnome.org/GNOME/gtk/-/issues/6558
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3404
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3697>
Right now we store touch serials on their according MetaWaylandTouchInfo
entries. These entries are gone as soon as the touchpoint ended though, and
it's not unlikely that clients will respond to that touch-end event after we
removed the touchpoint.
In this case we currently can't match the client provided serial to any of
our known touch sequences, which causes xdg_popup grabs that get requested
shortly after the touch-end to fail.
Let's be a bit more gentle on clients here and store the latest touch-down
serial on the MetaWaylandTouch, so that it continues to be around after the
touch-end and we can match the serial of the xdg_popup_grab() as expected.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2946>
In the rare event that hotplugs destroy and then create a new EGLContext
with the exactly the same ID, this ensures we will forget the old program
which presumably wouldn't work in the new context. It will be recreated.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3304>
The EGL context can only import and blit an EGLImage if the
backing DMA buffer has a format modifier combination that is advertised
as supported and not marked as "external_only".
When the context can't blit the imported image, we can still paint using
it GL_OES_EGL_image_external using the texture target
GL_TEXTURE_EXTERNAL_OES.
However, treat drivers who doesn't support modifiers at all as if they
do support blitting, if the modifier is 'linear', to avoid regressions.
[jadahl: Make shader path a fallback to allow hardware to utilize copy
engines via blitting]
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/6221
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2247
Related: https://launchpad.net/bugs/1970291
now only falls back if modifiers are supported, and they mark linear as
export only.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3304>
This is a critical part of any OpenGL program. Mesa allowed us to get
away without it and provided a sane default of the full buffer, but
Nvidia seems to default to an empty/zero viewport so would refuse to
paint any pixels.
In the OpenGL ES 2.0 spec this is ambiguous:
> In the initial state, w and h are set to the width and height,
> respectively, of the window into which the GL is to do its rendering.
because the first "window" used is EGL_NO_SURFACE in
init_secondary_gpu_data_gpu. It has no width or height.
In the OpenGL ES 3.0 spec the ambiguity is somewhat resolved:
> If the default framebuffer is bound but no default framebuffer is
> associated with the GL context (see chapter 4), then w and h are
> initially set to zero.
but not entirely resolved because neither spec says whether
EGL_NO_SURFACE should be treated as zero dimensions (Nvidia) or ignored
and not counted as the first "window" (Mesa).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3304>
As mentioned in the OES_EGL_image_external spec, there is no implicit
sync between the EGLImage producer and consumer. And in this code path
we don't have meta_drm_buffer_gbm_new_lock_front on the primary GPU to
do it for us either. So synchronization has to be done manually or else
the secondary GPU is likely to get an unfinished image.
This problem has only been observed when the secondary GPU is using the
Nvidia proprietary driver.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3304>
When the monitors change meta_window_update_for_monitors_changed is
called which is responsible for updating window->monitor. It however can
go through the entire window placement and constraint machinery before
it's able to do so. In this period window->monitor points to the old
MetaMonitor where the monitor number doesn't reflect the index into the
MonitorManager anymore.
Avoid relying on the window->monitor->number and go through the Monitor
directly.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3402
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3691>
In every other device and circumstance, we update the logical focus
(i.e. the surface that would be focused, if no other circumstances
applied) but let the MetaWaylandInput figure out the surface that
should be effectively focused, if at all. This is where we apply the
actual compositor state (e.g. grabs) and result in an effective focus
window.
The only exception to this is `meta_wayland_seat_set_input_focus()`
where the logical focus for keyboards and related devices is set, but
also applied as the effective surface.
We should do the same as everywhere else, and let MetaWaylandInput
focus determine whether the logical focus is also the effective focus.
These replaced set_focus() internal calls will happen through the
default MetaWaylandEventInterface.focus implementation in MetaWaylandSeat.
This resulted in keyboard focus being set on windows in circumstances
it ought not to, like in the overview. Actual key events were never
sent in these circumstances, but changes to modifier state could.
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/7528
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3704>
If a modal dialog (i.e. with a "revolves around center of parent window" policy)
becomes fullscreen, we cannot neatly honor both the modal dialog going fullscreen
and the window staying around the center of its parent.
In order to make fullscreening work in this situation, allow fullscreen modal
dialogs to "snap out" of the parent. This rule will become again effective after
the window is unfullscreened.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3425
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3695>
By ignoring obcured surfaces we can generalize the single-pixel-buffer
background check to work with arbitrary surface trees, as long as the
two top ones are the background and content surfaces.
As a nice side effect of this, clients like Firefox now usually take the
shorter `meta_window_is_fullscreen (window) && n_visible_surface_actors == 1`
route.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3699>
Move the code out of cogl_onscreen_egl_swap_buffers_with_damage, and
call the new function from callers of the former.
v2:
* Use early return if the cogl context doesn't support timestamp
queries. (Sebastian Wick)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3689>
`mtk_rectangle_new()` allocates the object dynamically,
but in the "contains-point" test case the allocated object
was not freed. Fix that by creating the object on the stack.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3679>
Avoids using a static and make sure things are properly freed by
using g_autoptr. Also take the opportunity to add some missing NULL
initialization for auto-pointers variables
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3687>
In consistence with the code style, and in order to fix build errors
with older clang:
../src/backends/native/meta-onscreen-native.c:521:7: error: expected expression
521 | graphene_rect_t src_rect;
| ^
../src/backends/native/meta-onscreen-native.c:529:39: error: use of undeclared identifier 'src_rect'; did you mean 'dst_rect'?
529 | &src_rect);
| ^~~~~~~~
| dst_rect
../src/backends/native/meta-onscreen-native.c:522:20: note: 'dst_rect' declared here
522 | MtkRectangle dst_rect;
| ^
And warnings with newer clang:
../src/backends/native/meta-onscreen-native.c:521:7: warning: label followed by a declaration is a C23 extension [-Wc23-extensions]
521 | graphene_rect_t src_rect;
| ^
This should allow the build for coverity to succeed again.
Fixes: adc776d0d7 ("crtc/kms: Pass on src and dst rects to primary plane assignments")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3686>
This implements the explicit sync protocol linux-drm-syncobj-v1. This
works by importing a DRM syncobj timeline and importing/exporting fds
to/from the sync points on the timeline corresponding to buffer acquire
and release. We take fds for sync points provided during a surface
commit and use them to delay transaction application, and fetch fds
from Cogl to signal when we are done using a particular buffer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3300>
This adds the explicit sync wayland protocol to the list of build
dependencies.
This adds a copy of the linux-drm-syncobj protocol that we will use
privately for builds. This avoids the pain of requiring wayland-protocols
1.34 for distros. This commit can be reverted once we want to use
linux-drm-syncobj from the system wayland-protocols.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3300>
This keeps an internal sync_fd for the latest work submitted to
the GPU. This can then be fetched with cogl_context_get_latest_sync_fd
for use. This is intended for use with linux-drm-syncobj-v1
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3300>
On big endian architectures the mapping of drm formats to cogl
formats is significantly trimmed compared to on little endian
architectures.
meta_wayland_init_shm tries a bunch of formats, including some
that just aren't mapped on e.g. s390x.
The code asserts the mapping will exist, however, leading to
crashes when Xwayland starts.
This commit makes failure to find a mapping non-fatal.
Suggested by Jonas Ådahl.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3672>
ClutterInputMethods content hint or purpose will be set again (even to the
same values) whenever a wayland client sends us a new content type/purpose.
Gtk appears to always set a content purpose on wl_text_input changes, so we
currently set and notify the "content-purpose" property on every change in a
gtk text field.
Since the OSK is gnome-shell listens to this property and re-generates its
entire layout when the content-purpose prop gets notified, this is currently
causing lag/freezes on every keypress in the OSK in gnome-shell.
So ensure to not notify these properties in case they're equal and set the
properties in the same way as we usually set them instead of going via
GObject.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3645>
When logging in from gdm to gnome, the main plane is deactivated, and
leads to the screen going blank before gnome is able to enable it
again.
Using the new CloseFB ioctl, allows to keep the gdm login screen
displayed until gnome-shell replace it.
Signed-off-by: Jocelyn Falempe <jfalempe@redhat.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3659>
This also gets rid of the MetaFrameSyncMode enum and instead issues a
VRR update when the requested state differs from the CRTC state.
Fixes: fee33299 ("onscreen/native: Allow requesting frame synchronization")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3646>
This includes checking the vrr_capable property on the connector as well
as the VRR_ENABLE property on any CRTC the connector might get assigned
to. Also takes into account when a GPU is tagged for broken VRR support.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3646>
As we are getting close to have a Wayland-only build, guard those usages
so we don't 'regress'.
Once Kiosk figures it use case, we can revert both this commit and the
previous one.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3652>
When multiple configure requests sent to the wayland client within a sort period, maximized windows may end up in wrong position and cover struts. To avoid this, queue a resize when the resize event sent by the wayland client results a changed size or position to ensure that the final size and position will be always right.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3601>
Using `meta_wayland_single_pixel_buffer_from_buffer()` can fail if
the buffer resource is already gone. Given that we release the buffer
immediately, that is actually expected - but the client this was tested
with did/does not do so.
This made the single pixel black bar optimization fail for
GstWaylandsink, while working for the mpv Wayland backend.
Access the object directly, which is both faster and works as long as
the buffer is alive.
Fixes: ed50cbbfe4 (window-actor/wayland: Update scanout candidate check)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3662>
Gestures are independent from each other when they are on different
actors and when they don't have a conflict over input with each other: For
example a gesture on one window and a gesture on another window will
recognize at the same time perfectly fine right now.
For those gestures (let's call them "independent gestures") we don't want
to control how the conflicting input should be handled, i.e. whether one
gesture wins over another or whether both can be recognizing using the
same touchpoint (e.g. zoom+rotate on the same actor). Instead we want
to control whether they are allowed to start while another one is running.
For now, make it impossible for two gestures to recognize globally at
the same time in all cases. This helps prevent subtle bugs and makes life
easier for users of the API. We can introduce API for fine grained control
over the behavior later.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2389>
With the next commit, we'll need a list of all gestures that currently are
active globally. Since actors and actions (and therefore also gestures) in
Clutter are bound to a stage, it makes sense for this list to exist on the
ClutterStage level.
The list itself is a simple GPtrArray (to allow for quick searches) that
doesn't reference the gestures and is not manipulated by the stage itself.
All manipulation of the array is left to ClutterGestures.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2389>
Quite often there are situations where multiple gestures try to
recognize, keeping track of the same set of points (for example an edge
drag gesture on the stage and a click gesture somewhere in the
scenegraph). Usually what's wanted here is that the first gesture to
move to RECOGNIZING wins over all other active gestures and "claims" the
point for itself.
We implement this by introducing a concept called "influencing". It
works by making all gestures operating on a shared set of points aware
of each other using ClutterAction->register_sequence().
ClutterGesture uses this vfunc to keep track of all other
ClutterGestures that are potentially conflicting, and keeps a list
(priv->cancel_on_recognizing) of those. As soon as the move to
RECOGNIZING happens, all gestures inside this list get moved to
CANCELLED.
To allow fine-grained control over this behavior, two APIs are
introduced:
1) on the implementation level (should_influence() and
should_be_influenced_by()): This is a vfunc that gets called as soon as
a potential conflict is detected. It's helpful when a specific gesture
always behaves the same towards another gesture, for example to make
sure a LongPress gesture never cancels a DragGesture.
2) on the gesture user level, clutter_gesture_can_not_cancel() is
introduced: This allows control for the user of a gesture to specify
that a specific instance of a gesture won't cancel another gesture.
Calling this twice so that both gestures can't cancel each other allows
for things like simultaneous recognition of a pinch-to-zoom and rotate
gesture.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2389>
Introduce ClutterGesture, a new ClutterAction subclass and the successor
of ClutterGestureAction that brings the necessary tools to handle
sequences of events and abstract touch and mouse gestures from those.
The big difference compared to ClutterGestureAction is that ClutterGesture
provides the implementation with point_added/moved/ended and
sequences_cancelled events and expects the implementation to move the
ClutterGesture through the ClutterGestureState state machine. This state
machine is then used internally by ClutterGesture to coordinate with other
gestures.
With the next commits, ClutterGesture will handle relationships between
conflicting gestures completely by itself, allowing the implementation or
the user to specify the details of the relationship between two gestures.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2389>
clutter_event_get_source() is still valid for the case of crossing events,
just like clutter_event_get_related(). The latter is not deprecated, so the
former shouldn't be either.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2389>
ClutterStage will unref an action in the middle of its own event handler in
case the action causes its own actor to be destroyed. In this case the
action would get freed underneath our feet. To avoid it, take a ref on the
action while calling its handle_event() vfunc, just as we do in
clutter_actor_event() while emitting an event to an actor.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2389>
The gnome-shell magnifier listens to the `visibility-changed` signal
and calls meta_cursor_tracker_set_pointer_visible(false) when the
cursor became visible.
This leads to a reentrance in meta_cursor_tracker_set_pointer_visible()
and clutter_seat_uninhibit_unfocus() gets called twice, once from the
meta_cursor_tracker_set_pointer_visible(false) by the magnifier and then
the original meta_cursor_tracker_set_pointer_visible(true) continues,
after the first call has set is_showing to false again. This breaks the
inhibitor counting and the ability to use the cursor while using the
magnifier.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3661>
Near window destruction, there might be cases where the surface
actor does no longer have a surface, yet it's still in the stage
and eligible for picking. In that situation looking for modal
dialogs attached to this surface will evidently fail, so avoid
this check on a NULL surface.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3332
Fixes: 93a9e7f3f ("core: Move code ignoring events on windows with modals to Wayland")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3660>
`eglGetProcAddress()` used to not work for core API in EGL versions
below 1.5. The workaround in place in turn can fail - notably for setups
with a local Mesa build in /usr/local.
EGL 1.5 is almost 10 years old and similar workarounds don't seem to be
in place for toolkits we rely on, notably GTK4. Thus remove it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3655>
Since commit e30eb78891 `ClutterFrameClock` assumes that a valid CPU time
implies timestamp query support, which is also checked in
`cogl_onscreen_egl_swap_buffers_with_damage()`.
Unconditionally setting the CPU time on direct scanout meant that the
compositing path would be stuck on the last (direct scanout optimized)
result on GL implementations without timestamp query support since.
be0aa2976e (clutter/frame-clock: Avoid rapidly toggling dynamic max render time)
Fix that by explicitly marking the gpu rendering duration as valid when
querying the GPU timestamps is supported and check for it ClutterFrameClock.
Fixes: 56580ea7c9 ("backends/native: Assume zero rendering time for direct scanout buffers")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3655>
This makes sure that xdg-output.logical_size and
xdg-output.logical_position are only sent when they actually changed.
There should be no behavior change in wl_output_transform_from_transform
but it now uses the same technique of tracking the protocol state and
comparing it to the current state to compute which properties have
changed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3622>
X11 server side focus changes, such as when a focus change was requested
by mutter for a globally active window, did not go through
meta_display_set_input_focus(), which is responsible for emitting the
`focus-window` signal. Since this signal is what triggers the display
server specific code to handle focus changes, this was leading to a
problem on Wayland where the focus remained on the last active Wayland
window when the focus got changed to a globally active XWayland window.
This commit now changes handling X11 server side focus changes to also
go through the code path that emits the signal while making sure to not
trigger another focus change and keeping the same serials as the
previous code to not interfere with future focus changes.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3328
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3651>
MetaSelectionSourceMemory currently uses GBytes for its underlying data.
This can cause memory overhead when large items, such as HD images, are
stored in the clipboard. This commit changes the underlying data
structure to a MetaAnonymousFile object, which writes to memfd instead
of heap. When reading, MetaSelectionSourceMemory will create a
Gio.UnixInputStream from the file descriptor generated by
MetaAnonymousFile. We subclass the UnixInputStream as
MetaUnixInputStream, to override the stream's close_fn function so
that it invokes meta_anonymous_file_close_fd when the stream terminates.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3551>
Which got introduced in wl_compositor version 6.
Note that if the surface is visible on multiple monitors with different
transforms, we pick the transform of the monitor which we choose for the
scale as well. This doesn't really matter at the moment, as the
transform is only really relevant for direct-scanout - which we
currently only support for fullscreen clients.
Once we support direct-scanout for partially visible clients we'll
likely want to introduce a more sophisticated algorithm.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3580>
Some clients - notably the Gstreamer vapostproc element when using Intel
GPUs - only support BGRA, not BGRx. We already assume that we can
support this format for window screen casts, and even in case of failure
we now have a re-negotiation fallback in place. Thus it's pretty safe to
support it for all screen cast types.
The possible duplication in case of window screen casts doesn't seem to
be a problem for either Pipewire or existing clients like OBS.
Note that the implementation lays the foundation to make it easy to add
more formats in the future.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3617>
Currently, ClutterFrameClock uses g_source_set_ready_time() to determine
the usec timing of the next frame. That translates into a poll() with a
millisecond timeout if no trigger occurs to break the poll() out early.
To avoid spinning the CPU, GLib always rounds *up* to the next millisecond
value unless a timeout of 0 was provided by a GSource.
This means that timeouts for the ClutterFrameClock can easily skew beyond
their expected time as the precision is too coarse.
This applies the same concept as GNOME/glib!3949 but just for the
ClutterFrameClock. That may be more ideal than adding a timerfd for every
GMainContext, but we'll see if that lands upstream. I wanted to provide
this here because it could easily be cherry-picked in the mean time if
this is found to be useful.
From a timer stability perspective, this improves things from erratically
jumping between 100s and 1000s off of the expected awake time to single
or low double digits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3636>
The POINTER_EMULATED flag was a convenience to filter either
side of smooth/discrete events that we should ignore based on
the source.
This distinction was challenged, first by v120 mice that use
Clutter smooth events to deliver semi-discrete changes, second
by commit e0c4b2b241 ("backends/native: Mark the emulated smooth
scroll event as such") which made the smooth events be flagged
as emulated, and the discrete whole-step events marked as
real.
This distinction feels convenient for the time being, since
upper layers might be confused by real smooth scroll events
without finish flags. Adapt to this change at MetaWaylandPointer
so that we drop the POINTER_EMULATED check, and the events are
perhaps filtered based on their source and the preferred
wl_seat version of the client that we are talking to.
This handles the whole grid of combinations:
- wheel sources with wl_seat >=8 result in wl_pointer.axis_value120
from "emulated" smooth scroll events, with value120 information.
- wheel sources with wl_seat < 8 result in wl_pointer.axis_discrete
from "real" discrete scroll events.
- finger/continuous sources prefer smooth events. Previously, always
non-emulated for those.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3642>
The wl_pointer.axis_discrete axis (applicable to wl_seat <= v8) is
meant to be sent together with wl_pointer.axis events in the same
frame. And the wl_pointer.axis_value120 event replaces it in
wl_seat >= v9, but has the same relation with the other information
available in a frame.
This emission should not be conditional to anything, so drop the
various checks leading to maybe sending wl_pointer.axis or not.
This fixes emission of wl_pointer.axis in conjunction with discrete
events, for some combinations of versions and (non)value120 mice.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3642>
Unveiled by commit e0c4b2b241 ("backends/native: Mark the emulated smooth
scroll event as such"). The sudden "lack" of smooth scroll events (Used by
Clutter to forward v120 events) made it evident we silently ignore Clutter
discrete events, as we don't send wl_pointer.axis_value120 for these.
Fix this by assigning a value120 value to discrete scroll events. This
makes wl_pointer.axis_value120 events actually sent on non-v120 mice.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3642>
The default value will change from `false` to `true` in future Meson
releases, so let’s be explicit.
We don’t want to check the exit status of the program in this case, as
we parse the `--help` output instead.
Signed-off-by: Philip Withnall <pwithnall@gnome.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3640>
meta_window_handle_ungrabbed_event() triggers the show of the window menu in
gnome-shell via meta_window_show_menu() on hold of Meta + right mouse button
click.
Since meta_display_handle_event() was refactored lately and now forwards a
lot more events to Clutter (including the one triggering the window menu),
gnome-shell now sees this event after the menu has opened, figures that the
source-actor is outside of the menu, and immediately closes the menu again.
This is the correct behavior from the PopupMenuManager on the gnome-shell
side, it is the responsibility of the event handler that opens the menu (aka
meta_window_handle_ungrabbed_event()) to return CLUTTER_EVENT_STOP and stop
event propagation.
So fix this issue by adding a return value to
meta_window_handle_ungrabbed_event() and stopping event propagation in case
the event opened the window menu.
While at it, also return CLUTTER_EVENT_STOP for events triggering window
drags, so we can drop the extra check for that in
meta_display_handle_event().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3623>
For each libinput scroll event we generate two clutter events
(continuous and discrete), one of them marked as emulated. libei
explicitly specifies that emulation of scrolling must be done in the
client (if desired) so drop the emulated one.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3637>
`meta_window_set_user_time()` will not update the window's
user time if it timestamp in the argument is before the
currently saved timestamp. However, when trying to work around
problematic timestamps, this is exactly what needs to be done.
So force the update to happen by setting the "is user time set?"
flag to false.
Fixes: 8f3da9f68a ("Use meta_window_set_user_time for setting user time consistently")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3027>
priv->current_drag is the property that
meta_compositor_get_current_window_drag() and therefore also
meta_display_is_grabbed() bases its check on. We use
meta_display_is_grabbed() to select which cursor to use (window cursor
vs root cursor) when updating the cursor in MetaCursorTracker.
Since meta_window_drag_begin() sets a new cursor, and therefore triggers the
cursor tracker to update the current visible cursor, we should set
priv->current_drag before the call to meta_window_drag_begin(). This makes
sure the cursor tracker sees that there's a window drag and changes the
cursor right away when the window drag begins (instead of doing so on
subsequent pointer events).
Fixes: 525ed1166c ("wayland/pointer: Unset current surface during window drags")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3630>
This is used as the minimum refresh rate in the variable refresh rate
range.
This value is expected to be found in the DisplayPort and eDP EDID of
every monitor that supports variable refresh rate.
It is also found in the HDMI EDID of some monitors that support
variable (FreeSync), but most likely not all of them. The rest require
parsing the AMD vendor extension which libdisplay-info doesn't
support.
No fallback is implemented for cases where libdisplay-info support is
disabled.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3576>
Find a surface actor that meets the criteria to drive the refresh
rate for the view. If a new surface actor is found, add a
reference to it and request frame synchronization to be enabled
for the relevant MetaRendererViewNative.
Whenever the surface actor schedules a repaint or an update, and
in case frame synchronization is enabled for the
MetaRendererViewNative, schedule the update to occur "now". This
effectively makes the surface actor's frame rate drive the refresh
rate of the monitor.
If the actor is frozen or destroyed, it is no longer expected to
schedule repaints or updates and should stop driving the refresh
rate.
When there is no surface actor to drive the refresh rate, request
frame synchronization to be disabled for the MetaRendererViewNative.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
This makes all stage updates that result from applying the pending
state of a Wayland surface emit an "update-scheduled" signal in the
context of the relevant surface actor.
A common case where an "update-scheduled" signal is needed is
when applying "empty" client commits. In this case a
"repaint-scheduled" signal would not be emitted since the commit
doesn't trigger a repaint. However, it is still important to add
handling for such updates with variable refresh rate when the
releavnt actor is also driving the refresh rate.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
This function allows scheduling a stage update in the context of a
surface actor and emit the "update-scheduled" signal. This signal is
similar to "repaint-scheduled", but can be used to be notified of
updates that do not necessarily result in a repaint.
With variable refresh rate, any update potentially affecting the
frame pacing of a surface actor needs to be handled differently
depending on whether that surface actor drives the refresh rate or
not.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
Frame synchronization is enabled for a view as long as it's
applicable to be enabled. It is considered applicable if it's both
requested for the onscreen and if the onscreen uses a CRTC which is
configured with a variable refresh rate mode.
When frame synchronization is enabled, it enables both the the variable
scheduling mode of ClutterFrameClock and the variable refresh rate
property for the CRTC.
Changes in the frame synchronization mode are applied asynchronously,
before the next frame is drawn.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
Add variable refresh rate output modes for connectors that are VRR
capable when VRR is not disabled for the GPU.
Variable refresh rate output modes are sorted before their fixed
refresh rate counterparts. They are also marked as the preferred mode
for the output between the two.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
When VRR is not disabled for a GPU, create two variants of every
display mode: one with fixed refresh rate and another with variable
refresh rate.
The variable refresh rate modes are not used yet. They will be used
in a following commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
This can be used to disable VRR in specific drivers and hardware
combinations where it is found to be problematic.
No default rules are added for now to encourage testing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
Require the "variable-refresh-rate" keyword under the
"experimental-features" gsetting to enable the feature for now.
It would no longer be required once the experience with variable
refresh rate is good enough for general use and handles all common
use cases well.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
Use a sufficiently low refresh rate to calculate the CRTC deadline
when variable refresh rate is enabled. This is done to avoid cursor
updates from driving the monitor refresh rate.
It's not great solution and is sometimes not enough, but it avoids
stutter in the main content as a result of cursor movement in most
cases.
The unfortunate downside of this approach is that cursor movement
would usually only update with the main content and would not be
smooth when the main content updates are not frequent enough.
A better solution may use an approach similar to LFC (Low Framerate
Compensation) to insert cursor-only updates between updates of the
main content, but achieving adequate results with an approach of this
nature requires more research and experimentation.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
It is not trivial to accurately estimate the expected presentation
time with variable refresh rate, and not doing so only affects debug
prints.
No change in behavior for now because the expected presentation time
is always calculated. A following commit will introduce a case where
it is not.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
Add functions to update and monitor the value of the "VRR_ENABLED"
KMS property.
This requires the addition of functions to process CRTC property
updates in both the atomic and the simple KMS backends. The
implementation is similar to the implemention of processing
connector updates.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
A new variable scheduling mode is introduced which allows lower
priority updates to be scheduled on a timeout which represents a lower
refresh rate, while allowing high priority updates to be scheduled to
occur as soon as possible.
This mode will be used by following commits to implement
synchronization of page flips to the update rate of specifc surface
actors.
High priorty updates are either scheduled to occur "now" if they
arrive at a rate which is lower than the maximum refresh rate, or
according to the measured maximum render time if they arrive at a
rate which meets or exceeds the maximum refresh rate. This approach
allows achieving low input latency in both scenarios.
Seperate handling for low priority updates is needed to avoid visible
stutter in the content of the surface that drives the refresh rate. An
example for a low priority update is cursor movement when the KMS
deadline timer is disabled.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1154>
This is conditionally toggled by grabs on the current key focus depending
on whether the current key focus actor would receive events according
to the grab or not. Which means it's no longer a reliable method for an
actor to know it does have focus, without asking the stage about it.
Avoid this check and ask the stage for the key focus, in order to make
key focus actors able to unset themselves despite the presence of grabs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3629>
Commit d48129f5ee broke the displaying of our own drag cursor during window
drags, as now the window cursor is always used, even during stage grabs
(window grabs are just a kind of stage grab).
To fix it, while not regressing on the intention of the other commit, unset
the MetaWaylandPointer surface in case a window drag is active (instead of
all kinds of grabs) by checking via meta_display_is_grab().
Fixes: d48129f5ee ("wayland: Fix pointer cursor during Wayland grabs")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3316
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3624>
Move the bulk of the implementation inside MetaWaylandPointer
files, like it happens in other places (e.g. MetaWaylandTabletSeat).
This avoids MetaWaylandPointer struct peeking from outside.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3627>
The _MUTTER_FRAME_EXTENTS need to have the correct size before the
window is shown, otherwise the client window size will effectively have
the headarbar height subtracted when initially syncing the frame/client
geometries.
This reverts commit f10b3eac62.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3628>
The min height reported by ClutterText when ellipsize and line-wrapping are
enabled was too small to fit the text.
The coordinates from `pango_layout_get_line_extents()` are baseline relative,
so the `y` coordinate means that the highest ascent would be `-y` above the
baseline and height gives the span between the ascent above the baseline and
the descent below it, so `logical_line_rect.y + logical_line_rect.height`
gives us the size of the descent. This is the wrong height to use for the
height of the actor.
The coordinates of the layout extents don't seems to be related to the baseline
and are just for offsets when rendering, that's probably how this bug got
initially introduced.
Therefore, the `y` coordinate from the layout is the correct offset to use,
even though, when looking at `pango_layout_get_extends_internal()`, it appears
that `y` is always set to 0.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3610>
Upstream the annotations used in Gala, make more Cogl methods available since
things got moved to a GObject base and thus works natively with the introspection.
Add all the public API to the introspection.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3619>
QueryDeviceString can successfully return NULL. The convention however
is that when NULL is returned, the error will be set.
This commit makes the returned string an output parameter which allows
us to return the success status and have the error set accordingly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3614>
The convention is that when a function returns FALSE or NULL, the error
will be set. In this file we call set_egl_error but it might not set an
error. Code of the form
if (egl_do_thing (..., &error))
use_error (error);
will crash in those cases.
This commit makes sure we always set the error even if EGL doesn't give
us an error.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3614>
We depend on Gio for many introspection features (GSettings, GFile...)
without this we'd end up not having many symbols being exposed.
And errors such as
Warning: Meta: meta_display_add_keybinding: argument settings:
Unresolved type: 'GSettings*'
As per previous commit we now have this dependency implicitly, but still
it's better to make it clearer in case Clutter would drop it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3620>
This is added as a hint so that the caller can optionally use it, e.g.
in the case that a window drag is driven by the mouse. This will be used
to more accurately follow the drag position, other than starting at the
current pointer position.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3546>
Pointer constraints obey to logical pointer focus changes, and looking
up the current surface in order to maybe enable them is getting ahead
of itself, since the pointer focus might differ from the current surface
due to other factors (e.g. grabs).
Change the code checking whether constraints should be enabled to again
check the pointer logical focus, this will be influenced by the
MetaWaylandEventInterface mechanism, and correctly reflect the logical
state accounting for those factors.
Fixes: 125ba92169 ("wayland: Port pointer constraints to using MetaWaylandEventInterface")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3303
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3618>
Currently, we don't handle too well the removal of a MetaWaylandEventInterface
during meta_wayland_event_interface_invalidate_all_focus(), since the
MetaWaylandEventInterface may be freed at an intermediate point in that function
while handling the focus change for all input devices.
Turn this invalidate_all_focus() function into a MetaWaylandInput method, and
always ensure to use the currently effective MetaWaylandEventInterface when
resetting the focus for each device.
This fixes the situation through handling reentrancy naturally, a focus
sync (say, triggered by a grab) would reset a device focus (say, pointer), which
would remove an event interface (say, a pointer constraint), which would
invalidate_all_focus() again underneath using the new effective
MetaWaylandEventInterface. When that is done, the initial invalidate_all_focus()
call would re-apply the same focus to the same currently effective
MetaWaylandEventInterface, resulting in a no-op for the remainder of the function.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3618>
Even though the logical focus is typically business that only the
MetaWaylandEventInterface mechanism minds about, there are some pointer
subsystems that want to look this up, as opposed to the current surface.
Add a getter to make this easier, without struct peeking.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3618>
This also removes some g_return_if_fail's because the test suite expects
to be able to create arbitrary KmsUpdates even if they don't make sense
for the real state. We just get lucky that the test suite isn't
constructing updates with color space and hdr changes, yet.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3552>
Currently querying support for some output features is done partially
through the OutputInfo and partially via KMS CRTC and Connector objects.
Let's be consistent and use OutputInfo always which works with all
backends and backend types.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3552>
We're currently pretending the gamma LUT has another size. This becomes
a problem when we try to reset the LUT to passthrough, create an
identity LUT for it and it has a size that the kernel doesn't accept.
We do track the size and have utility for creating the LUTs, so let's
just use them.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3552>
The kernel doesn't let us set gamma to passthrough with the legacy API
so we have to trick a bit and create an identity LUT, and also when we
read the KMS state, detect when an identity LUT is active.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3552>
The frame size_allocate() will set it correctly once show() is called by
the window tracker.
This is less code and also help reduces the chance of a brief visual
glitch in fullscreen games during startup. If the window is initially
still decorated the gray area would still show up until the next redraw,
which due to loading times can take a while.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3608>
Relying on the content size_allocate() to determine the content position
can fail in situations where the position of the content has changed,
but not its size.
This happens for example when the window initially is sized fullscreen
height + headerbar height while not considered fullscreen yet. Then when
the window is resized to just the fullscreen height and marked as
fullscreen, the content size has not changed and size_allocate() is not
called on the content. Thus the previous position which assumes the
presence of a headerbar still applies. As a result the window is shifted
down, revealing a headerbar sized area showing the gtk window background
color.
This issue can be avoided by using the frame's size_allocate(), which
gets called in response to all relevant events, such as any headerbar
size changes, headerbar visibility changes, window resizes and
fullscreen status changes.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2937
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3608>
Allows to mark a wayland client window as a DOCK window. The reason for
this is that in Gala (elementary OS's window manager) we would like to
continue using GTK apps as panel and dock on wayland.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3612>
Assigning the corresponding stack layer of DOCK windows is currently X11
specific, because there is no way for wayland clients to set the DOCK
window type. This is about to change, so move the code to the generic
layer handling.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3612>
adaita-icon-theme cleaned up its cursor set, and now only provides
names defined by GTK/CSS. Update the cursor-hotplug test to not
use legacy cursor that will fail with a recent cursor theme.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3613>
meta_window_update_monitor() can emit "::highest-scale-monitor-changed",
and we connected to that signal right before. Let's avoid calling
meta_wayland_surface_notify_highest_scale_monitor() twice and move the
g_signal_connect() for that signal and the initial call to
meta_wayland_surface_notify_highest_scale_monitor() to happen after
meta_window_update_monitor().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3598>
Turns out there is a better solution: Almost always, MetaWindow already has
an idea on which monitor it will be, even if it isn't positioned yet. Since
the last commit we're now using that monitor for setting the
highest-output-scale of the window, so this fallback is no longer necessary.
While we could keep this fallback around and also return a valid scale in
case the surface is not even mapped yet, this means we report fractional
scale twice for new surfaces: Once from
wp_fractional_scale_manager::get_fractional_scale() (here we'll enter the
fallback), and a second time (this time with correct scale) right after
creating the MetaWindow.
Note that wp_fractional_scale_v1 doesn't specify that a preferred_scale
event must be sent immediately after
wp_fractional_scale_manager::get_fractional_scale(), so we can safely remove
the fallback.
This reverts commit 8cfbdb4313.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3598>
MetaWindow always tries to have a main monitor: If the frame rect is empty
and the window has not been positioned, in meta_window_constructed() we fall
back to asking the backend for the current monitor, and in
meta_window_wayland_update_main_monitor() we fall back to
meta_window_find_monitor_from_id(), which then falls back to the primary
monitor.
In general this means that window->monitor is always set as long as there is
a monitor around.
For getting the highest-scale-monitor the window is on, we currently rely
completely on the frame rect. If the frame rect is empty, we set the
highest-scale-monitor to NULL. Since we usually know though which monitor
the window is, or will be on, and window->monitor is even set to that, we
can just fall back to window->monitor for the highest-scale-monitor.
This makes sure ::highest-scale-monitor-changed is emitted right after the
window is created, and it's set to the correct monitor that the window will
be on. This in turn means that we can send a correct wp_fractional_scale
fraction_scale event to clients right away.
https://gitlab.gnome.org/GNOME/mutter/-/issues/3262
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3598>
The existing comment tells us this is necessary:
> there may be drawing between the last damage event and the
> XDamageSubtract() that needs to be flushed as well.
But the commit message for 551101c65c also tells us that
synchronization is necessary before-update. Assuming both are correct
then it needs to be done in both places.
I did try optimizing out the second sync to only do it if damage
arrived during the update, but that doesn't seem to be the issue.
The damage event is arriving before the update starts and it's some
secondary changes within the damage region running late that need
flushing. So this means the client is reporting damage more frequently
than the frame rate and we're ignoring the secondary damage reports
for efficiency (XDamageReportBoundingBox), which is still a good thing.
Fixes: 551101c65c ("compositor-x11: Move synchronization to before-update")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2880
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3590>
Do not include it at header side as it is not part of the installed headers.
Only keep it in cogl-gl-headers.h as it is a private header.
Add it to all the source files that depend on it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3593>
This change adds modifier-aware screencasting support to Mutter.
Implicit modifier support is kept for backward compatibility and the
code fallbacks to implicit modifiers in case any new functionality added
for explicit modifier support fails.
The advertised modifiers are retrieved by a call to
eglQueryDmaBufModifiersEXT() function. The "external only" modifiers are
excluded as Mutter uses the buffers created with the explicit modifiers
as renderbuffers. Support for implicit modifiers is checked with a test
allocation since there are drivers that do not support them.
This change also removes various implicit modifier support checks that
disable DMA-BUF screen casting support globally as they are no longer
needed. DMA-BUF support for screencasting is determined by the available
formats and modifiers case-by-case now.
It also effectively enables DMA-BUF screencasting on NVIDIA hardware as
well since GBM buffer objects with linear modifiers are no longer used
by default to create a renderbuffer object for screencasting.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3537>
meta_render_device_gbm_allocate_dma_buf() function is updated to take a
list of modifiers. If no modifiers are specified, the modifier is
selected by the allocator, and implicit modifiers are used to import the
created DMA-BUF.
Co-authored-by: Jonas Ådahl <jadahl@gmail.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3537>
Currently, we blindly apply the transformation matrices of all parent
actors when calculating the absolute coordinates. This means if this
function is called while the window actor containing the surface is in
the middle of a transition (e.g. window open animation), it may return
incorrect values. As this function is used for calculating pointer
confinement bounds for a specific surface, this will result in incorrect
bounds value being used if pointer constraints are applied by the
application at the same time the window is created and the mouse is
inside the surface's bounds when it's created.
Fix this by only applying transformation matrices up to the window actor
of the surface and then calculating the absolute coordinates by adding
the position of the window actor.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3585>
They are float in libdisplay-info and our own EDID parsing also returns
a float but when then converted both to an integer. Especially the min
luminance can be <1.
We also don't need a variable for indicating presence of a CTA Static
Metadata block. The values are all zero if it is absent.
Found by Dor Askayo.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3574>
Some features depend on libdisplay-info, and making it a feature
option should increase its visibility and adoption.
This makes it required when building with "-Dauto_features=enabled",
unless explicitly disabled with "-Dlibdisplay_info=disabled".
If "-Dauto_features=enabled" is not set, everything remains the
same.
In the future, the libdisplay_info option can be made "enabled" by
default so that it would always be required unless explicitly
disabled.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3582>
The function was used only once so just move it content where it is
called. It allows us to drop more cairo paths from the API surface even
if it is not part of a public api
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3581>
If we don't have a monitor for a surface - e.g. because the surface is
not mapped yet - return the highest scale of all outputs. This makes us
send a preferred scale before a client draws its first frame. The highest
scale is always correct in single monitor cases and arguably a good
option otherwise as scaling down usually looks better than scaling up.
Note that this is currently only used by the fractional scale protocol,
but will also be used for the core `send_preferred_scale()` once we
implement it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3217>
With all early "goto out" paths bypassing wayland, we can pretty
much avoid the goto and use early returns in this function. This
will hopefully improve readability.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
We've so far returned FALSE (i.e. PROPAGATE) here, somehow
oblivious of the fact that the core event handler would stop
all non-gesture events directed to windows.
Incorporate this knowledge there, in order to be able to
streamline this piece of event handling in core/ code.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
This is Wayland specific code, handle it directly in MetaWaylandPointer.
This also fixes issues with the crossing event itself managing to reach
the window occluded by modals.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
With Wayland popups and drag-and-drop using grabs, we
should let window cursors prevail when there is one
in effect.
Also, resort always to the actor as known by the
stage. This fixes the cursor lookup right after crossing
events induced by grabs, e.g. right clicking on the
gtk4-demo textview without motion would keep the I-beam
cursor, now results on the right actor/cursor for the
menu being picked.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
Enable grabbing input for popups, and drag-and-drop. Since the very
switch to using ClutterGrab underneath Wayland grabs will challenge
assumptions in existing code, these had to change in one go. A notable
one is that meta_display_windows_are_interactable() is not 100% true
anymore for xdg_popups, at least not the same.
Another change happening in lockstep is MetaDnD no longer having
to funnel events to Wayland, since the grab triggered by Wayland DnD
is now a cause of "compositor grabs", and will naturally receive events
as long as it hold. while "modal".
A number of ad-hoc checks for grabbing state has also been dropped
from src/wayland/ internals, since again Wayland grabs are a reason
for Clutter grabs, plus the mechanism itself will already take care
of focus loss and restoration.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
Add the mechanism to integrate MetaWaylandEventInterface with grabs,
callers may now specify whether a grab is required, in which case
one is created, shared by all the event interface stack.
ClutterStage grab state is also tracked, so the MetaWaylandEventInterface
in charge will focus or unfocus depending on whether input should be
handled (if ungrabbed, or grabbed by the MetaWaylandInput itself), or
not (if grabbed by something else).
At the moment nothing uses this mechanism yet, later commits will add
the first users.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
Users of Clutter grabs may listen for notify::revoked changes in
order to know that their grab is no longer in charge of event
propagation, without the use of crossing events.
Since a ClutterGrab may stay in the stack and regain effects,
this notification also happens the other way around.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
Now that the backend handles 0-size regions naturally and MetaWaylandPointer
avoids sending wl_pointer.motion on unchanged coordinates, we can use the
default motion handler for the locked pointer constraint.
And since that is the only difference with the pointer constraint event
interface, we can unify them both into a single MetaWaylandEventInterface
handling focus for them both.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
The small catch is that MtkRegion (and pixman regions) "optimize away"
0-size rectangles, so a 0-sized region will always be seen as having
a 0,0 origin. We don't want that, so transfer the origin separately from
the region.
While at it, make the Wayland pointer lock use one such 0-size region,
to avoid the 1x1px wiggle room that it currently has (accounting for subpixel
motion).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
While every kind of input is seen as coming from the Virtual
Core Pointer in the X11 case, we can largely abstract away from
that fact, and lock XDnD pointer input to the most plausible
source (e.g. a device with a pressed button), instead of only
working with pointer input.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
This collection of event handlers is the most special of them all, as
they want to unset any pointer/touch/stylus/keyboard/pad/etc focus,
and handle events from a selected device/sequence combination through
the MetaWaylandDragDest interfaces.
The same interfaces also replace the MetaWaylandKeyboardGrabInterface
in effect that handled DnD action changes.
On the XDnD special grab side, we mainly need to let the current
client (i.e. the drag source) keep receiving input events, as they
drive the DnD operation from the X11 realm.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
This is again a grab interface that mostly wants to meddle with focus,
logically setting a NULL surface if the surface client does not match
the popup client.
Since popups are meant to naturally work with any input device, the
code has been refactored to not involve the MetaWaylandPointer directly
in MetaWaylandPopup creation or getting the top popup surface (memory
management was shuffled), or compressing multiple grabbing xdg_popups
together (the existing grab maintains a single MetaWaylandEventHandler
for all).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
Besides the pointer locking/constraining mechanism, these grab interfaces
were more of a focus tracking mechanism, revoking the constraints when
the conditions didn't meet.
This can be handled pretty similarly to keyboard grabs with the new
interface, with the added bonus that we can chain up to let the
parent/default handler handle the events themselves, without poking at
MetaWaylandPointer API.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
This is implemented at the MetaWaylandSeat level, and it governs
focus and event delivery for all devices, falling through each
of the MetaWaylandPointer/MetaWaylandKeyboard/etc components.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
MetaWaylandInput is an object that will become in charge of handling
input events on their way to the Wayland socket. It keeps a stack
of event handlers, and propagates events and changes across them in
order to have them emit Wayland events, or change focus.
Each of these event handlers has a MetaWaylandEventInterface, this
is a vtable meant to replace MetaWaylandPointerGrabInterface and
MetaWaylandKeyboardGrabInterface in an unified manner, with the
following methods:
- get_focus_surface: to return the focus surface for a device/sequence.
Since several handlers will want to delegate logic on previous
handlers, it is optional to chain up with
meta_wayland_event_handler_chain_up_get_focus_surface().
- focus: To trigger a focus change for a device/sequence, since
event handlers are daisy chained by default, it is mandatory to
chain up with meta_wayland_event_handler_chain_up_focus(), either
with the given surface, or passing NULL to let later handlers
unset their state.
- press/motion/release: Unified handlers for pointer/touch/stylus
input, they chain up like event handlers do.
- key: Key event handler, propagates like event handlers do.
- other: Fallthrough for other events (pad, scroll, ...), propagates
like event handlers do.
Since there is a variety of expected behaviors, and the possibility
of stacking for some of the existing Wayland "grabs", this provides
the mechanism for that to happen.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
That would be the surface under the tool that is being currently used
on the device. This will be used by MetaWaylandSeat to implement the
default MetaWaylandEventInterface.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
The MetaWaylandPointer used to put this together through
MetaCursorTracker cursor visibility, and ClutterSeat-level
inhibition API, applying the pointer focus changes due to
visibility logically to Wayland clients.
In order to make this work over all Clutter widgetry
instead of just Wayland clients, make the ClutterSeat-level
inhibition API control this feature at the ClutterStage picking
level, and leave/enter the seat pointer as appropriate.
By default, the seat pointer has (un)focus inhibited. The
MetaCursorTracker has been made another player in unfocus
inhibition, simply asking for the pointer to get its focus
while the cursor is visible.
This in practice means that picking code may return a NULL
actor, some asserts and preconditions had to be changed to
handle this, plus some test code slightly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
Do not jump at MetaWaylandSeat across the MetaWaylandTabletSeat to
poke at the tablet tools, and chain up the checks through a
MetaWaylandTabletSeat method instead.
While at it, use g_autoptr to manage the tool list, and fix a leak.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3420>
After negotiation of DMABUF transport mutter will silently allocate SHM
buffers if the allocation in the add_buffer callback fails. It's cleaner
to renegotiate the supported formats without announcing DMABUF
capabilities in this case.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2557>
To fixate the format or renegotiate after a DMABUF allocation failed we
need to rebuild the EnumFormat params.
The function meta_screen_cast_query_modifiers will return false if no
modifiers are supported, thouse we can drop the check and remove the
macro guard.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2557>
This function contains a stub, which returns support for implicit
modifiers, if modifiers are supported preserving the current
capabilities. The stub has to be replaced with a query to the cogl
renderer to support explicit modifiers.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2557>
Bleeding edge glib was required at some point last cycle, but
right now the last stable release is good enough.
Relying on the packaged version also avoids the need for an
updated gjs, as glib now provides a newer API version of
GIRepository.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3575>
This check was originally added because `window` was actually used.
While technically correct, there's no reason to keep it around.
Fixes: 4736f873f2 ("compositor/native: Add support for direct scanout per view")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3177>
Until now we only supported direct scanout to the primary plane if the
buffer size perfectly matched the display size.
Since display controllers usually support scaling and cropping buffers
highly efficiently, try to let them do the job. This is usually helpful
if wp_viewporter is used by the client or Mutter uses fractional
scaling.
This has several advantages:
- Games (e.g. SDL2 based ones) can almost always hit direct scanout
paths in fullscreen mode. Notably when fractional scaling is used or
the game renders in a non-native resolution (or both).
- Video players using YUV buffer formats and wp_viewporter can easily
hit direct scanout paths, making displaying video very power
efficient as the 3D engine is not used at all.
Note that this still only uses the primary plane, no overlay or underlay
planes, making this change comparatively low risk.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3177>
In a following commit we will start supporting scaled and croped
surfaces, thus, in preparation, update the logic to three common cases:
1. only one surface, fullscreen (most apps)
2. a content surface and a black background surface which the client
does not want to unmap, fullscreen
3. top-level subsurface covers the whole window and is opaque (Firefox)
The remaining currently supported cases should be fairly uncommen and
and harder to compute.
Note that we already check that the window cover the stage view in
MetaCompositorView.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3177>
This allows us to pass on the related data from CoglScanouts.
If dst_rect does not match the mode, we assume that not covered areas
are opaque black - usually black bars around a centered surface.
While such driver behaviour does not appear to be documented (well) yet,
it seems to be followed by all known existing drivers and is used in a
similar way in ChromeOS.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3177>
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>
The new CLUTTER_FRAME_CLOCK_STATE_SCHEDULED_NOW state is almost
identical to CLUTTER_FRAME_CLOCK_STATE_SCHEDULED, with one important
difference being that it avoids updates from being repeatedly
rescheduled "now" when multiple calls to
clutter_frame_clock_schedule_update_now() are done before the source
is actually dispatched.
Such repeated calls to schedule an update "now" may actually postpone
the dispatch if the CPU is very busy and the source dispatch is
delayed, defeating the purpose of scheduling a frame "now".
It also allows rescheduling "now" when the frame clock is uninhibited
after being inhibited while an update was scheduled "now". This may
be important in cases where the frame clock is inhibited for very
short periods in which it would otherwise lose the state of being
scheduled "now".
Scenarios such as this would become more common with the introduction
of variable refresh rate since it makes scheduling "now" a commonplace
occurrence.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3561>
Currently, the paint-volumes/redraws debug flags displays the actor
debug
names on top of the paint volume making it very unusable. Especially
that you can easily get the relevant actor from looking glass.
The motivation is to reduce the usage of pango (through the text node)
in order to possibly move all the fonts bits to gnome shell
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3571>
This removes an incorrect implicit assumption in
calculate_next_update_time_us() that a frame may only be scheduled
once in the duration of a refresh cycle. It accomplishes this by
setting last_next_presentation_time_us on presentation feedback
instead of calculating it every time an update is scheduled.
Specifically, it corrects the intended scheduling logic in scenarios
like the following, when all of the below occur in the context of a
single refresh cycle:
1. Frame update (1) is scheduled normally, and
"is_next_presentation_time_valid" is set to TRUE
2. Frame update (1) is dispatched but ends up being "empty" (no
presentation necessary)
3. Frame update (2) is scheduled "now" and
"is_next_presentation_time_valid" is set to FALSE
4. Frame update (2) is dispatched but ends up being "empty" (no
presentation necessary)
5. Frame update (3) is scheduled normally, and since
"is_next_presentation_time_valid" is set to FALSE, the
"early presented event" logic is unintentionally skipped in
calculate_next_update_time_us().
6. Frame update (3) is dispatched and ends up being a "non-empty"
update, but its update time was calculated incorrectly because
some logic was skipped.
Scenarios such as this would become more common with the introduction
of variable refresh rate since it makes scheduling "now" a commonplace
occurrence.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3560>
Except meta_window_x11_get_group, which is still used by GNOME Shell
and we can't make it a private API for now.
Will need further investigation and could be done as a future
step
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3555>
The macro used to call into a bunch of other macros so let us turn it
into a single function.
This would simplify things for the next commit that puts the MetaGroup
usage behind a X11 ifdef
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3555>
If such a failure is followed by a successful frame then the Cogl frame
queue would have size 2, leading to an assertion failure in
`meta_onscreen_native_notify_frame_complete`:
```
g_assert (!cogl_onscreen_peek_head_frame_info (onscreen));
```
Notifying on the failure however keeps the Cogl frame queue limited to
a size of 1 and we recover gracefully with only a missed frame and a
warning message.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3278
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3565>
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>
I haven't seen this cause a problem but it looks like the worst case is
that it would have put the wrong refresh rate value in CoglFrameInfo
for multi-monitor systems of differing frequencies. But even that seems
unlikely given `_cogl_xlib_renderer_output_for_rectangle` chooses the
output with the greatest overlap of the partially-correct rectangle.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3569>
In some circumstances, we may end up with outputs with the same
vendor/product/serial, in which case we have a hard time finding the
right one to map tablets to, since configuration only has these 3
pieces of data.
Add the handling of a 4th argument containing the output name
based on the connector (e.g. HDMI-1), so that it can be used to
disambiguate the output if necessary.
This only kicks in if there actually are multiple outputs with the
same EDID data. A goal of the configuration as it was stored was to
remain useful if the user changed how the device is physically
connected to the computer, this remains true for the vast majority
of users having a single thing of each.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3556>
Our hashtable stores tools by the serial but our stylus tool and eraser
tool share the same serial - they only differ by the tool type.
This results in only one tool being created and this tool re-used for
the other type tool. Fun side-effects of this are that the stylus ends
up using the eraser pressure curve (or vice versa).
Hack around this by bit-flipping the serial for the eraser to
make it distinct - this is the only place we need to wrorry .
Closes#1884
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3542>
For barrier validation, check_barrier() would start from the
(presumably) left-most monitor and walk the neighbor monitors to the
right.
This is assuming that there is always a monitor at (0.0), which is not
necessarily the case. If the first monitor on the left is not aligned at
the top, there is no logical monitor at (0.0) causing a NULL pointer
derefence.
Instead of starting from the monitor at (0,0), start from the primary
logical monitor, as there is necessarily one.
Fixes: 85885c6 - Check barriers don't extend into nonexisting monitors
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3272
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3562>
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>
For secondary GPU rendering contexts we currently might choose an EGL
config with a format which is not supported on all primary planes. The
renderer is created when a GPU is detected and lighting up outputs and
thus assigning CRTC and primary planes can happen at any point after
that. This means we have to make sure that all possible plane
assignments will work with the rendering context when we create it.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3235
Fixes: cc7bca073 ("crtc/kms: Dynamically assign primary and cursor planes")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3512>
It can be used to force a specific RGB range. Some monitors don't follow
the specification and expect a signal different from what we send. This
property allows to force a mode which hopefully then works correctly for
the sink.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3535>
A ring will naturally go from 355 degrees to 5 degrees (or vice versa),
giving us the illusion of a direction change. Avoid this by assuming
that any change larger than 180 degrees is actually the equivalent
smaller change in the other direction.
Closes#1885
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3545>
BTN_STYLUS is the lower one and traditionally (read: in X) maps to
middle button (2), BTN_STYLUS2 is the upper one and traditionally maps
to right button (3).
This is also what GTK does and our desktop actions too map MIDDLE to
BTN_STYLUS and RIGHT to BTN_STYLUS2.
See also gtk!6168
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3541>
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>
adwaita-icon-theme updated its cursor metaphors and changed all DnD
cursors to use arrows instead of hands, except for the grab related
ones. Mutter was using "grabbing" as default DnD cursor, which now
does not match the other DnD cursors ("copy" and "no-drop") anymore.
Change this to the "default" cursor.
Additionally, because the "no-drop" cursor now puts a stronger emphasis
on the crossed out symbol also prefer "default" for
META_CURSOR_DND_IN_DRAG and only use "no-drop" for things that
explicitly don't accept a drop.
Related: https://gitlab.gnome.org/GNOME/adwaita-icon-theme/-/merge_requests/63
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3532>
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>
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>
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>
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>
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>
In various public APIs, Clutter used to return a PangoDirection
while we have a text direction enum defined in Clutter.
This allows us to drop pango dependency from meta making it specific
to cogl-pango & clutter
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3531>
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>
Set the deadline timer state as "inhibited" in case a permission error
is returned while attempting to arm the deadline timer. This makes each
device enable its deadline timer again after a VT switch.
Also print a note in this case instead of a warning as such errors are
expected during a VT switch and should not raise concerns.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3259
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3534>
For now, this function only enables the deadline timer in case it was
inhibited. This would result in an attempt to use the deadline timer
again after a device is resumed.
If the conditions that resulted in the timer becoming inhibited
remain, it is expected to return to this state after the next frame
and before being armed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3534>
The "disabled" state indicates that the deadline timer is disabled
for the lifetime of the device, while the "inhibited" state indicates
that it is disabled temporarily for the device.
This distinction is needed to handle each state differently in a
following commit. For now, only "disabled" is used.
No change in behavior.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3534>
This allows GNOME Shell to communicate the user desired XKB model
to the compositor instead of sticking with the pc105 default.
Particularly useful for those with a custom keyboard layout/irregular
keyboards.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2760>
Transient dialogs are meant to be placed centered over their
parent. However as we don't use the DIALOG window type on
wayland, this currently only works for modal dialogs.
To fix this, also apply the policy to NORMAL windows for
wayland clients.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3533>
Some panels only support fixed resolutions and fixed refresh rate with reduced blanking:
Established Timings I & II: none
Standard Timings: none
Detailed Timing Descriptors:
DTD 1: 2560x1600 120.001823 Hz 8:5 203.283 kHz 552.930000 MHz (345 mm x 215 mm)
Hfront 48 Hsync 32 Hback 80 Hpol P
Vfront 3 Vsync 6 Vback 85 Vpol N
DTD 2: 2560x1600 48.000295 Hz 8:5 81.312 kHz 221.170000 MHz (345 mm x 215 mm)
Hfront 48 Hsync 32 Hback 80 Hpol P
Vfront 3 Vsync 6 Vback 85 Vpol N
...
Minimum Pixel Clock: 552922 kHz
Maximum Pixel Clock: 552922 kHz
When using mirror mode, resolutions like 2560x1440 120Hz can be too high
to meet the pixelclock limitation, so 2560x1440 90Hz is selected
instead. However, the panel only supports 120Hz so using 90Hz result to
failed mode set.
So add reduced blanking to fallback mode, so correct refresh rate can be
used.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3449>
Since StDrawingArea in gnome-shell is the only user of ClutterCanvas,
it is possible to move ClutterCanvas completely out of Mutter to
gnome-shell. This allows to remove another Cairo dependency from
Mutter.
This patch removes ClutterCanvas code from Mutter.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3470>
COGL_DEBUG_ENABLED is a macro to check if a debug flag is set.
COGL_ENABLE_DEBUG is set by the build system if it's a debug build. The
check `#ifdef COGL_DEBUG_ENABLED` always evaluates to true. Use the
appropriate macro to guard some debugging code.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3446>
Since StDrawingArea in gnome-shell is the only user of ClutterCanvas,
it is possible to move ClutterCanvas completely out of Mutter to
gnome-shell. This allows to remove another Cairo dependency from
Mutter.
This patch makes clutter_actor_create_texture_paint_node() function
public to be used by StDrawingArea in gnome-shell that replaces
ClutterCanvas.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3507>
To paraphrase jadahl: we have a dedicated KMS thread now, which also
has realtime scheduling enabled unconditionally. realtime scheduling
on the main thread isn't too great of an idea, considering GC can
take a hot minute.
And to quote rmader: we most likely won't be able to make the main
thread rt as long as we use GJS and thus have GC.
So let's get rid of it! It's just been breaking things anyways.
This just ignores the setting; we'll fully remove it when GNOME 46
comes around.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3296>
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>
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>
On one hand this avoids crashes early after startup if the very first
pointer event is a scroll event, since the stage did not pick an actor
for the pointer device yet.
On the other hand, scroll events have some likelihood to change the
actor under the pointer even though it doesn't move. We still want to
cross towards the new actor under the pointer ASAP, without waiting
for later events.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3112
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3517>
If we happen to be changing focus to a window *while* taking focus
away from Clutter widgetry, we would unintendedly trigger reentrance
in a way that the old focused window remained in focus, by asking
to focus the default focus window in an untimely manner.
To handle this reentrancy, delay dropping the Clutter key focus
until the window focus changed, so that the focus change will look
up the default focused window in the workspace, and find the up to
date one.
Fixes: ae102ee301 ("x11: Refactor ClutterStage key focus management")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3467>
Trying to get the xwindow of a wayland only window would fail when
casting to a x11 window. Which happens as
meta_x11_display_set_input_focus is called whenever the focused
window changes, whether it is a wayland or x11 one
Fixes: bc9cd123e ("window: Move xwindow to WindowX11")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3506>
Eventually we want to have all the high-level code documentation in the
component API reference documentation. However, gi-docgen is currently
missing support for mermaid so we just keep the files in `doc/` and link
to them from `code-overview.md`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3505>
For tablet device, the tool was created when the "Wacom Serial IDs" prop
changed values. This property does not exist on the xf86-input-libinput
driver but v1.5.0 of that driver has a different property for the serial.
The serial is constant (the driver creates one X device per serial), so
we can fetch it after device creation and set it then. For earlier
versions of the driver we assign the random serial 0xffffffaa - good
enough to have at least a tool.
This fixes the crash in #3120 - clutter_event_motion_new()
overrides event->device to the tool's device (if any). Without a tool
motion events use the Virtual Core Pointer instead and our source device
is never added to the stage's priv->pointer_devices.
When we generate an crossing event (which uses the source device) we
fall afoul of an assert in clutter_stage_update_device() that expects
our source device to be in priv->pointer_devices.
Fixes https://gitlab.gnome.org/GNOME/mutter/-/issues/3120
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3383>
The slightly different styles in the different build files make it
harder to reason about or share c_args.
This notably ensures we never set any extra c_args for plain builds and
fixes the cc.get_supported_arguments() check in Cogl, Clutter and Mtk.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3333>
Simply reinterpreting the bytes differently is a strict-aliasing
violation if the type of the object isn't char or the target type of the
reinterpretation. None of that is the case here, so we have to resort to
a memcpy.
Fixes: 60c082caa ("cogl/bitmap-conversion: Support packing fp16 formats")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3499>
DRM's kms atomic code was updated to include an API to set the mouse
cursor hotspot. This has historically been missing in the atomic kms
which meant that the virtualized drivers which require mouse cursor
hotspot info to properly render had to be put on a deny list and
had to fallback to the legacy DRM kms code.
Implement the new hotspot API by checking whether the device supports
hotspot properties and if it does set them on the cursor plane. This
enables atomic kms on all virtualized drivers for kernels where
mouse cursor hotspots are in drm core.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3337>
Add META_KMS_PLANE_PROP_HOTSPOT_[X,Y] properties
to the MetaKmsPlaneProp enumeration, and
properly initialise them.
Also, add a convenience method in meta-kms-plane
(i.e., `meta_kms_plane_supports_cursor_hotspot`)
to check whether a plane supports hotspot
property setting.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3337>
Based on the pressure curve control points sample a bezier curve and
then look up the pressure at that point of the curve.
We sample 256 points and do linear interpolation in between, this
strikes a balance between having to calculate the point for all
8K pressure points a modern pen supports while still giving us
reasonable detailed curves.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3158
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3399>
This keeps the existing ClutterBezier implementation but changes
the visible API to match the needs of the tablet tool pressure curve:
a bezier defined within a [0.0/0.0, 1.0/1.0] box,(sampled
into a set of x->y mappings for each possible pressure input x, and
a lookup function to get those values out of the curve.
This patch moves the internally-only functions to be statics and changes
meta_bezier_init() to take only the second and third control point, as
normalized doubles. Because internally we still work with integers, the
bezier curve now has an integer "precision" that defines how many points
between 0.0 and 1.0 we can sample.
The meta_bezier_rasterize() function calculates the x->y mapping for
each point on the bezier curve given the initial scale of the curve.
That value is then available to the caller via meta_bezier_lookup().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3399>
The ClutterBezier code was removed in
580d62b9b clutter: Remove unused Path related types
because it wasn't used anywhere. We do need a bezier curve for the
tablet tool pressure curve calculation though so let's move it
to the native backend and rename it to MetaBezier in the process.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3399>
Let's try to consolidate our documentation in doc/ in the repo. This
includes some documentation from README.md, the HACKING.md coding style
and the gitlab wiki.
The README.md file now links to all top-level topics (i.e. not reachable
via other topics).
This also includes a few small changes to make things more consistent.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3465>
This addresses the following race condition:
1. Window+MetaFrame are created non-fullscreen, _MUTTER_FRAME_EXTENTS
is initialized through widget measuring, accounting for frame.
2. Window and MetaFrame become fullscreen.
3. MetaFrameContent gets first size allocation, already fullscreen.
4. Borders were initialized to 0,0,0,0, become set to 0,0,0,0 correctly
reflecting fullscreen, however notify::borders is not emitted.
5. _MUTTER_FRAME_EXTENTS stays accounting for the frame extents.
It sounds sensible to have the borders initialized to a meaningful value,
so account for the first time the border would be set due to the content
being (re)sized, and let this first value trigger notify::borders resulting
in _MUTTER_FRAME_EXTENTS updates.
Since all later _MUTTER_FRAME_EXTENTS changes happen through content
resizes, we only have to cater for this initial handover between the
frame/content initialization paths done through widget measuring and
the later paths done through MetaFrameContent resizes.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2937
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3476>
The test makes sure the YCbCr formats create the expected image and we
don't accidentally break it.
Like all wayland tests, this is now part of mutter/wayland, mutter/tty,
and mutter/kvm and will use either shm or dma-buf depending on which
suite is chosen.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3371>
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>
The assumption is that all planes are always contiguous, and we don't
have any multi-plane formats where the first plane is subsampled.
The stride of the entire buffer is then just the stride of the first
plane and the stride of the other planes is derived from that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3371>
Instead of forcing every user of WaylandBuffer to create a listener and
destroy the wl_resource and the WaylandBuffer object, provide a default
listener which does it for the user.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3371>
The texture coordinates of all planes should be the same in theory so
using the coordinates of the first plane works.
The reason for this change is that Cogl somehow doesn't manage to get us
the correct coordinates for the 3rd plane in some circumstances. This is
really a workaround but not wrong in any way.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3176
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3469>
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>
Calculating the mipmap width as half of the texture width leads to a
mipmap width of zero for textures with width of 1 which leads to an
early exit instead of a mipmap texture.
Fixes: 16fa2100d ("shaped-texture: Stop using MetaTextureTower and use GL mipmapping instead")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3468>
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>
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>
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>
Integrate it into the code, instead of depending on MetaDisplay
notify::focus-window for it. Now, instead of focusing explicitly the
stage window, we focus a NULL window, and let the MetaX11Display
determine whether focus should go to the stage window if there's
a focused actor, or the no_focus_window if nothing has focus.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3269>
We currently offer the mechanism for GNOME Shell to implement, and
while this is not exercised often (our entries are typically surrounded
by a ClutterGrab ensuring key events, so this is reserved to grab-less
entries, probably there are some in extensions), this is arguably
something Mutter should cover by itself without GNOME Shell guidance.
This is only necessary on the X11 backend, although it is conceptually
more tied to the MetaX11Display connection, so perform the focus
tracking there only if not running as a Wayland compositor (i.e. --x11).
This avoids the only case where the low-level
meta_x11_display_set_input_focus_xwindow() function is used, or rather
makes it completely a MetaX11Display implementation detail, leaving
only the MetaDisplay API as the high-level entry points to handle
window key focus.
The public API that allowed GNOME Shell to implement these mechanisms
is also gone in this commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3269>
There might not be a single plane that is "for" a CRTC, so remove the
API that made it appear as if it did. The existing users only cared if
there was some plane for said CRTC, so replace the getters with API that
just checks the existance at all.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3428>
When there are a set of primary planes, and a set of CRTCs, where each
plane can be used on multiple CRTCs, we need to make sure that when we
mode set and page flip, each CRTC gets assigned an individual plane that
isn't used with any other CRTC.
Do this during the monitor resource assignments that sets up later to be
applied configurations of the mode setting objects, but with the new
hooks into the backend, so that we don't need to teach the monitor
config manager about planes.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2398
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3428>
This means that it doesn't necessarily mean what transform / rotation
the hardware resource gets, e.g. it instead represents the logical
transform related to the configured mode. This allows us to postpone
checking the plane capabilities until later (as rotation capabilities
depends is a plane property), when a plane has been assigned.
This was in practice already handled when configuring the
transform-via-offscreen case, handled when creating the view, and the
mode setting configuration.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3428>
When we're configuring monitors, allow backends to add backend specific
assignments during resource assignment (mapping connectors and CRTCs
etc).
This will later allow the native backend's KMS monitor resources to
assign a primary plane and optionally a cursor plane during
configuration. This will then dictate what plane will be used for
primary plane updates, as well as cursor updates, until reconfigured
again.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3428>
A set of primary planes can be usable with a set of CRTCs, meaning we
can't have general purpose functions that gets a plane for a CRTC, as
there is no such one to one relationship.
For tests we still want to have helpers that makes writing tests easier,
so to prepare for those functions going away, make the tests do the
equivalent themselves.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3428>
clutter_actor_get_transformed_position() would write the uninitialized
values of v2 when clutter_actor_apply_transform_to_point() fails in
_clutter_actor_fully_transform_vertices() because the actor has not been
added to the stage yet.
When called from JS this would overwrite the zero initialized values
passed in from gjs. If the uninitialized values now happen to correspond
to one of the NaN float values used by mozjs to represent a pointer
type, this would lead to seemingly random crashes in mozjs code later
on.
Avoid this by using _clutter_actor_fully_transform_vertices() directly,
which allows us to check if it failed.
Related: https://gitlab.gnome.org/GNOME/gnome-shell-extensions/-/issues/469
Related: https://gitlab.gnome.org/GNOME/gjs/-/issues/591
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3453>
Musl doesn't seem to include this by default so explicitly including it
should fix compilation on Musl.
Tested with Clang 16/17 and GCC 14.
Error:
src/backends/meta-fd-source.c:70:3: error: call to undeclared function 'close'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
close (fd_source->poll_fd.fd);
^
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3078
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3455>
Otherwise a tablet in relative mode will never have a tool set and
nothing happens on motion events - meta_wayland_tablet_seat_update()
simply exits early for tablet proximity, button or motion events.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3410>
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>
Changing the MultiTexture may require a different set of pipelines when
the texture format is different. We keep track of the attached
MultiTextureFormat just like we do for the width and height.
This fixes misrendering when a client attaches buffers with different
MultiTextureFormats to the same surface.
Fixes: 3dd9f15eb ("shaped-texture: Start using MetaMultiTexture")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3456>
Although they're in the same units, `radius` is easier to understand than
`sigma` and makes the public API independent of the blur algorithm used
behind the scenes. We now only keep the `sigma` terminology where the
implementation is Gaussian-specific.
The assumption that `sigma = radius / 2.0` is actually not new here. We
just move it from `_st_create_shadow_pipeline` (gnome-shell!1905) into
`clutter_blur_new`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1908>
The commit also moves certain functions from the private header
as they are not used anywhere else and removes COGL_EXPORT
when the function is supposed to be private and is not used outside of
cogl
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3437>
fatal: detected dubious ownership in repository at '/builds/whot/mutter'
To add an exception for this directory, call:
git config --global --add safe.directory /builds/whot/mutter
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3434>
We want to run those tests in VKMS later with the same reference image.
To make the tests as close to each other as possible we use the same
resolution for the VKMS and the virtual output which is 640x480.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3350>
This test currently only works because the monitor has the same width
and height. Generalize it to arbitrary monitor sizes by taking into
account that width and hight are swapped for some rotations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3350>
Most tests use the draw_surface function to draw a solid color to a
surface. This moves it from the shm-only path to WaylandBuffer which
makes all of those tests usable via dma-buf.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3350>
The custom drawing requires adjusting the test. Instead of poking at
memory directly, we can just draw a color at certain coordinates which
makes it independent of the pixel format used.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3350>
It abstracts away the kind of buffer so clients can be tested with both
shm and dma-buf paths. We'll make use of it in the future by adding the
wayland tests to the TTY and KVM test suits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3350>
When capturing the view we have to make sure the stage is actually
updated. In direct scanout mode the stage is unmodified and we can't
find the content we want to test.
Currently the ref-tests are all running on non-native setups where
direct scanout is impossible but we will change that soon!
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3350>
Some actors have a well-defined layout manager other than FixedLayout.
If they do, we can handle the layout manager creation at the
ClutterActor instantiation, like GTK does for widget layout managers.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3445>
We will also require GL_OES_texture_half_float and GLES 3.0 to enable
COGL_FEATURE_ID_TEXTURE_HALF_FLOAT. This gives us float types and makes
it possible to read pixels from framebuffers with internal floating
point formats (into float, half is never supported).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3441>
This implementation is copied from mesa.
It uses x86_64 assembler on CPUs where it's supported, otherwise falls
back on a software implementation (cogl-soft-float.c). It's intended to
be used for packing and unpacking f16 format bitmaps.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3441>
The implementation comes from mesa, which came from Berkeley SoftFloat
3e Library.
The software float implementation will be used to convert from and to
half floating point pixel format bitmaps, when no the CPU isn't capable
of the conversion.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3441>
The type of the intermediate medium for storing pixel channels is
changed from "is 8 or 16 bit" to an enum, and switch cases. This doesn't
add support for anything, but will make adding a "float" medium type
less intrusive.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3441>
There is no internal fp16 format which has no alpha which means we would
get garbage alpha when reading the framebuffer directly. We have to use
the packing/unpacking to always get the alpha of 1.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3441>
Which gltype and glformat are allowed for a given gl framebuffer depends
on the internal gl framebuffer format. Either the format we want to read
into matches the gltype, glformat and doesn't need CPU packing from
another format or we have to use a tmp buffer with a format that matches
the GL requirements.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3441>
Sometimes it's useful to combine the description from multiple places.
For example, a subsequent commit will add the output name as the first
thing to the frame clock span descriptions, leaving the rest of the
description with complex checks unchanged.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3417>
Tracy can filter its statistics by user text, in our case by span
description. It's useful to filter by actor type and name, and not so
much by the pointer. So, remove it, and also reduce the amount of
punctuation.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3417>
Transfer none was achieved using a stack GArray in the stage which
would get resized to 0 at the end of every frame to "free" it.
In the case of direct scanout however, painting the next frame only
happens after leaving fullscreen again. Until then the array just kept
growing and because GArrays don't reallocate when shrunk, this memory
remained allocated even after leaving fullscreen.
There is no cache benefit from storing paint volumes this way, because
nothing accesses them after their immediate use in the calling code.
Also the reduced overhead from avoiding malloc calls seems negligible as
according to heaptrack this only makes up about 2-3% of the temporary
allocations.
Changing this to transfer full and removing the stack array simplifies
the code and fixes the "leak".
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3191
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3442>
It was for a failed expriment that tried to mmap() dmabuf memory and
find damaged regions to decrease the amount that was eventually used to
write to an onscreen, but mmap:ing is only fast enough on intel, and
it's only relevant on various server GPUs. For it to be achievable, we
need to render to system memory in a way that we don't need to copy it
out of OpenGL, but that's currently not possible.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3443>
This is a step in cleaning up the Clutter context management. By making
it a GObject it's easier to add e.g. properties and features that helps
with introspection.
For now, this means the context creation is changed to go via a
"constructor" (clutter_create_context()). This is so that the global
context singleton can be mantained outsid of ClutterContext, until it
can be removed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2016>
The original purpose of being able to report errors is no longer
relevant, since the Clutter backend is now practically a thin wrapper
around the actual backend, which has already dealt with error reporting.
Thus move this to the regular constructor path.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2016>
The intention for ClutterContext is to be more or less the MetaContext
or CoglContext equivalent. Lets rename the type so that it becomes more
consistent with the other similar types.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2016>
The current usage of MetaWindow::unmanaging may result in confused
focus window lookups while undoing the MetaWindowDrag grab (i.e.
still pointing to the window that is now being unmanaged).
The meta_window_unmanage() function itself takes care of changing
focus outside of the window being unmanaged, so postpone the
MetaWindowDrag undoing to a point after that is done.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3073
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3376>
Since meta_kms_impl_device_get_sync_file always returns the same
file descriptor referencing the same sync_file, this means the atomic
ioctl doesn't need to wait for any fences to signal. This is fine
because we already waited for the buffer to become idle before applying
the Wayland surface state.
Fixes the atomic commit ioctl spuriously synchronizing to the screen
cast paint (at least with the amdgpu driver), which could result in
the page flip missing its target scanout cycle.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3148
v2:
* Rename local variable to signaled_sync_file for consistency with new
function name
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3389>
It returns a file descriptor which references a signaled sync_file.
v2:
* Change function name and add Doxygen comment to hopefully make its
purpose a bit clearer (Ivan Molodetskikh)
v3: (Jonas Ådahl)
* Create sync_file from scratch via a syncobj, no buffer needed anymore
* Initialize priv->sync_file = 1 and use g_clear_fd in finalize
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3389>
Using sized internal formats is required to make sure we actually get
the precision that we want.
The formats should also be renderable, otherwise they can not be used as
a framebuffer attachment. For GLES we have to check for a bunch of
extensions and fall back to internal formats with more bits when they
are not available.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3429>
In GLES 2.0 the required color-renderable formats to support to not
include 8bpc RGB/RGBA formats. The extension adds this requirement and
makes GLES 2.0 actually useable. Not specifying a gl internal format in
Cogl so far has hidden the problem and let the implementation fall back
to RGB585 for example.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3429>
Also be more strict about what we consider RGBA1010102 support. Before
GLES 3.0, using ReadPixels on a framebuffers with format RGB10_A2 was
not possible with type GL_UNSIGNED_INT_2_10_10_10_REV_EXT and thus our
code to read back pixels could fail.
Users of cogl should check those feature flags before using FP16 and
RGBA1010102 pixel formats via CoglFeatureIDs:
* COGL_FEATURE_ID_TEXTURE_RGBA1010102
* COGL_FEATURE_ID_TEXTURE_HALF_FLOAT
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3429>
We might pick an actor that needs relayout. I've seen this happen inside
hiding / unmapping in particular. In this case, calculate_clear_area ()
will call clutter_actor_get_abs_allocation_vertices () which in turn
will force a relayout. However, this is not what we want, because:
1. We don't want to run layout during picking.
2. If the actor needs an allocation, then the pick stack could not have
used an up-to-date allocation, because it is not computed. Therefore
this clear area would use a potentially completely different
allocation than the one stored in the pick stack.
Thankfully, clear area seems to be used as a cache/optimization, so
let's just avoid computing it if the actor is not allocated.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3425>
The debug controller can optionally, when passing --debug-control,
enable manipulating debug state, so far enabling/disabling HDR, via
D-Bus.
It's always created, in order to have a place to store debug state and
emit signals etc when it changes, but so far, it doesn't have its own
state it tracks, it just mirrors that of the monitor manager.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3432>
While we should ideally have a sensible cursor theme, handle the
case of cursor themes that lack certain cursor names, and fallback
to the 'default' cursor in those cases.
The 'grey rectangle' fallback is still left, in case we even fail
to load a 'default' cursor.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3295>
When a stream is destroyed by a consumer, mutter won't be able to
recognize that.
For mutter, the stream just paused, but did not disconnect, because the
connection state of a PipeWire stream only represents, whether the
respective PipeWire context is connected to PipeWire.
In addition to that, it may be the case, that the stream consumer just
recreates the stream.
So even if mutter would be able to know, when the stream consumer
destroyed a stream, but not the whole screencast or remote-desktop
session, then mutter would not know, whether the stream will be resumed
eventually or not.
So, add an explicit API call to the screencast interface to stop a
stream.
For virtual streams, this also means, that the respective virtual
monitor is destroyed.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2889
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3307>
When a virtual stream is destroyed, its respective virtual monitor is
destroyed too. When the virtual monitor is destroyed, mutter reloads
the monitor manager.
However, at this point, the virtual stream is not completely destroyed
yet. The viewport of the virtual monitor still exists at this point and
when the monitor manager reloads, it will try to fetch the logical
monitor of the now destroyed virtual monitor, which will fail and thus
gnome-shell will run into a segfault.
Fix this situation by reloading the monitor manager in an idle callback.
When the monitor manager reloads, the virtual monitor is completely
gone, since the viewport of the virtual monitor is destroyed after the
virtual monitor itself.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2864
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3307>
There is no way to set any gamma luts, or do anything other color
management related. Eventually we'll probably want to, but that requires
bringing color management plumbing to PipeWire.
Doing this is also needed when running a headless session, as when
headless, polkit doesn't let us create colord devices without explicit
user permission, meaning we'll spam the session with useless dialogs
each time a session is started.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3423>
Given destruction order, the display goes away before the stage, so
this lingering signal connection may trigger unintended crashes.
Fixes: 05eeb684d1 ("window: Postpone focusing until grab ended if uninteractable")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3422>
`count_mode_setting_devices` was incorrect in both name and in function.
What it was actually doing was counting GPUs that had been registered with
the backend so far (during the `init_gpus` loop). What it was intended to
do was to count the number of `MetaRenderDeviceEglStream` instances, which
is the thing we're limited to only one of. So `count_mode_setting_devices`
would return zero whenever the first GPU initialized happened to be a
`MetaRenderDeviceEglStream`, which would in turn prevent
`MetaRenderDeviceEglStream` from successfully initializing. Seems it only
ever worked in the case of a hybrid system where the first GPU initialized
was GBM-based.
Now we count `MetaRenderDeviceEglStream` instances (zero or one) externally.
This allows initialization to succeed when it happens to be the first (or
only) GPU. And so `MUTTER_DEBUG_FORCE_EGL_STREAM=1` now works.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2905>
For scanout on a secondary GPU, for the time being try only formats
which are guaranteed to be renderable with GLES3, which notably excludes
10 bpc formats without alpha channel.
v2:
* Use separate format array for 10 bpc formats without alpha.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3139>
If the EGL_KHR_no_config_context extension is supported, use it to
choose a format per onscreen which is compatible with the scanout CRTC
and the GL rendering API used.
Suggested by Jonas Ådahl.
v2:
* Drop code which checked for GLES3 renderability. Makes no sense for
various reasons, in particular that EGLconfigs are about EGLSurfaces,
whereas secondary GPU contexts use an FBO for blitting.
* Use error parameter directly for meta_renderer_native_choose_gbm_format
call (Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3139>
Preparation for the following commits, no functional change intended.
v2:
* Pass through MetaEgl pointer
v3:
* Make it return gboolean (Robert Mader)
v4:
* Add debug logging and corresponding purpose parameter
v5:
* Fix excessive function parameter indentation (Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3139>
If the EGL_KHR_no_config_context extension is supported, pass
EGL_NO_CONFIG_KHR to eglCreateContext. This will allow binding the
context to surfaces created with different configs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3139>
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>
In profilers with a timeline or flame graph views it is a very common
scenario that a span name must be displayed in an area too short to fit
it. In this case, profilers may implement automatic shortening to show
the most important part of the span name in the available area. This
makes it easier to tell what's going on without having to zoom all the
way in.
The current trace span names in Mutter don't really follow any system
and cannot really be shortened automatically.
The Tracy profiler shortens with C++ in mind. Consider an example C++
name:
SomeNamespace::SomeClass::some_method(args)
The method name is the most important part, and the arguments with the
class name will be cut if necessary in the order of importance.
This logic makes sence for other languages too, like Rust. I can see it
being implemented in other profilers like Sysprof, since it's generally
useful.
Hence, this commit adjusts our trace names to look like C++ and arrange
the parts of the name in the respective order of importance.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3402>
Now that the monitor screencast records to DMA-BUF buffers immediately
(since bc2f1145d8), and we know which phase of the paint rountines we
are (since last commit), we have the opportunity to bring back the
blitting technique.
Bring back blitting. This time, instead of simply failing if the blit
fails, add a fallback path that does a stage paint if something goes
wrong. Unlike the previous implementation of blitting, this one only
blits the current view - it does not blit all views that intersect
with the screencasted monitor.
Embedded cursors should still be fine because hardware cursor is
inhibited while embedded cursor screencasts are running.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3406>
Track where we are in terms of the paint cycle. Do this through an
enumeration that is passed through the paint vfuncs of screencast
sources.
Right now, this information is not used by any one of the sources,
but next patch will use it to prevent blitting when detached from
the paint cycle.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3406>
NULL actor pointers seem to arise sometimes in `clutter_stage_update_device`
when using a touchscreen, but that's only fatal with `CLUTTER_DEBUG=event`.
So just handle NULL where it was crashing: `_clutter_actor_get_debug_name`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3413>
With the existing ClutterInputMode terminology (inherited from XI2),
hardware devices may be "physical" (i.e. attached to a logical device),
or "floating" (i.e. detached from all logical devices).
In the native backend, tablet devices are closer to "floating" than
"physical", since they do not emit events relative to the ClutterSeat
pointer logical device, nor drive the MetaCursorTracker sprite. This
is in contrast to X11 where all tablet devices drive the Virtual
Core Pointer by default, along with every other pointing device.
Change this mode in the Wayland backend to be more coherent. The
existing checks on the ClutterInputMode along Mutter seem appropriate
for handling these as floating devices, since they mainly care about
logical vs non-logical.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3393>
The window actor can be mapped every frame, e.g. when it is dragged in
the overview. This commit keeps track when the geometry changed and we
didn't managed to sync the geometry yet and need to sync it at a later
time.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3404>
At the end of the sync_actor_geometry function the window buffer_rect
and the WindowActor position and size are the same and consistent.
Call the virtual method at the end and let the implementations look at
either the buffer_rect or the actor position/size itself.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3404>
In some cases the window is not mapped when the geometry changes.
Without the mapped window the surfaces are not mapped either and don't
have a sensible allocation.
This patch makes sure we abort syncing the geometry if the window is not
mapped and also make sure we sync geometry when the actor eventually
does get mapped.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3152
Fixes: 8f4ab53bd ("window-actor/wayland: Ensure to use allocation for black background check")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3401>
Scoped traces are less error prone, and they can still be ended
prematurely if needed (this commit makes that work). The only case this
doesn't support is starting a trace inside a scope but ending outside,
but this is pretty unusual, plus we have anchored traces for a limited
variation of that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3396>
Allow only specific files to use those deprecated APIs making
it easier to find where deprecated APIs are still in use
and avoid introducing new usages without being noticed
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3400>
Group all the three config files from clutter/cogl/meta into one
and also remove unnused configurations and replace duplicated ones
This also fixes Cogl usage of HAS_X11/HAS_XLIB to match the expected
build options
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3368>
clutter_keyval_name() returns a pointer to a static array, not
newly allocated memory. Add a transfer annotation to indicate
that callers must not free the returned memory.
While add it, make the return value const to stress further that
callers shouldn't touch the returned memory.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3386>
Currently, nothing uses the dumped json of the paint nodes tree. So
let us drop them in a separate commit so it can easily be reverted
if someone ends up wanting to build a tool to consume and inspect
the JSON.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3354>
Currently, json-glib is used for two things:
- For loading scripts, nothing seems to use that in real life other
than some tests
- For debugging paint nodes
For now, the PR drops the first use case and only require json-glib
if it is a debug build
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3354>
We don't actually need the host to be a container, so simply work on
actors saving us a few casts.
This'll simplify dropping ClutterContainer entirely later, and
StViewport/ShellWindowPreviewLayout will also need to be updated for the
new signatures
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3384>
Sticky keys configuration changes reset the pressed modifier state mask,
even though the XKB state might already match with the expected new
state. In those cases we can avoid the XKB state mask update completely.
This also fixes a crash at initialization with sticky keys toggled on,
since configuring the device a11y settings will trigger a XKB state
mask merely reassuring the initial state with no modifiers pressed,
while the connection between the ClutterSeat and the impl object has
not been set up yet. This crash was introduced by commit 00bb4190b
("backends/native: Drop device_native->seat_impl field").
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3392>
Tests that
* the window under the cursor has focus
* focus_default_window won't move the focus away from a previously
focused window even if the cursor is somewhere else
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3258>
When the desktop is focused instead of a window we want to tell the core
about this to handle focus-mode mouse. This is handled by looking for
CLUTTER_LEAVE events where the newly focused window is NULL.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3258>
Take a reference to the window to make sure the MetaFocusData->window
pointer is not pointing to a freed object.
Also make sure that the window that we want to focus is not currently
unmanaging.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3258>
On X meta_window_handle_enter was called when the desktop window was
entered. On wayland the "desktop" is no window anymore. We still want to
inform the core that the desktop is focused, so it can unfocus windows
if focus-mode is mouse.
This commit prepares the core for handling a NULL windows to mean the
desktop.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3258>
The following commits will make it possible to pass a NULL window to
display_handle_window_enter/leave to represent the cursor entering the
desktop. This means it can't be a method of the window class anymore.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3258>
If we have an existing focused window that may have focus, default focus
will leave the focus there. An unmanaging window for example must not
have focus and default focus will continue to select another window in
this case.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3258>
The layout manager takes the generic ClutterActor expand/align
properties into account. Everyone should already use those instead
of the custom layout/child properties, so removing them should have
little fallout, while making for a nice cleanup.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3382>
It tests various sub-surface corner cases. It would fail without the
previous fixes.
v2:
* Fix draw_descendant parameter formatting
* Make toplevel window fullscreen (Robert Mader)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3364>
To make sure a new wl_subsurface created for the same wl_surface won't
inherit the position from the wl_subsurface being destroyed.
v2:
* Move into permanently_unmap_subsurface
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3364>
Instead of duplicating the code in both.
v2:
* Rename helper to make it clear(er) that it permanently unmaps the
sub-surface
* Move transaction allocation & commit into the helper
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3364>
wl_subsurface_place_above/below need to hook it up to the parent
surface's pending state, so that it gets picked up next time the parent
is committed.
v2:
* Adapt to wl_subsurface_destructor not resetting sub-surface position
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3364>
This is simpler, and makes sure it's called only once for each surface
in the transaction as committed.
v2:
* Use array for surface states with placement ops (Jonas Ådahl)
v3:
* Use GPtrArray instead of GArray (Robert Mader)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3364>
This is necessary to make sure the unmapped sub-surface is no longer
visible.
v2:
* Use META_IS_WAYLAND_SUBSURFACE (Jonas Ådahl)
* Use same sequence of assignments in both cases
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3364>
meta_display_ping_window() does nothing when check-alive-timeout is set
to 0, but meta_window_check_alive_on_event() was relying on it to reset
the events_during_ping. Without this events_during_ping was just
counting up until the threshold was reached and the window was marked as
not alive, preventing further pointer events from being sent to the
window.
Fix this by not doing anything in meta_window_check_alive_on_event() if
check-alive-timeout is 0, similar to meta_display_ping_window().
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3142
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3367>
Locked modifiers should probably not have an effect on keybindings
while toggled. this is most relevant for modifiers that can be
either/both pressed or locked (e.g. Caps Lock key), if used in
keybindings.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3369>
This struct contains the pressed/latched/locked set of modifiers applying
to the event, and may be filled in by backends generating those events.
Other places where we forward modified key events, state may be normally
obtained from the original event.
Since this constructor is used in a variety of places, this commit
updates them all in one go.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3369>
This function may be used on key events to obtain the fully
detailed pressed/latched/locked modifiers that apply when the
event is received. No events have this detailed information
yet.
This API call may be compared to the clutter_event_get_state_full()
that existed in the past, although this getter has a stronger
predilection to it applying exclusively to key events.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3369>
If multiple sync events are send in the same dispatch, a further call to
wait_for_sync_event will get stuck. Fix this by keeping track of the
latest sync event serial in the display and always compare against that.
This also means sync event sequences must start at 0 and increase by 1.
The wayland-x11 interop test is the only one where that wasn't already
the case.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3358>
Currently, we use cairo_region_t despite it being a thing wrapper around pixman_region_32
In order to push for a cairo-less and wayland only build in the future, replace
cairo_region_t with a thin wrapper that is almost a copy of the upstream cairo implementation
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3292>
Whenever a MetaWaylandTestClient exists without success the calling test
will fail. This fixes a bunch of cases where the test would get stuck
waiting for some event from the client when it already died and won't be
able to send the event.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3348>
The test and dist CI jobs run wrap the meson calls in dbus-runner to
avoid setting up dbus servers and mocking services for every test but
the dbus-runner invocation from meson test didn't actually skip all the
setup.
This nested mocking also doesn't work because the system bus is assumed
to be the host system bus and not a mocked one.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2618>
This means one can run meta-dbus-runner.py effectively mocking
everything relevant except logind itself, meaning one can run from a TTY
and get permission to mode set etc, while still mocking things like
gsd-color, colord, etc.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2618>
Make CoglBuffer an abstract class and inherit the various Cogl*Buffer types from it.
As none of the subclasses is overriding the vtable functions, they were not turned into
vfuncs but plain function pointers in CoglBuffer.
We still use _cogl_buffer_initialize until we port the various params into actual construct-only
properties, similar to the previous commit for CoglTexture.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3193>
- Make Texture a parent GObject class and move the vtable funcs as vfuncs
instead of an interface as we would like to have dispose free the TextureLoader.
- Make the various texture sub-types inherit from it.
- Make all the sub-types constructors return a CoglTexture instead of their respective
specific type. As most of the times, the used functions accept a CoglTexture,
like all the GTK widgets constructors returning GtkWidget.
- Fix up the basics of gi-docgen for all these types.
- Remove CoglPrimitiveTexture as it is useless: It is just a texture underhood.
- Remove CoglMetaTexture: for the exact same reason as above.
- Switch various memory management functions to use g_ variant instead of the cogl_ one
Note we would still want to get rid of the _cogl_texture_init which is something
for the next commit
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3193>
These were leftovers from times when GL(ES) 1.x was still supported.
As we require HAVE_COGL_GL and/or HAVE_COGL_GLES2 to be defined, all
cases for checking both are now redundant.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3338>
My motivation was at first to replace cairo_rectangle_t with graphene_rect_t
but noticed nothing uses it anywhere: Shell/Kiosk/Gala; so it is safe
to just drop and people could still use the new_to_framebuffer ctor
and handle setting up things themselves if needed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3240>
The helper doesn't do anything that makes it worth
to be exposed as public API. End-users, such as GNOME Shell could have
an in-tree helper if they end up using it that much.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3086>
The shell might raise and make windows recent for another workspace when
an app gets activated on another workspace. Making the windows only
recent on the current workspace thus results in inconsistent focus when
another window of the same app is closed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3315>
There are existing extensions that implement desktop icons as
a combination of a GTK program and a small extension to make
the wayland window behave as if it was of type DESKTOP on X11.
That's quite painful, as it requires reimplementing WM behavior
that is already implemented in mutter itself (stacking, stickiness,
skip-taskbar, ...), as well as modifying gnome-shell to consider
the window in addition to "real" DESKTOP windows (workspace-switch
animations, ctrl-alt-tab, ...).
In addition to that, other extensions may also have special handling
of DESKTOP windows, and their code cannot easily be monkey-patched
to handle "alternative" desktop icons.
This whole game of whack-a-mole can easily be avoided by allowing
desktop-icons extensions to mark their desktop windows as DESKTOP,
so do just that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3305>
Assigning the corresponding stack layer of DESKTOP windows is
currently X11 specific, because there is no way for wayland
clients to set the DESKTOP window type.
This is about to change, so move the code to the generic layer
handling.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3305>
Change the order of events to adhere to the Wayland specification for
wl_keyboard.enter, which mandates:
> The compositor must send the wl_keyboard.modifiers event after
> this event.
Mutter currently sends the modifiers event before the enter event,
which may break applications that require information about the focused
surface in order to properly handle the modifiers.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2231
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3341>
These functions ends-up calling gdk-pixbuf for loading textures/bitmaps
from a file and they don't seem to be used anywhere.
These changes are only useful with the following up commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3097>
Create a fake monitor region right of the right-most monitor and if a
horizontal barrier extends into that region, fail the barrier. Barriers
are aligned on the top/left edge of the pixel so the most natural
barrier of (e.g. 0-1024) is also wrong - it's one pixel into the next
monitor.
Check this for nonexisting screens on the right too to avoid clients
suddenly failing when multiple monitors are present.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3319>
Assuming two 1920x1080 screens next to each other: a horizontal barrier
starting at 1920 going east is always outside the left screen.
Assuming two 1920x1080 screens on top of each other: a vertical barrier
starting at 1080 going south is always outside the top screen.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3319>
When running headless, only the invalid modifiers are advertised.
That breaks with the NVIDIA proprietary driver which then rejects the
buffers created with the invalid modifier, and that kills Xwayland,
meaning that running Xwayland on top of a mutter based compositor
headless is not possible.
The reason the modifiers are not sent is because AddFb2 is not supported
when running headless.
Other compositors (weston, wlroots) would still send the modifiers even
without AddFb2, and Xwayland works fine on those compositors when
running headless.
Remove the requirement for AddFb2 to send the modifiers, so that
Xwayland can work fine on top of mutter headless with the NVIDIA
proprietary driver.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3060
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3320>
`clutter_actor_destroy()` queues a stage update. Under certain
circumstances - i.e. when run in a very slow container - this can race
with the stage update triggered by the following
`clutter_virtual_input_device_notify_button()`, occasionally resulting in
`wait_stage_updated()` to return before the
`on_event_return_propagate()` callbacks ran, making the test fail.
This notably became more common since
8f27ebf87e (clutter/frame-clock: Start next update ASAP after idle period)
landed.
Thus wait for a stage update to happen after `clutter_actor_destroy()`,
preventing the race.
Fixes: f6da583d06 (tests/clutter/event-delivery: Add tests for implicit grabbing)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3332>
For frame updates in response to sporadic user interaction, this results
in input → output latency somewhere between the minimum possible and the
minimum plus the length of one display refresh cycle (assuming the frame
update can complete within a refresh cycle).
Applying a max_render_time based deadline which corresponds to higher
than the minimum possible latency would result in higher effective
minimum latency for sporadic user interaction.
This was discovered by Ivan Molodetskikh, based on measurements
described in https://mastodon.online/@YaLTeR/110848066454900941 .
v2:
* Set min_render_time_allowed_us = 0 as well, to avoid unthrottled
frame events. (Robert Mader)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3174>
Instead of g_get_monotonic_time. This makes sure last_presentation_time_us
advances by refresh_interval_us.
Doesn't affect test results at this point, but it will with the next
commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3174>
When more than one refresh interval has passed since
last_presentation_time_us.
I honestly can't tell if the previous calculation was correct or not,
but I'm confident the new one is, and it's simpler.
v2:
* ASCII art diagram didn't make sense anymore, try to improve
(Ivan Molodetskikh)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3330>
Every `mtk_x11_error_trap_push()` must be paired
with an `mtk_x11_error_trap_pop[_with_return]()` call
otherwise all future errors will be caught and ignored
even if they shouldn't be.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3328>
Certain kernel drivers can take an unreasonably long time to
complete mode setting operations. That excessive CPU time is charged
to the process's rlimits which can lead to the process getting killed
if the thread is a real-time thread.
This commit inhibits real-time scheduling around mode setting
commits, since those commits are the ones currently presenting as
excessively slow.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3037
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3324>
At the moment if a thread is made real-time there's no going back,
it stays real-time for the duration of its life.
That's suboptimal because real-time threads are expected by RTKit to
have an rlimit on their CPU time and certain GPU drivers in the kernel
can exceed that CPU time during certain operations like DPMS off.
This commit adds two new ref counted functions:
meta_thread_{un,}inhibit_realtime_in_impl
that allow turning a thread real-time or normally scheduled. At the same
time, this commit stores the RTKit proxy as private data on the thread
so that it can be reused by the above apis.
A subsequent commit will use the new APIs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3324>
Most of the code writes "real-time" as "realtime" not "real_time".
The only exception is one function `request_real_time_scheduling`.
This commit changes that function for consistency.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3324>
If we queued a mode set, but didn't end up compositing all frames, we'll
have pending mode sets in a hash table waiting to be applied. If we
before all monitors again try to reconfigure things we should drop the
old pending mode sets and start fresh.
We already do this when we're doing so when generating views, but when
just unsetting modes, we didn't, so fix that.
Related: https://bugzilla.redhat.com/show_bug.cgi?id=2242612
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3318>
We had a function called "reset_modes()" on MetaRendererNative, but what
it expected to do was to unset all modes on all CRTCs. Despite this, it
had code to unset modes on unconfigured CRTCs, probably because it was
used for multiple things in the past.
Make this a bit easier to follow by renaming the function
"unset_modes()" and fold the function doing the unsetting into the
function itself.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3318>
Intel has started to advertise big gamma LUT sizes on some hardware
because the hardware supports segmented LUTs. This means they have a lot
more precision at certain segments then others. The uAPI can't expose
this functionality meaningfully so they chose to expose a huge number of
TAPs to sample from to their segmented LUT.
This increase in uAPI LUT size resulted in stack overflows because we
allocated the LUT on the stack. This commit moves it to the heap
instead.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3064
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3322>
1. Centralize stride calculation in one function.
2. For dmabufs query the stride instead of assuming a certain value.
3. For system memory buffers use the pixel format to calculate the
stride.
4. Stop negotiating `SPA_PARAM_BUFFERS_size` and
`SPA_PARAM_BUFFERS_stride`.
2. fixes an actual bug where we reported wrong max buffer sizes,
resulting in crashes in Gstreamer when doing area screencasts on AMD
GPUs.
The reasoning for 4. is that the values were possibly wrong for
dmabufs as the negotiation happens before we create any buffers.
Further more neither Mutter nor the common consumers required it.
The later either ignore the values (OBS), always accept (gstpipewiresrc)
them or calculate the exact same possibly wrong values (libwebrtc).
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/6747
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3311>
With EI support wired to XTEST, and oeffis being enabled in Xwayland
means that XTEST will always go through the XDG portal.
While this the intended behavior for the general use case of Xwayland
running rootless on a desktop compositor, that breaks when Xwayland is
running on a nested compositor, because the portal is for the entire
session and not limited to the nested Wayland compositor.
Enable XDG portal support in Xwayland only when we managed to connect
to the GNOME session manager, which means we are running in a full
desktop session, and not in any form of nested mode.
This is determined by simply using the status returned by set_gnome_env()
which will fail if not connected to a GNOME Session manager.
See-also: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1586
See-also: https://gitlab.freedesktop.org/xorg/xserver/-/merge_requests/1170
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3047
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3303>
The function set_gnome_env() is used to pass environment variables
though DBus using the "org.gnome.SessionManager".
If that fails, it means we are not running in a full environment, which
might be useful to determine whether Xwayland should enable the portal
support.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3303>
Enabling portal support unconditionally breaks when running nested, so
Xwayland has now a new command line option to explicitly enable support
for XDG portal at runtime.
Check whether the version of Xwayland we are using has such an option.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3303>
This makes sure the new update takes effect over the pending update for
any common properties. It matches the other users of
meta_kms_update_merge_from.
Fixes: 27ed069766 ("kms/impl-device: Add deadline based KMS commit scheduling")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3316>
When a configuration has a fractional scale, but we're using a physical
monitor layout, we can't use the scale, but if we do, we end up with
wierd issues down the line. Just discard the config if we run into this.
Eventually we probably want to store the layout mode in the
configuration so we can handle more seamless switching between physical
and logical layout mode, but first do this.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3057
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3299>
Mutter so far very rarely use the name `self`, so lets describe that in
the conventions document so that we can stay consistent.
Also mention how to deal with abstract/generic object instance pointers
vs sub type ones.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3297>
The surface actors may not have a valid allocation when running the
test. The preferred height, which `clutter_actor_get_size()` returns
in that case, can be wrong in certain cases, making us not add the black
background when it's actually needed.
Query the allocation instead, even at the expense of additional
relayouts.
While on it, sneak it some small cleanups.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3024
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3310>
The absolute modelview contains OpenGL coordinates, which have a higher
chance to not be invertible or, when doing so, introduce rounding
errors. These again often result in relative transforms becoming 3D
instead of 2D, making us miss optimized code paths down the line.
Thus cache stage-relative matrices instead, improving correctness and
possibly performance.
While on it also add some fast paths for cases where we can skip
calculating inverted matrices altogether and change variable names to be
more precise.
Fixes: dfd58ca8f1 ("clutter/actor: Extend caching in apply_relative_transformation_matrix")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3286>
There's two aspects from its documentation
(https://xkbcommon.org/doc/current/group__state.html#gae56031a8c1d48e7802da32f5f39f5738)
affecting us here:
1. "This function is similar to xkb_state_key_get_syms(), but intended for
users which cannot or do not want to handle the case where multiple
keysyms are returned (in which case this function is preferred)."
We are indeed in that field, and have been for a long time.
2. "This function performs Capitalization Keysym Transformations."
This is unlike the xkb_key_get_syms() function that we use, and
convenient here for parity with X11 since it behaves exactly that
way.
Fixes cases where the keysym for some keys is not properly capitalized
when caps lock is toggled, due to the output of capslock+key being
different from shift+key. An example of this is 'é' in french(azerty)
layout (bound to the '2' key). Even though shift+2 outputs '2',
capslock+é should output 'É'.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3058
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3306>
This is similar, but reserved for the crossing events induced by the
input shape changes on our overlay window. The mechanism in the previous
commit does again protect against this, so this mechanism may go away.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3267>
Focus follows mouse is meant to avoid focusing windows that happened
to pop up under the pointer, e.g. due to mapping, workspace changes,
etc... On X11, this has been done since ancient times through a
moderately complex synchronization mechanism, so mutter would know
to ignore crossing events caused on those situations.
This mechanism is much prior to XInput 2 though, where we may know
this in a more straightforward way: If the sourceid of the crossing
event is a logical pointer (i.e. equals deviceid), the crossing event
was triggered logically, and not through user input.
Perform this simpler check, and drop the existing mechanism to
ignore logically induced crossing events.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3267>
The virtual stream source with CURSOR_MODE_EMBEDDED uses
META_STAGE_WATCH_AFTER_PAINT as the callback for recording its frame. In
this stage of the paint though, there is no ClutterPaintContext anymore
(there only is a paint context during the paint, not afterwards).
The callback (actors_painted()) tries to get the redraw clip from the paint
context, and we end up with a NULL pointer crash.
We actually do still have a redraw clip at this point, so because everyone
uses the paint context to get the redraw clip anyway, just pass the redraw
clip to the stage watches directly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3283>
After an event has been handled such that it bypasses both Clutter and
Wayland, e.g. when handling a keybinding, bypass_clutter would get
unset in the presence of a wayland grab. This means that the event is
handled both as a keybinding and by Clutter.
In the case of switcher popups in gnome-shell in the presence of a gtk4
autohide popover this meant that instead of selecting the next element,
it would select the one after that. If there are only two elements, as
is common with input sources, this would mean going back to the current
one, preventing switching them with a single press of the keybinding.
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/6738
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3054>
When a device is added, libei does not allow adding additional regions
for that particular device, as it is already advertised to the EI
client.
As a result, mutter currently effectively only adds the first region to
a device, but not the others.
This makes input in multi monitor sessions only possible on one monitor,
as the EI client cannot look up the other regions, since they were not
advertised to it.
Fix this situation by not adding and resuming the device, when a shared
device is used.
Instead, for shared devices, always add all regions first, and then
after that, add and resume the device.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3273>
Use the previously added API to release acquired mapping ids, when the
corresponding stream is destroyed.
Otherwise, the remote desktop session would maintain a whole bunch of
unused mapping ids, as their corresponding streams are already
destroyed, but maybe not the session.
Such situation would be a remote multimonitor session, where the amount
of used virtual monitors changes multiple times during the session.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3273>
The remote desktop session currently provides a mechanism to acquire
mapping ids.
However, when they are not used anymore, they currently cannot be
removed and thus just linger around.
So, add an API to release these acquired ids.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3273>
Avoid passing the MetaSeatImpl, since it may be potentially null at
MetaSeatNative construction time. An example of this triggering issues
are mousekeys, since those work on an emulated pointer device created
indirectly after a keyboard device is added (and the right settings are
enabled) at a time that the MetaSeatImpl is still being created, so the
MetaSeatNative cannot yet have a reference to it.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2869
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3278>
These objects already have a pointer to the ClutterSeat that has a
pointer to the MetaSeatImpl in its native implementation. This data
may be considered pretty much immutable (a pointer to the seat is
held, and the native implementation will shut down the implementation
thread within ClutterSeat finalization.
Avoids some awkward code, since the MetaInputDeviceNative needs to
be aware of the Clutter object implementation and the implementation
object.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3278>
vc4's implementation of `drmModeAtomicCommit` seems to require a few
milliseconds advanced notice or else it will miss the frame deadline.
That's too high for our deadline evasion threshold which is measured
in microseconds. Let's stop trying to use deadline timers on vc4 to
avoid this conflict without having to disable atomic KMS.
Suggested-by: Jonas Ådahl <jadahl@gmail.com>
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2953
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3279>
When CLUTTER_ENABLE_DEBUG is not defined, then CLUTTER_NOTE is defined
as an empty block of code. As a result of that, jitter_us is in that
situation unused, and a compiler warning about this unused variable
appears.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3275>
It's the 10 bit equivalent to NV12 and uses the same layout as P016, i.e.
16 bit components with the lowest 6 bits set to 0 (padding), allowing us
to use 16 bit "subformats".
Thus adding support is quite trivial as we can reuse the NV12 shader.
The format is widely supported in decoding and display hardware (on Intel
since Kaby Lake), as well as modern codecs (AV1, VP9, HEVC) and has
visible quality advantages over NV12.
Note that the additional colors are lost if composited to a 8 bit RGB
framebuffer. Switching between direct scanout and compositing can thus
cause quality differences. This is no new phenomena, however, as the
same is the case already for e.g. GL clients using 10 bit formats -
including video players.
Also note that P012 and P016 could trivially added as well - it's not
done here as they are uncommen and thus hard to test.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3244>
These shaders can be used for similar formats with other component
sizes since the values are represented as floats. So whether the source
value was stored in 8bit, 10bit or 16bit doesn't matter - the driver
will covert it for us.
Thus use a Weston-inspired, more general naming scheme.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3244>
They are needed as "subformats" for higher bit YCbCr formats, such as
P010, and we don't plan to use or expose them otherwise. Thus don't
implement any conversion or packing features.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3244>
Realizing a cursor will assume view related state objects are valid so
they can mark them as dirty. This assumption broke when there were a
scale changed that happened with multiple CRTCs, as we'd create view
object by view object as we realized the texture. Realizing the texture
would trigger a signal that had the handler assuming the validity of all
view objects, but if we only had gotten to the first, the second view
would not be there yet, thus we'd be doing a NULL pointer dereference.
Creating the view objects first, then handling the updating avoids this
problem by making the already done assumption valid on hotplugs.
The test case added tests exactly this series of events, and uses a
virtual monitor as a cheap trick to make the KMS CRTC based view the
first one, and an arbitrary view the second that previously had its view
object initialized too late.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3012
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3262>
The cursor surface is decided by the "current" surface; if that alone
changed (e.g. current surface was destroyed), we didn't update the
cursor, meaning it either got stuck, or got hidden if the client exited
completely.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3262>
It's hard to tell why turning on HDR mode failed without these log
messages. It could be missing support in the sink (EDID/DisplayID) or
missing support in the driver/display hardware (connector properties) or
just a failure turning it on.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3251>
Or else `glXQueryDrawable` will fail per the `GLX_EXT_buffer_age` spec:
> If querying GLX_BACK_BUFFER_AGE_EXT and <draw> is not bound to
> the calling thread's current context a GLXBadDrawable error is
> generated.
This mistake went unnoticed until `mtk_x11_error_trap_push` was introduced
(55e3b2e519) because for some reason it is incapable of trapping
`glXQueryDrawable`. Prior to that it seems
`cogl_onscreen_glx_get_buffer_age` would trap and so always returned zero.
This means we're reenabling clipped redraws on X11 here for the first
time in a long time.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3007
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3255>
Apart from a few edge cases we can avoid walking the tree and transform
to the ancestor coordinate space by multiplying the actor stage-relative
matrix with the inverse of the ancestor's stage-relative matrix.
Since the stage-relative matrices are cached, this reduces the number of
matrix multiplications we do in many situations considerably.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3259>
This used to be the HW device that triggered the crossing (i.e.
the mouse moving the pointer, etc), or the logical device if the
crossing event happened through other means than input device
events, e.g. relayouts.
The move to ClutterEvent constructors went a bit too far in
the simplifications and broke these expectations for input-generated
crossing events.
Make this event constructor behave like the other events: receive
a source device, and figure out the corresponding logical device from
there. Also pass the source device as it'd be expected, in the
input-induced crossing event generation paths.
Fixes: a8c62251f8 ("clutter: Port stage crossing events to new constructors")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2981
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3256>
When the actor gets a new "main" surface assigned, it adds the
new surface to the stack of surface actors, but forgets to remove
the old one.
This stale pointer in the array may cause invalid reads and crashes
after the assigned surface is disposed, e.g. when destroying the
MetaWindowActor tries to disconnect signals from all accounted
surface actors.
Fixes: 9a2c8b2592 ("window: Add suspend state")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3252>
Commit 3bfcb6d1 fixed the check for tiling via keybindings, but
ignored a subtle edge case when tiling with the pointer: The
monitor used for tiling is the monitor with the pointer, which
is not necessarily the one that contains the largest part of the
window.
That is, the correct monitor to check against depends on the
context where the function is called. We can either figure
it out automatically via the current window drag, or make it
a parameter.
The latter is clearer, because the callers already decide which
monitor to use for tiling anyway.
Fixes: 3bfcb6d1b9 ("window: Fix portrait orientation check for tiling")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3248>
These snippets are retrieved anew every time a window is resized. But
callers never modify them, they're effectively read-only so cache them
at the place of creation.
This is required to convince the pipeline hash that each reuse of the
same snippet really is the same snippet and so the pipeline is unchanged.
`CoglPipelineSnippetList` only does shallow comparisons and there's no
need right now to reimplement it as a deep comparison.
This eliminates the log message:
> Over 50 separate %s have been generated which is very unusual,
> so something is probably wrong!
which isn't actually a leak but more a warning about wasting time.
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/6958
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3224>
GInitable initialization is failable, currently, it may fail before error
traps are initialized, but error traps would be invariably deinitialized on
finalize() of the failed object. This results in an assert hit, on top of the
original failure to initialize the backend.
The libX11 error handlers are a pure client-side construct, and not a server
request, they just need XInitThreads() called to set up the library-side locks
protecting access to the global variable. This is done beforehand already at
meta_backend_x11_init(), so initialize the error traps around that time too.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3242>
Commit 9c3b130f67 changed slightly destruction order to handle use-after-free
situations, but missed a small new one introduced by the order change: The
MetaX11Display may schedule callbacks through MetaLaters, which depend on the
MetaCompositor, which is now freed before the MetaX11Display.
Since there is no winning move here, make the MetaX11Display aware of this
by avoiding to remove the callback if the MetaCompositor is already gone.
The MetaLaters infrastructure is already fully freed at this point (incl. the
data it contained), so this shouldn't be a leak.
Fixes: 9c3b130f67 ("display: Fix destruction order")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3247>
Use work area from the monitor that the window is currently on to
determine if tiling should be allowed.
Window tiling is disabled for monitors with portrait orientation, but
the work area we use to detect portrait orientation is taken from the
monitor that currently has the mouse pointer.
This works fine for edge tiling using the mouse, but this is broken when
using keybindings for window tiling because your mouse pointer could be
on a different monitor that has horizontal orientation.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3199>
While adjusting the monitor layout of my docked laptop, mutter got a
segfault while attempting to dereference the frame_info struct. This
happened on gnome-shell 44.4-1.fc38.
cogl_onscreen_peek_head_frame_info() just forwards the call to
g_queue_peek_head() which returns NULL in the event that the queue is
empty. If finish_frame_result_feedback() is expected to always be called
with a non-empty queue there's still a bug somewhere, but regardless
this API can legitimately return NULL so it should be checked for prior
to dereferencing.
Fixes: 61801a713a ("onscreen/native: Avoid freezing the frame clock on failed cursor commits")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3229>
If we are making an update that only disables CRTCs, we would not
actually post it, but just drop it then post nothing, as it wasn't ever
added to the mode set update hash table. This resulted in hotplugs where
we loose the all the connectors we had, where we want to disable all
CRTCs and enable nothing, to fail to disable said CRTCs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3073>
This removes the implicit dependency on `display->stack_tracker`
existing and being valid in `on_stack_changed()` because
now it is the stack-tracker's responsibility to subscribe
to the "changed" signal of the stack and handle the changes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3202>
The _NET_WM protocol, written before the birth of XInput 2.x, does have
no notion of different input devices whatsoever. Anyways, in a X11 session
it is safe to assume this refers about the Virtual Core Pointer since
every input device by default drives it (incl. touchscreens through the
"pointer emulating sequence", and styli).
This assumption falls apart in a Wayland session with non-pointer input,
since we do actually distinguish between all the distinct pointer devices
and touchpoints, and do not let them emulate mouse input.
We do need to specify a device/sequence there to drive the window
move/resize operation. The _NET_WM_MOVERESIZE message just gives us the
x/y root coordinates the resize was started from, so work from there
into guessing what is the most likely device/sequence that did trigger
the request on the client side.
Conversely, on Wayland we do not need to check for possible race
conditions in the pressed button states since we have larger guarantees
about not missing these events if we checked for the button modifier
mask beforehand, so make that race condition check specific to the
X11 sessions.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2836
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3059>
The stage has the knowledge about input that is ongoing over it
(incl. things like styli and touchpoints). Add an iterator API
for these devices/touchpoints, so they can be used for calculations
and heuristics in other places of the code.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3059>
There can be events which don't not have source devices set on them, because
they are not backed by real hardware and rather generated by us, for example
IM events coming from the shell's OSK.
So don't assume all events have a source device in
update_pointer_visibility_from_event() and rather ignore those without one,
as we are only interested in events coming from "real hardware" here.
This fixes an issue where the mouse pointer would appear on devices without
any input from actual mice/touchpads on OSK key presses.
Fixes: 6aa42d6dad
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3236>
So far, we expected all events to have input devices set on them, IM events
lost theirs with commit 6aa42d6dad. This somewhat made sense, because IM
events are not backed by any actual device, they are generated by us in
response to eg. an OSK key press.
To fullfil the assumption that all devices at least have a logical input
device set, pass the seat to the clutter_event_im_new() constructor and then
set the device to the logical keyboard device. The source_device we leave
empty, since there is no actual physical device that this event came from.
Fixes: 6aa42d6dad
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3236>
The assertion for !implicit_grab_cancelled in the
`grab_actor == old_grab_actor` case of
clutter_stage_notify_grab_on_pointer_entry() is meant to do a simple
sanity-check to ensure the grab machinery is working as intended: During a
seat grab, all input gets delivered to the tree inside the grab, and all
implicit grabs outside of that tree are cancelled.
When a new seat grab on the same actor as the existing one happens, we run
through the cancellation machinery for implicit grabs anyway, so we might as
well check that the assumption mentioned above holds true: By asserting that
no implicit grabs were cancelled, we know that no implicit grabs exist
outside of the existing seat grab tree.
This assertion is slightly over-eager though due to the way we set
implicit_grab_cancelled: We initialize it to TRUE in the
entry->press_count > 0 case and then only set it to FALSE once we find an
implicit grab that may remain active. If there are no implicit grabs though
(while entry->press_count is still >0), we never set implicit_grab_cancelled
to FALSE, triggering the assertion in question even though no implicit grabs
got cancelled.
There's two possible solutions for this: Either dropping the assertion, or
refactoring it so it observes the situation where the implicit grabs were
already undone. This commit implements the latter.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2700
Fixes: debbd88f8c ("clutter/stage: Cancel parts of implicit grabs when ClutterGrabs happen")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3216>
When we call get_grab_info() to get the sequence, device and coordinates for
a touch window drag, as the device we use the device from the
MetaWaylandPointer, assuming that it's set to the core pointer.
In the case where there is no pointer device present on the seat (so no
mouse nor touchpad), the wayland pointer remains disabled though, and
pointer->device is NULL.
This means touch window dragging on hardware without pointer devices
present is broken (because MetaWindowDrag assumes that there's a valid
device passed in meta_window_drag_begin()). Fix it by taking the core
pointer directly from ClutterSeat instead of going the extra detour through
MetaWaylandPointer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3238>
If meta_eis_viewport_get_position() returned FALSE, the variable
'has_position' would be initialized. This variable represents
exactly the return value of meta_eis_viewport_get_position(),
so just assign it to the variable directly.
Spotted by Coverity.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3237>
CLUTTER_SCROLL_SOURCE_UNKNOWN only generates continuous scroll events
and no discrete scroll events.
As a result, scrolling only works in applications, that support high
resolution scroll wheels, like GTK4 applications.
GTK3 applications, on the other hand, don't support high resolution
scroll wheel events, and such scrolling does not work in these
applications.
Fix this issue by using the scroll source CLUTTER_SCROLL_SOURCE_WHEEL.
Since commit 92a90774a4 ([0]),
CLUTTER_SCROLL_SOURCE_WHEEL generates discrete events to ensure that
scrolling in legacy applications still works.
[0]: https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2664
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3235>
We need to trigger a mode set when power-save changes to 'on' if it's
purely about power saving, but when they arrive as part of a hotplug
event, we'll handle all that later, in the monitors-changed handling,
that contains the new configuration.
This avoids a crash that happens due to the mode set being queued on now
disabled connectors.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2985
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3233>
We can change power save mode for two reasons: gsd-power told us to, or
we saw a hotplug event. Sometimes it's useful to be able to make the
distinction to why a power save mode changed, so add a reason to the
signal.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3233>
If the deadline timer is disabled (like on nvidia-drm or when
`MUTTER_DEBUG_KMS_THREAD_TYPE=user`), then we need to call
`meta_kms_device_set_needs_flush` on every cursor movement. But some were
getting skipped if they coincided with page flips, which resulted in some
cursor movements failing to schedule the frame clock. This resulted in
unnecessary levels of frame skips when using lower frequency input devices
which are less likely to provide another event within the same frame period.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3002
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3210>
Add a pair of calls to ensure the error trap infrastructure
survives for the MetaBackend. This will help on later commits that
largely operate on the MetaBackendX11 Display.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3230>
Keep a per-display list of error traps, so we don't mix them
together, and possibly deem unintended error traps outdated.
This means init/deinit calls are now stackable, and need to
happen evenly. In order to honor this, move the MetaX11Display
error trap destrution to finalize.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3230>
We currently have a bit of a disaster area wrt X11 error handling,
with all of Cogl/Clutter/Mutter offering implementations with different
expectations and different degrees of integration with other error
handlers.
It makes more sense to have a single X11 error handling implementation
that is used in all those places. Mtk seems like the reasonable place
to have this kind of general API, so adopt the more advanced code
at src/x11.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3230>
This forces not using the seat_proxy. But still allows the use of
session_proxy.
On tests, headless mode is explicitly set and it might not be available a
systemd session. To avoid test failing on this situation skip using
meta_launcher wich uses session_proxy and seat_proxy.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3093>
If we're a input-only remote desktop session, create libei regions on an
absolute pointer device corresponding to all logical monitors. This
allows absolute pointer motions without screen casting.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3228>
Sometimes it makes no sense to have a shared pointer device, for example
when they have no set region occupying the global stage coordinate
space. This applies to for example window screen cast based pointer
device regions - they are always local to the window, and have no
position.
We do need shared absolute devices in some cases though, primarily
multi-head remote desktop, where it must be possible to keep a button
reliably pressed when crossing monitors that have their own
corresponding regions.
To handle this, outsource all this policy to the one who drives the
emulated input devices. Remote desktop sessions where the screen casts
correspond to specific monitors (physical or virtual), we need to make
sure they map to the stage coordinate space, while for window screencast
or area screencasts, we create standalone absolute pointer devices with
a single region each.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3228>
We already have the remote desktop session ID, and we'll soon need the
actual remote desktop session in the screen cast session, so pass it on
construction.
The old screen cast type is set implicitly instead.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3228>
A MetaEisViewport represents an absolute region backend by e.g. a
pointer device. There are two kinds: a standalone viewport, which
corresponds to a viewport that has no neighbours, and a non-standalone,
which represents a region of a global coordinate space.
The reason for having non-standalone viewports is to allow to mirror the
logical monitor layout of a desktop, while the standalone are meant to
represent things that are not part of the logical monitor layout.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3228>
How EIS will be used depends on its context, meaning we'll have multiple
EIS contexts that expose different things. To prepare for this remove
the global socket since that won't work with multiple contexts.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3228>
This fixes a compiler warning:
```
src/x11/events.c:523:1: warning: ‘get_event_name’ defined but not used [-Wunused-function]
523 | get_event_name (MetaX11Display *x11_display,
| ^~~~~~~~~~~~~~
```
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3223>
This used to be the behavior, until commit 5d35138df0 changed the meaning
of the return value of MetaCursorRendererClass::update_cursor(). This
made the user of pure-overlay cursors (singular, MetaWaylandTabletTool)
miss their overlays.
Change the return value, so that it matches the desired behavior of
a backend-less overlay-only cursor renderer.
Fixes: 5d35138df0 ("cursor-renderer: Make 'handled_by_backend' state 'needs_overlay'")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3218>
We react on changes to has_hw_cursor, but always try to inhibit if
there is no cursor sprite. While this looks like a reasonable optimization
with the typical situation of one cursor renderer, it may fall into
inhibiting twice without knowing to unwind, e.g.:
1. has_hw_cursor: TRUE, cursor_sprite: !=NULL -> inhibit
2. has_hw_cursor: FALSE, cursor_sprite: NULL -> inhibit
3. has_hw_cursor: TRUE, cursor_sprite: !=NULL -> uninhibit, but once
And this may also result in the CLUTTER_PAINT_FLAG_NO_CURSORS flag
staying on for Tablet cursors, that (so far) always use overlay paths.
This results in invisible tablet cursors after using the mouse at
least once.
Fixes: e52641c4b6 ("cursor-renderer/native: Replace HW cursor with KMS cursor manager")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3218>
Under strange timings, the GTK frames client may implicitly queue
relayouts that end up disagreeing with the latest frame size as
given by Mutter, this results in GTK calling XResizeWindow, and
Mutter plain out ignoring the resulting XConfigureRequestEvent
received.
This however makes GTK think there's pending resize operations,
so at the next resize it will freeze the window, until enough
resizes happened to thaw it again. This is seen as temporary
loss of frame-sync ness (e.g. frozen frame, and other weird
behavior).
In order to make GTK happy and balanced, reply to this
XConfigureRequest, even if just to ignore it in a more polite
way (we simply re-apply the size Mutter thinks the frame should
have, not GTK), this results in the right amount of
ConfigureNotify received on the frames client side, and the
surface to be thawed more timely, while enforcing the size as
managed by Mutter.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2837
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3189>
This is meant for compatibility purposes with the shell extensions
avoiding to break a bunch of them in the last minute and we would
drop it in the GNOME 46 release.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3128>
Currently, Meta/Cogl/Clutter makes use of cairo_rectangle_int_t despite
the existance of MetaRectangle.
In order to make MetaRectangle usable in Cogl/Clutter as well, Mtk would
provide such base types that are shared across the various private
libraries
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3128>
This used to be the case before the refactor at commit 43cee4b6b6,
use_clipped_redraw would be unset before the larger check if has_buffer_age
was set, but clutter_damage_history_is_age_valid() was FALSE. This got
replaced by a check just on the latter, which will also be FALSE if
has_buffer_age is not present.
We have other means to achieve clipped redraws, so this slight change
culled all of them.
Fixes: 43cee4b6b6 ("stage-impl: Do clipped redraws when drawing offscreen")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2771
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3221>
This is at best pointless, since the relayout will change pointer
picking conditions, and buggy at worst, since the actor being
relayout is not at an state that it can be asked during picking.
This specifically fixes warnings like:
Can't update stage views actor <unnamed>[<Gjs_ui_workspacesView_WorkspacesDisplay>:0x5601dd557050] is on because it needs an allocation.
Can't update stage views actor <unnamed>[<Gjs_ui_workspacesView_WorkspacesView>:0x5601edcf6aa0] is on because it needs an allocation.
Can't update stage views actor <unnamed>[<Gjs_ui_workspace_Workspace>:0x5601ee163dc0] is on because it needs an allocation.
Can't update stage views actor <unnamed>[<Gjs_ui_workspace_WorkspaceBackground>:0x5601ee1c85b0] is on because it needs an allocation.
Can't update stage views actor <unnamed>[<ClutterActor>:0x5601ee4db280] is on because it needs an allocation.
Can't update stage views actor <unnamed>[<Gjs_ui_windowPreview_WindowPreview>:0x5601ee1840c0] is on because it needs an allocation.
Can't update stage views actor <unnamed>[<Gjs_ui_workspace_Workspace>:0x5601ebe1d1b0] is on because it needs an allocation.
Can't update stage views actor <unnamed>[<Gjs_ui_workspace_WorkspaceBackground>:0x5601edbd3420] is on because it needs an allocation.
Can't update stage views actor <unnamed>[<ClutterActor>:0x5601ee3cd630] is on because it needs an allocation.
Can't update stage views actor <unnamed>[<StLabel>:0x5601ee685730] is on because it needs an allocation.
Can't update stage views actor <unnamed>[<ClutterText>:0x5601ed5b5f20] is on because it needs an allocation.
When trying to dismiss the overview with a 3fg touchpad gesture
while the pointer is over a window clone.
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/6935
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3220>
This changes how state is tracked by introducing an explicit state. We
need this since we use asynchronous calls to the out of process
component that handles actual inhibitation, including idleness.
This means if inhibitations changes rapidly, we might end up with an
incorrect state if we e.g. try to uninhibit while we're currently trying
to inhibit.
This is done by adding a state variable that accounts for the pending
state, as well as the active state, with a function that looks at the
current conditions to derive what state we should be in, and what state
we are in, to decide what the next action should be.
For example, if we're trying to inhibit, but now wants to uninhibit,
we'll wait for the inhibit call to complete, recheck what we want, which
would result in an async uninhibit call being made.
Fixes: 388b534062 ("wayland: Implement idle inhibit protocol")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3219>
This fixes the following
1. Minimize window; minimize animation starts
2. Do something that immediately destroys the animated actor (e.g. terminate)
3. This triggered the timeline of the animation to emit a "stopped"
signal while all transitions of the actor were destroyed
Previously we'd implicitly animate the scale again (set_scale(..)) which
created a new transition The hash table iterator didn't like this and
abort():ed
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3213>
The 'suspend state' is meant to track whether a window is likely to be
visible any time soon. The hueristics for this are as follows:
* If a window is hidden, it will enter the 'hidden' state.
* If a window is visible, and unobscured, it will enter the 'active'
state.
* If a window is visible, but obscured by another window, it will enter
the 'hidden' state.
* If there is a mapped clone of a window, it will enter the 'active'
state.
* If the window has been in the 'hidden' state for 3 seconds, it will
enter the 'suspended' state.
This will eventually be communicated to Wayland clients so that they can
change their behaviour to e.g. save power.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3019>
This are called when a clone directly attaches to an actor. While not
reliable to know whether a particular actor is cloned or not, one can at
least this way get notified about whether an actor in particular is
cloned.
This can be useful for e.g. the MetaWindowActor, as it's what's most
likely cloned, e.g. for alt-tab views etc.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3019>
Adds va_marshallers to the ClutterStage signals. This allows for better
stack traces to be retrieved when profiling. Additionally, since the
generic marshaller was using GBoxedCopy/GBoxedFree functions for the
GValue usage, the previous code was acquiring a global reader/writer
lock in GObject via g_boxed_free() usage.
With G_SIGNAL_TYPE_STATIC_SCOPE, the generated marshallers can avoid
the additional copy/free on the instance.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3204>
Previously it transformed a physical CRTC coordinate to a logical desktop
coordinate. But current and future users of the function all require
conversion from logical coordinates to physical coordinates. We would have
had to always invert the transform parameter which is a waste of time when
we can instead just invert the function behaviour.
We also simplify the parameters to show both the point coordinate and the
area dimensions are potentially transformed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3180>
Same reasoning holds than with touchpad gesture events in the
previous commit, this will be the first event seen from a specific
device and will require early handling, or crashes may follow.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3198>
With commit be3bca01a7 in place, we now possibly skip early calls
to clutter_stage_repick_device() happening early during initialization.
These were also indirectly the ones that eventually ended up in
the first call to clutter_stage_pick_and_update_device() actually
initializing the PointerDeviceEntry.
With this no longer happening, we may end up with no PointerDeviceEntry
implicitly set up after initialization, which may fail if the first
event received from the seat pointer does not in fact trigger the
clutter_stage_pick_and_update_device() call necessary to make things
work from there on.
And this does indeed happen on touchpads, since the first input event
obtained after CLUTTER_DEVICE_ADDED is CLUTTER_TOUCHPAD_* instead
of CLUTTER_MOTION. This finds an unset PointerDeviceEntry and crashes
since the pointer device does still have no "presence" on the stage.
Fix this by making CLUTTER_TOUCHPAD_* events also trigger a
device update, so the fist event handled does always trigger the
necessary device update.
Fixes: be3bca01a7 ("clutter: Check that pointer has coordinates prior to repick")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2978
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3198>
This data is commonplace enough that it makes sense to keep it
as generic event data. Make these fields take a hard refcount, like
the private data used to do, and drop these fields from the
ClutterEventPrivate struct.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3190>
When creating a new stream, check if the preferred format is
different from the default (COGL_PIXEL_FORMAT_BGRX_8888). If
it is, then also include it in the list of potential formats
for the stream.
COGL_PIXEL_FORMAT_BGRX_8888 is still passed around as it's
both the default, and the fallback for when things go wrong.
When creating buffers, use the negotiated SPA format instead
of a hardcoded value. We leave it to PipeWire to figure out
what's the best format, since clients may not support the
preferred format of the stream.
Due to how chaotic things got, this commit also cleans up
the create_pipewire_stream() to use an auxiliary array of
SPA formats, which is then iterated on in order to generate
the format pods.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3175>
In future commits, we will want to create DMA-BUFs with pixel
formats other than COGL_PIXEL_FORMAT_BGRX_8888. In preparation
for that, let's start passing a new pixel format parameter to
this function, and the corresponding winsys vfunc.
All callers of this function pass COGL_PIXEL_FORMAT_BGRX_8888
for now. Next commits will change that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3175>
Freeing the window opaque region rather than the frame one when was
leaking the frame opaque region and wrongly setting the window opaque
region to NULL.
Fixes: 82b2b7688 ("core: Add infrastructure to keep window frames' opaque regions")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3188>
This is something the compositor could now track by itself, instead of
being pushed through events. It also makes more sense to do this directly
when the grabbing conditions change, as opposed to the next event.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3185>
Except for the tests that launches `mutter`, use a custom shell
implementation. It's roughly a copy of default.c with some cleanups on
top. A custom shell allows for a bit more freedom when doing testy
things.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3185>
The device/sequence may not currently have a set of coordinates to return.
We correctly leave the out values uninitialized, but don't tell the upper
layers in any way.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3183>
F39 has been branched, so we can use it as base of our CI image
and reduce the number of custom built components.
This will also help if gjs adds support for import maps and
gnome-shell bumps its gjs dependency to use it, as F39
already includes the new mozjs version that gjs now uses.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3173>
Meson stopped using polkit for automatic priviledge elevation, and
will no longer attempt any priviledge elevation when not running
interactively.
Running the entire install command as root used to be problematic
in the past, as it could result in ownership changes of files in
the build directory that would result in build failures later,
but the aforementioned change leaves us with little choice.
Presumably those issues have been fixed, let's hope that's true.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3173>
This splits culling into two different phases to move unobscured region
culling to pre-paint to fix#2680. This is needed as direct scanout
skips the paint phase altogether, but the pre-paint phase always runs as
it's used for selecting the direct scanout surface.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3127>
We have a mechanism to trigger repick after animations on
clutter_actor_set_final_state(), but this will not happen if
animations are disabled.
In this case, shell transitions and other typically animatable
changes on the transform of actors will not naturally trigger
a pointer repick when those get instantly changed to the final
state, possibly preserving the cached state and missing the
just popped in actor altogether.
Trigger an instant repick on animation-less transform changes,
so that these situations are also handled correctly, and the
pointer drops the cached state and is able to find the new
actor.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2918
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3168>
The clutter_seat_handle_event_post() function wants to handle
CLUTTER_DEVICE_ADDED/REMOVED to perform signal emission, but
checks (and asserts) that every event going through it has a
source device.
This is no longer quite true for IM events (they are attached
to the ClutterSeat's keyboard, not a HW device), so the assert
can now fire off (of course undesiredly).
But anyways, for events built through
clutter_event_device_notification_new() (the ones this function
is interested in, after all), it is already a precondition check
that the device is proper at the time of creating the event, so
asserting for it here is redundant.
We can drop this overly generic assert, this is already ensured
for the events that matter, anyways.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3167>
With the ClutterEvent subtype structs sealed, this remains the only useful
struct type that is now usable on the Javascript side. Make all
ClutterActorClass event vmethods use ClutterEvent, and update all users
to this change.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3163>
In test situations we sometimes do not create a libinput context, so
our check on dispose to see if we need closing the input thread is off
if META_SEAT_NATIVE_FLAG_NO_LIBINPUT was provided.
Check the input thread existing instead, since that is the thing we
want to undo here.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
Same reasoning than gnome-shell@dbc9ebc6abc8 applies, it may be
useful to run a pipeline with matching branches in the other
project, without necessarily opening a merge request.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
All ClutterEvent users have been changed to use getter methods
instead of direct field access. We may now make the ClutterEvent
union/structs entirely opaque by moving the definitions out of
public headers.
All future usage of ClutterEvent data should be done through
getter methods.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
Move the string construction bits in the event logging happening in
MetaSeatImpl to a clutter_event_describe() call, so that it has more
freedom in fiddling with ClutterEvent internals, and may be potentially
reused in other places.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
Add methods, and change the API of some rarely used methods, in order
to make all event info currently held/necessary accessible through
ClutterEvent getters, instead of direct field access.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
The full decomposed modifier state is pretty much unused, relying
in keymap signals/properties to track latched modifiers state instead.
The events can be simplified without having this information, and if
it were ever needed in the future, it would be nicer that it had a
dedicated struct and didn't increase the number of arguments in
ClutterEvent constructors.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
Since the full decomposed modifier state is unused, and only the
effective modifier mask matters to users, the new constructors take
just this effective modifier mask. This means this helper went
unused in the port to the new constructors, so can be now dropped.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
Since the full decomposed modifier state is unused, and only the
effective modifier mask matters to users, the new constructors take
just this effective modifier mask. This means this helper went
unused in the port to the new constructors, so can be now dropped.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
We need to change at least event flags in the event we reinject after
it was let through by the IM, in order to avoid doubly handling.
Create a full event copy from the original event parameters.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
Avoid peeking in the stage for loosely related actors, since the same
event could be handled by different actors across the picking stack.
This getter is also unused, so there's wiggle room here for changes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
These constructors take all the necessary arguments to build the
event from the get go, so the callers are not left up with the task
of setting up the event struct data.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
This is done from the backend X11 connection, but needs directing at times
from the frontend X11 connection. Commit 5a8509f895 added a XEvent
argument presumably for possible future expansions that did never come.
Since this function is nothing about events, drop the XEvent argument and
make the name a little bit more ad-hoc (according to what it does, at
least).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
Despite the attempt to make this a generic interface, this was
pretty much used only by the X11 backend, and now it ported away
from it.
This now stands unused and may be removed, in favor of backends
each creating and injecting events as they please.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
This is about the only reason now to go through the ClutterBackend
translate_event vmethod. We can do that directly, and stop requiring the
generic vmethod that is actually just used for X11 events.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
There's no need for an XEvent filter, since this is already code close enough
to MetaBackendX11 XEvent handling and always required anyways. Make the a11y
configuration checks happen directly from MetaBackendX11 event handling.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
We are pretty much guaranteed that the first event will be handled after
the cogl renderer has been set up. We can avoid the loop through
ClutterBackend vmethods and X11 event filters, and call this directly
from the code that is already close to the MetaClutterBackendX11.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
Stop taking a ClutterEvent and pass the essentials here (x/y/evtime),
we don't have a ClutterEvent handy in all places that we call this
API, and it feels awkward to create one just for calling this vmethod.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
It is a bit backwards that events contain information about
the stage they are being handled by. It makes more sense to
specify in the ClutterEvent handling entrypoint the stage
that will handle the event.
As a first step, add this ClutterStage argument, even though
the information is still carried through the event in order to
keep satisfying calls to the getter function.
This entrypoint has been also renamed to clutter_stage_handle_event(),
so that its ownership/namespace is clearer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
Nowadays, all our MapNotify event handling happens already prior to
the MetaCompositorX11 handling of XEvents. It does not make sense to
channel these events again through the backend, at best all it could
lead to is double handling of the same events.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3153>
Because `meson dist` will fail in that case:
```
Dist currently only works with Git or Mercurial repos
```
Being away from the git repo would only happen on non-marge CI runs of
'deploy' and only if the MR contains meson changes. So the bug went
unnoticed when introduced in !3083 because it didn't contain meson changes.
To fix it we just revert one line of !3083.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2908
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3126>
These "features" are somewhat less featured, it's becoming too ugly
to handle all of them with a single API call. The clear outlier are
buttons, so move them to a separate function.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3005>
queue_update() in a previous iteration was called in two situations:
* A page flip was already pending, meaning if we would commit an
update, it'd fail with EBUSY.
* A update was marked as "always-defer" meaning it should only be
processed from the deadline callback (would there be one). These were
used for cursor-only updates.
In the latter, we had to arm the deadline timer when queuing a new
update, if it wasn't armed already, while in the former, we would
currently idle, waiting for the page flip callback. At that callback
would the deadline timer be re-armed again.
Since we're only handling the former now, we'll never need to arm the
timer again, so remove code doing so. The code removed were never
actually executed anymore, after the "always-defer" flag on updates was
removed.
Fixes: 27ed069766 ("kms/impl-device: Add deadline based KMS commit scheduling")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2940
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3150>
Primary plane updates were forgetting to do this in OnscreenNative, but
rather than do it for each post there we should simply do it for each
post.
This fixes cursor stutter in the fallback path (not using deadline timers)
where needs_flush_crtcs would remain populated but CRTC_NEEDS_FLUSH would
never be emitted, because handle_flush hadn't been called for the last
post.
This is safe as the current use of scheduled flushing is only for cursor
updates, and since cursor updates happen on the same thread as processing,
and due to the fact that we always use the most up to date cursor position
when flushing, we never risk leaving an old cursor state unflushed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3138>
Implement importing of multi-plane formats. For now, only support
importing planes individually using "sub-formats". This is the most
commonly driver-supported approach in the moment, used by other
Wayland compositors as well.
In the future we will additionally want to support importing the formats
directly and let the drivers handle conversion internally.
Co-Authored-By: Robert Mader <robert.mader@collabora.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2191>
So they can be derived from the DRM format as well.
While updating the users, ensure we don't announce support for
DRM formats in zwp_linux_dmabuf_v1 if the MetaMultiTextureFormat is
INVALID. This will be used for YUV subformats in following commits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2191>
In future commits, we want to be able to handle more complex textures,
such as video frames which are encoded in a YUV-pixel format and have
multiple planes (which each map to a separate texture).
To accomplish this, we introduce a new object `MetaMultiTexture`: this
object can deal with more complex formats by handling multiple
`CoglTexture`s.
It supports shaders for pixel format conversion from YUV to RGBA, as
well as blending. While custom bleding is currently only required for
YUV formats, we also implement it for RGB ones. This allows us to
simplify code in other places and will be needed in the future once
we want to support blending between different color spaces.
Co-Authored-By: Robert Mader <robert.mader@collabora.com>
Co-Authored-By: Sebastian Wick <sebastian.wick@redhat.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2191>
When we see a mode set, the cursor manager will update all the cursor
planes so they are set correctly as part of the mode set. KMS updates
are always per-device, and what was wrong was that it didn't filter out
CRTCs on devices that wasn't part of the mode set.
Reported-by: Michel Dänzer <mdaenzer@redhat.com>
Tested-by: Michel Dänzer <mdaenzer@redhat.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3130>
It turns out that gnome-shell's toolbox image is still broken
after commit 86b77f65e7.
Toolbox' entry point ensures that the calling user exists inside
the container, and makes their home available inside the container.
There are two ways the `useradd` command in our image may interfere
with that:
- by default, useradd uses the smallest available UID in the
normal user range (usually 1000); this is highly likely to
clash with the host user UID
- if the host's /home is a symlink (for instance to /var/home
on Silverblue), then toolbox recreates that layout inside the
container; it cannot do that if /home is already a non-empty
directory
Luckily we can address both issues without affecting the ability
to build and run tests as user: We can simply create the `meta-user`
with a UID and home directory that are unlikely to clash.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3134>
Measurements above 100% were originally allowed to show when frame skipping
was occurring so you didn't have to also check the frame rate. But that
also resulted in arbitrarily high jitter values being reported when
returning from idle. And those are frequent enough to look like a bug or
untrustworthy so let's not do that anymore.
Taking the remainder of a high jitter value is still a meaningful jitter
value.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2906
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3123>
clutter_frame_clock_compute_max_render_time_us clamps to the refresh
interval anyway, so the only effect a higher recorded maximum can have
is to delay it falling below the refresh interval again.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3090>
Instead of separate pairs of short- and long-term maxima for each
measured step of the frame update process.
This should result in the render time estimate rising less often:
Previously it did whenever the measurement of any of at least 3 out of
4 steps reached a new maximum, even if that didn't result in a new
maximum for the whole update duration. Now it's only in the latter
case.
This should also result in a lower render time estimate (and thus
input→output latency) in general, since the variability of
measurements for each 3/4 steps doesn't always add up anymore. The flip
side of this is that it might result in missing a display refresh cycle
more often.
v2:
* Fix coding style in maybe_update_longterm_max_duration_us.
(Robert Mader)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3090>
It returns non-0 if there are any hints in the WM_NORMAL_HINTS
property, 0 if there are none.
Fixes the mouse cursor changing to the resize shape over the decorations
of non-resizable windows.
Fixes: c7b3d8c607 ("frames: Push error traps around various X11 calls")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3096>
With libdecor, window moving/resizing only works with
the pointer, not with touch.
The meta_wayland_pointer_can_grab_surface checks for subsurfaces,
but the meta_wayland_touch_find_grab_sequence does not.
Add a similar subsurface check to
meta_wayland_touch_find_grab_sequence.
Closes: GNOME/mutter#2872
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3125>
Removing duplication, making it easier to add new formats and ensuring
that the native backend and Wayland clients can use the same formats.
Also improve related build files so the Wayland backend can be build
without the native backend.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3065>
1. Move into the new 'common' folder and build for Wayland as well
so we will be able to share the code in follow-up commits.
2. Rename to cogl-drm-formats to make it more obvious that the format
map is more than an utility these days.
3. Drop the unused CoglTextureComponents part (see also previous
commit).
4. Move the map to the header, simplifying some future use-cases.
5. Sync formats with MetaWaylandBuffer and MetaWaylandDmaBufBuffer and
also use newly introduced opaque formats where appropriate.
This avoids duplicated code, ensures that new drm-formats added to
the dmabuf protocol have an adequate representation in Cogl from which
information like alpha support can be easily derived and finally
ensures we don't crash if the mappings got out of sync.
6. Remove some likely untested formats. In case some of these are
actually needed on certain hardware, we can test whether we got
the correct mapping by also adding support for the corresponding
wl_shm_format in MetaWaylandBuffer by extending the gradient test in
https://gitlab.freedesktop.org/jadahl/wayland-test-clients
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3065>
The default cogl blend string is
`RGBA = ADD (SRC_COLOR, DST_COLOR*(1-SRC_COLOR[A]))` which is alpha
blending with premult fragment results. We do not clear the src
framebuffer and even if we did set alpha to 1 in the src fb, the
resulting alpha would be 1 and we want to check the alpha of the
fragment color.
Just turn off any kind of blending instead and write out the fragment
color to the fb.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3065>
To obtain a float between 0 and 1 we have to devide the integer by the
highest possible value instead of the number of values.
Fixes off by one errors in the tests on some hardware/driver
combinations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3065>
So we can properly handle matching DRM and WL_SHM formats in a unified
manner.
Add extensive testing between these and existing pre-multiplied alpha
formats, i.e. all formats we support on Wayland.
Note that unfortunately for some format combinations the value in the
alpha channel is not cleared as expected, likely because of fast-paths
in Cogl. If both source and destination format is opaque, it always
works, however. This thereby includes all cases where they are the same.
Co-Authored-By: Jonas Ådahl <jadahl@gmail.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3065>
On GLES2 reading and writing some Cogl formats is not supported
natively. In those cases we use another format to do the reading and
writing. When the internal format and the temporary format differ in
premultiplication, Cogl tries to adjust for it.
Opaque Cogl formats don't have the premult bit set but our internal
format is a premult format. Cogl tries to adjust for it but completely
misses that the opaque format doesn't have an alpha channel and it
should not do so at all.
So skip the premult adjusting when the Cogl format has no alpha channel.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3065>
We can schedule an update from the cursor manager, but that doesn't mean
there will be an actual plane assignment changed at the time of the
update processing, since for example we might have "touched" a CRTC, but
already left it before the processing started, meaning we have nothing
to change after all.
Add a test case that checks that this works properly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
This removes the old hardware cursor management code and outsources it
to MetaKmsCursorManager. What the native cursor renderer still does,
however, is the preprocessing i.e. rotating/scaling cursor that wouldn't
otherwise be fit for a cursor plane.
The cursor DRM buffers are instead of being per cursor sprite now per
CRTC, meaning we don't need to stop doing hardware cursors if part of
the cursor is on an output that doesn't support it. This is why the
whole scale/transform code changed from being per GPU to per CRTC.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
It can be quite slow to set up the test environment inside the VM, as
well as outside, leaving very little time for the test itself. While
it'd be nice to not run the mock env etc outside the VM, let's just bump
the timeout for now, to avoid unnecessary timeout failures.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
If we turn of a CRTC, we might have invalidated the cursor manager for
the same CRTC, but that should not mean a cursor plane is assigned when
turning off the CRTC.
Add a test case for this.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
This new manager object intends to take over management of the cursor
plane from the native cursor renderer. It's API is intended to be used
from the main thread, except for the _in_input() function, but mainly
operates in the KMS context, i.e. the KMS thread.
It makes use of an "update filter" that is called before each
MetaKmsUpdate is turned into a atomic KMS commit or a set of legacy
drmMode*() API calls. When the cursor position has been invalidated,
it'll assign the cursor plane in the filter callback, using an as up to
date as possible pointer position as the source for the cursor plane
position.
Cursor updates from the input thread schedules updates for the affected
CRTCs which will cause the filter to be run, potentially for cursor-only
commits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
This adds some plumbing to get the "default" paint flags for regular
stage painting, where one either wants to paint the overlay, or not.
If inhibited, the 'no-cursors' paint flag is used, otherwise the 'none'
flag. This will be used to allow having a per stage view hw cursor
state.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
This makes it possible to post KMS updates that will always defer until
just before the scanout deadline. This is useful to allow queuing cursor
updates where we don't want to post them to KMS immediately, but rather
wait until as late as possible to get lower latency.
We cannot delay primary plane compositions however, and this is due to
how the kernel may prioritize GPU work - not until a pipeline gets
attached to a atomic commit will it in some drivers get bumped to high
priority. This means we still need to post any update that depends on
OpenGL pipelines as soon as possible.
To avoid working on compositing, then getting stomped on the feet by the
deadline scheduler, the deadline timer is disarmed whenever there is a
frame currently being painted. This will still allow new cursor updates
to arrive during composition, but will delay the actual KMS commit until
the primary plane update has been posted.
Still, even for cursor-only we still need higher than default timing
capabilities, thus the deadline scheduler depends on the KMS thread
getting real-time scheduling priority. When the thread isn't realtime
scheduled, the KMS thread instead asks the main thread to "flush" the
commit as part of the regular frame update. A flushing update means one
that isn't set to always defer and has a latching CRTC.
The verbose KMS debug logging makes the processing take too long, making
us more likely to miss the deadline. Avoid this by increasing the
evasion length when debug logging is enabled. Not the best, but better
than changing the behavior completely.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
This is helpful when we add callbacks that should be dispatched in the
KMS impl thread.
This invalidates an assumption about callbacks not being in the impl
context, so some asserts for that are also removed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
This signal is emitted before terminating the thread, but also when
resetting the thread type. This is to allow thread implementations to
make sure they have no stale pending callbacks to any old main contexts.
This commit "terminates" the impl thread even if there is no actual
thread; this is to trigger the "reset" signal, also when switching from
a user thread to a kernel thread.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
This means we can add COGL_TRACE*() instrumentation that is grouped
correctly in sysprof. If kernel threading is enabled, they will end up
in a "Compositor (KMS thread)" group (ignoring translations).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
Real time scheduling is needed for better control of when we commit
updates to the kernel, so add a property to MetaThread that, if the
thread implementation uses a kernel thread and not a user thread, RTKit
is asked to make the thread real time scheduled using the maximum
priority allowed.
Currently RTKit doesn't support the GetAll() D-Bus properties method, so
some fall back code is added, as GDBusProxy depends on GetAll() working
to make the cached properties up to date. Once
https://github.com/heftig/rtkit/pull/30 lands and becomes widely
available in distributions, the work around can be dropped.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
Also add an API to inhibit the kernel thread from being used, and make
MetaRenderDeviceEglStream inhibit the kernel thread from being used if
it's active.
The reason for this is that the MetaRenderDeviceEGlStream is used when
using EGLStreams instead of KMS for page flipping. This means the actual
page flipping happens as a side effect of using EGL/OpenGL, which can't
easily be done off thread.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
This will be necessary in order to default to 'kernel' and then switch
to 'user' if the thread instance can no longer be properly multi
threaded.
To avoid having the same thread impl creating and destroying
GMainContext's, this also means always creating a GMainContext for the
thread-impl. When running in user-thread mode, the GMainContext is
wrapped in a wrapper source and dispatched as part of the real main
thread GMainContext, and when in kernel-thread mode, it runs
independently in the dedicated thread.
This has the consequence that the wrapper source will always have the
priority of the highest impl context GSource, but only after it has
dispatched once. Would we need it earlier than that, we either need a
way to introspect existing sources in a GMainContext and their
priorities, or manually track known sources in MetaThreadImpl.
The wrapper source will never be below 0, as that'd mean it could reach
INT_MAX priority if it had no more sources attached to it, meaning it'd
never be dispatched again.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
While doing this, rename the old synchronous functions to more clearly
communicate that they expect to actually process the update during the
call, not just post it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
While the default when passing NULL will be the main context of the main
thread, make it possible to specify another main context, so that
result handlers can be invoked on the right thread.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
Callbacks could be queued to be invoked either on the impl side or the
main thread side of the thread; change this to take a GMainContext,
which effectively means a callback can be queued to be invoked on any
thread that has a GMainLoop running on its own GMainContext.
Flushing is made to handle flushing callbacks synchronously on all
threads. This works by keeping a hash table of queued callbacks per
thread (GMainContext); when flushing (from the main thread), callbacks
on the main thread context is flushed, followed by synchronization with
all the other threads.
meta_thread_flush_callbacks() is changed to no longer return the number
of dispatched callbacks; it becomes much harder when there are N queues
spread across multiple threads. Since it wasn't used for anything, just
drop the counting, making life slightly easier.
Feedback to thread tasks are however always queued on the callers
thread.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
This means each test is run 4 times:
* with atomic mode setting using a kernel thread,
* with atomic mode setting using a user thread,
* with legacy mode setting using a kernel thread, and
* with legacy mode setting using a user thread.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
This isn't a problem for user space threads, as there are no race
conditions, but when kernel thread support is introduced, we must make
sure that e.g. the main loop is actually running before quitting it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
This uses the queue that was introduced when migrating impl task
management from MetaThread to MetaThreadImpl, with the exception that
it's now fully used as an actual queue. It now has a GSource that sits
on the right GMainContext that is dispatched whenever there are tasks to
execute.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
It's the impl side that wants to add impl side idle sources, or fd
sources, etc, so make it part of MetaThreadImpl.
This changes things to be GAsyncQueue based. While things are still
technically single threaded, the GAsyncQueue type is used as later we'll
introduce queuing tasks asynchronously, then eventually queuing across
thread barriers.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
It currently does exactly what MetaKms and MetaKmsImpl did regarding the
context separation, which is to isolate what may eventually run on a KMS
thread into a separate unit. It works somewhat like a "user thread",
i.e. not a real thread, but will eventually learn how to spawn a
"kernel thread", but provide the same API from the outside.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2777>
Remote desktop version 2 added a new method ConnectToEIS .
ConnectToEIS allows clients to requests a file descriptor from the
compositor which can then be used directly from libei.
Once established, the communication between compositor and application
is direct, without the need to go through the portal process(es).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2628>
Accessibility should be handled on the receiving end, if needed. Make
sure this is the case by listening on some signals, verifying they are
only triggered if we're not capturing input.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2628>
This adds the actual input capturing rerouting that takes events and
first hands them to the input capture session, would it be active.
Events are right now not actually processed in any way, but will
eventually be passed to a libei client using libeis.
A key binding for allowing cancelling the capture session is added
(defaults to <Super><Shift>Escape) to avoid getting stuck in case the client
doesn't even terminate the session.
The added test case makes sure that the pointer moves again after
pressing the keybinding.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2628>
When a relative pointer motion gets constrained (e.g. a monitor edge or
barrier), save the constrained relative motion delta too.
This will later be used to send the remaining motion delta to input
capture clients.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2628>
An input only grab is a ClutterGrab on the stage that doesn't have an
explicit actor associated with it. This is useful for cases where event
should be captured as if focus was stolen to some mysterious place that
doesn't have anything in the scene graph that represents it.
Internally, it's implemented using a 0x0 sized actor attached directly
to the stage, and a clutter action that consumes the events. An
input-only grab takes a handler, user data and a destroy function for
the user data. These are handed to the ClutterAction, which handles the
actual event handling.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2628>
Adding a barrier and later enabling the input capture session will
create MetaBarrier instances for each added input capture barrier.
The barriers are created as "sticky" which means that when a pointer
hits the barrier, it'll stick to the point of entry, until it's
released.
The input capture session is also turned into a state machine with
explicit state, to more easily track things.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2628>
This allows for a sticky barrier to hold the pointer until it is
released, but the owner of the barrier doesn't need a barrier event to
release it. It will be used to implement input capturing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2628>
A sticky barrier means that a pointer in motion intersecting a barrier
doesn't move once having hit it. The intention with this is to allow an
input capture clients to continue a motion once a barrier is hit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2628>
This API aims to provide a way for users to capture input devices under
certain conditions, for example when a pointer crosses a specified
barrier.
So far only part of the API is implemented, specifially the session
management as well as zone advertisement, where a zone refers to a
region in the compositor which edges will eventually be made available
for barrier placement.
So far the remote access handle is created while the session is enable,
despite the input capturing isn't actually active yet. This will change
in the future once it can actually become active.
v2: Remove absolute/relative pointer, keep only pointer (ofourdan)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2628>
A 2D actorless paint volume can't ever need `enlarge_for_effects` because
it has no depth. Clamping to the pixel boundary is sufficient in this case
and avoids extending volumes on the edge of the view into the next view.
Which then avoids unnecessary secondary monitor updates.
Paint volumes correctly become actorless where `clutter_actor_finish_layout`
calls `_clutter_paint_volume_transform_relative`.
Relates to: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/6819
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3112>
This was a temporary fix until there was a better crossing event
delivery mechanism that accounted for actor changes beneath the pointer.
We nowadays have that, and don't seem to need this extra kick to get
crossing events triggered (and cursor changes, etc) when windows appear
or disappear under the pointer.
This commit is effectively a revert of commit
a64dba4d7a.
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/6808
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3104>
With window_is_terminal gone, "strict" and "smart" focus mode have no
behavioural difference. Let's broaden the scope of strict focus mode,
such that windows never automatically focus unless they are an ancestor
to the transient.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3063>
As noted in the comments of window_is_terminal, this is a hack. This
code has not been touched for the better part of a decade. App res_class
tends to differ between Wayland and X11, so it is likely that none of
these apps have been recognised as terminals under Wayland ever. Also,
there are reports that strict focus mode also does not work under X11,
likely due to changes in these terminal apps over the years resulting
in different res_class than those manually specified in here. Let's remove
this hack and change strict focus mode accordingly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3063>
Instead of using `clutter_actor_get_resource_scale()`, we now deduce the
intended buffer scale from the window by dividing the unscaled size by
the final actor size. This is more correct as while the return value of
`clutter_actor_get_resource_scale()` depends only on the monitor where
the surface resides, the actual scale of the surface is determined
solely by the application itself. `get_resource_scale` will differ from
the actual buffer scale if the application only supports 100% scaling
(Xwayland), or is performing scaling with wp_viewporter (clients using
fractional_scale_v1).
This also fixes a mismatch between the calculated buffer sizes between
`meta_window_actor_get_buffer_bounds` and
`meta_window_actor_blit_to_framebuffer` which causes broken
screencasting for Chromium 114 and later when using the native Ozone
Wayland backend.
Additionally, this commit also changes
`meta_window_actor_blit_to_framebuffer` from using a simple translation
to using an inverted matrix transformation of the transformation matrix
between the parent of the window actor and the surface actor to ensure
maximum sharpness for fractionally scaled windows.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3053>
Previously, restarting mutter in an X11 session resulted in
the previously set color temperature not being applied.
Fix that by applying the color temperature right after
the org.gnome.SettingsDaemon.Color proxy has been created.
Furthermore, only call `update_all_gamma()` from `on_gsd_color_ready()`
when the temperature has actually changed. Otherwise there is no need
since the current temperature has already been (or will soon be) applied
to all ready color devices.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3008>
We'd get a re-entry like scenario when destroying the PipeWire stream
object, where PipeWire would call the stream process vfunc. When this
happened, we had already destroyed the stream, so don't try to dequeue
or anything, just do an early exit. Fixes the following crash in the
test case client:
#0 pw_stream_dequeue_buffer() in /usr/lib64/libpipewire-0.3.so.0.367.0
#1 on_stream_process() at ../src/tests/screen-cast-client.c:348
#2 do_call_process() in /usr/lib64/libpipewire-0.3.so.0.367.0
#3 flush_items() in /usr/lib64/spa-0.2/support/libspa-support.so
#4 loop_invoke() in /usr/lib64/spa-0.2/support/libspa-support.so
#5 impl_send_command.lto_priv.0() in /usr/lib64/libpipewire-0.3.so.0.367.0
#6 suspend_node.lto_priv.0() in /usr/lib64/libpipewire-0.3.so.0.367.0
#7 pw_impl_node_set_state() in /usr/lib64/libpipewire-0.3.so.0.367.0
#8 client_node_removed() in /usr/lib64/pipewire-0.3/libpipewire-module-client-node.so
#9 pw_proxy_destroy() in /usr/lib64/libpipewire-0.3.so.0.367.0
#10 pw_stream_disconnect() in /usr/lib64/libpipewire-0.3.so.0.367.0
#11 pw_stream_destroy() in /usr/lib64/libpipewire-0.3.so.0.367.0
#12 stream_free() at ../src/tests/screen-cast-client.c:530
#13 main() at ../src/tests/screen-cast-client.c:803
#14 __libc_start_call_main() at ../sysdeps/nptl/libc_start_call_main.h:58
#15 __libc_start_main() at ../csu/libc-start.c:360
#16 _start() in /home/jonas/Dev/gnome/mutter/build/src/tests/mutter-screen-cast-client
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3095>
Setting up the image with a custom default user broke gnome-shell's
toolbox images. While running tests as non-root user seems like a
good idea, keeping people's development environment working should
be figured out first.
This partially reverts commit 69cc65d15f.
Keep the image to have a local user and use it to run tests so that
we can ensure that permissions are respected
Co-authored-by: Florian Müllner <fmuellner@gnome.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3083>
If an actor's expand properties haven't been set explicitly, its
expand flags are computed by traversing its children.
However we currently also traverse into children when explicitly
setting "expand" to FALSE, because that is the default value and
the properties are only marked as explicitly-set when the value
actually changed.
Fix this, so propagating expand flags can be stopped without
hacks like
```c
g_object_set (actor, "x-expand", TRUE, NULL);
g_object_set (actor, "x-expand", FALSE, NULL);
``
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3088>
If the timelines don't get destroyed they keep references to frame
clocks. Later tests check for the destruction of those frame clocks and
then can fail if the frame clock is implemented slightly differenty.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3084>
In remote desktop sessions, streams can be created and destroyed
on-the-fly.
If a stream is gone, it is not necessarily an error.
So, don't treat that situation like an erroneous one.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2911>
As it is the only place where cogl depends directly on cairo minus
the whole cairo_region_t.
The motivation behind the removal of this helper is to reduce the usage
of cairo in libmutter is to potentially completely drop it in
certain places or replace it with pixman.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3079>
The previous logic didn't work correctly at least when priority-based
preeption wasn't supported by the DRM driver, such as in the case
of amdgpu. The call to glGetQueryObjecti64v would block on client
work which is already in progress (most likely for the next frame)
and delay notifying the ClutterFrameClock about presentation.
Conveniently, the Wayland transactions mechanism guarantees that all
fences of a dma-buf buffer are signalled before the buffer is
included in a frame, which means that dma-buf buffers are ready for
presentation when being directly scanned-out.
Direct scanout is only supported for dma-buf buffers too, which means
that all buffers going through direct scanout are effectively ready
and require no GPU rendering before presentation.
Assuming zero rendering time for dma-buf buffers going through direct
scanout simplifies the code and removes the need for
glGetQueryObjecti64v, thus avoiding the aforementioned issue where it
could block for longer than expected.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2766
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3080>
This is expected for the common case of direct scanout of Wayland
buffers where transactions guarantee that all buffer fences are
signalled before a buffer is included in a frame.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3080>
Dispatch jitter is how much the dispatch interval has changed between
frames. It's a measure of sampling smoothness for events that are occurring
at a higher rate than the screen is refreshing:
* Mouse movement
* Clients rendering at swap interval zero
* Keyframe animation position
Zero jitter is ideal but will practically never happen, and a jitter value
of several thousand microseconds will be visible to the naked eye as stutter
even if you're maintaining a perfect frame rate.
To make the numbers easier to interpret we also log the jitter as a
percentage of the refresh interval.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3082>
This could happen when moving the cursor over GUIs that only redraw
in response to cursor movement. Mutter would experience alternating
cursor-only updates and page flips, and so the `max_render_time_allowed_us`
would jump between pessimised and optimised resulting in inconsistent
frame pacing.
Aside from fixing the smoothness problem this should also provide
lower latency cursor movement.
Fixes: https://launchpad.net/bugs/2023766
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3074>
Depending on the ordering of the surface-associated resources
being destroyed, we may fall into the following situation:
- wl_surface is destroyed
- destruction notifications for the surface runs
- The MetaWaylandKeyboard attempts to synchronize the window
focus
- The MetaWindow is not destroyed yet, so the focused window
remains the same, and the MetaWaylandKeyboard keeps the same
focus MetaWaylandSurface.
- wl_surface finalizes destruction, MetaWaylandSurface now has
a NULL resource
- xdg_toplevel destructor kicks in, it unmanages the window
- The current focus window is again looked up, forced to look
a different window
- The MetaWaylandKeyboard focus now changes, tries to leave the
old surface, but it has a NULL resource already, and raises
a protocol error.
If the order is inverted, the window being unmanaged triggers a
focus change into a different window, the MetaWaylandKeyboard
triggers a focus change while the MetaWaylandSurface is still
intact, it succeeds, and the window gets properly destroyed.
In order to make this independent of the order, it makes sense
to make MetaWaylandKeyboard do like the other objects tracking
focus surfaces, and have it care of its own little parcel. The
surface destructor changed to simply unsetting the keyboard focus
to NULL (guaranteeing that the old focus is left while the surface
resource is still up), and leaving potential focus changes to
the xdg_toplevel_destructor->unmanage->update_focus paths.
Doing that alone is basically a revert of commit 228d681b, thus
is still subject to keyboard focus being lost after a popup is
destroyed. Change the approach to trigger the focus sync (and
new focus surface lookup) so it happens from xdg_popup_destructor
specifically to popups and alike xdg_toplevel.
Fixes: 228d681b ("wayland: Trigger full focus sync after keyboard focus surface is destroyed")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2853
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3077>
Current behavior pushes a window which receives focus to the start of
the MRU list on every workspace it is on. By focusing a sticky window
the default focus on all other workspaces changes as well. This is fine
for sticky windows explicitly marked as sticky by the user but if a
window is on a secondary output and workspaces are only on the primary
output the behavior is unexpected. Instead we want the window to be the
default focus only on the current workspace but also keep those windows
in a relative MRU order to each other on all workspaces.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2681
Fixes: 058981dc1 ("workspace: Focus the default window only if no window is focused")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2909>
This avoids use-after-free when handle_start() is called following
handle_stop() during the lifetime of the MetaProfiler. This happens
on repeated profiling sessions using Sysprof.
Fixes: e16d68372 ("profiler: Add API to register profiler threads")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3076>
We need to juggle with some things here to keep key event ordering
and accounting consistent.
The keyboard internal state changes (and maybe modifier event emission)
happening through meta_wayland_seat_update() should ideally happen
from the same key events that reach the client through wl_keyboard.key,
so that wl_keyboard.modifier events are emitted in the right relative
order to other key events.
In order to fix this, we need to decide at an earlier point whether
the event will get processed through IM (and maybe be reinjected),
thus ignored in wait of IM-postprocessed events.
This means we pay less attention to whether events are first-hand
hardware events for some things and go with the event that does
eventually reach to us (hardware or IM).
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/5890
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3044>
Given the presence of IMs and the different paths in event handling to reach
one of them, we cannot make guesses about whether should stick to the original
hardware-triggered event, or wait/prefer a second hand IM event that might or
might not arrive. We also have no say for other IM foci unrelated to wayland
(e.g. ClutterText) triggering the double event emission.
So go with it and maintain our own internal state for keys, we already kinda
do, but mainly for warning purposes, at the time of updating the
MetaWaylandKeyboard state.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3044>
Currently, we let the same function handle key event filtering as they
are passed to the IM, and the IM events resulting in actions like text
commit or preedit changes.
Split these two aspects into filter/process functions, and port
ClutterText to it. MetaWaylandTextInput still handles everything in
a single place, but that will be split in later commits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3044>
This means initializing the pointer position in MetaSeatImpl
synchronously too, otherwise it's not guaranteed querying the seat state
will result in the expected position.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3071>
I have a monitor which can report two preferred modes: 5120x1440@240
and 3840x1080@60. Since they are enumerated in this order by KMS,
init_output_modes would end up using 3840x1080@60 (and it was impossible
to select any 5120x1440 mode in the GNOME display settings).
Fix this by using meta_kms_connector_get_preferred_mode, which returns
the first KMS mode with DRM_MODE_TYPE_PREFERRED.
v2:
* Use meta_kms_connector_get_preferred_mode. (Jonas Ådahl)
Signed-off-by: Michel Dänzer <mdaenzer@redhat.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3055>
This will consist of device-added events, meaning before init finishes,
we can derive some state that depends on the set of input devices
available on startup, such as cursor visibility.
This avoids cursor visibility switching between hidden and visibility
during startup.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3070>
This opens up for a possibility to handle initial events (devices
discovered on startup) during initialization, meaning we can figure out
a more correct initial state that depends on available input devices.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3070>
This is different from "warping" as it doesn't necessarily result in a
pointer motion event. This can be helpful during initializing so we can
avoid faked pointer events that would otherwise need to be special cased
to not appear as actual pointer movements.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3070>
glGetIntegerv() with GL_RED_BITS/GL_GREEN_BITS/GL_BLUE_BITS/etc. is not
supported with the GL core context, so there is no point in falling back
to that without supporting COGL_PRIVATE_FEATURE_QUERY_FRAMEBUFFER_BITS,
as this will cause an GL error.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3047>
The function ensure_bits_initialized() in cogl-gl-framebuffer-fbo.c
checks for COGL_PRIVATE_FEATURE_QUERY_FRAMEBUFFER_BITS whereas the same
in cogl-gl-framebuffer-back.c simply checks for the driver being
COGL_DRIVER_GL3.
Change the later to use the COGL_PRIVATE_FEATURE_QUERY_FRAMEBUFFER_BITS
flag as well.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3047>
Cogl's feature COGL_PRIVATE_FEATURE_QUERY_FRAMEBUFFER_BITS is required
to use the GL_FRAMEBUFFER_ATTACHMENT_* queries.
Unfortunately, the test for the availability of the private feature is
actually inverted in ensure_bits_initialized() which causes that whole
portion of code to be ignored, falling back to the glGetIntegerv()
method which isn't supported in core profiles.
As Mesa has recently started to be more strict about these, this causes
the CI tests to fail in mutter.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3047>
The stage already maintains its own list of stage-views via
clutter_stage_peek_stage_views(), it's a bit superfluous to copy that list
around all the time into priv->stage_views of ClutterActor. Let's deal with
that by returning clutter_stage_peek_stage_views() when
clutter_actor_peek_stage_views() gets called for the stage.
In order to make sure ClutterActor::stage-views-changed still gets emitted
correctly for the stage, always emit that signal on the ClutterStage when
the stage views get invalidated. This now depends on the backend only
actually invalidating the views and calling
clutter_stage_clear_stage_views() when things have actually changed, but
that should be the case.
This needs a change in one of the stage-views tests, namely the one which
tests stage-view-changed emission on the stage: Here we now see an emission
of stage-views-changed, but that signal emission actually seems correct.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2679>
While we're now doing a fairly good job at not needing those matrices
all the time anymore, we still need it multiple times during every paint
cycle, so it definitely makes sense to introduce some caching here.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2679>
With commit 5a565b4258, we changed
clutter_actor_finish_layout() to be not only about updating stage views,
but also take care of updating the visible_paint_volume, for that we
started doing a full tree traversal of all mapped actors.
This can be quite a performance issue, apparently especially on certain
ARM devices, where the simple tree traversal can take as long as 2ms.
This is precious time we need to paint our next frame, so lets do a bit
more work to avoid those useless traversals.
Fixes https://gitlab.gnome.org/GNOME/mutter/-/issues/2459
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2679>
We're using clutter_stage_schedule_update() now from ClutterActor to kick
off the stage updating machinery when a redraw needs to happen.
This introduced a bunch of unnecessary calls to
clutter_stage_schedule_update() and thus
clutter_stage_view_schedule_update() when multiple actors request redraws
during the same stage update cycle, which is a very common case.
Cut off all those unnecessary calls by bailing out in
clutter_stage_schedule_update() when updates are already queued.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2679>
Using a list of heap allocated ClutterPaintVolumes adds quite a bit of
unnecessary overhead: It means for every single redraw clip we allocate a
list and a paint volume on the heap.
Let's avoid all those heap allocations by using a GArray with static
ClutterPaintVolumes instead.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2679>
So far our logic for queueing redraws goes like this: Actor notices that it
needs to redraw -> actor tells stage that it needs to redraw via
clutter_stage_queue_actor_redraw() -> stage collects more and more redraws
into a QueueRedrawList before the actual stage update happens -> when
that happens, the stage collects the actual redraw clips from the actors via
clutter_actor_get_redraw_clip().
The logic behind this QueueRedrawList was that by storing a list of
redraw entries on the stage, way we can avoid traversing the whole actor
tree one more time to build the redraw clip before the stage update.
These days we have clutter_actor_finish_layout() though, which is basically
exactly that, a whole actor tree traversal that happens before every stage
update.
Since we have that now, we might as well get rid of the whole dance back and
forth between ClutterStage and ClutterActor, and simply merge the logic to
queue redraws into the finish-layout step.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2679>
If no actors have changed their positions and we're only repainting
because a window needs a repaint, the paint volumes of all actors
remain unchanged. There is no reason to redo those paint volumes on every
stage update.
So introduce caching and invalidation logic for the visible_paint_volume
that allows us to avoid a ton of matrix multiplications that right now
are happening for the whole mapped actor tree on every redraw.
Note that this removes two places where the visible paint volume is set
to an empty paint volume: This is a compromise so that we can keep
around the cached pv when hiding and showing an actor, it does "regress"
one case though: When hiding -> moving -> showing an actor, we'll now
include the old paint volume of the actor in the redraw clip on show (even
though redrawing that old region is not necessary, the actor was hidden
after all). This results in a bit of overpaint in this very specific case,
but for the sake of simplicity let's not care about that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2679>
There were two issues with using the shape region to derive an input
region.
Firstly, the shape region is against the client rectangle, while the
surface actor needs it to be against the buffer rectangle. Fix this by
offsetting the shape region before passing it along.
Secondly, we can't just intersect the shape and input region, since that
leaves out the window decorations. Fix this by only intersecting the
input region covering the client part, and the shape region, and then
union that with the input region covering the rest.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3045>
When building mutter with -Ddebug=false, a warning appears, that
`ClutterStagePrivate *priv` is unused.
Simply remove this variable and directly use `stage->priv` in
`CLUTTER_NOTE` to get rid of this warning.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3046>
Don't try to handle things by threads enabling/disabling the main trace
context on-demand, just have a clear start/stop API. For the D-Bus API,
it becomes more straight forward, and for the persistent variant too, as
it avoids having to pass garbage input when it's known that arguments
will be discarded.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2998>
Persistent profiling was started via an env var, but that's rather hard
to discover and remember without grepping; change to use a command line
argument.
The profiler is started early, even during (though late in)
configuration, but configuration should ideally be instant and pointless
to configure.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2998>
Under X11 hiding the backend implies also unmapping the stage window, if
we do that after that we've closed the display we may end up in a
BadWindow error because such window seems to be destroyed together with
the compositor output parent (even though we are not notified about), so
to prevent this, reparent the backend window during compositor unmanage,
setting it back as a root window child.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2835
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3020>
Original commit message:
> ATM sending modifiers to clients prevents direct scanout for DRI3
> clients via Xwayland. Xwayland using the dma-buf feedback v4 Wayland
> protocol will solve that, but that might take a while yet to appear in
> the wild. Once that happens, this can be reverted.
>
> Direct scanout still works for native Wayland clients as well.
Xwayland got support for v4 in 23.1.0, thus let's reenable modifiers on
AMD.
This reverts commit 2f825f3a86.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3030>
We are now building and testing mutter as user, but the clone may happen
as root, before the docker image takes place.
This may create troubles to git, causing errors such as:
fatal: detected dubious ownership in repository at ...
And we can't fix this using safe.directory option because we have no
control on the system at this scope.
So, let's just handle the cloning manually so that the meta-user is
always the owner of the repository.
This fixes the dist job, but also other jobs that may fail because of
this reason.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3024>
Using the old replace syntax doesn't allow to extend dicts such as
variables or other values, but instead it replaces them.
So use a newer and safer syntax, given we don't need to replace any
parameter where used.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3024>
We currently lock the capability of the MetaOrientationManager to emit
the ::orientation-changed signal, but otherwise keep reading the current
orientation and returning it if we are asked politely through
meta_orientation_manager_get_orientation().
This may bring issues e.g. around suspend/resume, since there may be other
parts of the code trying to get the current orientation without receiving
::orientation-changed signals, this may result in the display orientation
being effectively rotated, then stay locked after that.
In order to fix this, make the MetaOrientationManager return a fixed
orientation while locked, only updated after changes in the lock state.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2600
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3021>
This is missed, leaving the cursor renderer disconnected from the stage
updates that could trigger further frame callbacks on the cursor, leaving
some clients like Xwayland stuck with cursors.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3025>
It's just closer to reality of an user session.
As per this:
- We need to bump the required CI template to use this feature.
- Use sudo in the actions that require it
- Replace pkexec with sudo (it wouldn't work otherwise)
- Ensure we don't rebuild during install not to break build dir
- Give permission to use /dev/kvm to our user (we do this during the
image creation because we don't have an user when $FDO_DISTRIBUTION_EXEC
happens)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3016>
We already test install as part of other jobs (such as
can-build-gnome-shell) and in general that's wrong because we may
add to the final install path artifacts that are required during tests,
hiding potential issues with meson test when those files are not
installed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3016>
It's not required and makes things hard to maintain, we can just rely on
the fact we're in a shell and just use `set -e` to prevent any
unexpected failure to go unnoticed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3016>
ClutterText paints selected text using the selection_paint()
function. This function has to main branches of execution:
when the position is in the selection bound, or not. In the
former, we are leaking the CoglPipeline created by copying
the default color pipeline.
Unref the copied pipeline after using it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3007>
For now, it goes the "easy" way of creating the root node and
immediately painting and destroying it. From now on, the main
root node is created only by ClutterStage itself.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3007>
We do in fact allow these combinations of configuration since the Settings
Wacom panel revamp. We no longer need to look up Wacom device features,
since this is allowed for all the devices that have these settings.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3012>
We avoided setting the device matrix applying to the tablet tool (used if the
tablet is in absolute coordinates mode) if the device is configured for relative
motion, but forgot to apply the matrix if changing the device back to absolute
mode, this made the device seemingly forget its attached display until later
configuration changes.
In order to avoid the hassle of looking up the right display again on unrelated
configuration changes, make the matrix be always set on the device, but only
actually used in absolute coordinates mode. This makes the device able to
seamlessly switch between modes and remain mapped to the right display.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3012>
This was somewhat ineffective since it was applied after figuring out
the x/y absolute coordinates. Change the order (filter first, then
figure out abs coords), and use coordinates from the correct device
while at it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3012>
Fix the following warnings reported by udevadm verify:
data/61-mutter.rules:2 Whitespace after comma is expected.
...
data/61-mutter.rules:116 Whitespace after comma is expected.
data/61-mutter.rules: udev rules check failed
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3011>
A window may become undecorated while the frame window is
frozen due to updates. In that case we would both miss a
reply for the frame window, and any other means to trigger
the window actor being thawed.
Check the frozen state after destroying the frame, so that
meta_window_x11_are_updates_frozen() may end up changing
opinion if the frame window was caught in this situation.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2639
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2996>
`meta_surface_actor_is_obscured_on_stage_view` currently fails to
account for non-identity scaling of actor size (e.g. window actor
geometry scale or surface pixel alignment).
Fix this by using the new `meta_region_apply_matrix_transform_expand` to
calculate the unobscured region in stage coordinates.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2726>
This allows MetaCullable to work with actors using arbitrary transforms
which will be needed for implementing surface pixel alignment for
fractional-scale-v1.
This also deletes meta_cullable_is_untransformed as it's no longer
necessary, and we can also stop manually scaling the region objects
while performing opaque region culling in surfaces since it's now
handled transparently by the new `meta_cullable_cull_out_children`
implementation.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2726>
Implement the stable rounding algorithm as described in the discussions
for the fractional-scale-v1 protocol.
This adds an override of the ClutterActor::apply_transform vfunc for
MetaSurfaceActorWayland that ensures the size and position of the
contents of the surface are rounded according to the stable rounding
algorithm.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2726>
Always ensure that the MetaSurfaceContainerActorWayland is aligned to physical
pixel boundary in preparation for fractional-scale-v1 protocol support.
This introduces an override of ClutterActor::apply_transform vfunc for
MetaSurfaceContainerActorWayland that always ensures the actor content is aligned
to physical pixel boundary.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2726>
The template already does this at the end, so this step is
pointless in the best case.
When building the x86-64 image, we install additional packages
afterwards, so the repo metadata is downloaded again.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3010>
The only consumer of this type of rect was the scissor clipping,
which was removed by the previous commit.
Remove window rects from CoglClipStack, and all dependent code.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3006>
A failing allocation is non-fatal here, however if it fails later in a
lazy allocation triggered by `cogl_framebuffer_create_timestamp_query()`
we end up crashing. Thus force the allocation early, like we already do
in other places.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3004>
While it's obviously good to trap possible errors from X calls, we are
mixing the Clutter error trap with the MetaX11Display one for these
calls.
This may result in situations where a X call within a Clutter error
trap fails, but it's actually handled in these sections using the
MetaX11Display error trap. This one will consider the serial out
of the "handled" parts and raise an error.
It is better to stay consistent here, and use the same error traps
than the rest of the X11 backend.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2796
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3002>
When the X11 display is actually XWayland there's no point to delay the
compositor selection, given that mutter itself is the compositor and
doing this may cause the first X11 client that starts not to receive the
right information (and in some cases misbehave).
Since some toolkits are not handling the compositor selection changes
properly at later times, let's make their life easier by just
initializing the selection as early as the other X11 properties, given
that in this case there's nothing to replace.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2472
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2970>
Since c390f70edc ("backend: Set up and use ownership chains")
the type of the ClutterInputDevice object's "meta-input-settings-xdevice"
data is `DeviceHandle`, but that commit failed to change the one place
where the object data is queried. As a consequence, that part still
considers it to be an `XDevice`, so everything that uses the return
value of `device_ensure_xdevice()` works with invalid data. Furthermore,
`device_handle_free()` incorrectly uses the `user_data` as the argument
for `XCloseDevice()` leading to a double free.
Fixes: c390f70edc ("backend: Set up and use ownership chains")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2995>
Otherwise binding new wl_output's might try to send enter to the
destroyed resource. Fixes the following crash:
#0 wl_resource_get_client at ../src/wayland-server.c:801
#1 handle_output_bound at ../src/wayland/meta-wayland-surface.c:1287
#3 signal_emit_unlocked_R.isra.0 at ../gobject/gsignal.c:3812
#6 ffi_call_unix64 at ../src/x86/unix64.S:104
#7 ffi_call_int at ../src/x86/ffi64.c:673
#8 ffi_call at ../src/x86/ffi64.c:710
#9 wl_closure_invoke at ../src/connection.c:1025
#10 wl_client_connection_data at ../src/wayland-server.c:438
#11 wl_event_loop_dispatch at ../src/event-loop.c:1027
#12 wayland_event_source_dispatch at ../src/wayland/meta-wayland.c:125
#15 g_main_context_iterate.isra.0 at ../glib/gmain.c:4276
#17 meta_context_run_main_loop at ../src/core/meta-context.c:482
Related: https://bugzilla.redhat.com/show_bug.cgi?id=2196527
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2992>
Otherwise drivers would be free to alter the buffer content. While no
driver is known to do so, it's probably good to make things explicit.
See also `import_simple_dmabuf()` in Weston.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2990>
The DMA buffer paths vs MemFd paths differ slightly in when content is
recorded. This was in some places done by trying to record but bail if
the dequeued buffer had the wrong type. This is problematic for two
reasons: we'd update the timestamp even if we refused to record, making
the follow-up attempt fail, and we'd dequeue and queue buffers that
didn't get any content, meaning the receiving end would see empty
buffers potentially with only cursor updates.
Fix this by keeping track if a stream is DMA buffer able or not, and
don't attempt to record at all in the places we would previously require
DMA buffers. This avoids both issues: we don't dequeue/queue pw_buffers
that we refuse to record to, and we won't update the recorded timestamp
when we didn't intend to record to begin with.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2783
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2987>
This aims to reduce the amount of pixels that have to be redrawed on the
screen on a clipped actor redraw in case using the union of two
different clips in a surface will substantially increase the redrawn
area.
This should not result in excessive memory consumption as callers of
`clutter_actor_queue_redraw_with_clip` are expected to ensure that the
redraw clip rectangles are adequately deduplicated.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2965>
mutter-clutter is a private library that is only used by the shell and
not meant to be ABI/API compatible in between versions, so there's no
need to add padding to classes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2903>
Hides libdisplay-info under a build time default-off flag,
provides provision to parse essential edid parameters with
APIs provided by libdisplay-info. This implementaion increases
readibility, avoids code duplication and decreases complexity
of edid parsing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2642>
We are attempting to show windows that do not yet have a
surface/buffer, this makes GNOME Shell avoid transitions
for these windows.
Since on Wayland X11 windows are also Wayland surfaces,
this check is also valid for these, and is thus made more
generic to also cater for these windows.
Eventually, meta_window_update_visibility() is called
when the surface gets its buffer, so the window can be
neatly animated.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2611
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2975>
This ATM triggers missed .commit events for the window in question,
to be addressed in Xwayland. Since the test does not seem to specifically
rely on this window being CSD, make it a regular window instead.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2975>
Going for the default GL renderer is known to trigger rendering
artifacts using the NVidia proprietary driver. Since we don't have
too many expectatives about frames being flashy (not to the point
of mandating GL), resort to the cairo renderer in the mean time.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2976>
Just like the HDR Metadata property the Colorspace property values only
indicate that the display driver supports signaling certain colorimetry.
It does not indidcate that the sink actually supports processing the
colorimetry. For this we have to look up the colorimetry support in the
EDID.
The default colorimetry is always supported. If we want bt.2020 we might
get either the RGB or YCC variant even if we ask for the RGB variant but
there is nothing we can do about it so let's just pretend it's a driver
issue.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2919>
If the used EGL backend supports it. In practice this should currently
only affect the nested backend.
Enabling modifiers can help with app development. An example is
`weston-simple-dmabuf-v4l`, which requires the linear modifier to be
available.
Note that Weston behaves similar already.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2972>
This will be used to extract the resolution and refresh rate from
strings like "1920x1080@60.0" or "1280x720". This aims to replace the
use of the locale dependent sscanf() function.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2902>
This is needed by GNOME Shell to remove itself as a input method
implementation during its shutdown sequence. We can't do it ourself
later because at mutters own shutdowns equence, the GNOME Shell
Javascript context has by that time already been teared down.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2934>
This function gets called when a surface state transaction is applied.
Applying a transaction can get delayed, so the Wayland resource may have
already been destroyed when we get here. In that case we cannot send
events, so there's nothing to do.
v2:
* Drop code comment, expand commit log instead. (Jonas Ådahl)
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2737
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2967>
Otherwise we'll have a cursor sprite backed by a surface that no longer
exist. This usually doesn't happen, but can happen in rare situations
related to pointer capability changes Wayland client cursor changes and
hotplugs.
Fixes the following crash:
#0 meta_wayland_buffer_get_resource() at ../src/wayland/meta-wayland-buffer.c:128
#1 realize_cursor_sprite_from_wl_buffer_for_gpu() at ../src/backends/native/meta-cursor-renderer-native.c:1649
#2 realize_cursor_sprite_for_gpu() at ../src/backends/native/meta-cursor-renderer-native.c:1869
#3 realize_cursor_sprite() at ../src/backends/native/meta-cursor-renderer-native.c:1887
#4 meta_cursor_renderer_native_update_cursor() at ../src/backends/native/meta-cursor-renderer-native.c:1100
#5 meta_cursor_renderer_update_cursor() at ../src/backends/meta-cursor-renderer.c:414
#6 meta_cursor_renderer_force_update() at ../src/backends/meta-cursor-renderer.c:449
#7 update_cursors() at ../src/backends/meta-backend.c:328
#8 meta_backend_monitors_changed() at ../src/backends/meta-backend.c:338
#9 meta_monitor_manager_notify_monitors_changed() at ../src/backends/meta-monitor-manager.c:3590
#10 meta_monitor_manager_rebuild() at ../src/backends/meta-monitor-manager.c:3678
#11 meta_monitor_manager_native_apply_monitors_config() at ../src/backends/native/meta-monitor-manager-native.c:343
#12 meta_monitor_manager_apply_monitors_config() at ../src/backends/meta-monitor-manager.c:706
#13 meta_monitor_manager_ensure_configured() at ../src/backends/meta-monitor-manager.c:779
#14 meta_monitor_manager_reconfigure() at ../src/backends/meta-monitor-manager.c:3738
#15 meta_monitor_manager_reload() at ../src/backends/meta-monitor-manager.c:3745
or the following on gnome-43:
#0 meta_wayland_surface_get_buffer at ../src/wayland/meta-wayland-surface.c:441
#1 meta_cursor_sprite_wayland_get_buffer at ../src/wayland/meta-cursor-sprite-wayland.c:83
#2 realize_cursor_sprite_from_wl_buffer_for_gpu at ../src/backends/native/meta-cursor-renderer-native.c:1612
#3 realize_cursor_sprite_for_gpu at ../src/backends/native/meta-cursor-renderer-native.c:1836
#4 realize_cursor_sprite at ../src/backends/native/meta-cursor-renderer-native.c:1854
#5 meta_cursor_renderer_native_update_cursor at ../src/backends/native/meta-cursor-renderer-native.c:1087
#6 meta_cursor_renderer_update_cursor at ../src/backends/meta-cursor-renderer.c:413
#7 meta_cursor_renderer_force_update at ../src/backends/meta-cursor-renderer.c:448
#8 update_cursors at ../src/backends/meta-backend.c:344
#9 meta_backend_monitors_changed at ../src/backends/meta-backend.c:354
Related: https://bugzilla.redhat.com/show_bug.cgi?id=2185113
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2968>
Under certain conditions a stage-view update does not trigger a kms
update.
In such cases we still want the next update to run within the same
refresh cycle, as otherwise we'd waste the remaining time in the
current one.
At the same time we currently use the `after-update` signal for Wayland
frame events, which again may result in more "empty" updates -
creating an unthrottled feedback loop. This can trigger excessive
load both in the compositor as well as in clients.
Introduce a new GSource that is dispatched once per refresh cycle at
maximum per stage view and use it to emit frame events. Do so by
computing the time from when on we can be sure that an update resulting
from a client commit would certainly get scheduled to the next refresh
cycle.
Note: this only works on the native backend. Given that chances are
small that we hit the corresponding issue on e.g. the nested backend,
stick to the previous behavior there for now.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2823>
It will be used to schedule Wayland frame events independently from both
update and presentation time, as the former may happen multiple times
frame and the later not at all.
For frame events we want a timing that is just late enough to ensure
that a following commit by a Wayland client will not get included into
the current frame any more.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2823>
When selecting the default focus window, is_focusable() was not
considering the new conditions for whether a window should be shown or
hidden that were added to meta_window_should_be_showing() in 39942974.
As a result the default focus window could end up a window already
hidden or hidden once meta_window_flush_calc_showing() is called by
meta_window_focus() when focusing the default window. This would cause
meta_window_focus() to fail, which is an issue if it prevents us from
unfocusing a window when it is getting unmanaged.
Fixes: 399429742 ("x11: Integrate frames client into Mutter")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2644
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2962>
create_and_send_dnd_offer() sets the compositor of the offer to the one
from the MetaWaylandDataSource. This then later gets used in
display_from_offer() when trying to get the context from the compositor.
meta_wayland_data_source_xwayland_new() however was not setting the
compositor, so this was causing crashes when dragging things from X11
windows on Wayland.
Fixes: 2731f0cda ("wayland: Setup and use ownership chains")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2723
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2956>
This ensures that applications are notified when a drag gets cancelled
because the user dropped or press ESC while in overview.
This fixes an issue with Chromium on Wayland refusing to acknowledge
wl_pointer::enter events after accidentally dropping a
Chromium-originated object in GNOME Shell overview.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2953>
Ensure the frame window is created at the right fullscreen state
before showing it and assigning it to the client window.
A peculiarity of this property on frame windows is that it is
typically single-handedly updated from the Mutter side, in synchronization
with client window state. It can only differ during creation, since
GTK still likes to apply its own state.
Also, the only relevant property seems to be _NET_WM_STATE_FULLSCREEN,
since the others are less relevant to the role of the frames client,
and get applied to the MetaWindow as a whole, instead.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2712
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2961>
After this got changed from gdk_x11_get_xatom_name() to XGetAtomName(),
this no longer returns a const char* and it now also needs to be freed.
Fixes: e66f4396e ("x11: Avoid GDK API in X11 selections")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2957>
This was pointlessly being converted between atom and string and back,
which with the switch from gdk_x11_get_xatom_name() to XGetAtomName()
also introduced a leak for every XGetAtomName() call.
Fixes: e66f4396e ("x11: Avoid GDK API in X11 selections")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2957>
The private format and type member variables were not being used by any
of the callers, so they can simply be removed. This also fixes a leak of
type which was introduced when switching from gdk_x11_get_xatom_name()
to XGetAtomName().
Fixes: e66f4396e ("x11: Avoid GDK API in X11 selections")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2957>
After this got changed from gdk_x11_get_xatom_name() to XGetAtomName(),
this no longer returns a const char* and it now also needs to be freed.
Fixes: 014cde646 ("wayland: Do not use GDK functions on XDnD implementation")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2957>
We might end up trying to apply a pending state late if it was delayed
by DMA buffers not being ready. Trying to discard the pending state from
the transaction when dismissing is hard, because we might be applying a
chain of transactions that would disqualify subsequent transactions if a
former one dismisses the popup, so lets just drop what the apply would
otherwise do, if we're not going to use it anyway.
This fixes the following crash:
0) meta_wayland_surface_get_window (surface=0x0)
1) meta_wayland_xdg_popup_apply_state (surface_role=0xf5ee80, pending=0xf662a0)
2) meta_wayland_surface_role_apply_state (surface_role=0xf5ee80, pending=0xf662a0)
3) meta_wayland_surface_apply_state (surface=0xf5e640, state=0xf662a0)
4) meta_wayland_transaction_apply (transaction=0xf56170, first_candidate=0x7fffffffcee8)
5) meta_wayland_transaction_maybe_apply_one (transaction=0xf56170, first_candidate=0x7fffffffcee8)
6) meta_wayland_transaction_maybe_apply (transaction=0xf56170)
7) meta_wayland_transaction_dma_buf_dispatch (buffer=0xf448a0, user_data=0xf56200)
8) meta_wayland_dma_buf_source_dispatch (base=0xf5f140, callback=0x0, user_data=0x0)
9) g_main_dispatch (context=0x41baa0)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2940>
If the popup was dismissed (i.e. has no MetaWindow anymore), it'll also
have no parent surface. With no parent surface, we'd try to fetch a
transaction from NULL and crash, but if we don't try if we were
dismissed, we won't reach here anyway.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2940>
With the frames client, we do no longer handle events for the
frame window inside Mutter. This means we do not get events
"for free" to handle focus on a just clicked frame window.
This results in a background window not ending up focused if
clicked on its frame.
In order to fix this, make the passive button grab extend to
the frame window if a window has one. This brings back
focus-on-click behavior, while treating windows further as
a unitary surface.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2727
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2954>
Pass the timestamp of the frame as the target timestamp of the
record. This makes the rudimentary frame throttling mechanism
inside MetaScreenCastStreamSrc work with the timing variability
that dynamic dispatch times introduced.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2804>
Instead of always, unconditionally scheduling an idle callback for
frame recording, try to record a DMA-BUF only frame, and only if
that's not possible, schedule the idle callback.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2804>
When a stream source subclass asks for a DMA-BUF only frame record,
it is legitimate to return FALSE in do_record_frame() - meaning that
a frame was not recorded - but not return an error - meaning nothing
actually failed.
This avoids spamming the journal with warnings on a legitimate case.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2804>
Add meta_screen_cast_stream_src_maybe_record_frame_with_timestamp()
which operates on arbitrary timestamps; and make the current function
meta_screen_cast_stream_src_maybe_record_frame() just call into the
new variant, passing g_get_monotonic_time() as the timestamp.
This will be useful later we start using the target timestamp of the
frame for screencasting.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2804>
This is not yet used, but next commits will need to assign a frame
to the paint context whenever painting onscreens.
Assigning a frame to the paint context is a one-way operation, and
treats multiple assignments strictly as a programming error.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2804>
This change will export the damaged regions (when available) out
to the pipewire client. This change is currently specific to
virtual streams only (where I was able to test the change) and
maintains the current behavior for other screencast stream types.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2775>
This change allows clipped redraws for offscreen. The net
effect of this change is to preserve the original redraw clip when
possible (rather than overwriting it with the full view redraw) in
the paint context.
This eventually helps in retrieving the fine grained updated regions of
the frame since last redraw and sending it to the pipewire client
(as shown in a subsquent CL).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2775>
Returning FALSE does not indicate an error, but a valid backlight
value of 0. Consumers expect a negative value to indicate no
backlight support, so return -1 in case of error, just like we
already do for invalid values.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2947>
The window tracker is filtering XEvents manually as it only requires a
subset of the ones that Gdk listens to in the root window, and this is
nice, but we were restricting the set a bit too much because due to this
we were not notified when an xsettings manager was available, and thus
in case gsd-xsettings was launched after meta-window-tracker (a normal
scenario under X11), no xsetting was actually applied to the decoration
windows.
As per this, the default settings were used for everything and never
updated, until a restart of the window-tracker.
In order to be able to monitor the XSettings changes at startup, we also
need to select the StructureNotifyMask as gtk always do by default.
See also:
https://gitlab.gnome.org/GNOME/gtk/-/blob/4.11.1/gdk/x11/gdkscreen-x11.c#L947-950Fixes: #2580
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2948>
The anchor position calculations are somewhat unnecessarily complex
based on root coordinates of pointer and frame positions. This requires
tracking both things, and we don't always get it quite right with the
latter (e.g. window repositions, resizes or overshrinks, leaving the
anchor position visually outside the window).
In order to improve this, capture the window-relative coordinates
when starting the window drag, and ensure the window is always repositioned
in that position, relative to its current size.
This avoids these glitches when unmaximizing a window (e.g. dragged from
the bottom through super+button1 press), or moving windows between monitors
with different scales.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2730
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2942>
Since we only track changes to window_drag->anchor_window_pos
during move operations through on_grab_window_size_changed(), this
rectangle is in essence the same than window_drag->initial_window_pos
all the time. Just use that and move away from the anchor_window_pos
rectangle.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2942>
scan_visible_region() scans through each value of a uint8_t array and checks
whether that value is 255. Right now it always checks one value too much
though, resulting in a buffer overflow. Fix that by checking the array
bounds before actually accessing the array.
Found by running gnome-shell with address sanitizer and starting
GIMP.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2856>
Both Clutter and Cogl use g_return(_val)_if_fail() to safeguard
introspected API. Release builds were dropping these checks, which could
result in a much more crashy experience, especially when considering
extensions, but also due to bugs in the shell code itself.
This won't affect any major distro, because they all use "plain" builds.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2930>
Both GRAB_OP_KEYBOARD_MOVING and GRAB_OP_KEYBOARD_RESIZING_* are
defined as GRAB_OP_WINDOW_BASE with FLAG_KEYBOARD set, but the
latter have additional bits set to indicate the direction.
That is, the GRAB_OP_KEYBOARD_MOVING bitmask cannot be used to
differentiate between move- and resize operations. Instead,
check that no direction bits are set.
https://gitlab.gnome.org/GNOME/mutter/-/issues/2684
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2908>
The unknown color space's only purpose is to signal that the current KMS
state has a unknown color space set. It is not one of the color spaces
that can be set. We already only try to set a color space if the default
color space is supported so we should use the default color space as a
fallback instead of the unknown color space.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2693
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2915>
With the move away from GTK3, and the indirect dependency on GTK4
grown in the GNOME Shell side, we've indirectly gotten a small sneaky
behavioral change: The GTK4 library will, right on dlopen, get
DESKTOP_AUTOSTART_ID for itself and delete it from the environment.
This happens before our own X11 session management code is
initialized, which confuses the hell out of it, into thinking
initialization is actually shutdown, gnome-session does not follow
along with this request, which leaves GNOME Shell into a confused
startup state where it never calls SmcSaveYourselfDone() and grinds
startup to a halt until gnome-session decides to move things forward.
In order to fix this, get the DESKTOP_AUTOSTART_ID before we lend
control to GNOME Shell bits and GTK4 is possibly initialized, and
feed it directly to our X11 session manager bits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2906>
We might get told to restore the old monitor configuration by the
monitor configuration prompt, in case the user pressed "revert" or
equivalent. This might be in response to a button press, and those
happen during frame clock dispatch. If we would restore an old
configuration during dispatch, it means we would reconfigure the
monitors including their stage views while dispatching, which means we'd
destroy the frame clock while it's dispatching.
Doing that causes problems, as the frame clock isn't expecting to be
destroyed mid-function. Specifically,
We'd enter
clutter_frame_clock_dispatch (clutter-frame-clock.c:811)
frame_clock_source_dispatch (clutter-frame-clock.c:839)
g_main_dispatch (gmain.c:3454)
g_main_context_dispatch (gmain.c:4172)
g_main_context_iterate.constprop.0 (gmain.c:4248)
g_main_loop_run (gmain.c:4448)
meta_context_run_main_loop (meta-context.c:482)
main (main.c:663)
which would first call
_clutter_process_event (clutter-main.c:920)
_clutter_stage_process_queued_events (clutter-stage.c:757)
handle_frame_clock_before_frame (clutter-stage-view.c:1150)
which would emit e.g. a button event all the way to a button press
handler, which would e.g. deny the new configuration:
restore_previous_config (meta-monitor-manager.c:1931)
confirm_configuration (meta-monitor-manager.c:2866)
meta_monitor_manager_confirm_configuration (meta-monitor-manager.c:2880)
meta_plugin_complete_display_change (meta-plugin.c:172)
That would then regenerate the monitor configuration and stage view
layout, which would destroy the old stage view and frame clock.
meta_stage_native_rebuild_views (meta-stage-native.c:68)
meta_backend_native_update_screen_size (meta-backend-native.c:457)
meta_backend_sync_screen_size (meta-backend.c:266)
meta_backend_monitors_changed (meta-backend.c:337)
meta_monitor_manager_notify_monitors_changed (meta-monitor-manager.c:3595)
meta_monitor_manager_rebuild (meta-monitor-manager.c:3683)
meta_monitor_manager_native_apply_monitors_config (meta-monitor-manager-native.c:343)
meta_monitor_manager_apply_monitors_config (meta-monitor-manager.c:704)
After returning back to the original clutter_frame_clock_dispatch()
frame, various state in the frame clock will be gone and we'd crash.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2901>
Commit 7e9d9c7eb9 added new API to replace GTK for accelerator
parsing.
Unfortunately there is another case in gnome-shell, where we have
to get the label from the logical binding name rather than the
modifier+keysym combination.
Add another small method to cover that use case.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2899>
Normally, mutter implicitly allows a window being shown to take
focus. This is normally desired, except it steals input from
GNOME Shell self. Avoid focusing the just shown window in those
situations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2878>
When a X11 application is started, typically what happens is:
- A startup notification token is created, with a _TIME%d suffix
- The application being spawned receives it through the environment
- (dbus piping, maybe)
- The application replies the startup notification token, and
fetches the timestamp from it
- The application makes a _NET_ACTIVE_WINDOW client message request
with this timestamp
- Mutter handles this client request and activates/focuses the window
Prevent this last step if windows are not interactable (e.g. there is
a compositor grab) and ignore the focus request. This specifically
applies to X11 clients requesting focus themselves, and unlike previous
approaches, doesn't try to prevent focus changes that do come through
interaction with Mutter/GNOME Shell.
This should only break if applications do not observe _NET_ACTIVE_WINDOW
and perform XSetInputFocus on themselves, but in that case the X11
keyboard focus is stolen from our hands already.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2878>
This is the mask that lets us receive among other events the rather
important CreateNotify, that tells us about new winows. This has went by
rather unnoticed except for cases where multiple windows show up very
quickly directly after the frames client spawned, because the drag icon
surface cache eventually already did select that particular mask.
Make things more reliably by explicitly setting the mask for the events
we rely on to function.
This fixes flaky stacking tests that map multiple X11 windows in a row.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2894>
2d8fa26c8e ("core: Pass "frame action" grab operations as an
"unconstrained" grab op") changed the behaviour to treat non-grab
related window moving that has the "user action" flag set to still apply
the "constrain_titlebar_visible" constraint.
The fact that it wasn't applied before was relied upon by some
extensions. While it should arguably exist a better API that for such
extensions to use that have a bit more predictable behavior, until that
is so, restore the old semantics.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2891>
Linear sampling can influence the value of surrounding pixels beyond
the scaled framebuffer extents calculated during stage view rendering,
resulting in flickering graphical artifacts due to unaccounted pixel
changes. This is exhibited in xfreerdp and wlfreerdp at 150% display
scaling.
Fix this by ensuring that all pixels that may be affected by linear
scaling is included in the framebuffer redraw clip by padding the actor
redraw clip.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2771>
We have the drm/InfoFrame encoding and our MetaOutputHdrMetadata
encoding. Check that we can correctly convert between each other by
doing a encode/decode and decode/encode roundtrip and then checking for
equality.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
The existence of the KMS property just means that we can send an
InfoFrame but we also have to make sure the sink actually supports the
metadata type 1 and the selected transfer function.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
Allows to prepare KMS updates to set the color space and HDR Static
Metadata on the output.
For some reason we need ALLOW_MODESET on commits which change the HDR
Static Metadata InfoFrame on AMDGPU. There is no technical reason why
one needs to mode set to send an InfoFrame and the driver should just
manage without ALLOW_MODESET. Until this is resolved in the kernel we
just prepare KMS updates which might mode set.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
The HDR Static Metadata InfoFrame contents are described in CTA-861.3
and the kernel maintains a representation of that in `struct
hdr_metadata_infoframe` in `include/uapi/drm/drm_mode.h`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
The Colorspace property informs the display about the colorimetry of the
content. Only variants supported by the sink are exposed in the
property. The strings representing the color spaces are undocumented but
can be found in the `hdmi_colorspaces` list in
`drivers/gpu/drm/drm_connector.c` in the Linux kernel (v 6.2).
The HDR_OUTPUT_METADATA property is a blob with the InfoFrame content.
We have to query support for the different values in the struct from the
EDID/DisplayID ourselfs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
This adds a new 'experimental-hdr' string property to the MonitorManager
which can be changed from looking glass.
Currently when the string equals 'on', HDR (PQ, Rec2020) will be enabled
on all monitors which support it. In the future support for more
transfer functions and color spaces as well as HDR metadata can be
added.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
The color space and HDR metadata are eventually sent as metadata to the
display. The color space informs the display of the colorimetry of the
frames we produce, the HDR metadata informs the display of the transfer
function and additional mastering display colorimetry and luminance to
guide tone and gamut mapping.
The only color spaces we support right now are the default color space
and Rec bt.2020 which is typically used for HDR content. Other supported
color spaces can be added when needed.
The default color space corresponds to whatever colorimetry the display
has when no further changes are made to the calibration of the display.
The colorimetry is communicated to sources via EDID/DisplayID.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
A Wayland client repeatedly requesting activation of its surface using
the xdg-activation protocol would make mutter constantly update the
cursor.
To avoid needlessly updating the cursor back and forth between busy and
default, add a timeout to delay the update.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2849>
When a client (either Wayland or X11) is started, the window activation
will update the cursor to the "busy" cursor.
Mutter will then set the X11 cursor on the X11 root window to match that
so that X11 applications which do not explicitly set a cursor inherit
from that default (busy) cursor.
Updating the X11 cursor too often can hammer the X11 connection and
cause a deadlock with Xwayland.
Reload the X11 cursor in a later handler to avoid that issue.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2849>
It's a common mistake that links to issues being resolved are not URLs,
and that Part-of was added manually by contributors. Make it clear what
the expectations are regarding this. Formalize on `Fixes:` for fixing
commit messages, and `Closes:` for closing issues while at it. Format
and alias for generating `Fixes:` tags come from mesa.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2847>
We relied on them being valid longer to keep track of used GPUs. If we
don't have the CRTC (or output) we don't have a way to fetch the pointer
to the MetaGpu that drives the associated monitor.
This avoids a crash when trying to fetch said pointer from what would be
the NULL MetaCrtc pointer.
Fixes: 08593ea872 ("onscreen/native: Hold ref to the output and CRTC until detached")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2667
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2887>
Ensure we preserve the fast zero-copy paths in Xwayland fullscreen
windows, instead of maybe rendering the client surface on top of the
frame surface, and providing the latter to the compositor.
To achieve this, additionally synchronize frame state when
recalculating features (e.g. after fullscreen/unfullscreen), and
account for this new condition when creating or destroying frames.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2797>
We no longer need gnome-desktop-3.0, and its dependency on gtk3 is
not desirable. Move to gnome-desktop-4 which actually does not
link to any GTK, conveniently for us.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2864>
We were relying on gdk_cairo_region() to convert a cairo_region_t
into a path ready to fill/stroke in a cairo_t. This is a small
and detached helper that we can do ourselves, so put it together
with all other region helper functions.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2864>
Do the few remaining things that GDK is doing for us:
- Open and close the X11 Display
- Set up a GSource on the Display FD to handle events
- Allocate and free the content of XGenericEventCookie,
to "unroll" the few XInput2 events that Mutter still
does handle.
And remove the GdkDisplay we've so long relied on.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2864>
From reading the comment in the top of the file, not for the first
time. Keep our own error handler and maintain our own list of
failable x11 sequences in MetaX11Display, so we can move away from
GTK's.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2864>
These are done on the backend X11 connection, so it is unclear
what is the interplay through the borrowed global XSetErrorHandler()
that triggers issues for us here.
Anyways, better to be explicit, and use error traps the MetaBackendX11
style, in coherence with the rest of the things happening in that
display.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2864>
This is a X request that may result in errors, so it is better
to have covered by an error trap.
It is thus far not, explicitly at least, which means other less
lenient error traps might not like what happens here. Make the
error trap threeway between backend, x11 and cogl happen less
by chance here.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2864>
When an onscreen is "attached" it means it has an active CRTC and output
it interacts with, e.g. listens to configuration changes to update gamma
and privacy screen state.
MetaOutput and MetaCrtc are rather short lived objects meaning they are
disposed of and regenerated each time the compositor reloads monitor
resources, and while MetaOutput are indirectly kept alive due to the
MetaMonitor holding on to them during reloading, the same does not apply
to MetaCrtc, so to avoid trying to disconnect our signals from
disappeared outputs and CRTCs when we dispatch, hold our own references
to these objects.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2665
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2885>
On exit, explicitly detach the onscreens during disposal. This means no
functional changes, but allows for doing more cleanup on detach that
doesn't need to be repeated on disposal.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2885>
As implemented in colord 1.4.6, cd_icc_load_handle() has three possible
results:
1. success, taking ownership of the profile;
2. failure because cmsGetProfileContextID returns NULL, *not* taking
ownership of the profile;
3. failure in cd_icc_load(), taking ownership of the profile.
The previous commit ensures that we are not in case 2.
In case 3 where cd_icc_load() fails, ownership was already given to
the colord CdIcc object, so it will be freed when the g_autoptr unrefs
the CdIcc, and we must not free it again: that would be a double-free,
potentially resulting in memory corruption.
Resolves: https://gitlab.gnome.org/GNOME/mutter/-/issues/2659
Signed-off-by: Simon McVittie <smcv@debian.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2877>
We want to avoid using too high scales too easily, which started to
happen 2f1dd049bf ("monitor-manager: Rework default scale factor
selection"). Instead of using the closest non-fractional scale, which
effectively is what we'd do, only round upwards if we're closer than
0.25 (25%).
Since there are some wiggle room for scales to make the logical
resolution on the integer pixel grid, make sure to compensate. This
compensation is done by adding an extra 0.2 to scale difference.
For example the following fractional scales will get these corresponding
integer scales:
* 1.25 -> 1.0
* 1.5 -> 1.0
* 1.75 -> 2.0
* 2.0 -> 2.0
* 2.50 -> 2.0
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2880>
Instead of testing headless start using the dummy backend, do so with
the real native backend, and use the drm-mock library instead to emulate
monitors being disconnected at startup.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2821>
This new filter allows test cases to manipulate what the kernel reports,
e.g. mark connected connectors as disconnected to emulate monitors
connecting and disconnecting.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2821>
As part of https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/525
(introduction of transactional KMS API), the logic determining whether a
GPU can have outputs was changed from whether any connectors existed to
whether any connected connectors existed. That effectively meant that we
wouldn't attempt to start at all if there were no monitors connected
while starting up.
This was unintentional, so lets revert back the expected behavior.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2821>
In order to make things more and more asynchronus and to each time we
paint be an isolated event, that can be potentially be applied
individually or together with other updates, make it so that each time
we draw, we use the transient MetaFrameNative (ClutterFrame) instance to
carry a KMS update for us.
For this to work, we also need to restructure how we apply mode sets.
Previously we'd amend the same KMS update each frame during mode set,
then after the last CRTC was composited, we'd apply the update that
contained updates for all CRTC.
Now each CRTC has its own KMS update, and instead we put them in a per
device table, and whenever we finished painting, we'll merge the new
update into any existing one, and then finally once all CRTCs have been
composited, we'll apply an update that contains all the mode sets for all
relevant CRTCs on a device.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2855>
MetaRendererViewNative is a MetaRendererView which contains logic
specific to views of the native backend. It will be used by following
commits.
In the future, per-view logic from MetaRendererNative can be moved to
MetaRendererViewNative where it makes more sense to have it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2855>
Add a tiny library that sabotages errors in drmMode*() API calls. This
will be used to artificially trigger arbitrary errors, e.g. cause the
next commit to fail with EBUSY.
The three mocked methods are added as they will be used in a future
commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
Instead of using the "discarded" page flip callback when the
"discarding" happened during actual immediate processing, communicate
the same via the KMS update feedback.
The "discarded" page flip callback is instead used only for when a
posted page flip is discarded. In the atomic backend, this only happens
on shutdown, while in the simple backend, this also happens when a
asynchronous retry sequence eventually is abandoned.
This allows further improvements making KMS handling fully async.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
At first it was called seal(), but then updates could be amended after
being posted, given a flag. That flag has been removed, so we can go
back to sealing, since it's once again acts more as a seal.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
We test direct client buffer scanout using a TEST_ONLY commit on atomic,
and with various conditions in non-atomic, but if we end up failing to
actually commit despite this, handle the fallout asynchronously. What
this means is that we'll reschedule a new frame immediately.
For this to work, the same scanout buffer needs to be avoided for the
same CRTC. This is done by using the newly added signal on the
CoglScanout object to let the MetaWaylandBuffer object mark the current
buffer as non-working for the onsrceen that it failed on. This allows to
re-try buffers on the same onscreen when new ones are attached.
This queues a full damage, since we consumed the qeued redraw rect. The
redraw rect wasn't lost - it was accumulated to make sure the whole
primary plane was redrawed according to the damage region, whenever we
would end up no longer doing direct scanout, but this accumulation only
works when we're not intentionally stopping to scanout. For now, lets
just damage the whole view, it's just an graceful fallback in response
to an unexpected error anyway.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
Stage view users can schedule updates at ease with
clutter_stage_view_schedule_update(), but couldn't schedule update
"now". Make that easy too by adding
clutter_stage_view_schedule_update_now().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
If we call schedule(), which will schedule an update some time in the
future, and then schedule_now(), we should reschedule the frame clock to
update immediately, and not some time in the future.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
If we get a "ready" page flip feedback, it means the page flip was
symbolic, i.e. not real, e.g. as a result of an update that didn't
change the state of the primary plane. Warn if there is a "next fb"
meaning we expected to have a new buffer that we flipped to.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
This will later be emitted when a scanout failed, e.g. by the not-test commit
failing for some reason, or drmModePageFlip() failing even if the
pre-conditions for scanout in the simple KMS backend passed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
This is intended to be used only for plane assignment, and CRTC like
changes, so that one can e.g. change a cursor plane on a pending update
that changes the primary plane, before it has been committed to KMS.
The kms-updates test overrides the get-state function MetaKmsCrtc. This
is needd to not have the update mechanism not clamp the gamma size to 0,
as vkms reports the gamma length 0. By pretending it's 3, we can test a
simple and small gamma lut is merged correctly when merging updates.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
This looks like a bug: There's no reason why windows which advertise
min-size hints that are exactly the size of the workarea should not be
allowed to maximize, so change the checks here to allow for that.
The commit message of 7f64d6b9 also makes the point that this was not
intended, as it says "larger than".
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2873>
Surfaces belonging to a screen-casted window should always be considered
visible even if they are not visible on any stage view - be it because
they are on a different workspace, minimized or occluded.
Doing this in an optimal fashion is highly complex right now -
interdependent with (and somewhat similar to) ClutterClones. Thus treat
stream-casted surfaces similar to those with clones, with the small
difference that even a fully invisible surface still gets a primary view
- the fastest one. This ensures that clients never refresh too slow for a
screen-cast, at the cost of sometimes refreshing too fast.
The later only happens on certain multi-monitor setups and should thus be
acceptable.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2789>
There is an increasing number of cases where we want the frame callback
logic to run for a stage-view and the complexity needed to avoid these,
combined with the likelyhood of bugs, arguably does not justify the
benefit any more.
Thus unconditionally schedule updates for all stage-views when frame
callbacks are requested.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2789>
Screen-casted windows need to be considered visible in various situations
but existing APIs such as `clutter_actor_is_effectively_on_stage_view()`
don't do so. Add new API that allows checking if a surface belongs to a
screen-casted window for the respective cases.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2789>
The introduction of the META_GRAB_OP_WINDOW_FLAG_UNCONSTRAINED
flag threw off some checks around keyboard-driven resize. This
was partly because there were some == checks that did not account
for that flag maybe being enabled, but also the handling
of META_GRAB_OP_KEYBOARD_RESIZING_UNKNOWN into a definite
resize direction was maybe unsetting that flag. Fix both things
at the same time.
Fixes: 2d8fa26c8e ("core: Pass "frame action" grab operations as an "unconstrained" grab op")
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2629
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2871>
From the scale factors available to it, Mutter will now try to select
the scale factor that makes the UI's size as close as possible to the
size it would be, w/o scaling, on a display at 135 PPI (for mobile
displays) or at 110 PPI (for stationary displays)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2653>
After !2489, the active workspace's MRU list is now used to pick
the next focus window instead of the stack order.
This list is currently only updated on focus, which can lead to
surprising behavior when closing a window after activating its
ShellApp in the shell.
That is because raising a window (as part of shell_app_activate())
will only change the stacking order, so when closing the active
app window, the focus will switch to whatever had focus before the
app was activated, not the app's next window.
In order to allow gnome-shell to address this, add a new
raise_and_make_recent() method that also adjust the MRU order.
https://gitlab.gnome.org/GNOME/mutter/-/issues/2540
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2866>
The pointer to the manager, the peer name and the ID are things that are
always metadata related to a session, so make them properties on the
interface instead of duplicating them. The implementations still need to
keep track of them, but their existance is shared.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2713>
This class is intended to be used as a base class for D-bus interface
implementations that deal with "session" objects, i.e. a D-Bus object
representing a certain session of some kind, e.g. a screen cast session.
It handles things such as hooking up to the D-Bus client watcher,
generates IDs, handles shutdown procedures.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2713>
It's currently not set by anything, and will only be used by
non-abstract implementations of a future D-Bus interface session
manager. When interface implementations gets ported to this new type,
their MetaDbusSession implementations will set this vfunc.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2713>
This means the MetaDbusSession interface takes a more active role
instead of being something that more or less sends signals to the
interface implementor. This will allow better control when using
MetaDbusSession to manage these sessions, instead of their non-abstract
variants.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2713>
Allows for creating LUTs at some fixed size which maintains enough
precision for concatenating or otherwise manipulating the LUT without
having to care about the precision of the hardware.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2861>
If the device supports the atomic API the property based API is used to
write gamma updates and the legacy API is used in the non-atomic case.
The current state is read from the legacy API always though which can be
different from the property API. This commit always uses the correct API
to update the state.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2861>
This API will be used by GNOME Shell to handle X11 events
in the relevant places, as a substitute to gdk_window_add_filter().
It is ATM still a bit ironic, since the Mutter X11 event handler
is itself a GdkFilterFunc, but it may move away from that eventually.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2779>
With detach meaning having the onscreen stop listening on configuration
changes on the corresponding backing mode setting objects. We need to do
this as there is a time between rebuilding the views, and that the new
mode sets are called, where the old onscreen is kept alive, but the
stage view is gone. At this point in time, if privacy screen or gamma
configuration changes, e.g. by the night light temperature changing, the
onscreen would attempt to schedule an update on the now gone stage view.
This commit also renames the "keep onscreen alive" to "detached
onscreens" to more clearly communicate that it's detached onscreens from
their corresponding mode setting objects.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2621
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2863>
Currently, when the rewrite option is passed, the script does not give
much choice on whether changes should be applied or not, it just does
"git comit -a --amend". However, uncrustify is not always entirely
right about the proposed style changes, or it might suggest changes
in distant/unrelated bits in the changed functions. Thus the developer
needs to be given some option.
Change the approach of the --rewrite option, so that it first does
"git add -p", so that individual changes may be decided upon, and
after all the chunks were gone through, uses "git commit --squash"
so that the changes may be reviewed before manually doing
"git rebase --autosquash" to merge the changes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2829>
For the coordinates of pointers or stylii, we translate the ones we store
using the viewport matrix already. For touch events otoh, we store coords
untranslated and translate them later only for event emission.
Let's be consistent here and store the coordinates of touch events
translated, just like we do for pointer events.
This fixes touch window dragging on rotated monitors. MetaWindowDrag calls
clutter_seat_query_state(), which uses those stored coordinates. So in case
of a touch sequence the coords returned by query_state() would be
untranslated.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2859>
In the case where we early-out from meta_window_drag_begin(), the
effective_drag_window might not be set yet. In this case, we might finalize
the object before effective_drag_window is set, leading to a NULL pointer
when accessing window->display in hide_tile_preview().
To avoid that crash, add a check whether the window is set already. If no
window is set, we can just skip hiding the preview anyway.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2858>
This fixes an issue when GLFW tries to change the display resolution
while fullscreen where the application window size doesn't get updated
according to the emulated resolution.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2725>
Streams are generally recoverable by the client and errors may happen
e.g. on negotiation failures. Right now we close the stream and
corresponding session, which is neither necessary nor expected by
clients.
Just disable the stream instead and let clients handle things as they
seem fit. This allows clients to e.g. try several Gstreamer pipelines
with limited caps on a single stream.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2850>
Doing it in dispose means the backend is actively tearing down itself,
meaning various components might or might not be there, depending on how
the tearing down is implemented. Make things a bit more robust by doing
any work that might rely on the backend being there before shutdown is
done in response to the 'prepare-shutdown' signal being emitted by the
backend.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2853>
Meson enforces a separate build dir, so we no longer have to care
about build artifects in the source tree. Same applies for all
the generated crap autotools like to spread around.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2840>
Reading upon the history of this code branch (commits 6891ce95dc
and 7a4c808e43 are most relevant), it seems this code is meant to
synchronize Mutter focus state taking the Xserver state as true.
That is, if Mutter tried to change the focus but something truncated
that action, Mutter focus will be changed to be in sync with the
Xserver again.
This sounds backwards in a Wayland session. Mutter focus should be
the canonical source, and not second-guessed from the current Xserver
focus window. These race conditions might still apply between X11
clients, so make these paths only apply in that case.
An example of this breaking can be reproduced with a Spotify and
Firefox window, moving the focus from the first to the second by
going to the GNOME Shell overview in between, and clicking the
Firefox window from there. The Firefox window will be raised, but
refuse to take focus.
It's unclear what made this an issue recently, perhaps commit
0e6395d932 since the now possibly ignored XI_FocusIn/Out events
affect this accounting of the Xserver focused window. Anyhow it
sounds better to ignore these paths for Wayland/native altogether.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2841>
The intention when the offset request was added to protocol was
that the attach request in a new enough protocol version should
require dx/dy to be zero, but ignore them otherwise.
The current code checks for 0, but then overwrites the existing
dx/dy with it, which renders an earlier wl_surface_offset() call
ineffective.
Fixes: #2622
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2843>
This protocol is intended to let special clients create transient-for
relationships between X11 and Wayland windows. The client that needs
this is xdg-desktop-portal-gnome, which will create e.g. file chooser
Wayland dialogs that should be mapped on top of X11 windows.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2810>
When modal dialogs are attached, and we set the parent/transient-for
after setting the modal type, the attachedness isn't updated. This is
(apparently) not the case for X11 windows, as they go through a
unmanage/manage dance avoiding the issue.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2810>
The script is a list of newline separated command lines that are sent to
the client one by one as if one would have used e.g.
meta_test_client_do().
It doesn't have error handling as it's expected to be used from tests,
and handling errors in tests that never expects to handle errors is
cumbersome.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2810>
The service channel D-Bus interface aims to be a "back door" for
services that needs special casing in Mutter, e.g. have custom private
protocols only meant to be used by that particular service.
There are currently no special casing implemented; only the basic
service channel infrastructure is added. There is a single method on the
interface, that is meant to eventually be used by
xdg-desktop-portal-gnome to open a Wayland connection with a private
protocol needed for the portal backend's rather special window
management needs.
The service channel Wayland client works by allowing one instance of
each "type", where each time needs to be defined to work in parallel. If
a new service client connects, the old one will be disconnected.
MetaWaylandClient's are used to manage the service clients, and are
assigned the service client type.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2810>
One can add a wl_global filter to a wl_display instance, which can be
used to decide what clients should see what globals. This has so far
been used to limit a Xwayland specific protocol extension to only
Xwayland. In order to expand the logic about what globals are filtered
to what clients, introduce a filter manager and port the Xwayland
specific protocol filter to this new manager.
Tests are added, using a new dummy protocol, to ensure that filtering is
working as expected.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2810>
It only tests indirect clients, i.e. not the subprocess part, so far,
but tests explicitly terminating by destroying the MetaWaylandClient
object, as well as the client self terminating and the signal being
emitted.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2810>
This API creates a "client" then later sets up a wl_client and returns a
file descriptor some Wayland client can connect to. It's meant to be
used as a method other than WAYLAND_SOCKET and process launching, e.g.
passing a file descriptor via a D-Bus API.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2810>
There will be two kind of client instances, lets move fields that are
only relevant to the current way of operation in an anonymous struct to
keep things a bit separate.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2810>
On X11, the stage itself is backed by an XWindow, and moving the
input focus elsewhere will bypass any Clutter-level grabs.
This effectively allows newly opened windows to steal the focus
from gnome-shell itself, which is clearly undesirable. To prevent
that, only allow moving the X11 focus to a Window when no grab is
in place, just like commit 50e89e376 did for the stage focus.
But particularly the updating of x11_display->focus_xwindow is not
prevented. Since it's more consistent to the MetaDisplay/MetaX11Display
dual focus tracking and across Wayland/X11 backends, ensure the X11
input focus is actually set on the last focus Window after the
grabs are gone and windows became interactable again.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2832>
This patch unfortunately results in situations where it is intended
that the focus change happens while a grab is present (e.g. Alt+tab
popup), resulting in confused focus state.
This commit is reverted in order to try a similar approach at a
different level.
This reverts commit 7531669b4f.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2832>
We used it to retrieve a Display, and convert between Atoms and
strings. We can just use the MetaX11Display's Display (It's the
same than GDK's anyways) and use XInternAtom/XGetAtomName for
these conversions.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2836>
We didn't always set an implementation, when the foreign toplevel wasn't
found, and when the importer tried to set the parent-child relationship,
the implementation was missing and we'd crash in wl_closure_invoke() in
libwayland-server.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2834>
Windows that are decorated may get configure requests before
the frames client created a corresponding frame window and Mutter
reparented the window.
Since the configure request results in the buffer size being
used to update the window size and the window does not have a
buffer yet, these requests could mistakenly result in the client
window being given a minimal size.
In these situations, do not use the buffer size but the given
size. The window still has to undergo frame creation and
reparenting before being shown for the first time.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2588
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2605
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2808>
This used to be implicitly done by popups using a META_GRAB_OP_WAYLAND_POPUP
MetaDisplay grab. Since commit a8cd488c6f Wayland popups no longer do that,
so the keyboard focus was simply unset if a popup was destroyed while having
the keyboard focus.
Trigger a full input focus sync, so the correct MetaWaylandKeyboard focus
surface is looked up from the focused MetaWindow.
Fixes: a8cd488c6f - wayland: Drop redundant MetaDisplay grab op
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2833>
On one hand, this used to be handled generically in all the paths that
changed the MetaWaylandPointer focus surface, induced by user interaction
or not.
On the other hand, just listening for crossing events is not sufficient
since those also do happen programmatically. We must only listen to
crossing events that have a physical source device, meaning this was
created through user interaction.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/888
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2828>
For motion-induced crossing events, this will be the device that generated
the motion. For code-induced crossing events (e.g. grabs or actors disappearing)
this will be none.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2828>
On Wayland sessions, this handling is unnecessary and even prone
to confusion (e.g. crossing serials are only ignored in X11-exclusive
paths, so this handling competes directly with that in MetaWaylandPointer).
Avoid it entirely there, so MetaWaylandPointer can figure out
sloppy/mouse mode focus for all Wayland/Xwayland surfaces.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2828>
In GTK this is only used for GTK clipboard/DnD selections, and
finding out whether there is a compositing manager in charge.
In Mutter, we manage our own clipboard/DnD selections, and don't
perform any rendering through GTK in the Mutter process.
So there's no special reason to let these events go through GTK,
and (related to xwayland-on-demand?) there may be race conditions
in the handling of the second feature.
There's a chance this race condition may be in Mutter, but it
does not sound worth to chase this race condition when we can
let GTK ignore these events. And it does not make sense to "fix"
gtk3 for this Mutter-only condition, when we intend to eventually
avoid it.
So, take the easy path and ignore these events.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2617
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2831>
Marking the the depth/stencil as discarded before swapping buffers for
the screen signals the GPU that we don't need to keep them around for
the future.
This helps performance by reducing memory bandwidth usage in some GPUs
which may optimize to not write those buffers back to memory at all
after rendering, when they would just be cleared right after that
anyway.
It is not necessary to mark buffers as discarded after swapping buffers.
This should have no effect according to the spec (since that is going to
be followed by new rendering commands which make the buffer valid again)
and removing that has shown no impact in performance tests.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2091>
The first monitor in stacking tests is the primary monitor but that
doesn't have to stay this way forever. Instead of special casing the
name "primary" to refer to whatever monitor happens to be the primary
monitor, we add an `assert_primary_monitor` command to verify that the
monitor that should be the primary monitor actually is.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2748>
New add_monitor command for adding secondary monitors. Support setting
the workspaces-only-on-primary preference.
The stacking test tests the focus and stacking for multiple monitors
with workspaces-only-on-primary=true. The default_focus changes
previously broke this.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2748>
The X-GNOME-Bugzilla-* entries in the desktop file were for use by
bug-buddy, a GNOME 2 technology that's been gone for over a decade.
These entries are obsolete and the desktop file can be removed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2621>
bind_output() creates output interface resource, but does not
set implementation for it when wayland_output->monitor is NULL.
However, when the wayland library is running wl_closure_invoke(),
it expects the implementation to be non-NULL, and if not, it just
segfaults mutter by NULL pointer dereference.
This commit tries to address this issue by setting an implementation
when wayland_output->monitor is NULL. This could help prevent crash
when resuming from suspend or hotplugging displays.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2570
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2827>
The order of dependencies influences the order of -L arguments to gcc/ld,
we should put our private library first, so that introspection prefers
looking up libraries in private paths than public ones.
This could bring problems in API updates of the libmutter-test library,
since introspection would still prefer the old installed one.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2826>
ClutterActions now no longer receive their events via
clutter_actor_event(), instead they get special treatment by the stage
now. Make the MetaGestureTracker work with this and stop emitting events
directly to Clutter via clutter_actor_event(), but instead let them get
through to Clutter (but still not to Wayland).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
Since the last commit, ClutterStage automatically cancels an implicit
grab (including all its ClutterActions) when a conflicting ClutterGrab
appears.
This means we no longer have to look out for GRAB_NOTIFY crossings in
ClutterGestureAction and can instead depend on the sequence_cancelled()
vfunc for this.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
A ClutterGrab takes precedence over implicit grabs, so when one happens,
let's check which part of the implicit grab tree is inside the new
ClutterGrab. Cancel and remove the parts which aren't, and if nothing
is in there anymore, cancel the whole implicit grab.
Emitting crossing events correctly here is getting quite tricky:
- When the implicit grab didn't get cancelled by the ClutterGrab, we
simply want to emit all GRAB_NOTIFY crossings to the implicit grab, as
we do with all other crossings.
- When the implicit grab did get cancelled and the new ClutterGrab wants
to emit ENTER crossings, we want those to be emitted to the actual
targets, so cancel the implicit grab before emission.
- In the last case where the implicit grab did get cancelled and the new
ClutterGrab wants to emit LEAVE crossings, those should be emitted to
the implicit grab again, so we cancel the grab only after the emission
of those.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
Now that we have two kinds of grabs, the intricacies of event delivery
got slightly more complicated. So this seems like a good point to
introduce a new GRABS debug flag that gives an overview of which grabs
are currently in effect.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
We're almost there, everything is in place to notify ClutterActions
about a sequence getting pulled away under its feet.
The only thing that's missing is the actual notification to actions now,
so let's do that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
Another baby step just like the last commit: This commit takes care of
the opposite case: An action handling a sequence event stops further
emission of events to actors.
Since sequences remain around for longer than the context of just a
single event, it makes sense to provide a way to "claim" those sequences
even when outside of event handling context, so introduce API for that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
As soon as any event of a sequence is handles/stopped during emission,
all actors and actions that would have gotten to see it afterwards have
a big problem: If that event was a TOUCH_END event, the actor/action is
forever going to think that this touch is still active.
For ClutterActions, we're going to handle this by introducing a way to
send them a notification when stuff like this happens.
As a baby step towards all that, make event emission exclusive to actors
as soon as any actor stopped an event.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
We'll soon introduce a new gesture tracking framework which heavily
depends on ClutterActions seeing all events of a sequence. For this to
work, a larger change to event delivery is needed: Implicit grabbing of
all events for button and touch press->motion->release sequences to
ensure ClutterActions continue receiving events for the whole sequence.
This commit takes care of that: At the start of an event sequence we
collect all the event-handling actors and actions to a GArray that lives
in the PointerDeviceEntry, and then deliver all events belonging to
that sequence to the same actors/actions until the sequence ends.
To avoid events getting pulled from under our feet when mutters event
filter returns CLUTTER_EVENT_STOP, this also introduces private API
(maybe_lost_implicit_grab()) on ClutterStage so that we can't end up
with stale sequences.
Note that this also slightly changes behavior when it comes to event
delivery to actions: Because we now store actions separated from their
actors, any action returning CLUTTER_EVENT_STOP now stops event
propagation immediately. That was different before, where we'd emit
events to all actions of the actor and only then stop propagation.
Note that this isn't handling ClutterGrabs correctly right now,
this will be a little tricky, so we'll take care of that in a future
commit.
To handle actors getting destroyed or unmapped during a grab, listen to
notify::grab on the deepmost actor in the implicit grab tree. This gives
us a notification when any actor inside the tree goes unmapped.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
A fairly small refactor, move the emission of events to actions from
clutter_actor_event() to stage level.
We do this because in the future we'll need to know on stage level
whether events were handled by an actor or by an action.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
_clutter_actor_handle_event() currently allocates a new GPtrArray on the
heap for every single event emission, let's avoid this by keeping an
array around in ClutterStage and reusing that.
This is moving the last few bits of event emission into ClutterStage,
which will be useful when we introduce implicit grabbing in subsequent
commits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
There's no real reason to keep those events exclusive to the stage, some
actors or actions might want to get notified about proximity events too,
so propagate them like any other event.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
Right now and due to loads of refactorings lately, the event emission
paths are a bit cluttered (ha ha ha) around in Clutter. For example the
event target actor gets set in clutter-main.c, but event emission is
actually managed by ClutterStage these days.
Since we'll introduce implicit grabbing of touch/button-press sequences
soon, let's shuffle things around a bit to make that easier:
Move event emission to the stage, it now gets a ClutterEvent without any
extra context like the target actor from clutter-main. The stage then
looks up the target actor itself and emits the event to the appropriate
actors in the scenegraph. A special path is introduced for emitting
crossing events, because here the event-receiving actors don't follow
the "capture+bubble from pointer actor to grab actor" rule.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
Crossing events should never be stopped during event emission. We
already have a check that enforces this in clutter_actor_event(), but
ClutterActions still sometimes try to stop crossing events from
propagating.
Improve that situation and return CLUTTER_EVENT_PROPAGATE when handling
crossings in ClutterActions, too.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2342>
The previous logic tried to keep the position of the top left corner of
the window relative to the top left corner of the monitor. This allowed
the window to move out of the target monitor. This change keeps the
proportions of the distance between the window and the monitor borders
instead if possible. Otherwise it keeps the relative position of the
center of the window clamped to [0,1] to make sure the window lands on
the right output.
This also slightly changes what monitor is considered to be on: the
monitor which contains the center of the window and, if the center is on
no monitor, the monitor wich overlaps the most with the window.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2591>
This partly reverts f9857cb8 but leaves an exception for cursor
surfaces in place, as some apps/toolkits will likely not get updated
anytime soon to ensure cursor themes comply with the Wayland spec.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2815>
So we can remove the additional `next_fb` and `current_fb` pointers from
`MetaOnscreenNativeSecondaryGpuState`.
Some non-scanout buffers also need to be held in the case of GL blitting
which completes in the background. Those are referenced from the scanout
buffers themselves to ensure the source buffers live just as long.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2087>
There's still a possibility that some events remain within the
`ClutterMainContext` when it's being unref-ed for the last time (as seen
on asan logs). Make sure they get freed by using
`g_async_queue_new_full()` and specifying the appropriate destroy
function.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2817>
Calculating a timestamp from the past distorts the dispatch lateness
calculation, leading to an inflated max_render_time, which again
increases the likelyhood of next_update_time being in the past.
Fixes 99850f4645
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2819>
As with GAMMA_LUT, track whether privacy screen state has been pushed to
KMS in the onscreen. This leaves MetaOutput and MetaCrtc to be about
configuration, and not application.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2814>
As with CRTC GAMMA_LUT, we're moving towards making the entity managing
KMS updates aware if there are any changes to be made, and whether KMS
updates are actually needed or not, and for privacy screen changes, this
means we need to communicate whether the privacy screen state is valid
or not. This allows the caller to create any needed MetaKmsUpdate.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2814>
We're moving towards making the entity managing KMS updates aware if
there are any changes to be made, and whether KMS updates are actually
needed or not, and for GAMMA_LUT changes, this means we need to
communicate whether the GAMMA_LUT state is valid or not. This allows the
caller to create any needed MetaKmsUpdate.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2814>
We may fall through these paths on --nested too, resulting in us poking the
wrong internals from the wrong MetaRenderer subclass. Fixes launching of
clients using wl_drm in --nested.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2818>
Running each stacking test as a separate installed-test is analogous to
what was done for build-time tests in c6d1cf4a (!442) and should make it
easier to track regressions, by being able to see whether a regression
is specific to one .metatest script or applies to more than one.
Signed-off-by: Simon McVittie <smcv@debian.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2773>
While completely untested, at least this makes it work "in theory"
again. Before it'd listen to signals on the stage, but have an incorrect
type signature to handle the test paint procedures, meaning it'd
probably crash or cause memory corruptions.
What was needed was a signal which in the callback the test could call
some cogl functions to paint on the framebuffer. While there is no such
signal on the stage, and the ClutterActor::paint signal (which they
probably used in the past) is long gone, lets add a "test actor" that is
just a wrapper that adds that paint signal with a paint context.
The tests that need it are changed to add this actor to the stage, and
to listen to the paint signal on the actor instead of incorrectly
listening on stage signals.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2807>
At least indirectly, this is set as object qdata while the
window drag is ongoing, and reset/reconstructed if needed.
Consequently, this edge data does not need to be stored in
the MetaDisplay struct anymore.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
Even though the data is still stored in the display, add a "high
level" meta_window_drag_update_edges() call, so that the cached
edges may be updated while a window drag operation is ongoing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
This is a public API change. Add device/sequence parameters to this
operation, so that window dragging and resizing can stick to one
set of pointing events of them all.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
Since MetaWindowDrag took a lot of this code to handle window drags
internally with less interactions with the rest of the stack, this
code in display/window/keybindings is unused.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
Flip the switch in using MetaWindowDrag, leaving display grab
ops and a bunch other code unused. Some places checked the grab op
and/or window in complex ways, others just checked for grab existence
and should now look for clutter ones, and others already were already
doing this in addition.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
This helper object (and the whole window drag operation) will be
requested to the compositor instead of created directly, and only
one of those can exist at a time, so the compositor will also
safeguard that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
Since SSD X11 windows require synchronization between frame and client
windows on resizes, updates do not always happen immediately but in
control of external factors (i.e. when both windows become to have
a coherent size).
This method will be used to update the window position between
resize/sync operations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
This compositor-side object will single-handedly drive a window
drag operation. Currently, this largely copies meta_display_begin_grab_op
and meta_display_end_grab_op, except grabbing is done through a
ClutterGrab instead of direct meta_backend_grab_device() calls. This
also means that the switch from passive to active keyboard grabs is
handled differently.
Currently, this object is dormant. It requires moving more code from
other places to become a fully functional replacement.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
We only allow partial grabs in the case of a keyboard-type MetaGrabOp
happening while the pointer cannot be grabbed. In that case, it's not
a big stretch to unconditionally ungrab the pointer device at the time
of undoing the grab, as it will be always ineffective (not even implicit
grabs on frame windows can happen now, inside Mutter).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
This is no longer necessary, since the SSD frames are no longer
part of Mutter process, so it is not the MetaX11Display connection
which holds the implicit grab when a mouse button is pressed over
a window frame (say, to start a drag).
As the SSD frames client communicates the same way than CSD windows
for window operations, it is also expected to undo its implicit
grab before requesting a window move/resize operation.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
The final effect of this boolean can now be expressed through the
META_GRAB_OP_WINDOW_FLAG_UNCONSTRAINED flag to MetaGrabOp. Use that
in the relevant places, and drop the argument.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
Now that it is called from a single place, there's a few arguments
that are unnecessary:
- button and modifiers are unused
- already_grabbed was originally added to handle grab transitions between
window menus (GtkMenus, back in the day) with display grabs. It's no
longer necessary now
- frame_action can be passed through the META_GRAB_OP_WINDOW_FLAG UNCONSTRAINED
flag
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
Leave meta_window_begin_grab_op() as the only public API to initiate
a display grab. There's no longer grab operations that don't attain
windows, and ending these grabs usually happen through user interaction
when the right circumstances happen.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
There is no longer reason to call meta_display_begin_grab_op() except
for window grab operations, and meta_window_begin_grab_op() is a
perfectly fine entry point for all window grab operations.
Move away from meta_display_begin_grab_op().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
Currently, it is thought out to be called with META_GRAB_OP_KEYBOARD*
grab op parameters. Make it more generic so it can also be called for
pointer operations (avoiding pointer warping in that situation).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
Unlike the comment suggests, this piece of event handling manages
the ungrabbing of a window on button press in the following 2
conditions:
- If a keyboard grab operation was triggered, the window does
additionally follow the pointer, and first button press ends
the grab.
- If a button-press grab is ongoing on the window, but more buttons
are pressed.
We can simplify this to just happen every time a button press event
is received while a window grab op is ongoing. The only case where
this might diverge a bit is same button presses from different
pointer devices, and it's not a big stretch to also undo the grab
in that situation.
This also happens to make the "button" argument in
meta_display_begin_grab_op() completely unused.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
The frame_action boolean is only used by constraints.c code, in order to
determine whether a moving window should be able to move past the top
bar or not.
We can avoid the special casing by passing this information as a
META_GRAB_OP_WINDOW_FLAG_UNCONSTRAINED flag passed with the grab op.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
This is no longer necessary to prevent the bits we wanted to be
prevented by the presence of this grab. We can drop this, and
let it work through the MetaWaylandPointerGrab interface.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
The whole reason for META_GRAB_OP_WAYLAND_POPUP to exist is to
avoid windows from being activatable/movable/resizable when a
grabbing xdg_popup is active.
Use the meta_display_is_grabbed() method which can tell this
from existing MetaWaylandCompositor grabs, so that this remains
true after dropping META_GRAB_OP_WAYLAND_POPUP.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
Make this public API check just return a boolean about whether
there is an existing grab, instead of exposing MetaGrabOp.
It is desirable to avoid exposing details like
META_GRAB_OP_WAYLAND_POPUP, so that MetaDisplay and wayland
grabs can port to ClutterGrab at their own pace, but also
this further information is unused.
This is likely to be temporary API anyways, after both
MetaDisplay and wayland grabs port to Clutter, it will be
possible to check the ClutterStage for all of them.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
Rewrite this codepath so it handles the grab ops that it cares
about, and ignores the rest. This way the code works despite
possible future modifications to MetaGrabOp (e.g.
META_GRAB_OP_WAYLAND_POPUP removal).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
This piece of event handling only applies on windows receiving events while
the display is ungrabbed (i.e. for raising it, or beginning a move/resize
operation).
Move the checks on the current grab operation outside of window.c and into
events.c, so all checks about the current grab operation move closer to the
main event handler.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2683>
When sysprof-4 and libsysprof-capture-4 are installed into different
prefixes, such as with Nix package manager, the D-Bus interfaces
are likely not discoverable from the latter package.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2572>
Currently subschemas in org.gnome.mutter are marked as child schemas,
but subschemas in org.gnome.mutter.wayland are not. There's no harm
with that, it's just nice to be consistent, and having schemas show
up in `gsettings list-children` is a nice touch.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2809>
The GQueue node for transactions are inlined in the transaction struct,
meaning we should never let the GQueue API free the node itself, as that
actuall frees the transaction itself.
We did this during tear down if there were left-over transactions,
meaning we ended up with use-after-free issues after having popped
transactions from the queue.
Fix this by just popping the link itself, which won't attempt to free
it. It is effectively freed when freeing the transaction itself so we
won't leak any memory.
Fixes: 56260e3e07
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2805>
During grabs, it is expected that the X11 focus does not correspond
to the display's focus window, as focus should be on the stage's
XWindow instead.
This still messes up the keyboard focus even after we stopped moving
the X11 focus, because we end up with a presumed X11 focus window
of None, and as a result the stage is considered unfocused.
https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/5932
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2776>
When the pointer crosses monitors, we account for a single motion event
resulting in the pointer moving across more than 2 monitors, in order
to correctly account each monitor scale and the distance traversed
across each monitor in the resulting relative motion vector.
However, memory on the direction is kept short, each iteration to
find the target view just remembers the direction it came from. This
brings a pathological case with 4 monitors with the same resolution
in a 2x2 grid, and a motion vector that crosses monitors at the
intersection of all 4 in a perfect diagonal. (Say, monitors are
all 1920x1080 and pointer moves from 1920,1080 to 1919,1079).
In that case, the intersection point at the crossing between 4
monitors (say, 1920,1080) will be considered to intersect with 2
edges of each view. Since there is always at least 2 directions to
try, the loop will always find the direction other than the one
it came from, and as a result endlessly jump across all 4 possible
choices.
In order to fix this, consider only the global v/h directions,
we already know if the pointer moves left/right or up/down, so
only consider those directions to jump across monitors.
For the case at hand, this will result in three monitors visited,
(either bottomright/bottomleft/topleft, or bottomright/topright/topleft)
with a total distance of 0,0 in the middle one, effectively
resulting in a correct diagonal motion.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2598
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2803>
Refactor code so that variables don't depend the on motion line
content, but the other way around. This makes it clearer what each
vector means.
This has no functional changes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2803>
Some tests expect warnings to be logged, and handle that using
g_test_expect_message(). However, if debug topics are enabled, this
causes g_logv() to expect expected messages to also contain entries with
the debug level 'message' or higher to be listed in the expected message
list. Since meta_topic() always logged using g_message(), enabling debug
topics caused any test that used g_test_expect_message() and had debug
logging somewhere along the code path to fail.
Fix this by changing the log level of meta_topic() to 'debug' if we're
in a test. This doesn't mean they won't be visible, they still will
since debug log entries are printed by default during testing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2800>
That means before-update, prepare-paint, before-paint, paint-view, after-paint,
after-update. While yet to be used, it will be used as a transient frame
book keeping object, to maintain object and state that is only valid
during a frame dispatch.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2795>
It was missing a cairo_region_t. This also needs adapting the test case,
since prior to this, we didn't actually bump the paint counter when
painting.
When a scanout test isn't waiting to go from compositing to scanout, but
from scanout to compositing, we should not early out when we actually
composited, since that's what we're expecting to see.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2795>
Fix a silly copy-paste mistake. Since `GObject` is the parent class,
chaining up to `dispose()` from within your `finalize()`
implementation just leads to a little memory leak and nothing worse.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2799>
We might end up with a NULL opaque_region here in some circumstances
(client deleted _NET_WM_OPAQUE_REGION, or passed invalid data or a
region with 0 rectangles), account for that when freeing the variable.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2758>
These frames client will use a visual with alpha information, and
report the opaque frame shapes through the _NET_WM_OPAQUE_REGION
window property. We can use this information in the Mutter side
for accurate opaque shapes, despite X11 windows with frames now
being seen as possibly transparent.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2758>
Since the windows created by the frames client will have a RGBA visual, we
no longer can perform simple tests about whether the window is opaque. For
that, we will need to additionally know whether the client-side window has
a visual with an alpha channel.
Reviewed-by: Michel Dänzer <mdaenzer@redhat.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2758>
This does nothing wrt making race conditions shorter in the
X11 window manager switch case, but is a nice to have in order
to ensure an orderly shutdown of X11 stuff.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2796>
Restarting a X11 window manager is a busy process, trying to leniently
quit the main loop may result in old and new instances each having a
frames client up and running, and the window handover to be less clean
than it should due to the frames client that is about to exit still
being able to react to the batch of events resulting from the window
manager switch that is already undergoing.
In order to avoid extending this transition period any long, make
the frames client exit() the process immediately when SIGTERM is
gotten from the parent process.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2796>
Writing to fields (in this case the MetaColorDevice::pending_state) in
response to an asynchronous operation that was cancelled means we'll
write to an arbitrary memory location, potentially causing segmentation
faults or memory corruption.
Avoid these segfaults or memory corruption by only updating state if we
weren't cancelled. Also avoid trying to dereference the device pointer
if we're cancelled.
The memory corruption due to this has been causing test flakyness in the
monitor unit tests due, which should now hopefully be fixed.
Fixes: 19837796fe
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2794>
I hit this rare error running the "x11" test from the suite locally:
(mutter:194027): Gdk-ERROR **: 18:21:52.525: The program 'mutter' received an X Window System error.
This probably reflects a bug in the program.
The error was 'BadDrawable (invalid Pixmap or Window parameter)'.
(Details: serial 663 error_code 9 request_code 143 (DAMAGE) minor_code 1)
(Note to programmers: normally, X errors are reported asynchronously;
that is, you will receive the error a while after causing it.
To debug your program, run it with the GDK_SYNCHRONIZE environment
variable to change this behavior. You can then get a meaningful
backtrace from your debugger if you break on the gdk_x_error() function.)
The only call from the Damage extension in use by Mutter that could
return BadDrawable is XDamageCreate(), and it's likely to be this
call. Wrap this X11 in an error trap, in order to catch possible
failures.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2793>
Flushing the input thread might implicitly iterate the mainloop, and thus
update the stage while still inside the clutter_test_flush_input() call.
This means the stage update has already happened when we call
wait_stage_updated(), and that's why we call clutter_stage_schedule_update()
there currently.
This clutter_stage_schedule_update() call is not necessary though, instead
we can flush the input thread from inside wait_stage_updated() after
setting was_updated to FALSE.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2792>
If the window is unmapped or otherwise unmanaged while still existing,
we would fail to let the frames client follow up in destroying the
frame for the window.
Delete the _MUTTER_NEEDS_FRAME property, so that the frames client
can react to meta_window_destroy_frame(), this avoids stale invisible
frame windows for clients that simply unmap windows to reuse them
later.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2791>
Recent versions of Xwayland can allow or disallow X11 clients from
different endianess to connect.
Add a setting to configure this feature from mutter, who spawns
Xwayland.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2785>
This define was dropped by commit 0e8aaebc00 (xwayland: Make
XSetIOErrorExitHandler() mandatory), but some #ifdef checks were
brought back by commit 36f30341ac (wayland: Add a prepare-shutdown
signal).
Since there's no define anymore in config.h, these pieces of code
were unintentionally disabled, and a meta_get_display() call be
also left over. Remove the ifdefs and update the code to build
again.
Fixes: 36f30341ac - wayland: Add a prepare-shutdown signal
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2786>
This XChangeWindowAttributes call was never surrounded by an error trap
and was not really expected to fail with BadWindow since the frame window
would be owned by Mutter itself.
This however is no longer true, and we might be getting a BadWindow from
the frame window given the right timing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2745>
Commit 4e0ffba5c attempted to fix initialization of keyboard a11y,
but mousekeys do attempt to create a virtual input device at a
time that it is too early to try to create one.
Defer this operation until keyboard devices are added, so that
we are ensured to already have the seat input thread set up.
Fixes: 4e0ffba5c - backends/native: Initialize keyboard a11y on startup
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2778>
Quoting Carlos:
The META_PRIORITY_EVENTS ± 1 happening below are in order to set these idles
and timeouts in a priority that is relative to the literal GDK event priority,
making those diverge is a likely way to sneakily break things.
But that's unlikely to happen, and decoupling mutter from GTK further
should make it moot, so perhaps it's alright after all.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2407>
Clutter has an API to get the text direction but used to depend
on gtk3's translation domain. In order to avoid broken i18n
in case gtk3 is not installed, move the transtalable string to
clutter itself.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2407>
Now that dynamic max render time uses a new algorithm and takes dispatch
lateness into account, this seems worth a shot. We'll see how it works
out in the wild.
The net result compared to before these changes is still slightly higher
(by ~0.5 ms) minimum latency for me, as measured by
weston-presentation-shm. It should be less vulnerable to frame drops
though.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2500>
Store only two values per kind of duration: The short term and long term
maximum.
The short term maximum is updated in each frame clock dispatch. The long
term maximum is updated at most once per second: If the short term
maximum is higher, the long term maximum is updated to match it.
Otherwise, a fraction of the delta between the two maxima is subtracted
from the long term maximum.
Compared to the previous algorithm:
* The calculcations are simpler.
* The calculated max render time has a slow exponential drop-off (by at
most a few milliseconds every second) instead of potentially abruptly
dropping after as few as 16 frames.
This should fix https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/4830
since the short term maximum should always include a sample from the
clock's second tick.
v2:
* Use divisor 2 instead of 4.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2500>
Dispatch lateness is the difference between when we wanted frame clock
dispatch to run and when it actually started running. This can be up to
1ms even under normal circumstances due to process scheduling
granularity, or even higher under load.
This keeps track of dispatch lateness of the last 16 frame clock
dispatches, and incorporates the maximum into the dynamic render time
estimate.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2500>
There are two tests; one checks that clearing with a color that cannot
be represented using 8 bits per channel doesn't loose precision when
painted, then read back using glReadPixels(). Would the texture backing
store have 8 bits per channel instead of 10, we'd get a different value
back.
The other test checks that painting from one fbo to another also doesn't
loose that precision.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2461>
OpenGL requires more hand holding in the driver regarding what pixel
memory layouts can be written when calling glReadPixels(), compared to
GLES2. Lets move the details of this logic to the corresponding
backends, so in the future, the GLES2 backend can be adapted to handle
more formats, without placing that logic in the generic layer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2461>
COGL_PIXEL_FORMAT_ABGR_2101010 is defined to mean the 2 A bits are
placed in a 32 bit unsigned integer on the bits with highest
significance, followed by B on the following 10 bits, and so on, until R
on the 10 least significant bits.
UNSIGNED_INT_2_10_10_10_REV_EXT is defined to represent color channels
as
```
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
-------------------------------------------------------------------------------------
| a | b | g | r |
-------------------------------------------------------------------------------------
```
As can be seen, this matches COGL_PIXEL_FORMAT_ABGR_2101010, meaning
that's the format we can directly read and write.
In Cogl, when finding the GL formats, we get the tuple with the GL
format given the format we pass, but we also get returned "required
format" (CoglPixelFormat). This required format represents the format
that is required when reading actual pixels from GLES. In GLES, the
above mentioned format is the only one supported by the
EXT_texture_type_2_10_10_10_REV extension, thus for other types, we need
to do the CPU side conversion ourselves. To achieve this, correctly
return COGL_PIXEL_FORMAT_ABGR_2101010 as the required format.
The internal format should also be GL_RGB10_A2, not GL_RGBA.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2461>
Commit bf84b24 created meta-enums.h but it's pretty empty so far, the
vast majority of enum definitions is still in common.h.
Move the Meta enum definitions to meta-enums.h as one would expect them
to be found.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2467>
This replaces the v1 implementation, which is now renamed to
legacy-xdg-foreign. Both implementations use the same data structures
internally, so that protocol version mismatches between
the importer client and exporter client don't fail.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2770>
Sysprof's build options have changed recently. This both bumps the sysprof
version and updates the configuration options for the subproject.
If now is not a good time to bump this, that is totally fine, but I wanted
to give you a MR with the necessary changes all in one.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2428>
Things like meta_compositor_destroy() and meta_compositor_add_window()
isn't intended to be used externally, and if they was, things would
probably fall apart rather quickly.
MetaCompositor also isn't introspected, meaning things that technically
belong to the compositing parts isn't easily available via some object,
but much take detours via other objects like MetaDisplay.
So move the API intended for internal usage to compositor-private.h, and
leave API that is meant to be expose in the public compositor.h.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2718>
The "later" API is used to queue actions in relation to compositing,
thus is owned by the MetaCompositor instance. Make users of this
functionality get MetaLaters instance from the compositor, and stop
using the global meta_later() API.
display: Use non-singleton MetaLater API
tests: Use non-singleton MetaLater API
meta/common: Make docs refer to context aware MetaLater API
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2718>
This means we can eliminate the use of scattered singletons that isn't
added by the tests or the test framework itself.
tests: Don't get backend from old singleton getter
Either use the ownership chain, or the explicit test context instance
pointer.
tests/wayland: Pass context to test client constructor
So that we can get the Wayland compositor directly from the context.
tests: Don't get display from singleton
tests/client: Make test client carry a context pointer
tests/runner: Have test cases carry a context pointer
tests/wayland/test-driver: Get backend from context
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2718>
Instead of passing around state using GINT_TO_POINTER() pass around a
state struct that also carries a pointer to the context. This allows
avoiding using old singletons for getting a window list.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2718>
On the path towards clear ownership chains and always using them to find
other components, do the same for X11 client support paths too.
x11-display: Don't get backend from signleton
x11/selection: Don't get display from singleton
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2718>
The API has no concept of user data, and requires the user to some how
get an instance without context, i.e. via static globals. Limit this to
the file where this is needed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2718>
As with the backend commit, this means all objects can reach the
MetaContext by walking up the chain, thus can e.g. get the backend from
the context, instead of the global singleton.
This also is a squashed commit containing:
compositor: Get backend via the context
The MetaCompositor instance is owned by MetaDisplay, which is owned by
MetaContext. Get the backend via that chain of ownership.
dnd: Don't get backend from singleton
window-actor: Don't get backend from singleton
dnd: Don't get Wayland compositor via singleton
background: Don't get the monitor manager from the singleton
plugins: Don't get backend from singleton
This applies to MetaPlugin, it's manager class, and the default plugin.
feedback-actor: Pass a compositor pointer when constructing
This allows getting to the display.
later: Keep a pointer to the manager object
This allows using the non-singleton API in idle callbacks.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2718>
As elsewhere, make sure objects that need to have a ownership up to the
context, and use this ownership chain to find relevant components, such
as the backend or the Wayland compositor object instance.
wayland/data-device: Hook up data devices to seats
They are tied to a seat - make that connection in struct fields too, so
that related objects can get to the context via it.
wayland: Don't get Wayland compositor via singleton getter
This means via the ownership chain or equivalent.
xwayland: Hook up manager to Wayland compositor
Same applies to the drag-n-drop struct.
xwayland: Make X11 event handling compositor instance aware
This avoids finding it via singletons in the callee.
xwayland: Don't get Wayland compositor from singleton
xwayland: Pass manager when handling dnd event
window/xwayland: Don't get Wayland compositor from singleton
xwayland/grab-keyboard: Don't get backend from singleton
xwayland: Don't get backend from singleton
wayland: Always get the backend from the context
This means traveling up the ownership chain or equivalent when
necessary.
wayland: Hook up data devices, offers and sources to the compositor
This allows tying them to a context without going through any
singletons.
wayland: Don't get display from singleton
xwayland: Don't get display from singleton
tablet: Don't get display from singleton
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2718>
As with other parts, make objects have the ability to walk up the
ownership chain to the context, to get things like the Wayland
compositor or backend instances.
Contains these squashed commits:
display: Don't get backend from singleton
window: Don't get backend from singleton
keybindings: Don't get backend from singleton
workspace: Don't get backend from singleton
display: Don't get Wayland compositor from singleton
selection: Add display getter
context/main: Get backend directly from the context
clipboard-manager: Don't get display from singleton
stack-tracker: Don't use singleton MetaLater API
startup-notification: Hook up sequences and activations to display
This allows using context aware API directly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2718>
This means objects have an owner, where the chain eventually always
leads to a MetaContext. This also means that all objects can find their
way to other object instances via the chain, instead of scattered global
singletons.
This is a squashed commit originally containing the following:
cursor-tracker: Don't get backend from singleton
idle-manager: Don't get backend from singleton
input-device: Pass pointer to backend during construction
The backend is needed during construction to get the wacom database.
input-mapper: Pass backend when constructing
monitor: Don't get backend from singleton
monitor-manager: Get backend directly from monitor manager
remote: Get backend from manager class
For the remote desktop and screen cast implementations, replace getting
the backend from singletons with getting it via the manager classes.
launcher: Pass backend during construction
device-pool: Pass backend during construction
Instead of passing the (maybe null) launcher, pass the backend, and get
the launcher from there. That way we always have a way to some known
context from the device pool.
drm-buffer/gbm: Get backend via device pool
cursor-renderer: Get backend directly from renderer
input-device: Get backend getter
input-settings: Add backend construct property and getter
input-settings/x11: Don't get backend from singleton
renderer: Get backend from renderer itself
seat-impl: Add backend getter
seat/native: Get backend from instance struct
stage-impl: Get backend from stage impl itself
x11/xkb-a11y: Don't get backend from singleton
backend/x11/nested: Don't get Wayland compositor from singleton
crtc: Add backend property
Adding a link to the GPU isn't enough; the virtual CRTCs of virtual
monitors doesn't have one.
cursor-tracker: Don't get display from singleton
remote: Don't get display from singleton
seat: Don't get display from singleton
backend/x11: Don't get display from singleton
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2718>
While already cleaning up API, if this should ever be more non-static
than a constant, it's better if its a function on the monitor manager
instance than something static.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2718>
Since the Wacom panel rewrite, the "output" setting is handled as
a kind of tri-state for display-integrated tablets:
- If the setting is unset, the device is automatically mapped
to an output
- If the setting is set and not empty, the device is mapped to
the output defined by the EDID data
- If the setting is ['', '', ''], the device is mapped to the
span of all displays, like opaque tablets do.
This distinction for the unset setting fell through the cracks,
so both "Automatic" and "All displays" options were handled as
the former.
Add this distinction, so that display-integrated tablets can
be used like opaque tablets of sorts with no limitations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2767>
These are the ones attached to a display, thus they are the ones that may need
help from this heuristic. Non-integrated tablets (e.g. Intuos) will default to
the span of all monitors.
Fixes mapping of opaque tablets if a display-integrated tablet of the same
brand is also plugged in.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2767>
This might happen when the workspace is not switched and
focus_default_window is called or when 'workspace on primary display
only' is enabled, a secondary display exists and the workspace is
switched.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2747>
There's 2 users of this, meta_display_sync_wayland_input_focus() which
does already perform these checks on its own, and MetaCursorTracker's
update_displayed_cursor() to determine whether it should go with the
Wayland client's cursor.
This second check should also consider the existing ClutterGrabs, so
make meta_display_windows_are_interactable() handle them for both
callers.
Fixes the cursor shown over windows while e.g. there are menus opened.
Close: https://gitlab.gnome.org/GNOME/mutter/-/issues/2553
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2754>
This code path is important for "empty" commits to ensure we schedule
frame callbacks even if previous commits didn't cause stage redraws.
There is, however, no reason to schedule updates on all stage views
instead of only those the actor is on.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2671>
Avoid some allocations, save some CPU cycles and make the code easier
to read.
Behaviourwise the only expected change is that now, if there are mapped
clones, we unconditionally choose the view with the highest refresh
rate the actor (or one of its clones) is on and don't check the
obscurred region any more.
Thus in some cases a client may receive a higher rate of frame callbacks
when obscured on a faster view while a clone is present on a slower
one. The assumption is that cases like this are relatively rare and
that the reduction of code complexity, the reduction of allocations in
`meta_surface_actor_is_obscured_on_stage_view()` whenever the actor is
not fully obscured and has clones on other views, as well as generally
fewer lookups and less code in most common cases, compensate for that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2671>
We only support feedback-actors, such as DnD-icons, in the compositing
path at the moment.
The approach is similar to how we handle certain shell elements.
Implementations need to ensure no references to the object keep
around longer that necessary.
Arguably this should be replaced by a more robust and implicit actor
hierachy detection in the direct scanout code at some point.
Closes https://gitlab.gnome.org/GNOME/mutter/-/issues/2470
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2677>
The fields of 'priv->video_format.max_framerate' are all of type
uint32_t. Multiplying by G_USEC_PER_SEC can overflow, and equally,
dividing a large numerical type by uint32_t can err too.
Since the variable holding the result is int64_t, cast all uint32_t
fields to int64_t before doing any maths on it.
Spotted while trying to investigating an issue with framerates on
HDMI screencasts.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2762>
Unused since https://gitlab.gnome.org/GNOME/mutter/merge_requests/932
AFAICT.
As a bonus, gets rid of these compiler warnings:
../cogl/cogl/cogl-pipeline-layer-state.c: In function ‘cogl_pipeline_get_layer_min_filter’:
../cogl/cogl/cogl-pipeline-layer-state.c:1279:10: warning: ‘min_filter’ may be used uninitialized [-Wmaybe-uninitialized]
1279 | return min_filter;
| ^~~~~~~~~~
../cogl/cogl/cogl-pipeline-layer-state.c:1274:22: note: ‘min_filter’ was declared here
1274 | CoglPipelineFilter min_filter;
| ^~~~~~~~~~
../cogl/cogl/cogl-pipeline-layer-state.c: In function ‘cogl_pipeline_get_layer_mag_filter’:
../cogl/cogl/cogl-pipeline-layer-state.c:1291:10: warning: ‘mag_filter’ may be used uninitialized [-Wmaybe-uninitialized]
1291 | return mag_filter;
| ^~~~~~~~~~
../cogl/cogl/cogl-pipeline-layer-state.c:1287:22: note: ‘mag_filter’ was declared here
1287 | CoglPipelineFilter mag_filter;
| ^~~~~~~~~~
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2757>
Add a helper function that ensures any queued virtual input events have
been flushed from the input thread. This works by posting a task to the
input thread, which will itself queue another callback back to the main
thread. Once the main thread callback is invoked, the flush call is
unblocked and the function returns. Upon this, any previously emitted
virtual input event should have already passed through the input thread
back into the main thread, however not necessarily fully processed.
For making sure it has been processed, one also have to make sure the
stage has been updated, e.g. via `meta_wait_for_paint()`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2727>
After the commit "wayland/subsurface: Implement
meta_wayland_surface_get_window()" subsurfaces are supported. Adjust
some comments and fix a warning that could occur when closing a window.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2717>
The intention here was to check if the subsurface belongs to a window.
Thus it didn't behave as expected for subsurfaces belonging to non-toplevel
windows.
After the previous commit we can use `get_window()` to check for what we
actually want here.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2717>
Subsurfaces are special regarding windows as they don't have a window,
but usually have an ancestor which does. All current users of
`get_window()` are either used for known surface roles, such as xdg-*
ones, or, as is the case for pointer constrains, would actually want to
get the ancestors window.
Thus implement `get_window()` to allow pointer constrains to work.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2223
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2717>
The meta_prop_get_motif_hints() function was only used in the
old MetaUI frames code. The remaining code in mutter accesses
directly the MetaPropValue when loading properties for a window,
and does not use this API call.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2741>
Since we use XCB in the Mutter side, but Xlib in the frames client,
we cannot share the same struct definition since both libraries
will expect different type lengths (respectively, 32-bit ints vs.
longs).
Revert the changes that made both executables share the same
struct, since not both of them can get it right (and retrieve
correctly the struct with the contained flags) in reading the
Motif WM hints.
This reverts commit 2fb3c5a4f5.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2741>
Adding the 'default-decoration' CSS class to MetaFrameHeader after
it is set as the headerbar makes it not account for the minimum size
correctly sometimes. This is a bit racy though - if the window opens
very quickly, it works as expected.
Adding the CSS class before the widget is used guarantees it'll
always report the correct size though, so do that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2753>
The Python docs recommends using run() for all use-cases it can handle:
https://docs.python.org/3/library/subprocess.html#using-the-subprocess-module
run() waits for the subprocess started to complete, so it's not
necessary to use wait() and communicate() anymore. This simplifies the
script.
Previously running "check-style.py -r" after each commit in an
interactive rebase failed, b/c the script did not wait for the amend
command to complete. Using run() instead of Popen() solves this issue.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2733>
Before this, new files introduced by the range of commits checked were
not considered for formatting b/c uncrustify was always turned off in
the beginning of the files, and never turned back on.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2733>
Add this CSS class both to the header bar itself, since it is what
actually contains the window controls, and to MetaFrameHeader too,
since it's what's directly attached to the window.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2740>
Previous commit added support for setting the GTK4 theme setting
according to the color scheme setting. That's cool. What it didn't
add, though, was initializing the GTK4 theme setting to the proper
value. That means if the desktop starts at dark style, you'd still
get a light titlebar.
Fix that by updating the GTK4 theme setting on init as well.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2740>
These are now referenced on the frames client side (in order to
track deletable state from the client window) and the mutter side
(pretty much everything else, like figuring out if a window wants
WM decorations).
It makes sense to make this a separate header, so that we don't
need to doubly define these flags/structs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2735>
We use this for tracking the deletable state of the client window,
but forgot to check that the MWM_HINT_FUNCTIONS hint is set in
hints.flags before checking hints.functions.
This resulted in windows that do not specify this flag (and thus
should go with the defaults) in being mistakenly removed the close
button, as the functions flags would be typically 0 in that case.
Fixes issues with Chromium and Electron applications missing the
close button, since Chromium does this on X11.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2735>
Recalculating window features is a busy thing on the Mutter side, the
different properties being (re)set will overwrite the current state
and cause some side work. Between that is the rewriting of the
_MUTTER_NEEDS_FRAME property on the window being recalculated, which
throws the frames client off, by thinking the window does actually
require a new frame.
It is not sufficient to trust that PropertyNewValue means the property
or the value are new, also double check that the window did not have
in fact a frame, and avoid the busy work if it did.
Besides the busywork that can be easily avoided, this also fixes the
window close button state being stuck if the window changed its
deletable state, since the frame being respawn managed to miss the
property change.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2735>
meta_screen_cast_window_stream_src_set_cursor_metadata() relies
entirely on meta_screen_cast_window_transform_cursor_position()
to return the correct relative cursor position.
However, this function actually does not return the expected
values, since it does not apply the resource scale to the
transformed position.
Actually apply the cursor scale when calculating the cursor
position.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2737>
meta_screen_cast_stream_src_set_cursor_sprite_metadata() receives
the cursor sprite, position, and scale, and with that it downloads
the cursor sprite by drawing it into a separate framebuffer, then
calls cogl_framebuffer_read_pixels() in it - this is the offscren
path that is very common when using screen capturing applications
such as OBS Studio.
There's a sneaky issue in this code path though: the 'scale' value
is a float. The cursor size is then determined by multiplying the
sprite width and height - two integer variables - by scale, and
this relies on standard float-to-int conversions. This is problematic
as sometimes the rounded values disagree with what is expected by
cogl_framebuffer_read_pixels(). If the packing of either the cursor
width or height is off by one, glReadPixels() will try to write into
off bounds, which crashes.
This can be reproduced by enabling fractional scaling, setting a 150%
zoom level, on a 4K screen, and opening any commit with an image diff
in gitlab.gnome.org, all while screencasting. When hovering the new
image, the cursor sprite will be such that it triggers this code path,
and reproduces this issue.
Fix this by always ceiling the cursor sprite sizes.
Closes https://gitlab.gnome.org/GNOME/mutter/-/issues/2542
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2736>
The uninitialized fields in this event causes use of uninitialised
data as seen in valgrind:
==71864== Syscall param writev(vector[0]) points to uninitialised byte(s)
==71864== at 0x5026EBD: __writev (writev.c:26)
==71864== by 0x5026EBD: writev (writev.c:24)
==71864== by 0x6482A3B: UnknownInlinedFun (xcb_conn.c:296)
==71864== by 0x6482A3B: _xcb_conn_wait.part.0 (xcb_conn.c:551)
==71864== by 0x6482BAF: UnknownInlinedFun (xcb_out.c:469)
==71864== by 0x6482BAF: _xcb_out_send (xcb_out.c:470)
==71864== by 0x6483DD7: UnknownInlinedFun (xcb_out.c:416)
==71864== by 0x6483DD7: xcb_writev (xcb_out.c:409)
==71864== by 0x53B79B4: _XSend (xcb_io.c:587)
==71864== by 0x53BBF38: _XReply (xcb_io.c:679)
==71864== by 0x53AFFC9: XQueryTree (QuTree.c:47)
==71864== by 0x4982A5F: query_xserver_stack (stack-tracker.c:508)
==71864== by 0x4EA1F5F: g_closure_invoke (gclosure.c:832)
==71864== by 0x4ECFD45: signal_emit_unlocked_R.isra.0 (gsignal.c:3796)
==71864== by 0x4EC0129: g_signal_emit_valist (gsignal.c:3549)
==71864== by 0x4EC03B2: g_signal_emit (gsignal.c:3606)
==71864== Address 0x287d5900 is 32 bytes inside a block of size 16,384 alloc'd
==71864== at 0x4849444: calloc (vg_replace_malloc.c:1340)
==71864== by 0x53A5FE8: XOpenDisplay (OpenDis.c:240)
==71864== by 0x6100E3C: _gdk_x11_display_open (gdkdisplay-x11.c:1565)
==71864== by 0x60CF675: gdk_display_manager_open_display (gdkdisplaymanager.c:462)
==71864== by 0x49D59F1: open_gdk_display (meta-x11-display.c:1041)
==71864== by 0x49D5D64: meta_x11_display_new (meta-x11-display.c:1156)
==71864== by 0x49564AD: meta_display_init_x11_finish (display.c:743)
==71864== by 0x495679D: on_x11_initialized (display.c:818)
==71864== by 0x4D67558: g_task_return_now (gtask.c:1232)
==71864== by 0x4D67782: UnknownInlinedFun (gtask.c:1301)
==71864== by 0x4D67782: g_task_return (gtask.c:1258)
==71864== by 0x495663C: on_xserver_started (display.c:788)
==71864== by 0x4D67558: g_task_return_now (gtask.c:1232)
==71864== Uninitialised value was created by a stack allocation
==71864== at 0x49D4A59: take_manager_selection (meta-x11-display.c:640)
==71864==
To fix this, fully initialize the event struct before sending it.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2535
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2724>
Since the frames are now rendered by a separate process, we no longer
can guarantee at this point that all updates were handled. Engaging
in a new synchronous resize operation will again freeze the actor,
so sometimes we are left with a not-quite-current buffer for the
frame+window surface.
In order to ensure that the right changes made it onscreen, delay
this next synchronous resize step until the moment the surface was
repainted. This avoids those glitches, while still ensuing the
resize operation ends up in sync with the pointer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
Let the frames client render its own shadow. In order to do that,
avoid double painting a shadow on the compositor side, and extend
the mask area of the frame, so it does unveil the (so far)
hidden frames-client-side shadows.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
There's two meanings of "frame" there! Since SSD frames are now
rendered by an external client, and there are no actual mechanism
that ensures the frame did already get painted when the client did
respond to its NET_WM_FRAME_SYNC_REQUEST request, there may be
artifacts when resizing windows.
In order to get always the best visual result, we should actually
synchronize rendering with both the client window and the window
frame window.
This commit adds these mechanisms, so a sync alarm update is
expected on both windows until further resizes are allowed, this
ensures window and frame stay in sync, even after moving rendering
elsewhere.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
It will become necessary to track properties and changes from frame windows,
and it will be more convenient to have this managed by the common property
tracking mechanisms.
Add this source_xwindow parameter so property handler functions can check
whether the property belonged to the client Window or the frame Window.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
Store the alarms in a different hashtable, and look up the MetaSyncCounter
right away. It so far avoids the MetaWindow middle man, but will also be
simpler when each window can possibly have more than one active alarms.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
Replace the in-process implementation of frames with the external
frames client.
When a client window is created and managed by Mutter, Mutter will
determine whether it is a window that requires decorations and
hint the creation of a frame for it by setting the _MUTTER_NEEDS_FRAME
property on the client window.
After the frames client created a window that has the _MUTTER_FRAME_FOR
property, Mutter will proceed to reparent the client window on the
frame window, and show them as a single unit.
Rendering and event handling on the frame window will be performed by
the external client, Mutter is still responsible for everything else,
namely resizing client and frame window in synchronization, and
managing updates on the MetaWindowActor.
In order to let the frame be managed by the external client, Mutter
needs to change the way some properties are forwarded to the client
and/or frame windows. Some properties are necessary to keep propagating
to the client window only, some others need to happen on the frame
window now, and some others needs to be propagated on both so they
are synchronized about the behavior.
Also, some events that were previously totally unexpected in frame
windows are now susceptible to happen, so must be allowed now.
MetaFrame in src/core/frame.c now acts as the wrapper of foreign
windows created by the frames client, from the Mutter side. Location,
size, and lifetime are still largely in control of Mutter, some
details like visible/invisible borders are obtained from the client
instead (through the _MUTTER_FRAME_EXTENTS and _GTK_FRAME_EXTENTS
properties, respectively).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
This small X11 client takes care of creating frames for client
windows, Mutter will use this client to delegate window frame
rendering and event handling.
The MetaWindowTracker object will keep track of windows created
from other clients, and will await for _MUTTER_NEEDS_FRAME property
updates on those (coming from Mutter), indicating the need for a
frame window.
This process is resilient to restarts of the frames client, existing
windows will be queried during start, and the existence of relevant
properties checked. Mutter will be able to just hide/show
SSD-decorated windows while the frames client restarts.
The frames are created through GTK4 widgets, the MetaWindowContent
widget acts as a replacement prop for the actual client window,
and the MetaFrameHeader wraps GtkHeaderBar so that windows can be
overshrunk, but otherwise a MetaFrame is a 100% true GTK4 GtkWindow.
After a frame window is created for a client window, the
_MUTTER_FRAME_FOR property will be set on the frame window,
indicating to mutter the correspondence between both Windows.
Additionally, the pixel sizes of the visible left/right/top/bottom
borders of the frame will be set through the _MUTTER_FRAME_EXTENTS
property, set on the frame window.
In order to make the frame window behave as the frame for the
client window, a number of properties will be tracked from the
client window to update the relevant frame behavior (window title,
resizability, availability of actions...), and also some forwarding
of events happening in the frame will be forwarded to the client
window (mainly, WM_DELETE_WINDOW when the close button is clicked).
Other than that, the frames are pretty much CSD GTK4 windows, so
window drags and resizes, and window context menus are forwarded for
the WM to handle.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
This check dates all the way back to commit ac2aa5337d. At the time, the
window switcher was an actual X window, that could generate crossing events
if popped up under the pointer. Checking for this kind of crossing events
made sense back at the time in order not to break focus-follows-mouse as
it's been behaving for long.
But now, this UI is all Clutter widgetry, which in the worst case (X11
sessions, of course) it will update the stage window shape to make these
parts clickable. This happens in other places of code that do already
check for ignoring crossing events.
Underneath, this looked up for a Mutter-local GdkWindow of type
GDK_WINDOW_TEMP, only the main MetaFrames window matches those characteristics
nowadays, notably no window switcher popups. Since the remaining window is
never unmapped (until perhaps shutdown), the paths were functionally dead.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
If the window is not managed, it's weird that it asks for _NET_FRAME_EXTENTS,
it's even weirder that mutter replies with a frame border that would only
apply if the window were managed. Stop doing the latter, and drop the
MetaUI call that calculates borders from the theme settings.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
Put the helper to use, in order to lift MetaWindow itself from this
accounting. As a bonus, the data itself now moved to the MetaWindowX11
private struct, since this may only happen with X11 windows (or its
Xwayland subclass).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
This helper struct takes care of the handling of requests and alarms
in order to satisfy NET_WM_SYNC_REQUEST. It will be necessary to
decouple rendering of windows and frames in future commits, so each
window may need its own synchronization and accounting.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
This may result in a view of the stack in MetaStackManager that does not correspond
to reality, since the window is already being unmanaged, there is no point either in
notifying the stack manager about it.
This slight divergence with reality in the MetaStackManager may produce a non-accurate
view if querying its state has to go through the predicted branches. Later synchronization
with the X11 stack may even this out, but the result really depends on when it is asked.
Fixes some intermittent failures in the stacking/closed-transient-only-take-focus-parents
unit test.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
The meta_window_show() method internally relies on window->mapped being
up-to-date, or attempting to focus it may fail since the window is not
mapped yet, resulting on the window being mapped, but not focused as
it would be expected.
This is moot so far, since windows with frames are created sort-of
synchronously and showing them will result in the focus attempt happening
when the window is already mapped, but things will break when this
becomes an asynchronous step.
Ensure to synchronize client state before showing, so any attempts to
focus the window are able to succeed despite the initial state when
calling meta_window_update_visibility().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
The test does simply "wait" which apparently is not enough to ensure the
client window did resize to the expected dimensions. Use "wait_reconfigure"
and assert that the size after resize is the expected, before going further
at testing its behavior after maximize/unmaximize; it might end up with the
unexpected size after the whole operation.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2175>
Move the use count from a separate MetaWaylandBufferRef struct to the
MetaWaylandBuffer class, and remove the former.
The buffer use count is now incremented already in
meta_wayland_surface_commit, since the Wayland protocol defines the
buffer to be in use by the compositor at that point. If the buffer
attachment ends up being dropped again before it is applied to the
surface state (e.g. because another buffer is committed to a
synchronized sub-surface before the parent surface is committed),
the use count is now decremented, and a buffer release event is sent if
the use count drops to 0.
Buffer release events were previously incorrectly not sent under these
circumstances. Test case: Run the weston-subsurfaces demo with the -r1
and/or -t1 command line parameter. Resize the window. Before this
change, weston-subsurfaces would freeze or abort after a few resize
operations, because mutter failed to send release events and the
client ran out of usable buffers.
v2:
* Handle NULL priv->buffer_ref in
meta_wayland_cursor_surface_apply_state.
v3:
* Remove MetaWaylandBufferRef altogether, move the use count tracking
to MetaWaylandBuffer itself. Much simpler, and doesn't run into
lifetime issues when mutter shuts down.
v4:
* Warn if use count isn't 0 in meta_wayland_buffer_finalize.
* Keep pending_buffer_resource_destroyed for attached but not yet
committed buffers. If the client attaches a buffer and then destroys
it before commit, we ignore the buffer attachement, same as before
this MR.
v5:
* Rebase on top of new commit which splits up surface->texture.
* MetaWaylandSurfaceState::buffer can only be non-NULL if
::newly_attached is TRUE, simplify accordingly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
Until all dma-buf file descriptors for all buffers in the transaction
are readable, which corresponds to when the client drawing to the
buffers has finished.
This fixes https://gitlab.gnome.org/GNOME/mutter/-/issues/1162 if the
GPU & drivers support high priority contexts which can preempt lower
priority contexts.
v2:
* Also remove dma-buf fds from transaction and try applying it from
pending_buffer_resource_destroyed. Avoids freeze due to leaving a
GSource based on a closed fd attached if a client destroys a wl_buffer
which is part of a transaction which was committed but not applied
yet. (Robert Mader)
* Tweak transaction cleanup logic in wl_surface_destructor.
v3:
* Adapt to meta_wayland_dma_buf_get_source.
v4:
* Adapt to new commits using transactions for (sub-)surface destruction,
drop code to remove destroyed surfaces from pending transactions.
v5:
* Use g_clear_pointer in meta_wayland_transaction_destroy.
(Georges Basile Stavracas Neto)
* Add spaces between type casts and values. (Carlos Garnacho)
* Use (gpointer *) instead of (void**). (Carlos Garnacho)
* Use gpointer instead of void * in
meta_wayland_transaction_dma_buf_dispatch.
v6:
* Use g_hash_table_remove in meta_wayland_transaction_dma_buf_dispatch.
(Carlos Garnacho)
v7: (Jonas Ådahl)
* Move include of glib-unix.h below that of meta-wayland-transaction.h.
* Split up g_hash_table_iter_next call to multiple lines in
meta_wayland_transaction_commit.
* Call g_source_destroy as well as g_source_unref when freeing a
committed but not yet applied transaction (during mutter shutdown).
v8:
* Drop dma_buf_source_destroy, can use g_source_destroy directly.
(Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
From xdg_surface_constructor_get_popup / xdg_popup_reposition (called
during Wayland protocol processing) to finish_popup_setup /
meta_wayland_xdg_popup_apply_state (called when the popup state is
applied).
This makes sure that the parent window frame rectangle is up to date in
meta_wayland_xdg_positioner_to_placement.
v2:
* Use meta_wayland_surface_state_new () in
meta_wayland_transaction_add_xdg_popup_reposition.
v3:
* Move xdg_popup_repositioned handling to
meta_wayland_xdg_popup_apply_state.
v4:
* Do not steal pending->xdg_positioner in
meta_wayland_xdg_popup_apply_state, fixes leaking the corresponding
memory.
* Drop MetaWaylandSurfaceState::xdg_popup_repositioned, just use
::xdg_positioner.
v5:
* Reformat meta_wayland_xdg_positioner_to_placement calls to stay within
80 columns. (Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
This makes sure that finish_popup_setup is called after any previous
transactions for the parent surface have been applied, so the parent
window geometry is up to date.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
Preparation for potentially calling meta_wayland_transaction_apply some
time after surface commit, in which case doing it in the former would be
too late: The client may legally destroy the attached wl_buffer
immediately after commit, in which case meta_wayland_buffer_attach would
spuriously fail and disconnect the client (or possibly even crash mutter
due to NULL error).
Requires splitting up the surface texture between protocol and output
state, and propagating from the former to the latter via
MetaWaylandSurfaceState.
v2: (Jonas Ådahl)
* Move meta_wayland_surface_get_texture call to separate line.
* Use g_autoptr for GError.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
And keep track of the hierarchy separately for the Wayland protocol and
for output. Protocol state is updated immediately as protocol requests
are processed, output state only when the corresponding transaction is
applied (which may be deferred until the next commit of the parent
surface).
v2:
* Directly add placement ops to a transaction, instead of going via
pending_state.
* Use transaction entry for the sub-surface instead of that for its
parent surface.
v3:
* Use transaction entry for the parent surface again, to ensure proper
ordering of placement ops, and call
meta_wayland_surface_notify_subsurface_state_changed only once per
parent surface.
* Drop all use of wl_resource_add_destroy_listener, transactions are
keeping surfaces alive as long as needed.
v4:
* Rebase on https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2501
* Drop ClutterActor code from meta_wayland_surface_apply_placement_ops.
(Robert Mader)
v5:
* Rename MetaWaylandSubSurfaceState to MetaWaylandSurfaceSubState, since
the next commit adds not sub-surface specific state to it.
v6:
* Move include of meta-wayland-subsurface.h from
meta-wayland-transaction.c to .h, since the latter references
MetaWaylandSubsurfacePlacementOp.
v7:
* Drop superfluous !entry check from meta_wayland_transaction_apply.
v8:
* Rename output/protocol fields to output/protocol_state. (Jonas Ådahl)
v9:
* Use meta_wayland_surface_state_new in
meta_wayland_transaction_add_placement_op.
v10:
* Fix a few style issues per check-style.py.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
Destroy Wayland protocol related state immediately when the Wayland
resource is destroyed, but keep the rest alive by any transaction which
references the surface.
This makes it easier and cleaner to deal with a surface getting
destroyed while it's still referenced by transactions.
v2:
* No more need to keep references for surfaces in the entries hash
table.
v3:
* Do not use surface->sub.transaction in wl_surface_destructor, just
destroy it.
v4:
* No need for wl_surface_destructor to use its own transaction.
v5:
* Use g_steal_pointer & (more) g_clear_pointer in wl_surface_destructor.
v6:
* Leave SURFACE_DESTROY signal emission in wl_surface_destructor.
v7:
* Use finalize instead of dispose callback.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
This keeps all surfaces referenced by a transaction alive until the
transaction is destroyed, and makes sure transactions are applied in
the same order as they were committed with respect to all surfaces
they reference.
v2:
* Guard against NULL entry in meta_wayland_transaction_apply.
v3:
* Keep single entries hash table.
v4:
* Unref the surface in the meta_wayland_transaction_merge_into while
loop only if the "to" transaction didn't already have an entry for it,
to prevent premature finalization of the surface (likely followed by a
crash).
v5:
* Unref the surface (implicitly via g_hash_table_iter_remove) in the
meta_wayland_transaction_merge_into while loop even if the "to"
transaction already had an entry for it, or we leak a reference.
* Use g_clear_object & g_steal_pointer to not leave behind a dangling
from->state pointer in meta_wayland_transaction_entry_merge_into.
v6:
* Add curly braces around
meta_wayland_transaction_add_placement_surfaces calls. (Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
Instead of cached_state.
surface_commit for a synchronized sub-surface either commits the
transaction or merges it into the parent surface's transaction (if
the parent is a synchronized sub-surface itself).
This should fix or at least improve the behaviour of nested synchronized
sub-surfaces.
Also change wl_subsurface_set_desync:
* Commit sub-surface transactions separately. This may allow some of
them to be applied earlier in some cases.
* Commit transaction only for descendant sub-surfaces which become
newly de-synchronized themselves.
v2:
* Drop unused function prototypes
v3:
* Use g_clear_pointer for surface->sub.transaction.
v4:
* Use g_steal_pointer instead of g_clear_pointer. (Sebastian Wick, Jonas
Ådahl)
v5: (Carlos Garnacho)
* Add spaces between type casts and values.
* Use (gpointer *) instead of (void**).
v6: (Jonas Ådahl)
* Use g_clear_object in meta_wayland_transaction_entry_merge_into.
* Use meta_wayland_transaction_entry_free in
meta_wayland_transaction_merge_into.
* Fix alignment of meta_wayland_transaction_merge_pending_state
parameters.
* Remove unused meta_wayland_transaction_add_state declaration.
v7:
* Use meta_wayland_surface_state_new in
meta_wayland_transaction_merge_pending_state.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
A transaction contains the committed state for a surface, plus any
cached state for synchronized subsurfaces.
v2:
* Handle sub-surface positions separately from surface states.
v3:
* Sync child states only for surfaces with state in the transaction.
v4: (Jonas Ådahl)
* Drop unnecessary g_object_new call from wl_subsurface_set_desync. (me)
* Fix indentation & formatting in meta_wayland_surface_commit.
* Add meta_wayland_surface_state_new helper function.
* Fix alignment of meta_wayland_transaction_apply_subsurface_position
parameters.
* Add curly braces around meta_wayland_transaction_sync_child_states
call in meta_wayland_transaction_apply.
v5:
* Make meta_wayland_surface_state_new an inline function.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
If multiple transactions have entries for the same surface, they are
applied in the same order as they were committed. Otherwise, they can
be applied in any order.
This is preparation for following changes, transactions are still
applied as soon as they're committed.
v2:
* Move GQueue for transactions to MetaWaylandCompositor (Jonas Ådahl)
v3
* Say "entry for" instead of "state for", since there can be transaction
entries with no state (for surfaces which are getting destroyed).
v4:
* Use a hash table to keep track of all candidate transactions which
might be newly ready to be applied.
* Use clearer function / variable names.
v5:
* Use custom single-linked list instead of hash table for candidate
transactions, ordered by the transaction commit sequence number, so
that they're attempted to be applied in the same order as they were
committed.
* Rename transaction->queue to transaction->committed_queue, and
simplify its handling.
v6: (Carlos Garnacho)
* Add spaces between type casts and values.
* Use (gpointer *) instead of (void**).
v7: (Jonas Ådahl)
* Use G_MAXSIZE instead of ULONG_MAX.
* Fix indentation of meta_wayland_transaction_apply &
meta_wayland_transaction_maybe_apply_one parameters.
* Refactor find_next_transaction_for_surface & ensure_next_candidate
helper functions out of meta_wayland_transaction_apply.
* Refactor has_unapplied_dependencies helper function out of
meta_wayland_transaction_maybe_apply_one.
* Make while (TRUE) loop in meta_wayland_transaction_maybe_apply
consistent with general usage.
* Drop unused value local from meta_wayland_transaction_commit.
* Store pointer to compositor object in transactions, instead of
pointer to the queue of committed transactions.
* Drop tautological g_assert from meta_wayland_transaction_apply. (me)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
v2:
* Use single hash table with struct which will contain all kinds of
state handled by a transaction.
v3:
* Add meta_wayland_transaction_destroy.
v4 (Georges Basile Stavracas Neto)
* Fix struct _MetaWaylandTransaction(Entry) formatting.
* Explicitly test against NULL.
* Use gpointer insteadof void * for
meta_wayland_transaction_entry_destroy.
v5: (Robert Mader)
* Use for loop in is_ancestor.
* Include meta-wayland-transaction.h first in
meta-wayland-transaction.c.
v6:
* Use g_autofree & g_clear_object.
v7: (Jonas Ådahl)
* Rename meta_wayland_transaction_entry_destroy to
meta_wayland_transaction_entry_free.
* Drop g_autofree use from meta_wayland_transaction_entry_free again.
* Make meta_wayland_transaction_entry_free take a
MetaWaylandTransactionEntry pointer.
* Rename meta_wayland_transaction_destroy to
meta_wayland_transaction_free.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
Need to deal with surface->resource == NULL and
surface->pending_state == NULL in some places.
v2:
* Avoid expanding conditions to multiple lines.
(Georges Basile Stavracas Neto)
v3:
* Use a single bailout condition in meta_wayland_client_owns_window as
well.
v4:
* Remove spare empty line in meta_wayland_surface_apply_state.
(Robert Mader)
* Add wl_resource_post_error calls in xdg-shell request handlers.
(Robert Mader)
* Drop checks in functions which can only be called if there's a valid
resource.
* Drop more checks which are unnecessary due to leaving the
SURFACE_DESTROY signal emission in wl_surface_destructor later.
v5:
* Move resource = surface->resource assignments to if (!resource) tests.
(Jonas Ådahl)
v6:
* Fix style issue per check-style.py.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
Creates a GSource which will call the specified dispatch callback when
all dma-buf file descriptors for the buffer have become readable.
v2:
* Hold a reference to the buffer in the source, to prevent the buffer
from getting destroyed before the source.
v3:
* Do not use check callback, handle everything in dispatch callback.
(Dor Askayo)
v4: (Georges Basile Stavracas Neto)
* Define and use MetaWaylandDmaBufSource & MetaWaylandDmaBufSourceDispatch
types.
* Fix meta_wayland_dma_buf_source_dispatch &
meta_wayland_dma_buf_source_funcs formatting.
* Use gpointer instead of void*.
* Rename meta_wayland_dma_buf_get_source to
meta_wayland_dma_buf_create_source. (Carlos Garnacho)
v5:
* Explicitly handle NULL return value. (Jonas Ådahl)
v6:
* Fix style issue per check-style.py.
v7:
* Fix code style harder. (Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
And call it from meta_wayland_buffer_realize. This makes dma-buf fds
available for EGL image type buffers as well.
v2:
* Move buffer->dma_buf.dma_buf assignment value to next line.
(Jonas Ådahl)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1880>
Add internal state (starting, running, stopping), and use this instead
of MetaDisplay struct fields to determine whether to start animations.
This fixes issues when we try to animate things when shutting down.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2716>
When we remove a child, we stop its transitions (animations), but we
didn't stop animations on grand children. What we did, however, was to
clear the stage views of the grand children, and this caused a bunch of
orphaned transitions (ClutterTimeline) and accompanied warnings.
Make it so that if we stop transitions, and clear stage views, also stop
transitions for the grand children. Detached children don't have a way
to continue animating anyway, since they have no stage view (thus frame
clock) to be driven by.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2716>
When a badly behaving ClutterActor implementation manages to invalidate
the allocation after the layout phase and before painting, we have no
idea where the actor should be painted without running the whole layout
machinery again.
For paint volumes in this case we pretend the actor covers the whole
stage and queue full-stage redraws. When updating stage-views, we're
also handling this case, but not in the most graceful way. Just like
with paint volumes, we should assume an actor without a valid allocation
is simply everywhere, so set priv->stage_views to all available stage
views in that case.
Related: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/6054
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2694>
On hotplug, the events we receive from the kernel are async, and
connectors in the kernel come and go as they please. In practice, this
means that calling drmModeGetConnector() twice more or less directly
after each other, there is no guarantee that the latter call will return
anything if the former did.
When updating the connector in response to hotplugs, we'd first update
the list of existing connectors, and following that, query each and
every one again for their current state, to update our internal
representation; only the former handled drmModeGetConnector() returning
NULL, meaning if unlucky, we'd end up doing a null pointer dereference
when trying to update the state.
Handle this by querying the kernel for the current connector state only
once per connector, updating the list of connectors and their
corresponding state at the same time.
Fixes the following crash:
#0 meta_kms_connector_read_state at ../src/backends/native/meta-kms-connector.c:684
#1 meta_kms_connector_update_state at ../src/backends/native/meta-kms-connector.c:767
#2 meta_kms_impl_device_update_states at ../src/backends/native/meta-kms-impl-device.c:916
#3 meta_kms_device_update_states_in_impl at ../src/backends/native/meta-kms-device.c:267
#4 meta_kms_update_states_in_impl at ../src/backends/native/meta-kms.c:604
#5 update_states_in_impl at ../src/backends/native/meta-kms.c:620
#6 meta_kms_run_impl_task_sync at ../src/backends/native/meta-kms.c:435
#7 meta_kms_update_states_sync at ../src/backends/native/meta-kms.c:641
#8 handle_hotplug_event at ../src/backends/native/meta-kms.c:651
#9 on_udev_hotplug at ../src/backends/native/meta-kms.c:668
Related: https://bugzilla.redhat.com/show_bug.cgi?id=2131269
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2709>
There is no need to use the 'bypass-*' method of event processing in the
changed function since in all cases the 'bypass-*' variable was set, any
following event processing functions would ignore the event anyway.
Simplify things a bit by just returning TRUE if the event is consumed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2714>
We can land inside meta_window_focus() in the middle of changing the
window workspace, because some signal handler of MetaWorkspace's
"window-removed" signal triggers a focus. This can cause a crash in
`g_assert (link)` when updating the MRU list because we still think
we're on the old workspace when actually we are already removed from
this workspaces MRU list.
To avoid crashes like this, bail out of meta_window_focus() when we're
in the middle of a workspace change.
Fixes https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/5368
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2691>
It's a bad idea to have data like this in the middle of a struct, as it
will easily cause everything behind it to be badly aligned and thus
increase memory access times.
So move all those bitfield booleans to the end of the struct.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2691>
It is generally assumed here and there that the pointer at all point in
time is within some logical monitor, if there is any logical monitor to
be within.
With the input thread, this was for a short amount of time not reliable,
resulting in crashes in combination with hotplugging or suspend/resume,
where monitors come and go quickly.
What happens is that the pointer at first is within a logical monitor,
but when that logical monitor is removed, while the new monitor
viewports are handed to the input thread, the constraining happens
asynchronously, meaning there is a time between between the new
viewports are sent, and before clutter_seat_query_state() starts
reporting the constrained position.
If a new client mapped a maximized window during this short time frame,
we'd crash with
#0 meta_window_place at ../src/core/place.c:883
#1 place_window_if_needed at ../src/core/constraints.c:562
#2 meta_window_constrain at ../src/core/constraints.c:310
#3 meta_window_move_resize_internal at ../src/core/window.c:3869
#4 meta_window_force_placement at ../src/core/window.c:2120
#5 xdg_toplevel_set_maximized at ../src/wayland/meta-wayland-xdg-shell.c:429
#6 ffi_call_unix64 at ../src/x86/unix64.S:105
#7 ffi_call_int at ../src/x86/ffi64.c:672
#8 wl_closure_invoke at ../src/connection.c:1025
#9 wl_client_connection_data at ../src/wayland-server.c:437
The fix for this is to make sure that the viewports are updated and
pointers constrained synchronously, i.e. the main thread will wait until
after the input thread is done constraining before continuing.
Related: https://bugzilla.redhat.com/show_bug.cgi?id=2147502
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2711>
We've been sending all events to clients immediately for quite some time
now, so this is only really impacting the Clutter scene graph, not
clients anymore.
That makes this behavior a somewhat unnecessary optimization (it was
useful at the time it was added, but it's not anymore), which will only
make our lives harder when we actually expect an event to be queued
(eg. in tests), so remove it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2697>
We put a DEVICE_ADDED or DEVICE_REMOVED event into Clutters event queue
here, so we should also wait for Clutter to process events once.
Just putting an event into the queue doesn't mean it gets processed
immediately (especially when the commit after this one is applied), so
wait for a stage update here.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2697>
Currently, we will notify the display about a new window being created
during the constructed phase of the GObject. During this time,
property-change notifications are frozen by GObject, so we'll emit a few
::notify signals only after the window-created signal, although
the actual property change happened before that.
This caused confusion in gnome-shell code where a notify::skip-taskbar =
true emission was seen when the property already was true inside a
window-created handler before.
In order to fix that that, we notify the window creation
post-construction
of the GObject on GInitable.init vfunc
Details
https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/6119#note_1598983
Fixes https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/6119
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2703>
As suggested by Carlos in a review of this MR, refactor the logic of
clutter_do_event() to have both adding and removing of devices from the
devices list in a single place.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2696>
Having the stage device list be responsible for delivering the
same events twice (first immediately to clients, then later to Clutter)
was expected to be tricky, a sneaky problem with it right now is the
following case:
While collecting events for a stage update cycle, we get three touch
events from the backend: TOUCH_BEGIN(seq=1) -> TOUCH_END(seq=1) ->
TOUCH_BEGIN(seq=1)
What we do right now when we see a TOUCH_BEGIN event is adding a device
to the stage right when it comes in from the backend. And when we see
a TOUCH_END, we remove the device from the stage not immediately but
only after it went through the queue.
In the case of the three events mentioned above, with the current
behavior, this will happen when they come in from the backend:
- TOUCH_BEGIN(seq=1): device gets added to the stage with seq 1, event
gets queued
- TOUCH_END(seq=1): Nothing happens, event gets queued
- TOUCH_BEGIN(seq=1): we try to add device to the stage, but seq 1 is
already there, event gets queued
Now when we go through the queue and see the TOUCH_END, the device with
seq 1 gets removed, but on the subsequent TOUCH_BEGIN, we won't add a
new device, so this event (and all events with seq=1 that are still in
the queue) is now ignored by Clutter because it has no device.
What we want to do here is to cut short once the TOUCH_END event comes
in: Process queued events immediately and make sure the device is
removed from the stage list before a new device can be added. Same goes
for any other events that will lead to devices getting removed.
Small note: Since this leads to clutter_stage_get_device_actor()
returning NULL, I was wondering why we never crash because of this:
Turns out _clutter_actor_handle_event() handles self = NULL just fine
without crashing...
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2696>
With commit 6c17aa66c6 we made sure no
stale device entries might land in the stage device list. The same can
happen for pointer devices too in theory, in practice we never really
filter them out, but it's good to handle them here anyway.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2696>
We'll call this function from a few more places for the
CLUTTER_DEVICE_REMOVED case, so move the check for which devices are
valid into the function itself to avoid having to check everywhere.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2696>
If two X11 windows were the last two, we'd remove them from the stack
while unmanaging them. That'd hit an assert in
meta_stack_tracker_restack_managed(), resulting in the following crash
when Xwayland exited unexpectedly with two or more X11 windows being the
only windows on the stack:
#1 g_assertion_message() at ../glib/gtestutils.c:3256
#2 g_assertion_message_expr() at ../glib/gtestutils.c:3282
#3 meta_stack_tracker_restack_managed() at ../src/core/stack-tracker.c:1210
#4 on_stack_changed() at ../src/core/stack.c:142
#5 _g_closure_invoke_va() at ../gobject/gclosure.c:895
#6 g_signal_emit_valist() at ../gobject/gsignal.c:3456
#7 g_signal_emit() at ../gobject/gsignal.c:3606
#8 meta_stack_changed() at ../src/core/stack.c:265
#9 meta_stack_remove() at ../src/core/stack.c:324
#10 meta_window_unmanage() at ../src/core/window.c:1542
#11 meta_x11_display_unmanage_windows() at ../src/x11/meta-x11-display.c:111
#12 meta_x11_display_dispose() at ../src/x11/meta-x11-display.c:141
#13 g_object_run_dispose() at ../gobject/gobject.c:1448
#14 meta_display_shutdown_x11() at ../src/core/display.c:831
The added test specifically checks that this scenario is handled
gracefully.
Related: https://bugzilla.redhat.com/show_bug.cgi?id=2143637
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2704>
Instead of having users of the test client manually deal with alarm
filters, let the test client automatically add itself as filters. This
changes the MetaX11Display a bit, to handle an array of filters instead
of a single filter.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2704>
The CRTC cursor sprite scale was incorrectly assumed to be always 1.0
when using the default not-scale-monitor-framebuffer mode. This is
harmless in most cases, as most clients provide HiDPI capable cursors,
but for the ones that didn't, we'd end up drawing their cursors
unscaled, when using the cursor planes.
Fix this by using the "texture scale" which is what is intended for
this.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2477
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2698>
Cursor planes tend to be ARGB8888 and support no other format (ideally
we should not hard code this, but un-hard-coding that is for another
day), and if we put e.g. a XRGB8888 buffer in there, it'll either result
in the gbm_bo allocation failing (it doesn't allow USE_CURSOR with any
other format) or mode setting failing if using dumb buffers directly.
In the former case, we'll fall back to OpenGL indefinitely, and in the
latter, we'll have failed mode sets as long as we try to set the invalid
cursor buffer as the cursor plane.
Change things to process all buffers that are not ARGB8888 using the
scale/rotate machinery we already have, turning XRGB8888 into ARGB8888.
Related: https://gitlab.gnome.org/GNOME/mutter/-/issues/2477
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2698>
Attaching a new buffer with a different size than the old one means
that the viewport needs to be recalculated.
Not doing this caused the viewport to be incorrectly applied when
viewport_src_rect remained the same after attaching such buffer.
Pipeline reset usually happens when applying a new viewport,
but it doesn't happen when the viewport values remain the same.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2689>
A client may provide a positioner that places the window outside of its
parent. This isn't allowed, according to spec, so we hide the window and
log a warning. This, however, leads these affected clients with an
incorrect view of what is mapped or not, meaning it becomes harder to
recover.
Fix this by sending xdg_popup.done when we hide the popup due to an
invalid position. Don't error out the client, let the bug slide, as
that's a less jarring experience for existing applications that
reproduce this than being disconnected, which practically feels like a
crash.
Related: https://gitlab.gnome.org/GNOME/mutter/-/issues/2408
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2645>
In constrast to x11, Wayland has sane handling for touch events and
allows the compositor to handle a touch event while the clients are
already seeing it. This means we don't need the REJECTED state on
Wayland, since we can also grab sequences after the client has seen
them.
So disallow moving sequences to the REJECTED state on Wayland.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2508>
Updating of the paint volume used for culling these days happens
during the finish-layout stage, not while painting. Also we have
geometry-based, not paint-based picking anymore.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1492>
Rename the `last_paint_volume` to `visible_paint_volume`: That avoids
confusion with the `had_effects_on_last_paint_volume_update` flag and
also makes it clear that this paint volume is the currently visible one.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1492>
Rename the paint_volume_valid flag to has_paint_volume in order to
better reflect what it's for.
The name "paint_volume_valid" implies that the paint volume can be
invalidated and thus sounds like it's involved with some kind of
caching. The flag that's actually involved with caching is
"needs_paint_volume_update", while "paint_volume_valid" is only meant to
store whether the actor has a paint volume to work with.
So rename paint_volume_valid to has_paint_volume to avoid confusion
about which flag is used for caching.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1492>
For clarity and for further improvements, introduce a separate function
to update the paint volume instead of doing that inside
_clutter_actor_get_paint_volume_mutable().
Also add a FIXME comment for a possible bug I noticed while working on
it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1492>
Since ClutterActor now properly caches its paint volume and ClutterText
tries hard to invalidate its own cached paint volume on every redraw
anyway (that's more often than ClutterActor invalidates its own paint
volume), we can simply rely on the caching of the paint volume done by
ClutterActor and invalidate that on every redraw.
So remove the private cached paint volume from ClutterText and all its
invalidation machinery.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1492>
The function _clutter_paint_volume_get_stage_paint_box() actually
doesn't modify the paint volume that's passed to it, so make that a bit
more clear by passing a const paint volume as the argument.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1492>
These days it's possible to chain up into the default get_paint_volume()
implementation again, which renders
clutter_actor_get_default_paint_volume() unnecessary. So remove that
function and move clutter_actor_update_default_paint_volume() back into
real_get_paint_volume() where it belongs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1492>
When modifiers are enabled on mutter on some of the older i915 systems observed
Black-screen on 2nd monitor during multi-head use cases, upon debugging observed
that disabling modifiers on these systems resolved the Black-screen issue:!1618.
This issue depends whether we have enough DBuf space to provide required bandwidth
for the userspace demands. Those platforms which have less Display Buffer, will
just have more chance to face lack of it. However it still depends on various
factors like amount of planes(i.e the more planes we have, the more we divide the buffer),
refresh rate, bpp and so on.
This affects watermark calculations and the minimum blocks required for at least
wm level 0. If we don't have sufficient ddb at least for wm0 for all planes in
the configuration then it is rejected.
Until we have TEST_ONLY commit solution is built we could make sure to disable
modifiers support on these older i915 systems based on udev rules defined in this commit.
This commit makes sure that modifiers are still usable on latest i915 systems.
List of PCI-IDs are referred from:
f8bf2a9a15/include/pci_ids/iris_pci_ids.h
Related: https://gitlab.gnome.org/GNOME/mutter/-/issues/1618
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2641>
It's not really a backend thing, and we'll want to profile e.g. loading
the backend too, so create it very early and destroy it very late and
let MetaContextMain own it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2678>
This change fixes the issue where the cursor is always
embedded in the frames even when the client has requested
the cursor information be sent as metadata in the stream.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2629>
This reverts commit eac227a203.
Currently, Flatpak applications can bypass the X11 permission setting
and access the X server through abstract sockets because X11 authentication
is not enforced for the current user ID.
Fix this by always requiring X11 authentication for Xwayland. This also
means applications without XAUTHORITY set to the file with Mutter's
Xwayland credentials cannot connect to X, including apps launched from
VT or SSH.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2633>
When deciding if a window should be unredirected because it was causing
fullscreen damage in the past, it was not considered whether the window
is still fullscreen. This could result in a floating window being
unredirected if it was chosen for unredirection because of
_NET_WM_BYPASS_COMPOSITOR = 1 and was previously fullscreened for >= 100
frames, long enough to change does_full_damage, before getting
unfullscreened.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2434
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2651>
Back in 2014 sending pressed keys to Wayland clients caused issues,
because at least Xwayland didn't handle that gracefully, causing issues
like ghost-pressed keys. A way it was reproduced was quickly alt-tab:ing
to and from a Firefox window, which would cause the File menu bar
incorrectly appearing.
While this was reported to the Xwayland component back then, it was,
probably by mistake, assumed to be an issue in mutter, and mutter
stopped sending pressed key events on enter.
The following year, Xwayland was eventually fixed, but the work around
in mutter has been kept around until it was again noticed as an
inconsistency between compositor implementations.
Lets remove the work around, and follow the spec, again.
This reverts commit c39f18c2d4.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2457
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2657>
We'd set the capabilities to 'none', meaning all previously enabled
device classes would be disabled. That means we shouldn't re-disable
them directly after.
This ensures '..disable()' is only called once for every '..enable()'.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2657>
We traverse the whole screnegraph anyway these days in finish_layout(),
so no need for the whole "set the flag on parents even though we don't
need it" dance anymore.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2667>
With high frequency mouse devices, we would send very many configure
events per each update cycle, which had the end result that some clients
constantly re-allocating and redrawing their buffers far too often, if
they did this in direct response to xdg_toplevel configure events.
Lets throttle the interactive resize updates to stage updates, to avoid
having these clients doing the excessive buffer reallocation.
This also removes some old legacy X11 client resize throttling, that
throttled a bit arbitrarily on 25 resizes a second; it is probably
enough to throttle on stage updates for these clients.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2652>
Some mice send a value slightly lower than 120 for some detents. The
current approach waits until a value of 120 is reached before sending a
low-resolution scroll event.
For example, the MX Master 3 sends a value of 112 in some detents:
detent detent
| | |
^ ^ ^
112 REL_WHEEL 224
As illustrated, only one event was sent but two were expected. However,
sending the low-resolution scroll event in the middle plus the existing
heuristics to reset the accumulator solve this issue:
detent detent
| | |
^ ^ ^ ^
REL_WHEEL 112 REL_WHEEL 224
Send low-resolution scroll events in the middle of the detent to solve
this problem.
Fix https://gitlab.gnome.org/GNOME/mutter/-/issues/2469
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2668>
There do indeed seem to be places in our own code that trigger grabs on
actors before they are realized. It was not the intention to change the
practical preconditions for GNOME 43, so make it an even lower minimum
that every caller ought to match: That the actor is attached to the stage.
Further constraining of these preconditions will have to wait until
branching for new development.
Fixes: 9c79c7234 (clutter: Only allow grabs to be created on realized actors)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2670>
The bare minimum that we can ask to an actor before creating a grab
on it is that it is realized (and thus, attached to the stage). Bail
out if that is not the case when creating a grab.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2669>
If an actor is being unrealized or otherwise unparented, it's a good
indication that its grabs are now stale and possibly harmful. Ensure
these are dropped when the actor is unparented.
This is now an unlikely event, since there is code to also dismiss
grabs when a visible grabbed actor goes unmapped. But that may be
prevented from happening, or the ordering of circumstances allow a
grab to be created and an actor destroyed without going unmapped
first. This grab dismission on unmap stays as it matches the UI-level
expectatives that an actor must be visible to be grabbed.
Related: https://gitlab.gnome.org/GNOME/mutter/-/issues/2475
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2669>
Previously, when scroll was received in a remote session, it was handled
as continuous scroll.
This generated issues with clients without high-resolution scroll
support as the code path in charge of accumulating scroll until 120 is
reached was not used and therefore discrete scroll events were not being
generated.
Handle scroll generated in a remote session as discrete scroll when the
source is CLUTTER_SCROLL_SOURCE_WHEEL to fix this issue.
Fix https://gitlab.gnome.org/GNOME/mutter/-/issues/2473
Fixes: 9dd6268d13 ("wayland/pointer: Send high-resolution scroll data")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2664>
In fcfe90aa, multiple for loops were replaced with
META_WAYLAND_SURFACE_FOREACH_SUBSURFACE.
However, this substitution was not side-effect free, and introduced a
null-pointer dereference risk as shown in the example below:
Old:
for (n = g_node_first_child (surface->subsurface_branch_node);
n;
n = g_node_next_sibling (n))
{
if (G_NODE_IS_LEAF (n))
continue;
meta_wayland_surface_update_outputs_recursively (n->data);
}
n is checked for NULL during each loop in the condition expression.
Therefore, when `G_NODE_IS_LEAF (n)` is called, `n` is guaranteed not to
be NULL. Note also that g_node_first_child is also NULL-safe since it
performs a NULL check internally.
New:
META_WAYLAND_SURFACE_FOREACH_SUBSURFACE (surface, subsurface_surface)
meta_wayland_surface_update_outputs_recursively (subsurface_surface);
=
for (GNode *G_PASTE(__n, __LINE__) = meta_get_first_subsurface_node ((surface)); \
(subsurface = (G_PASTE (__n, __LINE__) ? G_PASTE (__n, __LINE__)->data : NULL)); \
G_PASTE (__n, __LINE__) = meta_get_next_subsurface_sibling (G_PASTE (__n, __LINE__)))
In the new logic `subsurface` is still checked for NULL in the loop
condition. However, in the new loop init:
...
meta_get_first_subsurface_node (MetaWaylandSurface *surface)
...
n = g_node_first_child (surface->subsurface_branch_node);
if (!G_NODE_IS_LEAF (n))
...
The above implementation performs a `G_NODE_IS_LEAF` call, which
performs a dereference on `n`, without first checking for NULLs.
This NULL dereference triggers the following gnome-shell crash:
Core was generated by `/usr/bin/gnome-shell'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0 meta_get_first_subsurface_node (surface=0x55d589623450) at ../src/wayland/meta-wayland-surface.h:399
#1 pointer_can_grab_surface (pointer=0x7f6dc4012700, surface=0x55d589623450) at ../src/wayland/meta-wayland-pointer.c:1306
#2 0x00007f6ddb94d509 in meta_wayland_pointer_can_grab_surface (pointer=<optimized out>, surface=surface@entry=0x55d589623450, serial=serial@entry=996) at ../src/wayland/meta-wayland-pointer.c:1321
#3 0x00007f6ddb950d05 in meta_wayland_seat_get_grab_info (seat=seat@entry=0x55d586c24f20, surface=0x55d589623450, serial=996, require_pressed=require_pressed@entry=0, x=x@entry=0x0, y=y@entry=0x0)
at ../src/wayland/meta-wayland-seat.c:467
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2655>
Unlike the multi-view path, the optimized/single-view one doesn't check
if the surface-actor is really present on the view. That is the case
whenever it's hidden - e.g. when the window is minimized.
Fixes 3b7137cb35
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2662>
The stage view list does not get updated when an actor gets hidden in
order to avoid unnecessary work, such as scale changes. However, we
still want `is_effectively_on_stage_view` to report `FALSE` in this
case.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2662>
The :input-purpose and :input-hints properties were added without
actually handling the get/set operations, whoops.
All code uses the (working) methods, so this only fixes expectations,
not an actual bug :-)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2659>
Fullscreen Wayland toplevel surfaces don't need to respect the
configured size in which case the window content get centered on a black
background which covers the whole monitor.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2338>
Fullscreen Wayland toplevel surfaces don't need to respect the
configured size in which case it should be shown centered on the monitor
with a black background. The black background becomes part of the window
geometry.
The surface container is responsible for correctly culling the surfaces
and making sure the surface actors are removed from the actor tree to
avoid destroying them.
The window actor culling implementation assumes all surfaces to be direct
children of said actor. The introduction of the surface_container actor
broke that assumption. This implements the culling interface in
MetaWindowActorWayland which is aware of the actor surface_container and
fullscreen state.
v2: Fix forwarding culling to surface even if there is a background.
v2: Don't alter passed geometry.
v2: Update window geometry code documentation to reflect these changes.
v2: Only use constrained rect if we're acked fullscreen.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2338>
Prepare for adding Wayland specific culling logic to the
MetaWindowActorWayland class by moving all the logic to the non-abstract
classes, since there will be no reason to keep the logic in
MetaWindowActor around.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2338>
This is helpful to know what current state a window actually have, in
contrast to the state in MetaWindow (e.g. MetaWindow::fullscreen) which
is the intended state, be it current or not yet so.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2338>
First make sure we call 'move_resize()' in all cases where the size or
position can change, then move the updating of the buffer rect to the
same place as we update the frame rect. This means keeping track of
surface size changes, in addition to geometry changes, and calling
finish_move_resize() whenever any of those changes, in addition to
acknowledged configurations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2338>
A "window rect" in most places refers to the rectangle the window
corresponds to when it comes to window management. MetaWindow::rect also
refers to this window management related rectangle. However in the
geometry sync functions, it instead called what was to be the rectangle
the actor should have as "window rect", which is arguably a bit
confusing. Fix this by renaming it "actor_rect" so that it becomes clear
that it's the rectangle the actor should get on the stage.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2338>
MetaWindowActor previously peeked at the number of child Actors to
determine the number of surfaces. The following commit rearranged the
tree such that MetaWindowActorWayland always has two Actors. This change
lets the subclass determine if the main surface describes the whole
window.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2338>
When a window configuration is constructed for a Wayland surface it
contains a position, size and a scale. The scale is the geometry scale
for the configuration, i.e. before the size is sent the passed dimension
is divided with the passed scale.
When moving between monitors with different scales, if we use the
existing geometry scale, this means we will send a configure event with
incorrect dimensions. Fix this by calculating the scale used in the
configuration given the rect we're configuring with as this will mean
the correct size will be sent to the client.
v2: Removed the fullscreen condition. Don't know why it was added to
begin with.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2338>
There were some magic conditions that decided when
meta_window_constrain() was to be called or not. Reasoning about and
changing these conditions were complicated, and in practice the caller
knows when constraining should be done. Lets change things by adding a
'constrain' flag to the move-resize flags that makes this clearer. This
way we can, if needed, have better control of when a window is
constrained or not without leaking that logic into the generic
to-constrain-or-not expression.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2338>
We have no way to sanely add safe modes if there are no modes we can
compare with, thus don't try.
Fixes the following crash:
#0 are_all_modes_equally_sized at ../src/backends/native/meta-output-kms.c:284
#1 maybe_add_fallback_modes at ../src/backends/native/meta-output-kms.c:310
#2 init_output_modes at ../src/backends/native/meta-output-kms.c:347
#3 meta_output_kms_new at ../src/backends/native/meta-output-kms.c:414
#4 init_outputs at ../src/backends/native/meta-gpu-kms.c:332
#5 meta_gpu_kms_read_current at ../src/backends/native/meta-gpu-kms.c:368
#6 meta_gpu_kms_new at ../src/backends/native/meta-gpu-kms.c:403
#7 create_gpu_from_udev_device at ../src/backends/native/meta-backend-native.c:461
#8 init_gpus at ../src/backends/native/meta-backend-native.c:551
#9 meta_backend_native_initable_init at ../src/backends/native/meta-backend-native.c:632
Fixes: 877cc3eb7d44e2886395151f763ec09bea350444
Related: https://bugzilla.redhat.com/show_bug.cgi?id=2127801
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2646>
This just checks for any chromaticity being zero and gamma being in
range but we could do a better job at detecting bad data in the future.
Also check the return value of cmsCreateRGBProfileTHR which can be NULL.
Fixes gnome-shell#5875
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2627>
ATM sending modifiers to clients prevents direct scanout for DRI3
clients via Xwayland. Xwayland using the dma-buf feedback v4 Wayland
protocol will solve that, but that might take a while yet to appear in
the wild. Once that happens, this can be reverted.
Direct scanout still works for native Wayland clients as well.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2637>
Make sure that if we wiggle a scan-out capable surface a bit, it won't
scan out if it's not exactly in the right position. Do this by first
making the window not fullscreen, then moving it back and forth,
verifying the correct scanout state for each presented frame.
This test addition reproduces the issue described in
https://gitlab.gnome.org/GNOME/mutter/-/issues/2387.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2624>
If we have a window that match the size (i.e. will pass the "fits
framebuffer" low level check), that doesn't mean it matches the
position. For example, if we have two monitors 2K monitors, with two 2K
sized windows, one on monitor A, and one on monitor both monitor A and
B, overlapping both, if the latter window is above the former, it'll end
up bing scanned out on both if it ends up fitting all the other
requirements.
Fix this by checking that the paint box matches the stage view layout,
as that makes sure the actor we're painting isn't just partially on the
right view.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2387
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2624>
Painting the swap region with CLUTTER_DEBUG_PAINT_DAMAGE_REGION happens
on the view framebuffer, so don't transform the region we paint to the
onscreen.
Fixes the swap region painting on rotated monitors.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2590>
Gnome-shell uses meta_display_focus_default_window() when shell elements
loose focus which is the case with Alt+Tab window switching. Globally
active input clients don't immediately gain focus though so if
meta_display_focus_default_window focuses a wrong window stacking and
focus don't behave as expected.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2489>
New commands to set the number of workspaces, activate a workspace, with
and without focus, move windows to specific workspaces, and check the
stacking on a specific workspace.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2489>
When switching workspaces we previously focused on whatever window is on
top of the stack. If a window is marked as "always on top" then it would
always receive focus when switching workspaces.
Fixes#2240
Fixes gnome-shell#5162
Fixes#178Fixes#678
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2489>
We want to use the workspace MRU list to decide the default focus but
Globally Active Input clients don't call
meta_window_set_focused_internal and therefore don't update the MRU
list. Move the update to meta_window_focus instead.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2489>
The completed signal is only emitted if the timeline actually completed
but when an actor is destroyed or removed from its parent the timeline
is stopped and not completed.
The workspace switch effect removes window actors from the window group
which destroys the timeline so on_$EFFECT_effect_stopped is never
called and the pointer to the timeline is dangling.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2489>
The workspace switch animation moves the WindowActors out of the
WindowGroup so if we shut down while the animation is playing the
WindowActors will have queued a destroy but will be disposed only after
the compositor is destroyed, leaving the WindowActor with a dangling
pointer.
Fix the issue by killing the workspace switch animation on shutdown.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2489>
This is an old relic from when ClutterStageView was being added, and
tests were somewhat prepared to be able to test the "X11 style" of
things, with the nested backend some how managing to emulate that.
Lets drop that stuff, it isn't used by the test suite, and isn't useful
anyway; if we want to test X11 configurations, we should use the actual
X11 backend, which didn't make use of this anyway.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2619>
It exposed unnecessary public and private API, and used a global static
variable instead of a return value, none which was necessary. Remove
both API and use a return value for communicating to the caller.
This doesn't remove a public symbol, lets do that for GNOME 44.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2619>
This was used gala to implement hot corners, and the way the barrier API
works, there isn't really any practical reasons to not make it
derivable, since the backend is a separate type and object.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2626>
This adds a copy of the calibration test profile and sets up a test to
first add it as a system profile, then setting up the XDG_DATA_HOME
directory so that the duplicate profile is detected, added, and later
discarded.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2622>
We might fail with some part of the color profile construction and
initialization. For example there might be a system wide profile with
the same ID as one we attempt to add from a local ICC directory. When
this happens, we should drop these profiles, and use the ones from the
system instead.
Profiles may fail to initialize for other reasons too, e.g.
unpredictable colord errors, or other I/O issues.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2429
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2622>
If our profile wasn't fully initialized, we'd try to clean it up, in an
attempt to handle race conditions by finding synchronously then cleaning
it up, but don't attempt this if the profile is ready, as that means we
didn't create one in the first place.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2622>
This is instead of getting anything from the CdDevice. This avoids a
crash when CdDevice isn't successfully setup but something still tries
to look up the filename of the ICC profile.
This isn't a real bug fix for anything, but there is no reason having to
rely on CdDevice for this anyway, and as we don't really have control of
it, it's less reliable of containing something valid.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2622>
When creating a render device, we create a temporary EGLContext where we
then query the `GL_RENDERER` string to check whether the renderer is any
of the known software renderers. After we're done, we destroy the
context and move on.
This should be fine as according to specification eglDestroyContext(),
with the context being actually destroyed a bit later when it's no
longer current, but mesa, when running RK3399 (Pinebook Pro), this
results in a crash in a future eglMakeCurrent():
#0 in dri_unbind_context () at ../src/gallium/frontends/dri/dri_context.c:266
#1 in driUnbindContext () at ../src/gallium/frontends/dri/dri_util.c:763
#2 in dri2_make_current () at ../src/egl/drivers/dri2/egl_dri2.c:1814
#3 in eglMakeCurrent () at ../src/egl/main/eglapi.c:907
...
We can avoid this, however, by calling eglMakeCurrent() with
EGL_NO_CONTEXT on the EGLDisplay, prior to destroying, effectively
avoiding the crash, so lets do that.
Related: https://gitlab.freedesktop.org/mesa/mesa/-/issues/7194
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2414
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2612>
We'll never scan out, which is why ADDFB2 is required otherwise, and we
won't enable the DMA buffer extension if
'EGL_EXT_image_dma_buf_import_modifiers' is missing, so send modifiers
in this case.
This also happens to avoid crashing when the GPU is null, since we'd
otherwise attempt to dereference it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2578>
EGLStream is incompatible with atomic mode setting, but nvidia-drm when
using libgbm is not, so lets only deny using atomic mode setting when
the render device is an EGLStream based one.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2578>
The type of render device used for a specific GPU affects the mode
setting backend that can be used, more specifically, when the render
device is an EGLStream based one, atomic mode setting isn't possible, as
page flipping is done via EGL, not via atomic mode setting commits.
Preparing the render devices before KMS devices means can make a more
informed decision whether to deny-list atomic mode setting for when
a certain GPU uses a EGLStream based render device instance.
This also means we need to translate mode setting devices to render node
devices when creating the render device itself, as doing it later when
creating the mode setting device is already too late.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2578>
Doing an early out in a constructed() is a bit awkward, and unexpected,
and makes it tricky to call the parents constructed() method (which we
didn't), so clean that up.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2578>
Currently, the peripheral "output" setting will be unset if Mutter is
deciding automatically the mapped output of a tablet device. In that
case, gnome-control-center will have a hard time figuring out itself
the better output to show the tablet calibration UI, unless it's hand
held by Mutter.
Add this private D-Bus interface so that gnome-control-center can look
up the output as determined by Mutter to bring the missing harmony
between both. This interface consists of a simple method to get the
mapped output for a input device node.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2605>
The D-Bus runner used by tests, including installed tests, is made to be
reusable from GNOME Shell. To do this, install it and the templates in
the pkgdatadir (e.g. /usr/share/mutter-APIVERSION/tests/), generate a
custom runner for the installed tests that uses the installed script and
templates, and change the non-installed original runner to use the
non-installed templates.
The end goal is to reuse the D-Bus session runner and templates used for
mutter when test running GNOME Shell.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1354>
A few calculations and assignments are done unnecessarily when the
last next presentation time is invalid. This increases the cognitive
complexity of the function for no reason.
No change in behavior.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2486>
When moving to another monitor the window size may change in some
cases. While unconditionally notifying a size change is not always
correct, it animates the window when moved to another monitor in
GNOME Shell.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2558>
Following the EGL_KHR_swap_buffers_with_damage specification, the
surface damage used by eglSwapBuffersWithDamage does not need to
contain the damage history.
Rework that to initialize swap_region earlier, before appending the
damage history.
This may help optimizing the composition process in some cases (at least
on X11 when EGL_KHR_swap_buffers_with_damage is available) by not
accumulating additional regions as damaged unnecessarily.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2241>
This is what the protocol says we should do, and even though normally
an out of focus client should not have any reason to create IM requests,
there is a bit of a grey area around focus changes, as both the client
losing focus and the client gaining focus may respectively try to
disable/enable in an undetermined order.
Anyways, since in that situation the client losing focus is not aware
of the requests being ignored, the serial should always be incremented
in order not to break accounting of .done/.commit for that specific
client.
Fixes the IM focus being possibly "lost" after changing focus between
clients, if the race condition turned the odds in that direction.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2585>
Mutter can play sounds in some contexts and also provides an API
for libmutter users to do so using libcanberra internally.
In some specific use cases of Mutter, we would like to not depend
on libcanberra and not have any sound playing feature by default.
The changes keeps the sound player API but make it no-op if the
sound_player feature is disabled to not make it possible to break
a gnome-shell build.
See https://gitlab.gnome.org/GNOME/gnome-shell/-/merge_requests/2270
for relevant discussion
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2375>
If the vendor_name was previously successfully determined, we would end
up in the else case, overwriting it with "Unknown vendor" and leaking
the previous vendor_name.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2603>
This allows using two separate ICC profiles for one "color profile",
which is necessary to properly support color transform
calibration profiles from an EFI variable.
These types of profiles are intended to be applied using the color
transformation matrix (CTM) property on the output, which makes the
presented output match sRGB. In order to avoid color profile aware
clients making the wrong assumption, we must set the profile exposed
externally to be what is the expected perceived result, i.e. sRGB, while
still applying CTM from the real ICC profile.
The separation is done by introducing a MetaColorCalibration struct,
that is filled with relevant data. For profiles coming from EFI, a
created profile is practically an sRGB one, but the calibration data
comes from EFI, while for other profiles, the calibration data and the
ICC profile itself come from the same source.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2568>
We may want to use scanout even if the default framebuffer
of the stage view is an offscreen, for example when a Wayland
client provides pre-rotated buffers. The caller is responsible
to ensure this is correct - we already asserted on that before.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2468>
If a stage view uses an offscreen framebuffer exclusively for
rotation and a Wayland client provides pre-rotated buffers,
we should try to use scanout.
This saves us one copy more than scanout in the onscreen case,
i.e. two fullscreen copies in total.
Offscreen rotation is notably used for all 90/270 degree rotations
at the moment, as using display hardware for them is apparently
more complex than for x-/y-flips and can even have detrimental
effects on power consumption.
This can be tested with `weston-simple-egl`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2468>
This mocks gsd-colord to enable night ligth at a given temperature. The
test then verifies that the result exactly matches that of the gamma
ramps gsd-color generated for the same temperature and ICC profile.
There are two types of profiles tested; ones with VCGT, i.e. calibrated
profiles, and ones without. These are tested as the VCGT affects how the
gamma curve looks, while the non-VCGT profiles all only rely on
the blackbody temperature to generate a gamma ramp.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2165>
More or less copied from gnome-settings-daemon. The look up tables are
either calculated based on the VCGT (Video Card Gamma Table) and the
blackbody color for a given temperature, or the blackbody color for a
given temperature alone, if no VCGT is available.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2165>
This means that e.g. custom profiles or calibrated profiles will be
added and registered with colord. This does not use CdIccStore for two
reasons: don't want to generate duplicate entries for auto-generated
EDID or EFI profiles, and we want to store profiles as MetaColorProfile.
It also happens to be the case that CdIcc does synchronous I/O, which
should be avoided everywhere except on startup.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2165>
It will be used to generate gamma look up tables depending on
temperature.
The temperature comes from org.gnome.SettingsDaemon.Color and
depends on the current night light state.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2165>
It uses the org.gnome.SettingsDaemon.Power.Screen D-Bus API. Currently
brightness set if the proxy is not ready are ignored; whether the
brightness value should be cached and set once it appears or whether
color profiles should be reapplied is yet to be decided.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2165>
Instead of passing 4 arguments (red, green and blue arrays as well as a
size), always pass them together in a new struct MetaGammaLut. Makes
things slightly less tedious.
The KMS layer still has its own variant, but lets leave it as that for
now, to keep the KMS layer "below" the cross backend CRTC layer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2165>
In practice, for KMS backend CRTC's, we cache the gamma in the monitor
manager instance, so that anyone asking gets the pending or up to date
value, instead of the potentially not up to date value if one queries
after gamma was scheduled to be updated, and before it was actually
updated.
While this is true, lets still move the API to the MetaCrtc type; the
backend specific implementation can still look up cached values from the
MetaMonitorManager, but for users, it becomes less cumbersome to not
have to go via the monitor manager.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2165>
We created device profiles, that we manage the lifetime of in colord,
but color devices can be assigned profiles other than the ones it was
created for. For example, this can include the standard sRGB profile
provided by colord.
To achieve this, keep track of the default profile of the CdDevice as
the "assigned" color profile of the device. Given this profile
(CdProfile), construct a MetaColorProfile that can then be interacted
with as if it was generated by ourself.
The assigned profile (default profile in colord terms) does nothing
special so far, but will later be used to determine how to apply CRTC
gamma ramps etc.
The sRGB.icc file used in the tests was copied from colord. It was
stated in the repository that it has no known copyright restrictions.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2164>
This works similiarly to how MetaColorDevice works, by creating them
asynchronously then signalling the 'ready' signal when done. Also
similarly to MetaColorDevice, the on-demand sync cleanup on finalize is
added, to avoid race conditions when hotplugs happens very rapidly,
e.g. in tests.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2164>
Just as gsd-color does, generate color profiles. This can either be done
from EFI, if available and the color device is associated with a built
in panel, or from the EDID. If no source for a profile is found, none is
created.
The ICC profiles are also stored on disk so that they can be read by
e.g. colord. The on disk stored profiles will only be used for storing,
not reading the profiles, as the autogenerated ones will no matter what
always be loaded to verify the on disk profiles are up to date. If a on
disk profile is not, it will be replaced. This is so that fixes or
improvements to the profile generation will be made available despite
having run an older version earlier.
After generating, add some metadata about the generated file itself
needed by colord, i.e. file MD5 checksum and the file path.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2164>
Tests that test case EDID is setup correctly, and that color devices for
monitors are created.
tests/color: Add hotplugging tests
Checks that changing the number of connected monitors reflects the
number of current color devices, and that we end up with the correct end
state.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2164>
Ready means it has established the connection to colord and can operate.
Will be used by tests to make sure tests don't fail due to race
conditions when connecting to colord.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2164>
gsd-color provides this API, which exposes details about the night light
state. Currently, gsd-color also turns this state into CRTC gamma
changes, but this will eventually change, and this is a preparation for
this.
The proxy isn't yet used for anything.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2164>
Previously, gsd-color handled adding color devices. It got information
about those via the GnomeRR API, which is part of libgnome-desktop.
libgnome-desktop itself got this information from the
org.gnome.Mutter.DisplayConfig.GetResources() D-Bus method, implemented
by mutter.
Now, mutter itself will add all the monitor color devices itself,
without having to go via gsd-color.
We sometimes need to delete colord devices synchronously, in certain
race conditions when we add and remove devices very quickly (e.g. in
tests). However, we cannot use libcolord's 'sync' API variants, as it
has a nested takes-all main loop as a way to invoke the sync call. This
effectively means we end up sometimes not return from this function in a
timely manner, causing wierd issues.
Instead, create our own sync helper, that uses a separate context that
we temporarly push as the thread-default one.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2141>
This will be needed for adding colord integration without breaking
testing.
The test context is altered to make sure any left over color devices are
cleaned up before starting. This means it becomes possible to run a test
case multiple times without having to restart meta-dbus-runner.py.
Note: Don't use os.getlogin() to get the current username; as that
requires a controlling terminal.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2141>
It's not really about monitors, even though it is used for monitors.
Lets shrink MetaMonitorManager a bit moving it to the backend.
While at it, stop leaking it too.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2141>
What determines whether one MetaMonitor is the same as the other should
be whether the actual monitor is the same. The way to check this is
comparing the EDID vendor/product/serial fields. Whene these are
incomplete, fall back on the 'winsys ID'.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2141>
Same applies to MetaOutput. The reason for this is to make it possible
to more reliably know when there was EDID telling us about these
details. This will be used for colord integration.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2141>
We fairly consistently had multiple monitors with the whole
vendor,product,serial tuple identical. If we start relying on making
monitors a bit more unique, e.g. for colord integration, we need to make
two monitors connected distinguishable in order for tests to properly
reflect reality and excercise the correct colord integration paths.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2141>
As for the types of monitor, X11 and KMS are currently assumed to always be
physical, while the virtual ones are assumed to be virtual. In theory
X11 ones could be virtual, but lets not bother. KMS ones can be virtual
in the case of virtual KMS, but we typically use that for testing as if
it was physical, so lets leave it as such.
Will later be used to feed correct information to colord.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2141>
Create a color manager type that eventually will be the high level
manager of color related behavior, such as ICC profiles and
color "temperature" a.k.a. night light.
For now, it's only an empty shell. It's also constructed by the actual
backend, as at a later point, the X11 and native color management
implementations will differ.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2141>
Client connections may linger after the test driver is teared down;
handle this gracefully by unsetting the user data on the wl_resource,
and make the resource destructor a no-op, instead of where it would
otherwise remove itself from the resource list. This fixes this crash
seen in CI:
Received signal 11 (SIGSEGV)
#0 g_list_remove() at ../glib/glist.c:596
#1 test_driver_destructor() at ../src/tests/meta-wayland-test-driver.c:219
#2 destroy_resource() at ../src/wayland-server.c:730
#3 for_each_helper() at ../src/wayland-util.c:416
#4 wl_map_for_each() at ../src/wayland-util.c:430
#5 wl_client_destroy() at ../src/wayland-server.c:889
#6 wl_display_destroy_clients() at ../src/wayland-server.c:1482
#7 meta_wayland_compositor_prepare_shutdown() at ../src/wayland/meta-wayland.c:441
#8 meta_context_dispose() at ../src/core/meta-context.c:667
#9 g_object_unref() at ../gobject/gobject.c:3863
#9 g_object_unref() at ../gobject/gobject.c:3780
#10 glib_autoptr_clear_GObject() at /usr/include/glib-2.0/gobject/gobject-autocleanups.h:29
#10 glib_autoptr_clear_MetaContext() at ../src/meta/meta-context.h:32
#10 glib_autoptr_cleanup_MetaContext() at ../src/meta/meta-context.h:32
#10 main() at ../src/tests/wayland-unit-tests.c:707
#11 __libc_start_call_main() in /usr/lib64/libc.so.6
#12 __libc_start_main() in /usr/lib64/libc.so.6
#13 _start() in /builds/GNOME/mutter/build/src/tests/mutter-wayland-unit
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2601>
This reverts an attempt at saving texture memory that was introduced
recently in 16fa2100. It was misguided because the same texture may be
needed in the next frame if a window has multiple previews visible on
screen at once (gnome-shell's overview). Keeping the mipmaps around
seems to reduce the peak render times of the overview by roughly 5%-10%.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2598>
Although its atomic KMS support seems to work at first, mode sets to
anything other than the Xilinx preferred max resolution of 2048x1280
would result in a hang. The xlnx kernel driver is given:
`DRM_MODE_ATOMIC_ALLOW_MODESET | DRM_MODE_PAGE_FLIP_EVENT`
and it does complete the mode set without error, but page flip events
never arrive and so you're frozen on the first frame.
Revert to legacy KMS which has no such problem with non-default modes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2596>
I don't see how this makes sense at all, ClutterClickAction really
shouldn't mess with BUTTON_RELEASE events that are not part of a
gesture.
So propagate those events instead of stopping them.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2552>
The source field was removed from ClutterEvent with commit
b644ea1bce because the preferred way of
getting the event actor is now to use the device/sequence actor from the
stage directly.
With crossing events it's not that easy though, as crossing events
explicitly have a source and related actor that doesn't have to be the
same actor as the device actor. Since we kept around the "related" field
there anyway, let's also introduce a "source" field in
ClutterCrossingEvent and return that actor when get_source() is called
on a crossing event.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2551>
Just like we did with the ::captured-event signal, add detail to the
::event signal too. At the first glance this might not seem necessary
since there are individual signals like scroll-event or touch-event that
get emitted at the same time, but these don't exist for touchpad gesture
events and others.
As an easy solution for that, just make it possible to use detail on the
event signal as we did with the caputured-event signal.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2431>
The "activate" and "deactivate" signals on ClutterStage are used by
Cally to track the key-focus when the user is interacting with shell UI.
key-focus only gets tracked while the stage is activated.
Wayland has no concept of the stage receiving focus or not, so right now
the activation state is bound to whether there's a focus_window in
meta_display_sync_wayland_input_focus(). Since display->focus_window is
set pretty much all the time, this effectively binds activation state to
whether the stage holds a grab or not. This is almost good enough, but
it misses cases where key-focus is on the stage without a grab, for
example when keyboard-navigating the panel after using Ctrl+Alt+Tab.
It seems to make more sense to bind the activation state to whether
key-focus is set to an actor or to NULL, so let's do that instead.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2329>
In the timespan between an activation token being created and the
token being used by the activated application, the surface that started
the activation request may end up destroyed/disposed.
In that case, the token would be left with a stale surface pointer,
maybe causing crashes later on. Set up a destroy notification listener
so that we do know to unset the token surface if that situation arises,
this will result in Mutter not considering the token activatable, thus
maybe issuing the "Application needs attention" notification if the
activated surface did not immediately get focus. In any case this is
better than a compositor crash.
A typical situation where this may happen is "Open With..." dialogs,
since those don't live long after launching the application.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2390
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2592>
There is some surface tracking going on here, and all notify handlers
are possibly leaving the linked wl_listener behind. Ensure it is unlinked
in all destroy notification functions.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2588>
Since commit 1bf70334 "tests/runner: Make test runner use the headless
backend", tests are run with the native backend in headless mode, which
will attempt to open each GPU and show a warning (fatal during tests)
if it cannot.
However, in headless mode we might not be logged in on any seat (for
example we might be logged in via ssh instead), which means we might
legitimately not have permission to use any GPUs, even if they exist.
Downgrade the warning to a debug message in this case.
Resolves: https://gitlab.gnome.org/GNOME/mutter/-/issues/2381
Signed-off-by: Simon McVittie <smcv@debian.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2584>
The catch option makes test run via 'catch'[1], which will log
backtraces whenever an abort or segmentation fault happens in any of the
subprocesses. The aim is to enable this when running in CI to help
debugging crashes that only tend to happen in CI.
While it's possible to wrap the whole meson command in 'catch', doing so
doesn't cover the KVM tests, so this option is added instead that covers
both cases.
[1] https://gitlab.gnome.org/jadahl/catch/
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2561>
Until recently, mutter-test-runner called into libraries that
indirectly depend on (mutter's fork of) Clutter, but did not actually
call into Clutter itself. Commit 1bf70334 "tests/runner: Make test
runner use the headless backend" gave it a direct call into Clutter,
which means the runtime linker will fail unless the executable's
RUNPATH is sufficient to find Clutter.
For future-proofing, do the same for the other test executables.
Resolves: https://gitlab.gnome.org/GNOME/mutter/-/issues/2389
Signed-off-by: Simon McVittie <smcv@debian.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2581>
This is a followup to GNOME/mutter!1578
This commit does a couple of things to avoid creating multiple
pipelines per commit.
First, it avoid catch all `when: manual` rules, which might
end up matching custom variables set which might potentially
not be handled.
Secondly it reworks the `workflow:rules:` and the pipeline guard
rules to avoid duplicate pipelines as the gitlab documentation
suggests.
Last, it switches from yaml anchors to the new `reference` gitlab
keyword which is more flexible.
https://docs.gitlab.com/ee/ci/jobs/job_control.html#avoid-duplicate-pipelines
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2534>
This adds the 4 new connector types that mutter didn't know about from
drm_mode.h in the kernel.
Noticed because mutter kept crashing when plugging in a USB-C adapter to
use an external monitor.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2577>
When we e.g. generate switch configs (i.e. the ones from pressing the
Super+P or the switch-config key on laptops), try a bit harder to find a
"good" monitor scale.
With "good", it means pick a scale that was used in a previous
configuration. In practice, this means that if you for example have
configured your external monitor to use a specific scale, then pressed
e.g. "built in only", and then switched back to e.g. "external only" or
"linear", the generated configuration will use the scale that was
previously configured.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2479>
If two modes are roughly the same, they should probably use the same UI
scaling factor. I.e. for the same monitor, if a 4K mode was configured to
have a certain scaling factor, and we generate a new configuration with
a similar sized 4K mode, we should re-use the scale previously
configured; however if we e.g. go from a 4K mode to a FHD mode, we
shouldn't.
This allows implementing better hueristics when using the switch-config
feature, where we'd be less likely to loose the for a certain monitor
mode combination previously configured scaling factor.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2479>
This will eventually help with better hueristics for finding a good
scale. It currently doesn't change much, but the helper will later gain
more functionality that will also help when coming up with mirroring
configs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2479>
The resulting logical monitor was eventually marked as primary anyway,
but without the config being marked as such, various primary properties
was not set e.g. the one on the MetaOutput. Also, tests would fail.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2479>
This condition is inverted of how it should be. Since pad focus relies
on grouped devices lookups (e.g. pads not grouped with a tablet do not
focus surfaces), this fixes issues in pad focus and event propagation to
wayland clients.
Fixes: fff3654941 - wayland: Check input device capabilities in tablet seats
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2573>
This checks that an unmapped but created Wayland window correctly handle
monitor changes. This is specifically added to test an edge case causing
a crash with the following backtrace:
```
...
4) 0x00007ffff78a2a6b in g_assertion_message_expr ()
5) 0x00007ffff7defd5b in meta_window_update_for_monitors_changed () at ../src/core/window.c:3745
6) 0x00007ffff7899758 in g_slist_foreach () at ../glib/gslist.c:885
7) 0x00007ffff7dbe562 in meta_display_foreach_window () at ../src/core/display.c:3185
8) 0x00007ffff7dbe5fd in on_monitors_changed_internal () at ../src/core/display.c:3210
9) 0x00007ffff796f4ff in g_closure_invoke () at ../gobject/gclosure.c:830
10) 0x00007ffff7981316 in signal_emit_unlocked_R () at ../gobject/gsignal.c:3740
11) 0x00007ffff7987699 in g_signal_emit_valist () at ../gobject/gsignal.c:3495
12) 0x00007ffff7987bc2 in g_signal_emit () at ../gobject/gsignal.c:3551
13) 0x00007ffff7d89915 in meta_monitor_manager_notify_monitors_changed () at ../src/backends/meta-monitor-manager.c:3517
...
```
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2554>
If the window didn't have a size, it would still have a monitor, and
when we are asked to update, we must update, as the old monitor might
not be kept around, leaving us vulnerable to use after free.
Avoid not updating the monitor by using the stored IDs (preferred, or
previous) to find suitable logical monitors, with the primary monitor
being the last fallback unless we're completely headless.
This fixes the assert
!window->monitor ||
g_list_find (meta_monitor_manager_get_logical_monitors (monitor_manager),
window->monitor)
in meta_window_update_for_monitors_changed() being hit when a Wayland
window has been created, but not mapped, when a hotplug happens.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2554>
The function finds a suitable logical monitor given the window
rectangle; this wasn't all that clear from the name
"calculate_main_logical_monitor".
This is in preparation for finding a new logical monitor using things
other than the geometry of the window.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2554>
This will allow tests to change monitor resolution. The first argument
is the monitor ID; there is always one monitor added by default, and it
has the id 0. It's currently not possible to add more monitors, so
passing '0' is the only valid way to resize monitors.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2554>
This hasn't worked for a while, since the test always runs the nested
backend, meaning it's a Wayland compositor. To unblock testing window
management in combination to monitor changes, lets remove the
unreachable X11 WM paths, so that we can start using virtual monitors.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2554>
util-private.h includes glib-i18n-lib.h, which requires GETTEXT_PACKAGE
to be defined. The define comes from config.h,
but that cannot be included in headers, so we have to make sure
that any source file that pulls in util-private.h (or a header
that includes it) includes config.h first.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2566>
In some hardware configurations, presentation timestamps could be
missing from some page flip events, leading to a temporary loss of
vblank synchronization.
This occurs at least with AMD GPUs for atomic commits that only update
the cursor plane. [0]
In those cases, it's better to calculate the next update time
according to the last valid presentation timestamp instead of relying
on the dispatch lateness.
[0] https://gitlab.freedesktop.org/drm/amd/-/issues/2030
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2481>
The image build step is prone to race conditions, e.g. token changes. It
also tends to hit sporadic connection errors to FDO's gitlab. To
minimize the risk of these types of issues block pipelines, always retry
the image building step if it failed.
This has the unwanted consequence that changes to the image building
that results in the script actually failing, but right now there doesn't
appear to be a way to distinguish between actual build errors, and the
mentioned race conditions, as both cause the script to fail.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2564>
mutter needs GDK to use the x11 backend. It already calls
gdk_set_allowed_backends ("x11") for this purpose; however, if
GDK_BACKEND=wayland (or any other non-x11 backend possibly) happened to
be in the environment, GDK would fail to initialize at all. This would
result in mutter not registering as X11 window manager, and all X11
clients hanging.
Big thanks to Olivier Fourdan for figuring this out!
v2:
* Restore original value of GDK_BACKEND environment variable after
initializing GDK.
Bug: https://bugzilla.redhat.com/show_bug.cgi?id=2022283
Bug: https://bugs.debian.org/1008992
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2496>
Both the fragend and vertend shader state was called
"CoglPipelineShaderState", which was rather annoying, especially when
the type needs to be exposed outside of the .c file as part of moving
out unit tests. Make the types unique. This also avoids confusing what
type one is looking at.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2555>
It consists of only a macro and build description logic.
Adds a macro for simpler tests that doesn't require a context; unit
tests requiring a context should use the same framework as conform
tests.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2555>
All working tests have already migrated to the test suite using mutter;
move the old unported tests over too, and remove the conform test
framework, as it is no longer used.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2555>
This is in preparation of moving Cogl tests into src/tests, so they can
use the real backend, instead of the franken-backend it some how still
manages to use some how.
This makes them no longer installed. Most mutter tests are yet to be
installed, so leave that for later, since bigger changes are needed for
that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2555>
Now that we support Wayland buffer transforms in all cases, we can
properly report them to outputs.
Also make sure we resend the output geomerty on transform changes.
This partly reverts commit bda9c359
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/344>
This allows the GL fallback path to correctly paint the cursor
if clients pre-rotated the buffer using
`wl_surface::set_buffer_transform`, visually matching the
hardware cursor path.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/344>
They can be quite heavy, as they load up one virtual machine each. If
your system is already busy, this can easily cause them to time out
instead of finish in time, as they all fight over the same limited
amount of CPU and I/O time.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2556>
Failing in `wait_for_effects_completed()` or `wait_for_view_verified()
indicates client- or compositor-bugs. As hitting those is quite likely
during test development, print error messages to simplify debugging.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2246>
The "single pixel buffer" Wayland protocol extension provides a way for
clients to create 1x1 buffers with a single color, specified by
providing the color channels (red, green and blue) as well as the
alpha channel as a 32 bit unsigned integer.
For now, this is turned into a 1x1 texture. Future potential
improvements is to hook things up to the scanout candidate logic and
turn it into a scanout capable DMA buffer.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2246>
When taking the scanout path we still want to clear the
redraw-clip from the stage-view in order to ensure we skip
frames in `handle_frame_clock_frame()` if no new redraw-clip
was recorded.
This was not done previously as the accumulated redraw-clip was
needed for the next repaint, likely under the assumption that
scheduling a scanout repeatedly would be computationally cost-free.
This assumption does not hold in a VRR world.
In order to archive both, an accumulated redraw-clip for the next
paint and frame-skipping during scanout, introduce new API to defer
and accumulate redraw-clips until the next repaint.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2480>
To Wayland clients, it doesn't matter how we configure our onscreen
buffers, since they don't necessarily have the same bandwidth issues
related to mode setting, whichis the primary reason why we disable
modifiers using the udev rule, so simply check whether importing with
modifiers will work at all and advertise modifiers if so is the case.
This might help avoid issues using legacy non-modifiers path in drivers.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2546>
We disable modifiers for two reasons: an udev rule saying so, or the
lack of a working drmModeAddFB2(). However, to the users, this is not
granular enough. While the current user, whether to enable modifiers in
MetaRendererNative, doesn't need more granularity, we want to send
modifiers to Wayland clients even if the onscreen framebuffers should
still be allocated without modifiers.
Prepare for differentiating between how Wayland DMA buffers work and how
onscreen buffer allocation work by separating the relevant device flags.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2546>
Similar to the clutter commits
- Drop all the private structs documentations
- Make use of gi-docgen items linking as much as possible
- Use markdown formatting for code snippets
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2441>
This uses MetaCompositorViewNative to find a candidate surface for
scanout and to keep track of it separately for each view, effectively
allowing each CRTC to use a different buffer for direct scanout.
There are three parts for potentially assigning a buffer for direct
scanout at the compositor level:
1. Finding a candidate surface actor on the view (if any)
2. Attempting to assign the candidate's buffer for direct scanout
3. Updating references relating to the scanout candidate as needed
The three parts were moved in their entirety from being handled by the
MetaCompositorNative to being handled by the MetaCompositorViewNative.
As part of this transition, the logic was also slightly refactored so
that each of the three parts is handled by its own helper function.
This allowed to avoid the use of "goto" statements and hopefully make
the logic easier to read and follow.
The first part mentioned above was changed in this commit to make use
of the new meta_compositor_view_get_top_window_actor () API to get the
top window actor in the view instead of the top window actor on all
views.
The second part and third parts mentioned above weren't changed other
than being done in the context of a view instead of globally.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2526>
All of the checks this function performed internally were already
done before calling it, making it a simple wrapper function without a
meaningful purpose.
Removing this function also reduces the chance of additional checks
being added to the MetaSurfaceActor after it is already chosen as a
scanout candidate.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2526>
This class is meant to hold logic specific to the native backend
in the context of a MetaCompositorView.
Its addition requires making MetaCompositorView inheritable, and an
addition of a virtual function which allows each compositor to create
its own MetaCompositorView instance.
In the case of the MetaCompositorNative, a MetaCompositorViewNative
is created. In all other cases, a MetaCompositorView is created.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2526>
First, add logic in MetaCompositorView to find topmost visible
MetaWindowActor on its view, and expose it through a new API.
Then, queue an update to find the top MetaWindowActor of each
MetaCompositorView in the following cases:
1. The MetaCompositor is in its initial state.
2. The window stack order has changed.
3. A window has changed its visibility.
4. A "stage-views-changed" signal was emitted for a MetaWindowActor.
Finally, perform the queued update in meta_compositor_before_paint (),
and assert that an update isn't queued during painting. This ensures
that the top window actor in the MetaCompositorView remains up-to-date
and available to child classes of MetaCompositor throughout the entire
paint stage.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2526>
The idea is that the state of the MetaCompositorView shall be
up-to-date only in specific scenarios, thus allowing operations
performed on it to be queued and aggregated to be handled in the
right time, and only if they are still necessary.
For example, in a following commit, the top window actor in each
view will be planned (if needed) only once before painting a frame,
rendering the top window actor in the MetaCompositorView potentially
stale in all other times.
Similarly, if a MetaCompositorView is destroyed before the beginning
of the frame, a queued operation to update its top window actor can be
discarded.
As an interface segragation measure, and as part of an attempt to
avoid the use of g_return_if_fail () to check the validity of the
MetaCompositorView's state in multiple places (which is still prone to
human error), the interfaces through which a MetaCompositorView is
made available would only ones where it's state is gurenteed to be
up-to-date.
Specifically, this commit gurentees that the state of the
MetaCompositorView would be up-to-date during the before_paint () and
after_paint () vfuncs exposed to child classes of the MetaCompositor.
The frame_in_progress variable will be used in a following commit to
guarantee that the MetaCompositorView's state is not invalidated during
this time.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2526>
MetaCompositorView is a class which contains compositor logic
specific to ClutterStageViews.
Each MetaCompositorView is "attached" to a ClutterStageView as an
opaque pointer using g_object_set_qdata_full (), and is freed when
the ClutterStageView is destroyed. This ensures that the lifetime of
the MetaCompositorView can't extend beyond the lifetime of its
ClutterStageView.
In a following commit, MetaCompositorView will be expanded to allow
keeping track of the top MetaWindowActor located on each
ClutterStageView.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2526>
This can happen with the native backend if the previous frame clock
dispatch didn't result in any KMS update, e.g. because it was triggered
by an input event, but the HW cursor didn't need updating on the stage
view. (This is likely to happen on some out of multiple stage views,
but might be possible even with a single stage view if the cursor isn't
visible)
We would previously delay next_presentation_time_us by one refresh
interval in this case, which could result in spuriously leaving one
refresh cycle unused.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2498>
Use the dark variant for decorations if the color-scheme preference
indicates that it's preferred, and the client didn't explicitly
pick a variant via the _GTK_THEME_VARIANT hint.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2541>
Although mipmapping is still slower than not mipmapping, commit 16fa2100
simplified N synchronous draw calls per texture tower into just one. So
it's more efficient now, and four years have passed since the throttling
was introduced so people also have better hardware as well as mutter being
generally faster than it used to be. So I am happy to effectively revert
commit c9c32835.
This means antialiasing will remain consistent rather than popping in and
out of existence.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/403
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2538>
For this to pass, pass an explicit Wayland display name to avoid the
display conflict warning that may happen when there is an already
running Wayland display server.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2448>
Often, most of the output consists of a long list of exposed modes for
each monitor. If --short is passed, only pass modes that has properties.
In practice, this means "preferred" modes, "current" modes, and
similarly special cases, which significantly reduces noise.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2448>
Somewhat long overdue... We've been supporting more than a single
pointer for quite a long time now, let's make sure things don't break if
two pointer devices enter the same ClutterActor: Count the number of
pointers an actor has instead of using a simple boolean value.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2348>
In certain edge cases it's currently possible that an actor never
gets a valid allocation and paint volume.
One such case is adding an unmapped, hidden child to an unmapped
cloned parent and then showing the child. This happens currently
e.g. if a Wayland subsurface is added to a already mapped window
while the user is in the overview.
Ensure relayouts in two more such cases.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2530>
This one does proper error reporting. Via Javascript, barriers are
constructed directly via GObject construction, which currently can't
handle error reporting, but when calling from C we can. However, if we
initialize using GInitable, and use that in our constructor method, once
gjs gains support for construction using GInitable, including the error
reporting, we'll automatically get proper error reporting to Javascript.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2442>
In the past, barries were added to the window management X11 display
instance window table, and then special cased when iterating over the
list.
Since then, barriers, which are really part of the backend, has stopped
being added to the window hash table, instead being managed by the
backend. Lets clean up the left-over special casing that is no longer
needed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2442>
Prior to this commit, barriers were created with a MetaDisplay pointer,
despite being entities related and owned by the backend. In the X11
case, it was also not hooked up to the backend X11 connection, but the
clutter one, meaning for example that the logic was active (but dormant)
also for the Xwayland connection.
Fix this by moving X11 barrier management and event processing fully to
the backend. Also replace passing a display pointer with passing a
backend pointer. Keep the display pointer around for a release, but mark
it as deprecated.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2442>
It papered over wrong `meta_rectangle_transform()` behaviour for
non-flipped output transforms. Also there is no obvious reason
why we would need inverted values here.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2459>
- Drop bogus `meta_monitor_transform_invert()`. It papered over
wrong `meta_rectangle_transform()` behaviour for non-flipped
output transforms.
- Update `scale_and_transform_cursor_sprite_cpu` to match the GL
pipeline matrix in `MetaShapedTexture`, fixing several of the
flipped cases. Note: the rotation applied is the one a client would
need to apply to the buffer for a given monitor transform.
- While on it, drop a redundant `return`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2459>
With `META_MONITOR_TRANSFORM` values matching their `WL_OUTPUT_TRANSFORM`
counterparts, the definition from the Wayland spec applies: the
`META_MONITOR_TRANSFORM` value tells us how the output was rotated
and that the buffer was drawn by the client to compensate for that.
The matrix describes the transformation from surface- to buffer-
coordinates, so the operation we need here is the same one that
the client applied (not from buffer- to surface-coordinates, i.e.
the inverse).
While on it fix `FLIPPED_90` and `FLIPPED_270` to use the correct
axes: flip on the x-axis, rotation on the z-axis.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2459>
`meta_rectangle_transform()` is used in the stack to *compensate* for a
`MetaMonitorTransform` applied to a output, not to apply it again.
Change the function accordingly.
Context:
Experimenting with direct scanout on offscreen-rotated outputs revealed
that the 90/270 degree cases were actually interchanged.
Further digging revealed that we use `meta_rectangle_transform()` with
those values swapped in every single case, papering over the issue.
Either a unintuitive and unexplained `meta_monitor_transform_invert()`
was added, in which case "flipped" values would be wrong, or, in case
of Wayland buffer transforms, the values were swapped by interpreting
the Wayland enums accordingly, see commit 8d9bbe10.
Swapping the 90/270 degree values in `meta_rectangle_transform()`:
1. fixes hardware cursor positioning with flipped output transforms
2. fixes rendering issues with offscreen-rotated flipped output transforms
3. allows us to drop unexplained `meta_monitor_transform_invert()`s in
follow-up commits
4. allows us to make `META_MONITOR_TRANSFORM` and `WL_OUTPUT_TRANSFORM`
enums match again (reverting 8d9bbe10, as already done)
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2459>
The following implicit definition for `transform()` did not
correctly apply:
```
a * b = c
c * invert(b) = a
```
Crucially the following did not apply for `FLIPPED-90`
and `FLIPPED-270`:
```
a * invert(a) = identity
```
Fix this by applying the operations, first the flip, then the
rotation, in this order and add tests to ensure correct results
for the requirement above.
Also drop `relative_transform()` as it only had a single user and
can be replaced by `transform()`:
```
invert(a) * b = c
a * c = b
```
As this is not very intuitive, ensure in tests as well.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2459>
Checking for both bits at once means only one matching bit is
sufficient - very likely in case of `rotate-0'.
This fixes crashes on hardware that does not support 'reflect-'
bits when setting a flipped output transform.
While on it, also update the check for `reflect-y` instead of
`reflect-x` + `rotate-180`. They are logically equivalent,
however some hardware may support `reflect-y` but not both
other bits.
Fixes commit 4e3f3842a1
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2459>
As testing of direct scanout revealed, `META_MONITOR_TRANSFORM`
does actually match `WL_OUTPUT_TRANSFORM` enums. The fact that
things rendered correctly with 90/270 degree values swapped
was because other parts of the stack got the interpretation
wrong, most notably `meta_rectangle_transform()`.
Thus lets revert this change and fix the stack accordingly.
This reverts commit 8d9bbe109b.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2459>
The notification list in the GNOME Shell calendar popup triggers some
interesting interactions when closing a notification:
- Close button is clicked
- The notification animates to be hidden
- The next notification ends up hovered as a result of the animation
- The notification being hovered sets its close button as non-transparent
and reactive
- The pointer is now again over a close button
At this point the reactiveness change should trigger a repick, so that
the new notification's close button is picked, and future button presses
are directed to it, but we do not handle this situation.
To fix this, handle actors becoming reactive so that if the closest
reactive parent has a pointer, it will be repicked again just in case
the pointer is over the newly reactive actor.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2364
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2532>
The function that currently invalidates pointers over an specific actor
also asserts for the situations where this invalidation makes sense to
happen (i.e. the actor became unmapped, or non-reactive).
We want to have a function that is more forgiving, and that doesn't
enforce any guarantees about the pointer focus actually changing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2532>
We started to report resource changes using prediction when an update
had been successfully committed. While at it, gamma changes were
reported too, but this was problematic, as gsd-color will listen for the
MonitorsChanged D-Bus signal and naively set the gamma again, even if it
didn't change. There aren't currently any actual use cases for being
told when gamma changes from a prediction, so just ignore it and just
report privacy screen changes.
This avoids a feedback loop between mutter and gsd-color.
Fixes: 81b28a1d97
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2531>
Fixes memory leak:
==995170== 936 (40 direct, 896 indirect) bytes in 1 blocks are definitely lost in loss record 15,090 of 15,641
==995170== at 0x48445EF: calloc (vg_replace_malloc.c:1328)
==995170== by 0x4B211D0: g_malloc0 (gmem.c:155)
==995170== by 0x4A56693: meta_wayland_tablet_manager_new (meta-wayland-tablet-manager.c:109)
==995170== by 0x4A56693: meta_wayland_tablet_manager_init (meta-wayland-tablet-manager.c:126)
==995170== by 0x4A3FA95: meta_wayland_compositor_new (meta-wayland.c:626)
==995170== by 0x49C7FA7: meta_context_start (meta-context.c:412)
==995170== by 0x10F065: main (mutter.c:148)
Fixes: 745cb67988 ("wayland: Initialize the MetaWaylandTabletManager")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2527>
Fixes memory leak:
==995170== 383 (96 direct, 287 indirect) bytes in 1 blocks are definitely lost in loss record 14,733 of 15,641
==995170== at 0x483F7B5: malloc (vg_replace_malloc.c:381)
==995170== by 0x4B21178: g_malloc (gmem.c:125)
==995170== by 0x4B395C0: g_slice_alloc (gslice.c:1072)
==995170== by 0x4B0766D: g_hash_table_new_full (ghash.c:1071)
==995170== by 0x4A4A8B4: meta_wayland_compositor_update_outputs (meta-wayland-outputs.c:483)
==995170== by 0x4A4ABAB: meta_wayland_outputs_init (meta-wayland-outputs.c:716)
==995170== by 0x4A3FA65: meta_wayland_compositor_new (meta-wayland.c:620)
==995170== by 0x49C7FA7: meta_context_start (meta-context.c:412)
==995170== by 0x10F065: main (mutter.c:148)
v2:
* Use meta_backend_get_monitor_manager. (Jonas Ådahl)
Fixes: 9a4783e364 ("Integrate the monitor manager with wayland")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2527>
Fixes memory leak:
==995170== 288 (96 direct, 192 indirect) bytes in 1 blocks are definitely lost in loss record 14,607 of 15,641
==995170== at 0x483F7B5: malloc (vg_replace_malloc.c:381)
==995170== by 0x4B21178: g_malloc (gmem.c:125)
==995170== by 0x4B395C0: g_slice_alloc (gslice.c:1072)
==995170== by 0x4B0766D: g_hash_table_new_full (ghash.c:1071)
==995170== by 0x4A4F973: meta_wayland_init_presentation_time (meta-wayland-presentation-time.c:222)
==995170== by 0x4A3FB04: meta_wayland_compositor_new (meta-wayland.c:635)
==995170== by 0x49C7FA7: meta_context_start (meta-context.c:412)
==995170== by 0x10F065: main (mutter.c:148)
Fixes: dccc60ec3e ("wayland: Implement stub presentation-time")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2527>
Fixes memory leak:
==995170== 240 (48 direct, 192 indirect) bytes in 1 blocks are definitely lost in loss record 14,428 of 15,641
==995170== at 0x48445EF: calloc (vg_replace_malloc.c:1328)
==995170== by 0x4B211D0: g_malloc0 (gmem.c:155)
==995170== by 0x4A3CDB3: meta_wayland_activation_init (meta-wayland-activation.c:383)
==995170== by 0x4A3FB0C: meta_wayland_compositor_new (meta-wayland.c:636)
==995170== by 0x49C7FA7: meta_context_start (meta-context.c:412)
==995170== by 0x10F065: main (mutter.c:148)
Fixes: ec390b68c5 ("wayland: Implement the xdg-activation protocol")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2527>
Fixes memory leak:
==995170== 192 (96 direct, 96 indirect) bytes in 1 blocks are definitely lost in loss record 14,260 of 15,641
==995170== at 0x483F7B5: malloc (vg_replace_malloc.c:381)
==995170== by 0x4B21178: g_malloc (gmem.c:125)
==995170== by 0x4B395C0: g_slice_alloc (gslice.c:1072)
==995170== by 0x4B0766D: g_hash_table_new_full (ghash.c:1071)
==995170== by 0x4A3F3A4: meta_wayland_compositor_init (meta-wayland.c:477)
==995170== by 0x4E1F509: g_type_create_instance (gtype.c:1929)
==995170== by 0x4E03DFC: g_object_new_internal (gobject.c:2011)
==995170== by 0x4E0538C: g_object_new_with_properties (gobject.c:2181)
==995170== by 0x4E05D40: g_object_new (gobject.c:1821)
==995170== by 0x4A3F8C4: meta_wayland_compositor_new (meta-wayland.c:590)
==995170== by 0x49C7FA7: meta_context_start (meta-context.c:412)
==995170== by 0x10F065: main (mutter.c:148)
Fixes: 8df2a1452c ("wayland: Notify actively of xwayland window/surface associations")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2527>
Fixes potential use-after-free during mutter shutdown, e.g.:
==993876== Invalid read of size 8
==993876== at 0x4A4FCA3: meta_wayland_presentation_time_ensure_feedbacks (meta-wayland-presentation-time.c:373)
==993876== by 0x4A3F07F: on_presented (meta-wayland.c:282)
==993876== by 0x661B7E9: ??? (in /usr/lib/x86_64-linux-gnu/libffi.so.8.1.0)
==993876== by 0x661A922: ??? (in /usr/lib/x86_64-linux-gnu/libffi.so.8.1.0)
==993876== by 0x4DFF4BC: g_cclosure_marshal_generic_va (gclosure.c:1648)
==993876== by 0x4DFE948: _g_closure_invoke_va (gclosure.c:893)
==993876== by 0x4E17498: g_signal_emit_valist (gsignal.c:3406)
==993876== by 0x4E176BE: g_signal_emit (gsignal.c:3553)
==993876== by 0x51D9DB5: clutter_stage_view_notify_presented (clutter-stage-view.c:1226)
==993876== by 0x499ACD2: frame_cb (meta-stage-view.c:83)
==993876== by 0x499ACD2: frame_cb (meta-stage-view.c:43)
==993876== by 0x50CAA41: notify_event (cogl-onscreen.c:175)
==993876== by 0x50CAA41: _cogl_onscreen_notify_complete (cogl-onscreen.c:545)
==993876== by 0x4A877F5: meta_onscreen_native_notify_frame_complete (meta-onscreen-native.c:211)
==993876== Address 0x24b7be58 is 296 bytes inside a block of size 344 free'd
==993876== at 0x484217B: free (vg_replace_malloc.c:872)
==993876== by 0x4E1F88B: g_type_free_instance (gtype.c:2001)
==993876== by 0x49C793C: meta_context_dispose (meta-context.c:675)
==993876== by 0x4E037E0: g_object_unref (gobject.c:3636)
==993876== by 0x4E037E0: g_object_unref (gobject.c:3553)
==993876== by 0x10F145: glib_autoptr_clear_GObject (gobject-autocleanups.h:27)
==993876== by 0x10F145: glib_autoptr_clear_MetaContext (meta-context.h:32)
==993876== by 0x10F145: glib_autoptr_cleanup_MetaContext (meta-context.h:32)
==993876== by 0x10F145: main (mutter.c:126)
==993876== Block was alloc'd at
==993876== at 0x483F7B5: malloc (vg_replace_malloc.c:381)
==993876== by 0x4B21178: g_malloc (gmem.c:125)
==993876== by 0x4B395C0: g_slice_alloc (gslice.c:1072)
==993876== by 0x4B39C29: g_slice_alloc0 (gslice.c:1098)
==993876== by 0x4E1F544: g_type_create_instance (gtype.c:1901)
==993876== by 0x4E03DFC: g_object_new_internal (gobject.c:2011)
==993876== by 0x4E0538C: g_object_new_with_properties (gobject.c:2181)
==993876== by 0x4E05D40: g_object_new (gobject.c:1821)
==993876== by 0x4A3F864: meta_wayland_compositor_new (meta-wayland.c:585)
==993876== by 0x49C7FA7: meta_context_start (meta-context.c:412)
==993876== by 0x10F065: main (mutter.c:148)
Fixes: 2ce3a050f0 ("wayland: Wire up presentation-time machinery")
Fixes: 8cff3b84f7 ("wayland/compositor: Process frame callbacks on 'after-update'")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2527>
When we change the privacy screen, we added a result listener to the KMS
update object to notify the upper layer about the privacy screen state
change. This was slightly awkward as one might have changed the state
multiple times for a single update, thus it was necessary to remove any
old result listeners to an update before adding a new one.
Doing this will not be possible when updates are fully async and managed
by the KMS impl device.
To handle this, instead make the post-commit prediction notify about
changes that happens in response to a successfully committed update. We
already predicted the new privacy screen state, so the necessary change
was to plumb the actual change into a callback which emits the signal if
there actually was a privacy screen change.
This will then be communicated via the same signal listener that already
listens to the 'resources-changed' signal.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2340>
The name had a bit conceptual conflicts with MetaKmsUpdate, as it shared
its namespace but had no relation to it. Fix this by renaming it
MetaKmsResourceChanges (and the corresponding META_KMS_UPDATE_CHANGE_*
to META_KMS_RESOURCE_CHANGE_*). The term "resource" is used since that's
already used in the signal, and the fact that the changes partly comes
from changes in the DRM resource as retrieved by drmModeGetResources.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2340>
With this header we can set a flag to signal that the whole buffer is
corrupt and should be ignored. With this we can cater to all cases:
* Window buffer fine, but cursor broken:
Use the spa_meta_cursor properties like id or offset accordingly
* Window buffer broken, but cursor fine:
Use the chunk flags
* Both are broken / the dequeued buffer is not usable
Use the spa_meta_header flag
Additionally clients can now check if a buffer contains spa_meta_header
data and can thus only check for the new or the old behaviour.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2516>
Since the commit below, meta_crtc_kms_get_cursor_renderer_private has
returned a CrtcCursorData pointer, but this code was still treating it
as a MetaDrmBuffer pointer.
Fixes: fea8ebcca9 ("cursor-renderer/native: Store struct in CRTC private")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2524>
`clutter_actor_iter_destroy` will try to match up the iterator's `age`
with that of the parent ("root") actor:
```
g_return_if_fail (ri->age == ri->root->priv->age);
```
In a simple actor graph that's completely reasonable but somewhere in the
more complex graph of gnome-shell the parent's `age` was skipping ahead
faster than that of the iterator. This could happen in theory if the
destroy indirectly leads to more children being destroyed than the
iterator has visited.
So there's no evidence of actual corruption, only the age check might
fail in a `clutter_actor_iter_destroy` loop because the age check itself
can't handle all possible valid scenarios.
Since our only mandate is to destroy all children, we can do that reliably
without an iterator and thus without assuming anything about the parent's
`age` counter.
Fixes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/4747
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2074>
PipeWire supports flags to signal a corrupted buffer. We should use the
flag SPA_CHUNK_FLAG_CORRUPTED for `chunk->flags` instead of setting
`chunk->size = 0` since the size isn't well defined for arbitrary dmabufs
and should be set to 0.
Sadly clients like obs are using a chunk size of 0 to decide if a buffer
should be imported. Thus we should offer both until clients are using
the flag.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2323>
Change meta_seat_impl_notify_discrete_scroll_in_impl to receive 120
based values and report high-resolution scroll values as smooth scroll.
Notify discrete scroll only when the accumulated value reach 120.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1962>
In order to get the delta X/Y value of the
LIBINPUT_EVENT_POINTER_SCROLL_FINGER
or LIBINPUT_EVENT_POINTER_SCROLL_CONTINUOUS events the new function
libinput_event_pointer_get_scroll_value should be used instead of
libinput_event_pointer_get_axis_value.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1962>
Ignore deprecated LIBINPUT_EVENT_POINTER_AXIS events and handle
LIBINPUT_EVENT_POINTER_SCROLL_WHEEL,
LIBINPUT_EVENT_POINTER_SCROLL_FINGER and
LIBINPUT_EVENT_POINTER_SCROLL_CONTINUOUS instead.
The scroll source is now encoded in the event type making
libinput_event_pointer_get_axis_source and translate_scroll_source
redundant.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1962>
When building the list of formats to be sent as part of the scanout
tranche, avoid requiring modifier support by the DRM driver for
formats relying on implicit modifiers (DRM_FORMAT_MOD_INVALID).
Specifically, the previous check required the DRM driver to have
advertised some modifier support for the given format in its
IN_FORMATS KMS plane property, regardless of modifier it was. If it
hadn't, the format was left out of the list of formats to be sent
in the scanout tranche.
When no formats remained to be sent in the scanout tranche, the
tranche simply wasn't sent.
This resulted in the scanout tranche never being sent for GPUs where
modifiers aren't supported. In those cases, no formats are advertised
using the IN_FORMATS property, and thus the list of formats to be sent
in the scanout tranche remained empty.
Since Mesa doesn't use scanout-compatible buffers for native Wayland
clients unless specifically requested to do so using the "scanout"
tranche flag, it effectively means that direct scanout of native
Wayland clients wasn't supported for GPUs without modifiers support.
Sending a tranche with formats paired with the implicit modifier
(DRM_FORMAT_MOD_INVALID) is both allowed by the protocol and is
already done by default for GPUs with modifiers support, unless the
experimental support for explicit modifiers is enabled in Mutter.
So instead of requiring modifiers to be supported for each format
being evaluated for the scanout tranche, when processing formats
which rely on implicit modifiers, only check if the format in
question is supported by the DRM driver for scanout on the primary
plane.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2510>
While the check for `clutter_actor_has_mapped_clones` clearly indicates
an intention to take clones into account, the following code
does not do so, likely because it predates the introduction of
`clutter_actor_is_effectively_on_stage_view()`.
Switch to that newer API in order to take clones into account. This
avoids unnecessary `wl_surface_send_enter()` and `wl_surface_send_leave()`
events when entering the overview, reducing client work.
This also avoids unnecessarily allocating a `cairo_region_t`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2502>
`clutter_actor_set_child_at_index()` is far from a no-op, even if
the current index is equal to the new one - presumably for good
reasons. For the use-case here we want it to be a no-op though, so
skip calling it if the index already matches.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2501>
Unparenting the surface actor when the subsurface object is destroyed
has several issues:
- subsurface actors can be unparented while a close animation is
still ongoing, breaking the animation for e.g. Firefox.
- adding and removing the actor to/from the parent is not handled in
one place, making the code harder to follow.
- if the destroyed subsurface had children of its own, they potentially
stick around until a surface-tree rebuild. This makes the Firefox
hamburger menu not close with the "compositor" backend.
Move the unparenting back to
`meta_window_actor_wayland_rebuild_surface_tree()` and instead just
notify the parent of a state change, if it still exist. This will ensure
a correct mapping between the subsurface node tree and the flat surface
actor list. In case of the closing animation the parent will already be
removed and the call is skipped.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2501>
Since b443bd42ac, we unmanage a wayland window when clearing its
transient parent. That's to make sure that xdg-foreign doesn't
leave the dialog around after the imported surface was destroyed.
While that behavior is sound, it is problematic to implement it
by unmanaging the window, as that happens entirely behind the
client's back.
Instead, send a close event for the window. Unless the client has
good reasons, it should honor the request. (And if it has good
reasons - like unsaved work - then effectively hiding the window
from both the user and client is probably not the best idea anyway).
https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/5458
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2499>
Not all windows can be minimized: X11 clients can disable the
functionality, and so do we for windows that aren't shown in
the alt-tab popup or the shell overview, so there would be no
way of getting them back.
While we make sure that we respect that ourselves (keybinding,
window menu, etc.), we don't guard meta_window_minimize(), so
clients or extensions can still minimize a window that isn't
supposed to be minimized.
That can lead to all kinds of issues, from the hidden window
being lost (as far as users are concerned) to a crash when
the minimzed window has a transient parent.
Just add an explicit check to make sure the unexpected doesn't
happen after all, and print a warning if it does.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2491>
The cursor rendering code path used by the screen cast code relies on
the cursor tracker machinery to determine where to blit the cursor
texture, but at the moment the cursor position invalidation is behind
a check for whether the shell is using a Wayland backend. (This code
path used to be Wayland-specific before 00cbcb7ba1 but has been
backend-agnostic since).
This commit removes the check for a Wayland compositor, allowing
cursor drawing to function correctly on X11 when screen casting in
embedded cursor mode.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1780
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2474>
The test case checks that the stage views of hidden actors are
not updated when the views of the visible outer parent change.
The check for the outer parent's updated stage views currently
relies on ClutterFixedLayout not excluding hidden children in
its size request: As the container doesn't contain any visible
children at that point, its size would change to 0x0 and end
up on no stage view (rather than the assumed two).
Avoid that oddity by giving the outer container a fixed size,
so that the visibility of its child doesn't affect the test
when we fix ClutterFixedLayout.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2488>
This adds support for E-EDID extensions. Tags are allocated by VESA and
the CTA has such an extension defined in CTA-861.
The switch in `decode_ext_cta` is empty in this commit because we don't
parse any CTA-861 data blocks, yet.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2351>
The EDID code is copy from elsewhere, without adapting to conventions
regarding e.g. API and types. Clean this up a bit, as EDID information
will be kept around longer when possible, to be used e.g. by color
management.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2351>
The test aims to test that trying to fetch X11 clipboard content after
Xwayland went away doesn't cause issues. What happens though is that
sometimes the clipboard content doesn't have time to settle (i.e. fetch
mime types etc) before Xwayland gets terminated, which causes flakyness.
Fix this by waiting for the compositor side clipboard owners to finish
setting up before continuing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2484>
The pixel clock determines how fast pixels can be processed. When adding
non-native common modes, avoid adding modes that exceed the max pixel
clock frequency of the native modes. Avoiding these avoids potential
mode setting failures where the GPU can't handle the modeline since the
configured pixel clock is too fast. This replaces the "bandwidth" check
which used the number of pixels and refresh rate, which wasn't enough to
avoid incompatible modes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2492>
'screen-cast/monitor-src: Use clutter_stage_paint_to_buffer'
(6c818cd8d5) made the non-dma-buf path use
clutter_stage_paint_to_buffer() to avoid running into direct scanout
issues. At a glance, the dma-buf paths didn't have the same issue since
it explicitly handled dma-bufs by blitting them.
What it also did was move the recording to an idle callback, to avoid
paint reentry issues. A side effect of this, however, is that it also
broke the dma-buf paths, as they rely on the back buffer existing, and
the stage view direct scanout already being setup, which it isn't in an
idle callback.
Fix this by using the dma-buf variant of
clutter_stage_paint_to_buffer(): clutter_stage_paint_to_framebuffer().
This has some negative performance impact, but we can't use
cogl_blit_framebuffer() when using an idle callback for recording.
Potential performance improvements to make things work more as they did
before is to enhance 'cogl_blit_framebuffer()' a bit, making it a vfunc
that could be implemented by MetaOnscreenNative. A flag to say whether
to look at the back or front buffer would let MetaOnscreenNative know
whether to use the already committed-to-KMS buffer, or the current back
buffer.
Fixes: 6c818cd8d5
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2282
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2462>
Fixes leak:
==14889== 2,168 (16 direct, 2,152 indirect) bytes in 1 blocks are definitely lost in loss record 15,308 of 15,584
==14889== at 0x48445EF: calloc (vg_replace_malloc.c:1328)
==14889== by 0x4BAC1D0: g_malloc0 (gmem.c:155)
==14889== by 0x4AAFF60: meta_wayland_dma_buf_feedback_new (meta-wayland-dma-buf.c:298)
==14889== by 0x4AAFFE0: meta_wayland_dma_buf_feedback_copy (meta-wayland-dma-buf.c:317)
==14889== by 0x4AB16B6: ensure_surface_feedback (meta-wayland-dma-buf.c:1121)
==14889== by 0x4AB1848: dma_buf_handle_get_surface_feedback (meta-wayland-dma-buf.c:1169)
==14889== by 0x66F77E9: ??? (in /usr/lib/x86_64-linux-gnu/libffi.so.8.1.0)
==14889== by 0x66F6922: ??? (in /usr/lib/x86_64-linux-gnu/libffi.so.8.1.0)
==14889== by 0x5318750: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-server.so.0.20.0)
==14889== by 0x5313B99: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-server.so.0.20.0)
==14889== by 0x5316649: wl_event_loop_dispatch (in /usr/lib/x86_64-linux-gnu/libwayland-server.so.0.20.0)
==14889== by 0x4AA7C19: wayland_event_source_dispatch (meta-wayland.c:110)
Fixes: 64e6bedb6b ("wayland/dma-buf: Add support for scanout surface feedback")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2469>
The MetaKeyboardA11yFlags are used by gnome-shell to show a dialog
whenever a keyboard accessibility feature is switched using the
keyboard.
Unfortunately, commit c3acaeb25 renamed the Clutter flag to Meta and
moved them to a private header. As a result, gnome-shell do not show any
dialog anymore when a keyboard accessibility feature is activated.
Move the MetaKeyboardA11yFlags definition to a public header so that
gnome-shell can use it.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2306
Fixes: c3acaeb25 - backends: Move keyboard a11y into backends
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2463>
The min distance to the right/bottom edge depends on Wayland concepts
(wl_fixed_t) and eventually geometry scale. Move the logic the Wayland
side of the pointer constraints machinery to avoid the backend trying to
figure this out without the proper data.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2460>
There were some coordinate nudging to avoid running into Clutter
floating point math issues related to coordinate transformations. Over
the years these things have improved, especially with the move to
graphene, so remove the old work around.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2460>
The ImplDeviceAtomic converts the MetaKmsPlaneRotation back to the
concrete KMS value. The MetaMonitorTransform is always directly
converted to a MetaKmsPlaneRotation.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2379>
Updating the PropTable has the side effect that the parse callback now
also gets called on hotplug but it is used to initialize data. The parse
callbacks are moved to the read_state functions which are aware if this
is an initializing call or just an update.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2379>
* creating an actor will result in it being assigned a color state
with the color space sRGB
* creating an actor with a color state passed will result in that
color state being returned
* changing an actor's color state makes that happen
* changing an actor's color state to NULL ends up with it being
changed back to a color state with the sRGB color space
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2443>
ClutterColorState, that is a GObject. each ClutterActor would own
such an object, and it'd be set via a GObject property.
It would have an API to get the colorspace, whether the actor
content is in pq or not, and things like that.
if it is NULL, it will default to color state with sRGB colorspace.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2443>
This avoids the following error:
../src/tests/wayland-test-clients/dma-buf-scanout.c:100:5: error:
implicit declaration of function ‘close’; did you mean ‘pclose’?
[-Werror=implicit-function-declaration]
100 | close (buffer->dmabuf_fds[i]);
| ^~~~~
| pclose
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2458>
There are no 'features' left, the last one, GLSL shader support, was
moved to Cogl.
This also move the Cogl context creation to a more sensible place, as it
was hidden away in the feature initialization.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2015>
The Cogl feature was removed a while back, while Clutter just hard coded
it to TRUE. Lets remove the confusion that GLSL isn't supported and just
remove the (dead) fallback paths.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2015>
Add `sync_effects_completed()` and `verify_view()` in
order to allow Wayland test clients to trigger verifications
and add convenience functions to use them to client-utils.
Notes:
- `sync_effects_completed()` works in two stages in order
to ensure it doesn't race with window effects. By the time
`sync_effects_completed()` is processed, an effect could
already have ended or not yet been scheduled. Thus we
defer a check for pending effects to the next paint cycle,
assuming that by then they should have been scheduled.
- `meta_ref_test_verify_view()` internally triggers the
`paint` signal for the stage which is why it can not be run
in the after-paint signal handler.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1055>
Version 2 is required for buffer transform, however directly going
for the highest currently supported version doesn't break any
tests and makes more features available.
Also fix indentation below while on it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1055>
Our internal interpretation of output transforms is not in line with
the Wayland spec. Wayland describes them as the transform that a
compositor will apply to a surface to compensate for the rotation
or mirroring of an output device - counter-clockwise.
Mutter in turn interprets it the other way around. One could
argue it does the same but clock-wise - or it interprets the transform
from the viewpoint of the content, not the device.
In either way, the difference is that 90 and 270 degree values are
switched. Thus swap these accordingly when we translate from
`WL_OUTPUT_TRANSFORM` to `META_MONITOR_TRANSFORM`.
See: https://gitlab.freedesktop.org/wayland/weston/issues/99
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1055>
This launches Xvfb, using xvfb-run, and inside tests the following:
1. Launching 'mutter --x11' works
2. Launching a couple of X11 clients works (doesn't crash or result in
warnings)
3. Launching 'mutter --x11 --replace' works
4. Terminating works
It does this using a simple shell script.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2434>
We don't make use of the refresh rate in any useful way in the X11, and
in this case we just ended up with warnings since the refresh rate was
NaN. Fix this by making it 0.0 to mean "no refresh rate". This also is
what 'xrandr' itself reports.
Fixes warnings when launching 'mutter --x11' in Xvfb.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2434>
This adds a minimalistic fullscreen direct scanout test case, that runs
on vkms. It doesn't use EGL, and it uses uninitialized memory, thus it
lacks any kind of implicit synchronization, but it does test that the
scanout selection paths are working.
What is tested is:
* DMA buffer allocated using gbm on top of VKMS
* Buffer passes a mode setting TEST_ONLY check
* Paint is omitted
* Correct buffer active in KMS after presentation
What isn't yet tested:
* Implicit synchronization related behavior
* Presented pixel content
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2417>
They're just typedefs to CoglObject and CoglObjectClass, respectively,
and the latter is unused already. The former is used in a couple of
deprecated headers, which can easily be changed.
Change all CoglHandleObject instances to CoglObject, and drop the types
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2355>
There is no need to make CoglJournal a CoglObject, nor any kind of
object for that matter, since it doesn't require refcounting at all.
CoglJournal is entirely private to, and managed by, CoglFramebuffer,
and it only needs to create and destroy it.
Make CoglJournal a free-form struct, and adjust CoglFramebuffer to
call _cogl_journal_free() instead of cogl_object_unref().
Related: https://gitlab.gnome.org/GNOME/mutter/-/issues/2087
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2355>
Cleanup all the boilerplate, and port the function to use the auto
generated private helper. Remove the manual autocleanup declaration
since this is now done in the clutter-image.h header.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2355>
We passed the pointer to a GError * as user data on an async I/O call.
The callback function didn't make use of it, so it was never written to,
thus remained NULL, thus was dead code. Remove it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2446>
It works by using an X11 client to set the clipboard content, using a
mimetype that on purpose is not handled by the clipboard manager. The
test then makes sure we don't crash when trying to transfer data from
the old X11 selection source.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2364>
The Xwayland server can go away at any time; when this happen we might
have a test client running, and for it to tear down more nicely, make
sure to avoid trying to clean up X11 resources on the old X11 display.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2364>
Xwayland can disappear at any time, for example during a new_async() or
read_async() call. When we eventually finalize the stream, the X11
display it was created for is gone, thus can't clean up the X11
resources. Handle this by making the MetaX11Display pointer a weak
pointer, and ignore cleaning up if it disappeared. This is fine since
the X11 server it created those resources one is gone already.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2364>
The property doesn't necessarily exist when using drivers that doesn't
support atomic mode setting, and the way it worked will break night
light and other gamma related features. This makes things use the gamma
length; if it is higher than 0, it definitely supports it one way or the
other, i.e. GAMMA_LUT with the atomic backend, and drmModeCrtcSetGamma()
with the legacy/simple backend.
Fixes: 364572b95c
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2287
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2435>
It doesn't depend on whether the CRTC is active or not, so always read
it. This is also useful to know whether a CRTC supports gamma, before it
is being turned on, without relying on the existance of properties.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2435>
The COMPOSITOR_GRAB event route has effectively been replaced by
ClutterGrabs, which are no longer covered by the existing check.
So check for grabs as well to restore the old behavior.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2436>
The event-route is never set to COMPOSITOR_GRAB nowadays, so the
condition will never be met.
Furthermore, it is expected that ClutterGrabs only happen when
events are routed normally, so the remaining NORMAL check should
already fully cover the old COMPOSITOR_GRAB case.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2436>
We already bypass wayland if there is a ClutterGrab, so the case
that used to be covered by the event-route check is already handled,
and we can just remove the obsolete check.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2436>
Since the new ClutterGrab API replaced the old plugin-modal hook,
the event-route is never set to COMPOSITOR_GRAB.
The code in question already checks whether the stage has a grab,
so we can just remove old checks.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2436>
Initializing the event mask, SubstructureRedirectMask in particular,
before taking the manager selection fails with BadAccess. Fix this by
initializing said mask after taking the manager selection.
This fixes `--replace`.
Fixes: eb4307c350
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2432>
The code is already trying to avoid creating new laters when there
already is one for the queue type, but this wasn't working because the
ID of the later was never stored after creating a new one. This would
then result in as many laters as meta_display_queue_window() was called
and all of them would run the handler function, even if only the first
one had a non-empty window queue.
Similarly this was causing the later to not be removed if the window
queue got empty after meta_display_unqueue_window().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2414>
Currently the signal is getting emitted accidentally, because even when
removing a window from the queue, the later handler of that queue will
still get run due to a bug. This bug is going to get fixed in the next
commit, but some things might depend on the signal getting emitted when
the visibility of a window has changed.
This change affects the behavior in two ways. First the signal is now
emitted immediately rather than from an idle. And second it now
correctly includes the window in the should_show or should_hide list.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2414>
The compositor currently only updates the topmost window actor that is
visible to it after stacking changes. The visibility of a window actor
to the compositor however might only change via the display idle queue
after the stacking changes. This could then lead to the topmost window
actor being assumed to be NULL on Wayland after switching from an empty
workspace or when opening the first window on an empty workspace. The
result of this is direct scanout being disabled in these cases.
To fix this also trigger the update when the visibility of windows
changes.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2269
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2413>
Some windows span the entire screen but still use transparency, such as
the desktop window of Nemo. When these windows were used for direct
scanout, the transparent areas would turn black and nothing else would
be rendered.
In addition to checking the surface for opaqueness, for X11 windows also
the window actor itself has to be checked, because its opacity might
have been changed via _NET_WM_WINDOW_OPACITY.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2263
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2409>
This replaces the API to get the topmost surface actor with an API to
get the surface actor that could be a candidate for direct scanout. The
advantage of this is that it allows X11 and Wayland specific
restrictions for these actors.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2409>
It's not allowed to call eglQueryWaylandBuffer() if the call to
eglBindWaylandDisplay() failed, and will result in an assert being hit
in mesa if called.
Avoid that by keeping track whether we succeeded to bind, and only
attempt to realize a legacy EGL wl_buffer if binding succeeded.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2415>
Mutter makes use of a gsettings scheme that comes from
gnome-settings-daemon to check for the screen orientation.
In use cases where gnome-settings-daemon is not available,
this would lead to a crash as the key doesn't exists
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2398>
The monitors settings such as the privacy screen property is propagated
to the monitors via kms updates, however during initialization and
on monitors changes, we end up clearing the pending KMS updates because
such settings are added to the queue before the backend has fully
initialized the monitors, and this may lead to discarding all the
pending updates, including the one we've just planned.
To avoid this, move settings applications after we've both initialized
the backend and notified it about changes.
Also avoid to try set the settings during actual initialization, but
delay that after post-init.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2372>
Prior to 'compositor: Destroy actors when unmanaging', window actors
were destroyed when the compositor object was destroyed, long after the
windows were unmanaged, however, when this instead changed to happen
when unmanaging, with the original goal to avoid having these actors try
to interact with the disposed MetaCompositor instance, it caused an
issue where window actors would be indirectly destroyed as a side effect
of their parents being destroyed, which caused some fallout in the logic
handling window-close animation tracking, which relies on
meta_window_actor_queue_destroy() being called before a window actor is
actually destroyed.
Fix this by unmanaging windows before unmanaging the compositor.
From an X11 point of view, this should be harmless, since all it really
do is call XCompositeUnredirectSubwindows().
For the native backend and the common behavior, all unmanaging the
compositor instance does is destroy clutter actors, so doing so after
window actors were already cleaned up should not be a problem, as this
was the case before too.
Fixes: 35ac3a096d
Closes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/5330
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2403>
Quoting Ray Strode:
we don't expose a way to explicitly save the session in gnome anymore
afaik, and I don't think it's going to show on log out because
I believe we use the FORCE flag from the log out dialog.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2370>
A slightly annoying "feature" of Clutters debug messages is that it also
logs the filename and line of the current debug message. If you don't
have an ultrawide monitor, this can be very annoying and cause lots of
linebreaks in the debug logs.
So remove that debugging feature and no longer log the filename and
line number with debugging messages.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2378>
'kms/impl-device/simple: Get the buffer handle from MetaDrmBuffer'
changed how fb ids are generated, but it only made it fully work with
atomic mode setting. For legacy/simple mode setting, it only handled the
primary plane buffer, not the hardware cursor.
Fix this by making sure the fb id is generated also in the legacy mode
setting case.
Fixes: ea39142da2
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2250
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2397>
When an X11 window becomes an all-workspace window its `workspace` is
set to NULL before `meta_window_x11_current_workspace_changed()` is
called. The latter then checks for `workspace` being NULL (which also
happens when unmanaging) and then returns early. So this does not update
`_NET_WM_DESKTOP` to 0xFFFFFFFF. Instead it remains at the workspace the
window was on before. This was causing programs like `wmctrl` to switch
to this old workspace when activating such a window.
Fix this by checking if the window is unmanaging instead.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2242
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2387>
Simply signal preedit string changes from/to NULL once, in order
to avoid unwanted activity in the client side. We do still need to
send the preedit once each .done event, if there is one, in order
to behave according to the protocol when it matters the most.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2395>
With the unthrottled input emission, we ended up often getting the
cursor updates long before any damage had been posted, meaning that if
you moved around the mouse pointer where the mouse had a high enough
refresh rate, we'd effectively stall the screen cast stream by only
sending cursor updates and nothing else.
Fix this by scheduling an update when we get a cursor update, then
sending a cursor-only frame after any damage and relayout has been
processed, but only if there is no queued damage that will cause an
actual repaint.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2393>
This is a signal that will be emitted between the 'before-update' and
'before-paint'. It can be used to handle things when you know whether
there is an update, and you know whether a paint or not will happen, by
looking at the current damage.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2393>
This handle is used by the legacy KMS API; lets avoid having to have GBM
specific code where this is done by letting the MetaDrmBuffer API, that
already has this information, expose it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2275>
We'd put the message in a variable called `message`. If something passed
to meta_topic() was called `message`, it'd end up being `NULL` in the
log entry. Avoid this by making the local message variable a bit more
"on topic".
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2391>
Mutters event filter can prevent events from getting processed by
Clutter, this can also happen for TOUCH_END/CANCEL events. Processing
these events in Clutter is crucial for proper tracking of touch
sequences though, that's because Clutter adds a PointerDeviceEntry to
the stage on a TOUCH_BEGIN *before* going through the event filter, but
removes that entry on a TOUCH_END *after* going through the filter. So
Clutter really needs to see those TOUCH_END events, or else there will
be a stale PointerDeviceEntry on the ClutterStage.
Make sure those TOUCH_END/CANCEL events always get seen by Clutter by
removing the device entry immediately when those get filtered out.
Because there might still be events belonging to this sequence in the
event queue of the stage, we need to flush the queue before removing the
entry, too.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2350>
Unfortunately we cannot do this generically since the target of the
button/touch press does matter, e.g. tapping on the OSK, or clicking
the IBus candidates window. These situations should not trigger a
reset.
So be more selective about the situations where button/touch presses
trigger an IM reset, in the case of ClutterText these are still clicks
inside the actor, for Wayland's text-input it is when clicking the
surface that has text_input focus.
For all other situations where clicking anywhere else might make
sense to trigger an IM reset are covered by the focus changing paths,
that also ensure a reset before changing focus between surfaces/actors.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1961
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2384>
Focus changes should trigger an IM reset, as some engines do want
to maybe commit the preedit string before changing focus. In addition,
we do not want the preedit string to be able to move between
windows/applications.
Ensure that the commit string is committed when the IM deems so, and
ensure we send a .done event disntinct to the .leave event, so that
the client doesn't miss the commit.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2030
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2384>
Focus changes should trigger an IM reset, as some engines do want
to maybe commit the preedit buffer before changing focus. Since
the preedit string is also cleared on reset(), we can do without
that explicit call.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2384>
Right now we have a bit of a mixed bag between an active model where
input foci set the surrounding text without being asked for (e.g.
wayland's text_input), and a passive model where the IM engines ask
for content.
Make ClutterText take the same side than text_input, so that dealing
with those is at least consistent.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2384>
The clutter_text_delete_text() function used underneath expects character
offsets for both start/end position. Fix the end position passed an offset
instead of that, and compesnate for the cursor position being always -1
when the caret is at the end of the string.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2384>
I've overseen quite an important case in commit
98a5cb37d9: Repicking only when actors get
destroyed is not enough, we actually need to repick when actors go
hidden/unmapped.
While we could also listen to notify::mapped just like we listen to
notify::reactive, it seems better to avoid using property notifications
here due to the usage of g_object_freeze/thaw_notify() in ClutterActor.
It can lead to the stage receiving a notify::mapped with mapped = true
for a pointer actor, which really shouldn't happen (just like
notify::reactive with reactive = true shouldn't happen).
Fixes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/5124
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2333>
DMA buffers might be allocatable, but it doesn't mean the driver doesn't
fail when we try to allocate a buffer with an implicit modifier. Using
the proprietary NVIDIA driver for example, it will fail. Lets catch this
up front and avoid advertising DMA buffer support when we know it won't
work.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2383>
As of currently, we only emit .done() on actual changes coming from the
ClutterInputMethod/ClutterInputFocus. With the recent changes in the
interpretation of serials, it becomes more important now that the
compositor acknowledges every .commit done by the client, in order to
keep them feeding future IM state updates.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2365>
Compensate the protocol statelessness with our ClutterInputFocus
statefulness. This becomes more necessary now, since sending
consecutive .done() events is now considered acceptable behavior.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2365>
MetaCursorRendererNative only updates the cursor state when the
underlying texture changes. The cursor scale and transform do not
trigger updates. This results in wrong cursor orientations on rotated
displays. Use both texture changes and scale and transformation changes
to figure out when to update the cursor state.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2363>
We want all pointer events to be passed through the pointer a11y
processing before going through event filters: Once we go through event
filters, events might be dispatched to Wayland and get filtered out.
With the changes to immediately dispatch events to wayland, this changed
and the pointer a11y is now no longer seeing any events going to wayland
clients. Fix it by shuffling things around a bit and letting pointer
a11y take a peek at events earlier.
Fixes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/5192
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2332>
There's a bunch of crashes right now where the assertions in
clutter_actor_set_mapped() after calling the map/unmap() vfuncs are
failing. The only way this can happen is by re-entering
clutter_actor_set_mapped() during the map/unmap recursion.
The reason for those crashes is that the shell hides/shows some actors
in response to crossing events and key-focus changes. These in turn get
triggered by the newly introduced ungrabbing of ClutterGrabs when an
actor gets unmapped, which triggers GRAB_NOTIFY crossing events and
key-focus changes.
Since these situations are hardly avoidable (it's a valid use-case to
hide/show something in response to a crossing/key-focus event), catch
the set_mapped() call early while we reenter the mapping machinery and
log a warning instead of crashing.
Fixes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/3165
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2299>
With the introduction of untrottled event delivery to wayland clients,
we moved the _clutter_event_process_filters() call outside of
_clutter_process_event(). This also moved the processing of event
filters outside of the timespan where the event is added to Clutters
current_event stack, making Clutter.get_current_event() no longer
available to anything happening inside mutters event filter.
One thing that happens in mutters event filter is detecting and
triggering keybindings like the alt-tab switcher. Now the alt-tab
switcher has a special case where it finishes and activates a window
right when the keybinding gets activated, relying on the current event
time as the timestamp to activate the window.
Now since the current event time is no longer available from inside
mutters event filter, we'd pass 0 to meta_window_activate(), causing
mutter to send a notification instead of actually activating the window.
To fix this, also set a current_event for the ClutterContext when going
through event filters, this makes sure Clutter.get_current_event_time()
works when called inside keybinding handlers.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2327>
When switching between the existence and not of a stage ClutterGrab, we
would correctly attempt to synchronize key focus from the perspective of
the Wayland clients.
But this synchronization should do its own checks about existing stage
grabs before determining a client window has key focus or not.
Add that check, so that grabs correctly unfocus the keyboard in Wayland
clients, in addition to pointers and touch.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2194
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2366>
When we get passed a "snippet" to the shaped texture, it's added as a
pipeline layer snippet to change how the source texture is sampled. When
we draw from a texture tower however we have allocated regular textures
which doesn't need any special layer snippet, so create separate
pipelines for those that doesn't use that snippet.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/528
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2278>
In the right combination of circumstances, and given 2 actors (parent
actor P with an offscreen effect and child actor C), we may have the
following situation happening:
- A redraw is queued on the actor C, actors C and P are marked as
priv->is_dirty and priv->propagated_one_redraw.
- During paint() handling we paint actor P, priv->propagated_one_redraw
is turned off.
- We recurse into child actor C, priv->propagated_one_redraw is turned
off.
- A new redraw is queued on actor C, actors C and P are marked as
priv->is_dirty and priv->propagated_one_redraw.
- The paint() method recurses back, actors C and P get priv->is_dirty
disabled, priv->propagated_one_redraw remains set.
- At this point queueing up more redraws on actor C will not propagate
up, because actor C has priv->propagated_one_redraw set, but the
parent actor P has priv->is_dirty unset, so the offscreen effect will
not get CLUTTER_EFFECT_PAINT_ACTOR_DIRTY and will avoid repainting
actor C.
The end result is that actor C does not redraw again, despite requesting
redraws. This situation eventually resolves itself through e.g. relayouts
on actor P, but may take some time to happen.
In order to fix this, consider actors that did get a further redraw
request still dirty after paint().
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2188
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2353>
Simplify the function arguments (the origin is just the actor that
the function is originally called from), and make it also handle
marking as dirty the actor that got the redraw queued up explicitly.
This makes it a single place where priv->is_dirty is being enabled.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2353>
With Xwayland on demand, a number of maintenance X11 applications need
to be run first, before Xwayland starts accepting requests from the
normal clients, as soon as the WM_S0 selection is acquired by mutter.
On startup, mutter also sets a number of X11 properties that can be
queried by X11 clients.
Unfortunately, mutter acquires the WM_S0 selection before setting those
properties, so mutter and the first regular X11 client will race on
startup.
As a result, the X11 properties set by mutter on startup may not be
available to the very first X11 client when Xwayland starts.
To avoid that issue, make sure to take the WM_S0 selection last when
opening the display.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2176
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2336>
Currently, meta_xwayland_shutdown_dnd() is called from the handler
on_x11_display_closing() triggered from the signal "x11-display-closing"
hooked up from meta_xwayland_init_display().
Once the signal has been triggered, on_x11_display_closing() removes the
signal handler, disconnecting from the signal.
As meta_xwayland_init_display() is called from meta_display_new() which
is issued only once, the signal handler is not restored again.
As a result, meta_xwayland_shutdown_dnd() is not called anymore after
Xwayland has been restarted, but meta_xwayland_init_dnd() will check and
assert that the manager's DND object is NULL.
Basically, restarting Xwayland more that once will trigger an assertion
failure in mutter. That's even more of a problem with autoclose-xwayland
where Xwayland is expected to terminate when there is no meaningful X11
client remaining, which can happen multiple times during the lifetime
of a user session.
To make sure that meta_xwayland_init_display() is called for every new
instance of Xwayland, simply keep the signal hooked in place by not
disconnecting it when triggered.
This reverts commit 9a10b8ff94.
Even though, originally, this issue was first introduced with commit
b4fe1fdd95 ("xwayland: Make setup/teardown
a bit more symmetrical") which didn't actually kept 'x11-display-setup'
and 'x11-display-closing' connected.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2168
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2339>
For wayland meta_window_move_to_monitor sends a configure to the client
without actually moving the window, yet and the
meta_display_queue_check_fullscreen call won't detect any changes.
Checking for fullscreen in meta_window_update_monitor fixes the problem
because it is called whenever the window actually changed the monitor it
is on.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2325>
We rather confusingly still call a secondary display card that is
GPU-less (DisplayLink or other basic KMS device) a "secondary GPU",
so just because secondary_gpu_state is non-NULL doesn't mean we
can use it for rendering. The clearest indication of this is when
there is no EGL surface.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2341>
Since devices may be multiple things now, check all capabilities in order
to ensure all aspects of the device are correctly configured.
This change does the following observations:
- Devices that have TOUCHPAD | POINTER capabilities prefer the 'touchpad'
settings path. The regular pointer settings path is left for all
non-touchpads.
- Devices that are both a tablet and a touchscreen prefer the tablet
relocatable schema. This works for both aspects as the touchscreen
schema is a subset of the tablet one.
Other than that it's a rather boring, even if verbose search and replace.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2331>
We not just have X11 devices, but also virtual devices on both backends.
In the mean time, keep these working on top of a ClutterInputDeviceType,
but transform that into capabilities on device construction so users can
rely on the new flagset.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2331>
We do not need to open code the ClutterInputDeviceType fetching from a
libinput_device, since we already created a native ClutterInputDevice that
has the right type.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2331>
We use meta_workpace_focus_default_window() to sync the input focus back
to a window after it was on shell UI, this is not really necessary on
Wayland, but it is on X11. What this function does internally is ask
MetaWindowStack about the topmost window and focus+raise that window.
In gnome-shell we set the input focus to the default window every time
the key-focus changes to NULL (see shell-global.c ->
sync_stage_window_focus()). Now when closing the alt-tab switcher and
activating a window while there's an always-on-top window on the
workspace, meta_workspace_focus_default_window() will focus that
always-on-top window right after closing the alt-tab switcher, making it
impossible to focus another window using alt-tab.
To fix this, make meta_workspace_focus_default_window() check if there's
an existing focus_window first, if there is, use that, and if there
isn't, resort to just focusing the topmost one.
Fixes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/5162
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2328>
This fixes instances of:
```
*** BUG ***
In pixman_region32_init_rect: Invalid rectangle passed
Set a breakpoint on '_pixman_log_error' to debug
```
seen when navigating the overview and launching apps.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2349>
The ClutterGestureAction base code would correctly try to cancel a
gesture if it would receive GRAB_NOTIFY leave events (that would indicate
other portions of the actor tree stole input away from the gesture actor),
but it would mistakenly do so only if the gesture was already initiated,
possibly leaving stale point information if the gesture collected input
but didn't initiate yet.
This could be indirectly seen clicking with the mouse on OSK keys with
no motions in between, clicks would accumulate on the swipeTracker
gestures until the trigger point, so the third click could drag the
workspaces.
We do always want to unregister the related device/sequence here, do that
while still cancelling any already initiated gesture.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1907
Fixes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/4987
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2334>
This commit makes the KVM tests run in special VM runners tagged with
the 'kvm' tag. In order to avoid building the kernel image used for
running the tests each pipeline, it's built as part of the CI image
building.
For now, KVM tests are only run on the x86_64 architecture. The reasons
for this are two that the kernel image building script doesn't yet handle
any other architecture than x86_64 due to differences in how the image
is built and handled, as well as the fact that there only exists a kvm
tagged runner for x86_64.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2152>
This will allow us to reuse the keys and values more easily, as later
commits will rely on being able to iterate over the keys and values to
construct explict env strings for passing into special test cases.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2152>
Previously, we would only check for EXT_swap_buffers_with_damage which
generally will find an implementation. However, some EGL implementations
do not appear to support that naming of the extension, preferring to
only advertise KHR_swap_buffers_with_damage.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2316>
We use get_window_for_event() to check whether an event happened on top
of a window or on top of shell UI to decide whether to bypass delivering
the event to Clutter. In case of crossing events though, we can't just
use the device actor to determine whether to forward the event to
Clutter or not: We do want to forward CLUTTER_LEAVE events which
happened on top of shell UI. In that case the device actor is already a
window actor (the pointer already is on top of a window), but the shell
still needs to get the LEAVE crossing event.
Since the event source actor got removed from the detail of
ClutterEvent, the context we're looking for (which actor did the pointer
leave) is now the target actor that the event gets emitted to. Since the
last commit, we also made event filters aware of this context by passing
the target actor to them, so use this context now to determine whether
we're on top of a window or not.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2321>
We'll need the additional context of which actor the event will be
emitted to in mutters event filter (see next commit), so pass that
target actor to the event filters that are installed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2321>
Before scanning out the surface of a native client we have
to check the following attributes that influence the
relationship between buffer and the defined result on screen:
- buffer scale
- buffer transform
- viewport
In the future we can loose these checks again in cases where the
display hardware supports the required operations (scaling, cropping
and rotating).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2276>
Prior to 67033b0a mutter was accidentally including sizes for
configurations that were just focus state changes. This was not leading
to any known problems on the client side, but it was causing issues in
mutter itself when detecting whether a resize originated from the client
or the server.
Not including sizes in focus change configurations anymore however
revealed a bug in gtk. It was storing the window size when in a fixed
size mode (tiled/maximized/fullscreen), but not on any other server side
resizes. It was then restoring this stored size whenever there was a new
configuration without a size while in floating mode, i.e. the focus
change configurations generated by mutter after 67033b0a.
This change now addresses the issue 67033b0a was fixing in a way that
restores the previous behavior of always including the size whenever
sending a configuration.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2091
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2238>
Instead of having different coverage outputs for different architectures
and compilation flags, have each of those tests run coverage in order
to generate a JSON report, and have that merged at a final common job.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2315>
Running meson test from the build directory in a jhbuild environment
picks up libraries which have been installed previously because
LD_LIBRARY_PATH takes precendence over RUNPATH. Make the linker generate
the RPATH tag again.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2307>
If the remote desktop service emits absolute input events (e.g. absolute
pointer events) before the stream has started streaming, we don't have a
virtual monitor, as the size has not been negotiated yet. When this
happens, just drop the event. Remote desktop services should probably
make sure not to send events before the streaming has started, but them
doing so anyway shouldn't trigger a crash, which would be the case
otherwise.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2270>
This test resizes the stream by updating the PipeWire stream properties.
This triggers a format negotiation, that results in the buffers being
reallocated with the new size. The test makes sure we eventually
receive this new size.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2270>
Keep the virtual monitor around if it's being resized. This reduces the
number of unnecessary object rebuilding that happen during monitor
rebuilding.
This changes finalize() vfunc into a dispose() vfunc in the abstract
stream source object implementation, as it needs the abstract stream
source object to close the stream early, so that various signal
listeners get disconnected early.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1904
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2270>
We'll change mode's on-demand so using IDs identical to the virtual
monitor ID would mean IDs didn't change when changing mode, and that is
rather unintuitive. IDs don't mean much anyhow, just make them grow
within the realm of a 63 bit unsigned integer, as the 64th bit means its
a virtual mode ID. Making sure the ID is in the virtual mode namespace
is handled by meta_crtc_mode_virtual_new().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2270>
In some configurations (e.g. NVIDIA driver 470) Xwayland may use DMA
buffer for passing buffers around. When this is done, we might attempt
to scanout these buffers when they are fullscreen, and to do so we
import them using gbm.
However, for the mentioned configuration, there is no gbm device
available for importing. This was not handled, and resulted in a crash;
avoid this crash by checking whether we have a gbm device and fail
gracefully if we don't.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2098
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2318>
This aims to replace the x,y arguments in wl_surface.attach(); meaning
it can be used more sanely together with EGL, and at all when using
Vulkan.
The most common use case for the offset is setting the hotspot of DND
surfaces.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1905>
This implements the new 'bounds' event that is part of the xdg_toplevel
interface in the xdg-shell protocol. It aims to let clients create
"good" default window sizes that depends on e.g. the resolution of the
monitor the window will be mapped on, whether there are panels taking up
space, and things like that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2167>
We'd guess the initial monitor before it was actually calculated by
looking at the initial geometry. For Wayland windows, this geometry was
always 0x0+0+0, thus the selected monitor was always the primary one.
This is problematic if we want to provide initial more likely
configurations to Wayland clients. While we're not doing that yet, it'll
be added later, and this is in preparation for that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2167>
This will later be used to tell Wayland clients about a size they
shouldn't exceed.
If the window doesn't have a main monitor, this function does nothing
and returns FALSE.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2167>
It does not make sense that the event "source" (aka the target) is
both content and recipient of a message. Not doing so, events become
largely independent of the actor that is receiving/handling an
event. This is small step toward making events opaque and immutable.
Every user of these API calls in our code have ported away from
them, but other users may remain in extensions, so make these
functions work on top of the alternative API without accessing the
soon to be removed event field.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2311>
This is just "necessary" for --nested stages, since the pointer is
allowed to leave the stage in that case. Since the only side effect
is that there is still a pointer focus somewhere inside the stage,
simply drop this.
This is a small leftover of commit b8f92a6ce4, since we stopped
handling the double ENTER event there.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2311>
In addition to the presented callback time, it shows the time to the
reported presentation time (which can be earlier or later than the
presented callback), as well as the GPU rendering duration.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1928>
gnome-desktop is used to retrieve the monitor vendor name which in some
use cases is not needed as it brings a bunch of gnome-desktop unwanted
dependencies.
The change makes mutter fallback to an "Undefined" vendor name if it is
built without gnome-desktop
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2317>
Return in meta_egl_choose_all_configs() the actual number of
configurations returned by eglChooseConfig(), which are not
necessarily the same number as those from eglGetConfigs().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2303>
Since the introduction of ClutterGrabs, MetaDnd now no longer gets
notified about input events on the stage during grabs (for example while
the alt-tab popup is shown) and thus can't move the grab feedback actor
anymore.
To fix this, forward events to MetaDnD directly from
meta_display_handle_event() when a ClutterGrab is in effect.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2308>
We get the MetaWaylandCompositor a bunch of times, but we can do with
getting it only once and then also replace the is_wayland_compositor()
checks with a if (wayland_compositor).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2308>
The XDG activation support was missing interoperability with other
startup sequences, notably those coming from other means than XDG
activation.
In order to play nice with X11 startup sequence IDs, we not just
have to check for the startup ID being in the general pool, but
we also need to fallback into X11-style timestamp comparison so the
window ends up properly focused.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2314>
When a drag and drop occurs from an X11 client to a Wayland native
client, mutter uses an internal X11 window as a peer for the DnD drop
site.
That internal X11 window is moved and resized to match the Wayland
native windows as the drag destination moves.
When moving from one Wayland native window to another Wayland native
window, the same X11 window is used, and as a result no DND enter/leave
events is emitted.
In that case, the drop may occur on the wrong Wayland native window,
because no new XdndEnter/XdndLeave event were emitted.
To avoid that issue, use a pair of X11 windows instead of just one and
alternate between the two when repicking a new drop surface, so that
moving from a Wayland surface to another will always generate the
expected enter/leave events that we rely on.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2136
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2305>
When generating the action label, we expect both directions of these
features to have consistent settings (either both get a keycombo, or
they don't) or these just return NULL altogether.
Since one of the directions has an action associated, this is
misleading, so be more lenient at the time of generating the action
label.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2001>
The distribute_natural_allocation() function was copied over from Gtk to
Clutter 11 years ago with commit e636a0bbce.
Gtk only supports integers sizes in its layout machinery, while Clutter
does everything using floats.
Since this function sets the minimum_size (the size we allocate the
children in the end) to an integer, this means we're implicitly
typecasting floats to integers here, effectively floor()'ing all sizes
that we allocate the box children.
A bug this caused in gnome-shell was that a scrollView (like the one in
the endSessionDialog) was showing scrollbars even though the content
perfectly fit inside the view: Say the content and its scrollView parent
request a size of 63.9 px, but get allocated a size of 63 px by a box
layout. Now the scrollView notices that its allocated size is smaller
than the requested size and thus shows a scrollbar.
So fix that and use floats in distribute_natural_allocation() instead of
integers, as we do everywhere else in the layout machinery.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2284>
If we happen to handle a CLUTTER_TOUCH_BEGIN without a corresponding
CLUTTER_TOUCH_END at MetaWaylandTouch, we would still attempt to
reuse the older MetaWaylandTouchInfo, resulting in an assert triggered
as there is a stale touch reference on the previous surface.
Warn in place and create a new touch info struct to still fix the
broken surface accounting, instead of finding out the hard
way after the surface is destroyed. The assert is preserved to ensure
the accounting does not sneakily break anymore/further.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/584
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2251>
It passes a MetaLogicalMonitor, which isn't introspected right now, so
skip it completely. The entry point to the UI is handled via
MetaDisplay, so it isn't needed.
This fixes the following warning:
<unknown>:: Warning: Meta: (Signal)monitor-privacy-screen-changed: argument object: Unresolved type: 'MetaLogicalMonitor'
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2287>
Structure tests in a list of dictionaries, instead of requiring each
test to have its own executable(...) and test(...) statement. The
intention of this is to make it easier to add more test cases.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2262>
It already was built into it without any symbols exported, but also
duplicated in test cases that used it. Make it so that the built in
functions are exported, with prefixes, and make all tests use the
exported functions. While at it, make things go via MetaContext or
MetaBackend depending on how early in initialization things are run.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2262>
We update some details like the last used device and pointer visibility
from events, but this is done inconsistently on X11 since the
ClutterEvents are created and pushed from an additional place.
Make these updates happen on a private call, that will be called from
these places in X11.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/285>
Even though it's great that XI2 has an event to notify about device
changes, this is something we can let the MetaBackend code handle
consistently for all backends, since looking for the source device
works everywhere.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/285>
Instead of relying in the device being updated from different parts of our
machinery for different backends, hook this up to our own event dispatching.
This will allow dropping all other places where this is done.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/285>
We create a cursor renderer per device for those at
meta_seat_native_handle_event_post() with PROXIMITY_IN events, but
the MetaWaylandTabletTool handles the event before that, and goes
with a NULL cursor renderer.
Make MetaBackend::get_cursor_renderer() on the native backend create
those cursor renderers on demand, and only handle PROXIMITY_OUT in
handle_event_post() to dispose those. This makes MetaWaylandTabletTool
happily get a cursor renderer again.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/285>
We now only enable DMA buffer based PipeWire screen casting if a
format/modifier has been negotiated. This practically means a consumer
is aware about what is needed, and we should not try to predict that it
uses the DMA buffer the right way (i.e. not mmap:ing directly).
However, in case we're not hardware accelerated, we never want to
attempt to use DMA buffer screen sharing, as we want to avoid
compositing into a DMA buffer on such hardware as doing so can be very
slow.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2086>
meta_window_(un)queue() was implemented with global arrays in window.c
that managed MetaLater handle IDs and lists of window queues. In order
to rely less on scattered static variables and making it clearer that
we're dealing with per display window management and not something
specific to a single window, move the window resize/calc-showing queue
management to MetaDisplay.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2264>
It's still used by e.g. GNOME Shell to produce fallback icons for X11
applications that doesn't come with a .desktop file. Geometry stays in
the generic class because it's used for minimize animations and is
configured by the panel (e.g. the one in gnome-shell-extensions).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2264>
The comments in this function tells a story of C programmer self
reflecting about data types and Perl. While that can be nice, the rest
consisted mostly of repeating what the code line below did, with the end
result being that the function didn't fit on screen, resulting in worse
readability overall.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2264>
When handling lid state, we used to update the idle time right after
opening the lid. This behavior changed in commit 14b6c8780d due
to a typo/thinko, "if (lid_is_closed)" used to be an early return
condition before updating idle time, now it only updates in that
case.
Restore the original behavior, since this idle time update is key
in having gsd-power light up the display again, this presumably
fixes situations that required extra "light up" hints after suspend.
What it does surely fix is "ninja test" in g-s-d against recent
mutter, since the behavioral change induced a test timeout there.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2272>
The `ensure_x11_unix_perms` function tries to detect systems on which
/tmp/.X11-unix is owned by neither root nor ourselves because in that
case the owner can take over the socket we create (symlink races are
fixed in linux 800179c9b8a1e796e441674776d11cd4c05d61d7). This should
not be possible in the first place and systems should come with some way
to ensure that's the case (systemd-tmpfiles, polyinstantiationm …). That
check however only works if we see the root user namespace which might
not be the case when running in e.g. toolbx.
This change relaxes the requirements such that in the root user
namespace we detect and abort if a vulnerable system is detected but
unconditionally run in toolbx.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2261>
Add some debug logging that allows checking whether we're using DMA
buffers for screencasting or system memory buffers. This can be useful
for debugging screencasting performance and CPU usage.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2256>
With the ability to query the renderer for DMA-BUF support we can
announce support for implicit modifiers. This allows PipeWire to check
for matching modifiers while negotiation.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1939>
Returns TRUE if the active renderer backend can allocate DMA buffers.
This is the case hardware accelerated GBM backends, but FALSE for
surfaceless (i.e. no render node) and EGLDevice (legacy NVIDIA paths).
While software based gbm devices can allocate DMA buffers, we don't want
to allocate them for offscreen rendering, as we really only use these
for inter process transfers, and as buffers allocated for scanout
doesn't use the relevant API, making it return FALSE for these solves
that.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1939>
There may be situations where we may stack a ClutterGrab on top of a
wayland popup's. Since ClutterGrab should win over client grabs, we
mostly correctly figure out that it should start doing
bypass_wayland=TRUE and bypass_clutter=FALSE while the ClutterGrab
holds, however the late checks for the MetaDisplay event route can
still toggle bypass_clutter on, resulting in neither handling events.
This check for wayland popups in the display event route should just
enforce wayland handling if wayland is meant to be receiving events,
so ensure these don't mix together.
Fixes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/5020
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2271>
Without input device grabs in play, all functions that emit
pointer/key/crossing/touch events are pretty much the same. Remove this
duplication and use a common emit_event() function.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2100>
In case of misuse (e.g. passing NULL stage) this might result in crashes
before the precondition checks managed to kick in. Move this priv variable
initialization after these checks.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2099>
Wayland event processing and WM operations are themselves outside the
ClutterGrab loop so far. Until this is sorted out, these pieces of
event handling have got to learn to stay aside while there is a
ClutterGrab going on.
So, synchronize foci and other state when grabs come in or out, and
make it sure that Wayland event processing does not happen while
grabs happen.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2099>
Since we want these accessed from bindings this must be a boxed
type. This has the side effect of making ClutterGrab a refcounted
object, since we want to avoid JS from pointing to freed memory
and maybe causing crashes if misusing the object after dismiss.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2099>
Toggling the click action on when leaving the actor/action sounds weird,
this was presumably meant to toggle it off on leave, and back to in_held
on enter. This way, the CLUTTER_LEAVE handling also matches what we want
to do in case of grabs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2099>
The lack of handling of regular crossing events here is dubious, perhaps
to be fixed later on. So far, ensure gestures are cancelled whenever
a grab-inducted crossing event would leave this action in the blue.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2099>
This is (luckily!) unused, and it's inconvenient to have a toggle to
break the input model we are striving towards. Drop this function
and stick to the default behavior.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2099>
The experimental feature "autoclose-xwayland" requires a couple of
prerequisites:
1. Be able to (re)start Xwayland on demand, i.e. with systemd
2. Xwayland must support the terminate delay
Add a warning message if "autoclose-xwayland" was requested but any of
those prerequisites is not met.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2258>
Since commit 226afa24a - "Use Xwayland auto-terminate feature", the
callback function shutdown_xwayland_cb() does not check for the
autoclose-xwayland experimental feature anymore.
As a result, when running nested or outside of systemd,
gnome-shell/mutter would quit after 10 seconds unless some X11 window
was mapped.
But now that we rely on Xwayland's own terminate feature, there really is
no need to use any xserver timeout function anymore.
We do not need to keep track of X11 windows being created or unmapped, as
again, Xwayland does all that for us at the client level.
Remove all this code that we do not need anymore.
fixes: 226afa24a - Use Xwayland auto-terminate feature
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2258>
Because both code paths require the existence of `GL_TIMESTAMP[_EXT]`
which is only guaranteed if `ARB_timer_query` (included in GL core 3.3)
is implemented.
We know when that is true because `context->glGenQueries` and
`context->glQueryCounter` are non-NULL. So that is the minimum
requirement for any use of `GL_TIMESTAMP`, even when it is used in
`glGetInteger64v`.
Until now, Raspberry Pi (OpenGL 2.1) would find a working implementation
of `glGetInteger64v` but failed to check whether the driver understands
`GL_TIMESTAMP` (it doesn't).
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2107
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2253>
This looks like a relic of glReadPixels-based picking, the pointer
might well be outside redrawn areas, yet still require a device
update (e.g. in order to reflect the actor layout changes in the
"clear area" info).
Instead, always update all devices that are inside the view after
relayouts, the tracking on the need for that update is now done
on each ClutterStageView, instead of globally in the ClutterStage.
This theoretically fixes situations where pointers might miss
updating their "clear area" after the actor tree changed.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2117
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2257>
When using Xwayland-on-demand (default), if the (experimental) autoclose
features is enabled, we can rely on Xwayland's auto-terminate feature
instead of explicitly killing the Xwayland process.
With it, gone is the mechanism that was added to check the X11 clients
connected and their executable to check whether we can (safely) kill
Xwayland.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1794>
The connection to the Xserver for the X11 window manager part of mutter
even on Wayland may prevent the Xserver from shutting down.
Currently, what mutter does is to check the X11 clients still connected
to Xwayland using the XRes extension, with a list of X11 clients that
can be safely ignored (typically the GNOME XSettings daemon, the IBus
daemon, pulseaudio and even mutter window manager itself).
When there is just those known clients remaining, mutter would kill
Xwayland automatically.
But that's racy, because between the time mutter checks with Xwayland
the remaining clients and the time it actually kills the process, a new
X11 client might have come along and won't be able to connect to
Xwayland that mutter is just about to kill.
Because of that, the feature “autoclose-xwayland” is marked as an
experimental feature in mutter and not enabled by default.
Thankfully, the Xserver has all it takes to manage that already, and
is even capable of terminating itself once all X11 clients are gone (the
-terminate option on the command line).
With XFixes version 6, the X11 clients can declare themselves
"terminatable", so that the Xserver could simply ignore those X11
clients when checking the remaining clients and terminate itself
automatically.
Use that mechanism to declare mutter's own connection to the Xserver as
"terminatable" when Xwayland is started on demand so that it won't hold
Xwayland alive for the sole purpose of mutter itself.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1794>
This is to get support for the “-terminate delay” option in Xwayland
with libXfixes 6.0.
libXfixes 6.0 provides the new ClientDisconnectMode that clients can
use to declare themselves as to-be-terminated.
This can be used to let Xwayland terminate itself automatically when no
other (“non-terminatable”) X11 client is left.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1794>
Adding a <dbus/> element containing a boolean (yes/no) determines
whether org.gnome.Mutter.DisplayConfig ApplyMonitorsConfig will be
callable. The state is also introspectable via the
ApplyMonitorsConfigAllowed property on the same interface.
For example
<monitors version="2">
<policy>
<dbus>no</dbus>
</policy>
</monitors>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2030>
The test aims to verify that setting the following policy
<policy>
<stores>
<store>system</store>
</stores>
</policy>
only applies monitor configurations from the system level.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2030>
This adds a way to define a way, at the system level, to define a policy
of how monitor configuration files are loaded.
The intended use case is to e.g. either prefer system level monitor
configurations before user levels, or only allow system level
configurations.
Examples:
Prefer system over user level configurations:
<monitors version="2">
<policy>
<stores>
<store>system</store>
<store>user</store>
</stores>
</policy>
<configuration>
...
</configuration>
</monitors>
Only allow system level configurations:
<monitors version="2">
<policy>
<stores>
<store>system</store>
</stores>
</policy>
<configuration>
...
</configuration>
</monitors>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2030>
strncmp() always return 0 if the passed length is 0. What this means is
that whatever the first string check happens to be, if the parsed XML
cdata was empty (e.g. if we got <element></element>), the first
condition would evaluate to true, which is rather unexpected.
Fix this by making sure the string length is correct first. Also move it
into a helper so we don't need to repeat the same strlen() check every
time.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2030>
The way device backends implement power saving differ, and power saving
needs to contain nothing incompatible in the same update. Make it
impossible to e.g. mode set, page flip, etc while entering power save by
not using MetaKmsUpdate's at all for this.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2159>
When we're predicting state, i.e. when having posted an update while
avoiding reading KMS state, copy the predicted state, update the actual
state, and check that the predicted state matches the newly updated one.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2159>
It was a bit scattered, with it being split between MetaKms and
MetaKmsImpl, dealing with MetaKmsDevice and MetaKmsImplDevice
differentation. Replace this by, for now, single entry point on
MetaKmsDevice: meta_kms_device_process_update_sync() that does the right
thing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2159>
As other KMS tests, depends on being DRM master and vkms being loaded.
Currently consists of a sanity check that checks for the expected set of
connectors, CRTCs, planes, etc.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2159>
Right now gamma is set only via the D-Bus API (from gsd-color), but the
actual gamma isn't right after SetCrtcGamma(), meaning if one would call
GetCrtcGamma() right after setting it, one would get the old result.
Avoid this by getting the "current" CRTC gamma from the cache we manage.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2159>
In order to support dynamic imports, gjs added an implicit mainloop
that can drive the main context independently from other mainloops
like the one from GApplication or MetaContext.
That means that sources can now get dispatched to the main context
from the moment the plugin is started, resulting in a crash as the
association between compositor and plugin manager doesn't exist until
meta_plugin_manager_new() returns.
Make sure this doesn't happen by only starting the plugin after
meta_plugin_manager_new() has returned.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2242>
The code style in the remote-desktop API is currently a mess.
While, the overall code style in mutter and gnome-remote-desktop is to
use spaces as tabulator, the remote-desktop API is mixed with both
spaces as tabulators and actual tabulators.
In addition to that, the code style in the XDG desktop portal APIs uses
spaces as tabulators as well.
To unify the code style, replace all actual tabulators with spaces.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2255>
The code style in the screencast API is currently a mess.
While, the overall code style in mutter and gnome-remote-desktop is to
use spaces as tabulator, the remote-desktop API is mixed with both
spaces as tabulators and actual tabulators.
In addition to that, the code style in the XDG desktop portal APIs uses
spaces as tabulators as well.
To unify the code style, replace all actual tabulators with spaces.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2255>
When privacy screen is changed and this happens on explicit user request
(that is not a setting change) we should notify about this via an OSD.
To perform this, we keep track of the reason that lead to a privacy
screen change, and when we record it we try to notify the user about.
When the hardware has not an explicit hotkey signal but we record a
change we must still fallback to this case.
Fixes: #2105
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1952>
Privacy screen events on connector are handled as notification events
that won't cause any monitors reconfiguration but will emit monitors
changed on DBus, so that the new value can be fetched.
We monitor the hardware state so that we can also handle the case of
devices with hw-switchers only.
In case a software state is available it means we can also support
changing the state, and if so expose the state as unlocked.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1952>
When both a setting change and a monitor change happens we need to
ensure that the monitor settings are applied.
This is currently only related to privacy settings, but will in future
also handle other monitor parameters such as brightness.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1952>
Some monitors support hardware features to enable the privacy screen
mode that allows users to toggle (via software or hardware button) a
state in which the display may be harder to see to people not sitting
in front of it.
Expose then this capability to the monitor level so that we can get its
state and set it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1952>
In some cases mutter is started in the user scope from a TTY (for
example using toolbox). Using sd_pid_get_session fails because it's not
in the session scope so it falls back to the primary session
(sd_uid_get_display). We want to start mutter on the TTY we started
mutter on however. Instead of relying on the scope to figure out the
correct session we first look at $XDG_SESSION_ID which is set by
systemd_pam.so.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2254>
gtk-doc support was taken out about six years ago, but never replaced.
Add support for the newer gi-docgen, which should hopefully encourage
both C and GJS users of these libraries to improve the reference.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2248>
It is possible that we never create a cached state for a surface
even if it is synced. That is the case if `commit()` is never called.
We still need to call `apply_state()` in this case in order to run
e.g. `role_post_apply_state()` or `parent_state_applied` on subsurfaces.
So just ensure to initialize the cached state instead of bailing out.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2232>
Subsurfaces can be effectively synced indirectly via their ancestors.
Right now such indirectly synced surfaces don't apply their cached
state when their ancestor effectively becomes desync as by the time
we call `parent_state_applied()` on them, they are considered as
desync.
Thus sligthly reoder things so when the ancestors becomes desync
and applies its state, those surfaces still count as synced and
will thus apply their cached state as well.
While on it, add a check to prevent `set_desync()` to have side
effects when the target surface is not currently synced.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2232>
At first glance the `goto` looks like a loop, or potentially an infinite
loop. It's not a loop because the mode has changed at that point to
`META_SHARED_FRAMEBUFFER_COPY_MODE_PRIMARY`. But we can make it more
obvious and avoid the need for a goto.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2240>
The stage window is an interface, that added properties, that were only
then actually managed by MetaStageImpl. Shuffle things slightly, and let
the MetaStageImpl object deal with these things itself.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2014>
What the keymap eventually is after, are things handled by the actual
backend (MetaBackendX11), so let it keep a pointer to that. This
eliminates some usages of globals.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2014>
It tests that if we go from (x is the pointer cursor)
+--------+
| |
| X |
+--------+
to
+----------------+
| |
| |
+--------+ |
| | |
| X | |
+--------+----------------+
i.e. making sure that X ends up somewhere within the logical monitor
region.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2237>
These will be skipped by default, but can be run from a TTY for easier
debugging by doing:
dbus-run-session -- meson test -C build --suite mutter/native/tty --setup plain
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2151>
This commit makes it possible to run test executables in a test
environment constructed of a virtual machine running the Linux kernel
with the virtual KMS driver enabled, and a mocked system environment
using meta-dbus-runner.py/python-dbusmock.
The qemu machine is configured to use 256M of memory, as the default
128M was not enough for the tests to pass.
Using qemu is also only made possible on x86_64; more changes are needed
for it to be runnable on aarch64, so add a warning if it was enabled on
any other architecture.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2151>
This is needed if one wants to run the test suite parts that need KMS or
evdev access in a virtual machine.
However, only initiate these methods if the meta-dbus-runner.py program
was launched with --kvm, as it's only suitable for using while running
as root in a virtual machine.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2151>
When we test, we might not have a systemd session to rely on, and this
may cause some API we depend on to get various session related data to
not work properly. Avoid this issue by passing fallback values for these
when we're running in test mode.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2151>
There will be another mode added later, 'test'; prepare for this by
changing the existing "mode" boolean ('headless') to a mode, which is
either 'default' or 'headless'. Checking the is_headless variable is
changed to using the function is_headless(), except for one place, being
VT switching, which in preparation is only allowed on the 'default'
mode. Other places where it makes sense, the conditions are changed to
switch statements.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2151>
When running in KVM, the EGL driver supports querying the render node
path, but it returns NULL. Handle that better by falling back to
querying the device main device file, instead of falling back on v3 of
the protocol and logging a warning.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2151>
Some API will return NULL or the equivalent; sometimes it's an error,
and sometimes it's not, and the way to check that is by looking at the
return value of eglGetError(). When we check this, don't set the GError
if it returned EGL_SUCCESS, as that indicates that the return value is
expected behavior, and not an error.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2151>
As ClutterGrab is a stack, the backend only cares about some grab
existing currently or not. Make it sure that we grab whenever we
go to >=1 grabs, and ungrab whenever we go to <1.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2068>
Dissociate clutter_stage_set_key_focus() from the actors focused
state, so that it obeys stage grabs. The key focus actor state may
also change due to grab changes, add the code to notify about this.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2068>
Emit crossing events whenever a grab coming or going would cause a
pointer/touchpoint to become inactive on their position. Depending
on whether the pointer lies inside the old or new grab widgets,
enter or leave events would be generated.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2068>
We will want to be more specific about the portions of the actor
hierarchy that receive this event, separate creation and emission
so each place does what is relevant.
However, this commit brings no functional changes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2068>
These events are not meant to be ever silenced away, every actor
that is meant to receive one should do so. Make it sure that those
events cannot be stopped, despite the event signal handlers return
values.
This opens the debate about whether crossing events should be
ClutterEvents, since they are more and more uncommon at being one,
maybe this notification mechanism should be taken away from the
event machinery, but that's something for future refactors.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2068>
Hopefully, the one to make them all converge. This new ClutterGrab
represents a handle on a created grab. These are stacked, so grabs
can be overridden and remain inactive until there is a time that
they become active again, although undoing these early is optional.
These grabs are global, they do apply to all pointer, touchpoint
and keyboard foci.
At the moment, only the API to create and stack those is added,
the actual functionality is added in future commits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2068>
There was a sanity check that complained if there was still a "next
framebuffer" when disposing an onscreen. This is correct to complain
about under normal operation, as we always wait until receiving the page
flip callback before cleaning up the onscreen and their state.
However, when there are many hotplugs occurring, we might end up with
race conditions when the above sanity check is not valid: when we have
more than one monitor active, paint 1 one of them, but receive a hotplug
event before we paint the other(s), we will discard the already painted
onscreen before really issuing a page flip.
In this situation, we will have the "next framebuffer", but having that
is not a bug, it's a race condition, thus to not leak in this situation,
make sure to clean up the next framebuffer here too.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2081
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2225>
If there are any pending updates, for example if we painted one of
multiple monitors but without having posted the update due to waiting
for another monitor to be painted, but before we paint all of them and
post the update, another hotplug event happens, we'd have stale pending
KMS update. When that update eventually would be processed, we'd try to
apply out-of-date updates which may contain freed memory.
Fix this by discarding any update when we're rebuilding the views. We
can be sure not to need any of the old updates since we're rebuilding
the whole content anyway.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1928
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2216>
Following the EGL_KHR_swap_buffers_with_damage specification, the
surface damage used by eglSwapBuffersWithDamage does not need to
contain the damage history.
Rework that to reduce the amount of rectangles that get passed to
the backend.
Also rework some of the regions that were using fb_clip_region and
missing the last scaling to support fractional scaling.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2089>
A property for reversing the visible order of children is a bit odd.
It has also been unused by actual gnome-shell code since 2010, and the
somewhat related pack_start()/pack_end() API in GtkBox(Layout) is gone
in GTK4.
With that in mind, turn the property into a no-op and deprecate it,
so that it can be dropped next cycle.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2206>
When the before-paint function is executed, it's only purpose
is to check if there's any scanout queue, and immediately
record it if any.
However, since [1], we regressed in this specific case with the
introduction of an idle callback in the before-paint function.
The regression only happens when the PipeWire stream is using
DMA-BUF buffers, and it would operate as follows:
1. In before-paint, when there's a scanout available, we queue
an idle callback to capture the monitor. The idle callback
(almost always) executes after the scanout is pulled from
the stage view
2. meta_screen_cast_stream_src_maybe_record_frame() is called
by the idle callback. In the DMA-BUF case, it then runs
meta_screen_cast_monitor_stream_src_record_to_framebuffer()
3. In meta_screen_cast_monitor_stream_src_record_to_framebuffer(),
because the stage view doesn't have a scanout anymore, it
ends up calling cogl_blit_framebuffer() with the stage view
framebuffer. This is the regression bug.
This regression presents itself in the form of the screencast
stream showing the desktop when there's an unredirected fullscreen
application window running.
Revert before-paint - and only that - back to immediately capturing
any available scanout. Only record these frames when the target
buffer is a DMA-BUF handle. Nothing is captured on before-paint if
the stream is not using DMA-BUF, since the regular paint routine
will handle these frames regularly post-paint.
[1] https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1914
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2186>
Next commits will reintroduce a certain behavior of stage
capturing that can only happen with DMA-BUF buffers. To
control this, add a new flag tp MetaScreenCastRecordFlags
for this behavior.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2186>
If a Wayland subsurface is the topmost actor, consider in for
scanout as well. This will extend our scanout capabilities to apps
like Firefox
While on it, correct a unnessary type check to a NULL check.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2211>
On Wayland a window actor may have more than one surface actor,
most importantly when subsurfaces are used.
Add a new function to request the one which is at the top -
it will be used in the next commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2211>
If the EGL header is not new enough, it will not contain that relatively
new macro definition, so to avoid breaking compilation, define it
ourselves for now. Should be possible to remove after some time.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2214>
We save the window rect before going fullscreen to a dedicated variable,
so we can go back to the correct dimension. We also have a dedicated
variable for returning from other window states, e.g. maximized, and
this one we initialized when creating the MetaWindow. This meant that we
could always rely on this being up to date on X11 windows that were
mapped maximized or fullscreen.
What the commit that introduced the saved rect dedicated for going
unfullscreen missed was to initialize the new saved rectangle too when
creating the MetaWindow. This resulted in windows mapped as fullscreen
often ending up misbehaving when unfullscreening, as mutter would tell
them to unfullscreen to 0x0.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1786
Fixes: a51ad8f932
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2210>
Since every input stream now uses its own window, the X property used to
transfer the data no longer has to be unique, so we can stop generating
those unique names. This avoids creating a new atom for every transfer
since those are never freed, neither on the shell nor on the server
side. Also don't unnecessarily duplicate other strings that are
(almost) never used and get them from the atom in the rare case when
they are needed.
Fixes https://gitlab.gnome.org/GNOME/mutter/-/issues/1328
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1812>
When there are two (or more) concurrent XConvertSelection requests with
the same target, selection and window and the data is large enough for
SelectionNotify events to overlap. This can result in the affected streams
being considered completed without any data being transferred.
While regular mutter/shell code does not make use of concurrent
XConvertSelection requests with the same targets, some extensions might.
Such as for example a clipboard manager that like the built-in clipboard
manager tries to read the selection on owner-changed.
One potential solution would be to make sure the event is for the correct
property, but not all clients seem to support concurrent requests for the
same targets but different properties on the same window.
This commit instead changes the streams to use their own window which
seems to be more widely supported.
Fixes https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/4034
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1812>
This is a preparation for each input stream creating its own window. It
moves deleting the property from meta_x11_selection_input_stream_xevent
where it can run after the stream has been finalized to a spot where
the stream still exists. Use an error trap in case the property was not
set by the client, such as when the conversion failed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1812>
The only currently known issue with allocating scanout buffers with
modifiers is memory constraints in multi head setups. Heuristics for
handling that are not implemented, but since it doesn't apply to
anything but i915, remove the other drivers from the deny list.
The other drivers had modifiers disabled to marginally increase the
chance of becoming scanout-able when allocated by Xwayland.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2146>
This significantly increases the chance of a fullscreen surface buffer
being scanned out instead of being painted via composition. This is
assuming the client supports the DMA buffer feedback Wayland protocol.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2146>
This API can be used to construct a MetaKmsUpdate with plane assignments
that in isolation will be tested against the current KMS state. How it
is tested depends on the KMS implementation; in the simple / legacy KMS
backend, the tests are identical to the current scanout requirements
(dimension, stride, format, modifiers, all must match), and with atomic
KMS, it uses the TEST_ONLY on a real constructed atomic mode setting
commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2146>
Whenever a surface is promoted as a scanout candidate by
MetaCompositorNative, it'll get a CRTC set as the candidate CRTC.
When a client asks for DMA buffer surface feedback, use this property to
determine whether we should send a scanout feedback tranche.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1959>
This ensures they remain perfectly smooth regardless of how the
dispatch time has been adjusted/optimized/delayed/jittered.
Idea by Ivan Molodetskikh <yalterz@gmail.com>
For example, dragging a window on a 60Hz monitor:
BEFORE
delta(time_us) = 17014μs
delta(time_us) = 15998μs
delta(time_us) = 17006μs
delta(time_us) = 16975μs
delta(time_us) = 16001μs
delta(time_us) = 17002μs
delta(time_us) = 17006μs
delta(time_us) = 16004μs
AFTER
delta(time_us) = 16667μs
delta(time_us) = 16667μs
delta(time_us) = 16670μs
delta(time_us) = 16667μs
delta(time_us) = 16669μs
delta(time_us) = 16668μs
delta(time_us) = 16664μs
delta(time_us) = 16674μs
Caveat 1: Because we don't know a "next presentation time" on the first
frame, the interval between the first and second frame will usually be
different to the subsequent steady interval. So this change increases the
jitter of just frame 2, but eliminates jitter thereafter.
Caveat 2: `clutter_frame_clock_schedule_update_now` schedules updates
earlier than `clutter_frame_clock_schedule_update`. This means potentially
you could get multiple frames targeting the same "next presentation time".
That doesn't really change here though - we're dispatching at the same
times as we used to and just giving timelines a better vsync-aligned
timestamp now.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/25
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2161>
The final tick of a timeline is >= its duration, but when using ticks that
are slightly in the future ("next presentation time") this means the final
tick will execute and complete the timeline up to one frame interval before
the timestamp of that final tick.
For the single clock test we now just check if the overall duration is
within one frame of the expected timeline duration.
The dual clock (switching) test needs a threshold of two frames because
starting each new clock creates a phase shift (error) of up to one frame.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2161>
We can't just run a single 'meson' command in the script as we need to
launch multiple long running processes inside the D-Bus session:
* The meson test invokation itself
* wireplumber
This was handled for the 'test' stages, but not for the 'dist' stage,
which as a result would fail due to wireplumber not being launched
causing any test using pipewire to timeout.
Address this by making run-tests.sh a more generic run-meson.sh that
launches wireplumber before running the actual meson command.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2204>
We're in the destructor, it's pointless to unset the userdata as we'll
never ever see a request being invoked with it ever again, since the
resource itself will be destroyed or marked as destroyed.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2202>
When rendering to a buffer that is not the stage view buffer, we can not
know where the buffer will be displayed on the screen. As a result we
also can not know what translation would need to be applied to culling.
This was causing glitches when the gnome-shell magnifier was applying
offscreen effects. ClutterOffscreenEffect causes MetaWindowGroup to be
rendered to an offscreen buffer at an offset, because it draws to a
slightly larger texture with an accordingly translated origin. This
translation then later is canceled out again when the offscreen buffer
is drawn. To meta_actor_painting_untransformed() however which only sees
the translation used when drawing to the buffer this looked like the
window group was being rendered at the offset. This then lead to
redraw_clip getting translated accordingly, resulting in wrong
coordinates used for culling.
Similarly this was leading to issues when taking area screenshots while
at 1x zoom.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1678
Fixes: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/4876
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2080>
Refresh rates >60Hz become ever more common. In order to allow users
to keep hight refresh rates when not running at a natively advertized
resolution, add common refresh rates to our fallback modes.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2189>
Right now we often add a duplicate fallback mode that's almost
identical to the native mode. This adds unnecessary clutter to
UIs, thus filter out such modes.
In order to keep the code small, use `MetaCrtcModeInfo` directly
instead of recalculating the values. And to keep consistency, do
the same in the loop above.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2189>
This is so that it can unregister from it on tear down. The tracker owns
references to cursors too, but this cycle is already broken as the
backend calls 'g_object_run_dispose()' when tearing the cursor tracker
down.
Fixes a crash on shutdown.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2181>
An incorrect assumption that after mode set there would be no pending
page flips was made. This meant that if there was a mode set, followed
by a page flip, if that page flip was for a CRTC on a now unused GPU,
we'd crash due to the renderer GPU data having already been freed. This
commit avoids that by keeping it alive as long as the page flips are
still in the air. It fixes crashes with backtraces such as
0) meta_render_device_get_egl_display (render_device=0x0)
at ../src/backends/native/meta-render-device.c:320
1) secondary_gpu_state_free (secondary_gpu_state=0x1c8cc30)
at ../src/backends/native/meta-onscreen-native.c:560
2) meta_onscreen_native_dispose (object=0x1cb65e0)
at ../src/backends/native/meta-onscreen-native.c:2168
3) g_object_unref (_object=<optimized out>)
at ../gobject/gobject.c:3540
4) g_object_unref (_object=0x1cb65e0)
at ../gobject/gobject.c:3470
5) clutter_stage_view_finalize (object=0x1cbb450)
at ../clutter/clutter/clutter-stage-view.c:1412
6) g_object_unref (_object=<optimized out>)
at ../gobject/gobject.c:3578
7) g_object_unref (_object=0x1cbb450)
at ../gobject/gobject.c:3470
8) meta_kms_page_flip_closure_free (closure=0x1d47e60)
at ../src/backends/native/meta-kms-page-flip.c:76
9) g_list_foreach (list=<optimized out>, func=0x7fb3ada67111 <meta_kms_page_flip_closure_free>, user_data=0x0)
at ../glib/glist.c:1090
10) g_list_free_full (list=0x1cb4d20 = {...}, free_func=<optimized out>)
at ../glib/glist.c:244
11) meta_kms_page_flip_data_unref (page_flip_data=0x1c65510)
at ../src/backends/native/meta-kms-page-flip.c:109
12) meta_kms_callback_data_free (callback_data=0x227ebf0)
at ../src/backends/native/meta-kms.c:372
13) flush_callbacks (kms=0x18e2630)
at ../src/backends/native/meta-kms.c:391
14) callback_idle (user_data=0x18e2630)
at ../src/backends/native/meta-kms.c
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2147>
This ensures we don't have any left over cursor GPU buffers (via
gbm_bo's) after destroying the corresponding gbm_device (owned by
MetaRenderDevice).
Fixes crashes with backtraces such as
1) meta_drm_buffer_gbm_finalize at ../src/backends/native/meta-drm-buffer-gbm.c:450
4) invalidate_cursor_gpu_state at ../src/backends/native/meta-cursor-renderer-native.c:1167
9) update_cursor_sprite_texture at ../src/wayland/meta-wayland-cursor-surface.c:70
10) meta_wayland_surface_role_apply_state at ../src/wayland/meta-wayland-surface.c:1869
11) meta_wayland_surface_apply_state at ../src/wayland/meta-wayland-surface.c:832
12) meta_wayland_surface_commit at ../src/wayland/meta-wayland-surface.c:993
13) wl_surface_commit at ../src/wayland/meta-wayland-surface.c:1158
14) ffi_call_unix64 at ../src/x86/unix64.S:76
15) ffi_call at ../src/x86/ffi64.c:525
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2147>
During tear down, if anything teared down after the seat tries to get
the cursor renderer, we'd crash trying to get it as the seat would
already be gone. Avoid this by returning NULL when there is no seat.
It's assumed that any path that will happen during tear down that relies
on getting the cursor renderer will gracefully handle it not being
present, e.g. by relying on the cursor rendering cleaning up itself.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2147>
Most clients nowadays switched to buffer damage, most notably Mesa
and Xwayland. Thus lets avoid the extra cost of allocating three
`cairo_region_t`s and doing some calculations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2168>
If no viewport is set, the neutral viewport is the surface size
without viewport destination size applied - i.e. transform and
scale applied to the buffer size. Change it accordingly, giving
us the same values we'd return in `get_width` in this case.
As result, this only changes cases where a viewport destination
size but no viewport source rectangle is set.
The change fixes exactly such cases, e.g. the Gstreamer Wayland
sink. Can be tested with: `gst-play-1.0 --videosink=waylandsink`.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2168>
When checking panel orientation on logical monitors we should take
panel orientation transform to check it's properly applied, so ensure
that we're checking the right one.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2090>
Commit 2289f56112 ("monitor-manager: Don't apply unneeded orientation
changes") added an early return to handle_orientation_change () in case
the transform is unchanged.
But this did not take the correction of the transform for devices
with 90° mounted panels into account causing a desired orientation
change to get skipped if the new orientation matches the corrected
logical orientation from the previous transform setting.
Fix this by calling meta_monitor_crtc_to_logical_transform () on the
transform before comparing it, matching the
meta_monitor_crtc_to_logical_transform () call in
create_for_builtin_display_rotation ().
Related: https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1233
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2090>
When Xwayland was not initalized, we'd still clean things up. What this
accidentally meant was that the uninitialized display number 0 was
cleanud up, which very likely was main display of the host session.
What this meant in practice was that /tmp/.X11-unix/X0 was often
removed, causing every Flatpak X11 application to fail to start until
Xwayland was restarted nad the X0 socket file was restored.
Fix this in two ways: firstly only shutdown Xwayland if we ever started
it, i.e. if the X11 display policy was not 'disabled'. This should fix
the issue most of the times. Secondly only clean up the socket if it was
ever initialized. This should fix things if the socket creation failed,
as if it did, the name would be set.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2162>
Needs a couple of changes to make tests pass again:
* mesa-dri-drivers needs explicit installation to install swrast dri
driver
* wireplumber needs to be started inside the D-Bus session
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2142>
A popup surface can be remapped multiple times using the same
wl_surface, if a new xdg_popup object is created. To properly handle
this, we need to reset the 'dismissed_by_client' boolean to false, as
otherwise we won't allow new buffer commits.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1993>
Change to use the headless backend with a virtual monitor, instead of
the nested backend. This means tests can create and use virtual input
devices, which isn't possible with the nested backend.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1993>
The roundtrip in the handle-configure function could in theory (and in
practice in the future) happen to receive another configure event. If we
send yet another invalid geometry the second time, we log one time too
many in the server, which fails the test. Avoid this by ignoring the
second configure event; it's enough to pass through the error handling
path once.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1993>
This won't change anything for 60Hz displays but higher refresh rate
users will benefit.
Using Nvidia EGLStreams on a 240Hz monitor for example (refresh interval
~4.1ms), the maximum render time allowed before dropping to 120Hz is now
3.6ms whereas it was previously 2.1ms.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2158>
When Cogl gained support for importing pixmaps, I think there was a
misunderstanding that there is a difference in how it works in GLX and
EGL where GLX needs to rebind the pixmap in order to guarantee that
changes are reflected in the texture after it detects damage, whereas
with EGL it doesn’t. The GLX spec makes it pretty clear that it does
need to rebind whereas the EGL spec is a bit harder to follow. As a
fallout from Mesa MR 12869, it seems like the compositor really does
need to rebind the image to comply with the spec. Notably, in
OES_EGL_image_external there is:
"Binding (or re-binding if already bound) an external texture by calling
BindTexture after all modifications are complete guarantees that
sampling done in future draw calls will return values corresponding to
the values in the buffer at or after the time that BindTexture is
called."
So this commit changes the x11_damage_notify handler for EGL to lazily
queue a rebind like GLX does. The code that binds the image while
allocating the texture has been moved into a reusable helper function.
It seems like there is a bit of a layering violation when accessing the
GL driver internals from the EGL winsys code, but I noticed that the GLX
code also includes the driver GL headers and otherwise it seems pretty
tricky to do properly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2062>
We are using internal API that has the benefit of checking that the
focus surface still matches, but has the drawback that it does not
check the MetaWaylandKeyboard state.
In order to fix this, look for keyboard focus and serial matches
specifically when triggering activation.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2148>
The GBM support in the NVIDIA driver is fairly new, and to make it
easier to identify whether a problem encountered is related to using GBM
instead of EGLStreams, add a debug environment variable to force using
EGLStream instead of GBM.
To force using EGLStream instead of GBM, use
MUTTER_DEBUG_FORCE_EGL_STREAM=1
Related: https://gitlab.gnome.org/GNOME/mutter/-/issues/2045
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2132>
We created pipes for the stdout of the spawned mock services. This
resulted in the pipe being filled if enough things were logged, as
nothing was reading from it. Change this to allow for two modes:
verbose - where output is logged to the parent stderr, as well as non-verbose
(default) - where things are logged directly to /dev/null.
This fixes frozen tests when running with --repeat and a high enough
repeat count.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2139>
This is notably necessary with transformations, since these don't
trigger allocation machinery, but may affect the actor under the
pointer.
Visible e.g. with GNOME Shell's "Application does not respond"
dialogs.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
With Wayland handling all events as they come, this code now just
performs motion compression for events that will be handled by Clutter
widgetry.
The intent to opt tablets and styli out of motion compression was
early and fast client handling, since that is now covered in a generic
manner, this code is superfluous. We don't really need the extra events
for these devices in compositor widgetry either.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
We essentially create those at the time they need to be handled, and
use shortcuts that avoid the event from being queued up. It's too much
of a short cut though, these events are also of interest to the Wayland
event handlers, e.g. to handle pointer state changes (e.g. repicks due
to the pick actor being destroyed) immediately, instead of at the next
event.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
If we are still under the "clear area" of the pick actor, we forget
to update the coordinates. This is usually not needed, unless we
need to repick again for non-event circumstances (e.g. pick actor
is destroyed). This will ensure the right pointer coordinates are
used afterwards in those situations.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
Traditionally, the next repaint would also involve picking, which
would correct the actor under the pointer. This now does not happen
out of the box, so we really are waiting for the next pointer event
here.
To avoid the pointer/cursor to lag behind, trigger an immediate
repick here, that will look up the new actor under the pointer
coordinates.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
And resort to it first, unless we are told to ignore the cache
(e.g. after relayouts). This avoids further pick context operations
while the pointer is on the current actor.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
This safe area is the region (in stage coordinates) where the pointer
is ensured to stay within the current actor. This is not used yet, but
will be used for optimizations in pointer picking.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
These may be used for optimizations once we find the pick actor,
so picking can be avoided in areas we know didn't cross into
other actors. Nothing makes use of it yet though, just log these
so far.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
Since this signal is in a hot path during input handling, it makes sense
not to have this be a signal at all, currently most of the time spent in
it is in GLib signal machinery itself.
Replace it with a function/user data pair that are set on the sprite
itself. Only the places that create an sprite are interested in hooking
one ::prepare-at behavior per sprite, so we can do with a single pair.
This makes meta_cursor_sprite_prepare_at() inexpensive enough.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
Let the meta_cursor_sprite_realize() function return a boolean value
telling whether there was an actual change in the sprite cursor. E.g.
the surface/icon for it changed in between.
This is used in the native backend to avoid converting/uploading again
the cursor surface.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
Add a clutter_stage_pick_and_update_device() method that is the only
single entry point for updating a device position as seen by the
stage.
Also, update all callers to use it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
The clutter_stage_get_actor_at_pos() calls it almost 1:1 underneath
and is public API, we can have all callers use this, and stop using
this function outside of clutter-stage.c.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
As event handling goes:
1) Events get generated and queued by the seat (from another thread in
native, in the same thread in X11)
2) The MetaBackend gets those events and forwards them to Clutter
via clutter_do_event()
3) The events get queued in the ClutterStage
4) At the time of processing a frame, the input events are processed,
5) Motion events are throttled, only the last is effectively handled
6) Events are filtered, wayland and WM handling happens here
7) Events maybe reach to clutter
This commit moves 6 to happen between 2 and 3. The end result is that:
- Throttling only applies to Clutter event handling, The wayland event
forwarding bits will handle the event stream as soon as it comes, as
timely as possible.
- WM event handling is also unthrottled, but that's more of a side
effect.
- This all still happens on the main thread, so there's the possibility
that other busy areas (e.g. relayout) temporarily block this event
forwarding.
- Sending events unthrottled inherently means more CPU, probably
dependent on input devices' frequency. The impact is not measured.
This should bring the best of both worlds with e.g. 1000Hz mice, wayland
clients get unthrottled events, while GNOME Shell UI still behaves like
it used to do.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
If we wait till finalize, dispose will destroy the actor hierarchy
and cause untimely repicks. Ensure to free the pointer/touch info
first, so the hooked signal callbacks are gone when destroying the
actors.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
This test is injecting input events without checking the correct stage/
device state. Wait for the pointer to enter the stage, so the event gets
correctly forwarded across.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1915>
Listen to changes in MetaWindow::is-alive, so that the pointer
can logically leave the surface as soon as that happens. This
helps prevent flooding the client socket while it is stalled.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2122>
Change some things in these "app is alive" checks:
- The dialog timeout is separated from the ping timeout, in order
to show it again at a constant rate after dismissing, despite in
flight pings. It still shows immediately after the first failed
ping.
- As we want to tap further into is-alive logic, MetaWindow now
made it a property, that other places in code can fetch and
subscribe.
- Motion events trigger ping (as long as there was none other in
flight for the same window), and are counted between ping and
pong, in order to preemptively declare the window as not alive
before there is trouble with event queues being overflown.
This results in a separate logic between "the application does
not respond" and "we are showing the close dialog" so that the
former may get triggered independently.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2122>
If we were cancelled, it could mean we teared down, meaning fetching
manager instances will attempt to fetch past freed instances. Handle
this by waiting with the fetching until we know we weren't cancelled.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2140>
MetaBackend can now show whether it is in headless mode or not
using a vfunc is_headless.
Fallback of is_headless returns FALSE.
MetaBackendNative implements is_headless returning its
is_headless property.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2130>
This is a strange thing to do since MetaInputMapper also does take care of
devices with an output configured through settings, since we might have
devices that were configure through settings exclude other devices that
belong together with an output (e.g. a display-integrated tablet).
This was essentially here as a last resort to avoid matching two very
similar looking tablets to one of two very similar looking outputs. There
was a 50% chance already that the choice was wrong, and now these devices
can all be configured specifically through settings, so this shouldn't
be missed either.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2107>
Non-display-attached tablets (e.g. Intuos) may find no match, which
should mean "use the span of all monitors", not "pick one for me".
Reserve this fallback to touchscreen devices, since these might
still benefit from it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2107>
The matrix and aspect ratio of the tablet is irrelevant on pads, and
it actually triggers warnings when trying change that on those devices:
gnome-shell:42536): mutter-CRITICAL **: 17:22:41.994: meta_input_device_native_get_mapping_mode_in_impl: assertion 'device_type == CLUTTER_TABLET_DEVICE || device_type == CLUTTER_PEN_DEVICE || device_type == CLUTTER_ERASER_DEVICE' failed
This is unnecessary to do on pad devices, these just need to be moved
together with their respective stylus.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2107>
In the case a11y is required, the screen reader is very much
interested in getting an uninterrupted flow of key events. It attempts
so by setting a ::captured-event callback on the ClutterStage, but
that falls short with our MetaDisplay event handler, as clutter events
can be stopped before a11y gets a chance to see them.
This kind of selective amnesia wrt key events is not new, in X11 those
go unheard of by the WM as long as a client is focused and no grabs hold,
so it is clients' responsibility to talk with AT bridge.
This commit doesn't yet change that for X11, but we can do this right
away from the compositor on Wayland, and without any chance to be
tampered by clients.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1328>
Since the completion callback (on_switch_workspace_effect_complete) sets
priv->tml_switch_workspace1 to NULL, the unref was trying to unref NULL,
and the reffed ClutterTimeline was not getting unreffed.
This could be triggered by rapidly switching workspaces, switching again
before the animation of the initial switch was done.
Found while working on #2038.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2120>
This avoids the following crash, that could happen in certain rare race
conditions, e.g. in tests:
0) wl_closure_invoke (closure=0x2fbf9e0, target=0x2e5b3d0, opcode=0)
at ../src/connection.c:1014
1) wl_client_connection_data () at ../src/wayland-server.c:432
2) wl_event_loop_dispatch () at ../src/event-loop.c:1027
3) wayland_event_source_dispatch () at ../src/wayland/meta-wayland.c:104
4) g_main_dispatch () at ../glib/gmain.c:3381
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2127>
Passing a NULL workspace does not make sense, since it silently
returns no windows. Mandate that a workspace is explicitly requested,
and while at it check the other arguments as well.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2124>
The cursor renderer shouldn't assume all the CRTCs of a logical are KMS
CRTC's, as we'll end up checking hardware capabilities for CRTC's of
virtual monitors as well, when they were created to not embed the cursor
image directly in the framebuffer.
Instead, use the newly introduced API for checking CRTC cursor
capabilities. This fixes a crash with the following backtrace:
0) get_plane_with_type_for at ../src/backends/native/meta-kms-device.c:150
1) meta_kms_device_get_cursor_plane_for at ../src/backends/native/meta-kms-device.c:173
2) has_cursor_plane at ../src/backends/native/meta-cursor-renderer-native.c:678
3) foreach_crtc at ../src/backends/meta-logical-monitor.c:247
4) meta_monitor_mode_foreach_crtc at ../src/backends/meta-monitor.c:1920
5) meta_logical_monitor_foreach_crtc at ../src/backends/meta-logical-monitor.c:274
6) crtcs_has_cursor_planes at ../src/backends/native/meta-cursor-renderer-native.c:718
7) should_have_hw_cursor at ../src/backends/native/meta-cursor-renderer-native.c:881
8) meta_cursor_renderer_native_update_cursor at ../src/backends/native/meta-cursor-renderer-native.c:1085
9) meta_cursor_renderer_update_cursor at ../src/backends/meta-cursor-renderer.c:411
Related: https://bugzilla.redhat.com/show_bug.cgi?id=2000183
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1991>
On a KMS backed CRTC, hardware cursor are supported when there are
cursor planes to assign them to. Note that when using legacy mode
setting, fake cursor planes are added when adequate.
On virtual CRTCs, used with virtual monitors, the equivalent of hardware
cursor are always supported, as they are sent using embedded PipeWire
stream metadata.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1991>
This can happen if a texture was newly assigned to the actor, but the
unobscured region hasn't been updated yet. Without bailing here, the
actor would display correctly via direct scanout, but other parts of
mutter would continue considering it obscured, which would e.g. result
in no frame callbacks getting sent for its surface.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1636
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2112>
Previously we chose to only anti-alias texels inside the boundary
(`clip_radius - 1.0`) but zoomed in you could see it was slightly smaller
than the correct curve (#2024).
Similarly if you choose to only anti-alias texels outside that edge
(`clip_radius + 1.0`) then you'd get an overly convex curve that doesn't
match up with the straight line sections.
So now we anti-alias texels that intersect the circle boundary, regardless
of which side they are mostly on. For efficiency we define "intersect" to
mean any texel whose center is within 0.5 of the theoretical edge.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2024
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2102>
It was dropping to zero after the first frame because it hadn't been
incremented high enough. So the second frame would crash with:
```
#0 g_type_check_instance_cast
#1 META_DRM_BUFFER
#2 copy_shared_framebuffer_cpu
```
That's the CPU-copy path (fallback-fallback) that probably no one is using
but it does work after this fix. Exactly the same issue as was fixed
in `copy_shared_framebuffer_primary_gpu` by 36352f44f9.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2104>
A client can create a token without any seat, serial, or surface. In
this case, we'd still try to grab, which would run into some unforseen
code paths, potentially resulting in the following crash:
0) meta_wayland_tablet_seat_device_added (tablet_seat=0x55dff4271c90,
device=0x7f87b80655b0) at
../src/wayland/meta-wayland-tablet-seat.c:200
1) meta_wayland_tablet_seat_new (seat=0x0, manager=0x55dff3ec7b40) at
../src/wayland/meta-wayland-tablet-seat.c:283
2) meta_wayland_tablet_manager_ensure_seat (manager=manager@entry=0x55dff3ec7b40,
seat=seat@entry=0x0) at
../src/wayland/meta-wayland-tablet-manager.c:239
3) meta_wayland_tablet_manager_ensure_seat (seat=0x0, manager=0x55dff3ec7b40) at
../src/wayland/meta-wayland-touch.c:595
4) meta_wayland_seat_get_grab_info (seat=0x0, surface=0x55dff43ff5b0,
serial=0, require_pressed=0, x=0x0, y=0x0) at
../src/wayland/meta-wayland-seat.c:479
5) activation_activate (...) at
../src/wayland/meta-wayland-activation.c:261
Fix this by not trying to grab if not enough parameters was passed when
creating the token. Also add a test case that reproduces the above
crash.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2081>
When an activation times out, we'll be signalled two signals on the
startup sequence object: "timeout", and "complete".
Normally, the "complete" signal is emitted when a startup sequence is
completed succesfully by it being used for activation, and in this case,
the xdg_activation implementation should remove the sequence from the
startup notification machinery.
However, in the timeout case, we should not remove it, as the startup
notification machinery itself will deal with this. If we would, we'd end
up with use-after-free issues, as the sequence would be finalized when
removed the first time.
To avoid this, just clean up the Wayland side in the "timeout" signal
handler, leaving the "complete" signal handler early out if it was
already handled by it.
This avoids crashes like:
0) g_type_check_instance (type_instance=type_instance@entry=0xdd6740)
1) g_signal_handlers_disconnect_matched (instance=0xdd6740, ...)
2) meta_startup_notification_remove_sequence (sn=0x4cc890,
seq=0xdd6740) at
../src/core/startup-notification.c:544
3) startup_sequence_timeout (data=0x4cc890, ...) at
../src/core/startup-notification.c:504
4) g_timeout_dispatch (...) at ../glib/gmain.c:4933
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2081>
We set it via setenv(), and might not have the MetaX11Display at hand.
This fixes a crash when the stuck-client dialog (using zenity) appears
without any X1 client having appeared.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2081>
This test ensures that windows that were resized such that they extend
beyond the screen will be moved to be fully on the screen (if possible).
This has been working on X11 since forever, but on Wayland only since
the last commit.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2103>
The previous code was trying to detect client resizes by only
considering resizes without any pending configurations as client
resizes. There can however be pending configurations that do not involve
resizing, such as ones triggered by state changes. These may also stay
unacknowledged by the client until the next size change. This was
causing client resizes after showing the window (and therefore changing
its status to focused) to not be detected as client resize.
Fix this by checking whether the queue has any configuration with size
changes rather than just whether it is empty.
Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2023
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2103>
Used to log multiple line entries. Just make continue prefix things, no
need to mess with maybe-prefixing; it'll just complicate using some less
custom logging functionality.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2094>
If some connectors disappeared, but the rest didn't change, we missed
actually removing the ones that disappeared, as we incorrectly assumed
nothing changed. Fix this by only assuming nothing changed if 1) we
didn't add any connector, and 2) we have the same amount of connectors
as before the hotplug event. The connector comparison checking makes
sure we report changes if anything of the still available connectors
changed.
Fixes: a8d11161b6
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2007
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2097>
When a docking station is disconnected, a few previously existing DRM
connectors may now be gone. When this happens, getting them via the
libdrm API results in NULL pointers returning, and we need to handle
this gracefully by making sure the connector state is properly updated.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2097>
The 'stop_after_next' will execeute one command, then not return to the main
loop until a 'continue' command is passed. Commands will still be
processed between 'stop_after_next' and 'continue'.
This is intended to be used to induce race conditions.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2066>
Currently the stored unconstrained_rect is only ever updated if there
was a move, resize or state change according to the move_resize_internal
implementation. For Wayland windows however resizes or state changes
are done in two steps, first the new configuration is sent to the client
and then once client acknowledges it, it is set on the mutter side in
another move_resize_internal call. Only the second call would result in
the unconstrained_rect being updated.
This started causing problems when unfullscreening windows was
immediately followed by a strut change. These strut changes started
happening in gnome-shell due to the visibility of the panel now being
considered for the struts and the presence of a fullscreen causing it to
be hidden until unfullscreen. In this situation first the unfullscreen
would resize the window to its pre-fullscreen size as expected, but then
the strut change triggers another window resize. This window resize is
based on the stored unconstrained_rect, which is still at the fullscreen
size because the unfullscreen resize only has sent its configuration,
but it has not been acknowledged yet. As a result the strut change
causes a resize to the fullscreen size which due to the constraints now
looks like a maximized window.
To fix this always update the unconstrained_rect when the requested size
has changed, but not when a previous request has been acknowledged
unless it is originating from the client itself.
If this included the move_resize_internal call from acknowledging the
size as well, it would be possible for this to be delayed long enough on
the client side to overwrite an intermediate resize originating from
mutter. And if this did not include resizes originating from the client,
clients would not be able to set an initial window size.
Fixes https://gitlab.gnome.org/GNOME/mutter/-/issues/1973
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2066>
meta_window_wayland_finish_move_resize() is called for both, finishing
a resize that has been requested through/by mutter and for resizes
directly done by the client. This introduces a CLIENT_RESIZE flag to
differentiate the former from the latter. Having this distinction is
required to know what the last requested size by either the client or
mutter is while ignoring older requests that might only have been
applied now.
This excludes client resizes when there are still pending
configurations, because the resize is known to be only temporary.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2066>
Because POSIX sh was, with hindsight, not a particularly well-designed
programming language, if we don't 'set -e', then we'll respond to failure
of a setup command such as cd by carrying on regardless.
Signed-off-by: Simon McVittie <smcv@debian.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2009>
The assumption here seems to be that it's an overlay onto the
current environment which would make sense; but the implementation in
gnome-desktop-testing currently removes all other environment variables
(see GNOME/gnome-desktop-testing#1). This causes test failure when the
tests are run in Debian's autopkgtest framework, possibly because PATH
is cleared.
Signed-off-by: Simon McVittie <smcv@debian.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2009>
Systems with AMD GPUs do not take advantage of Mutter's zero-copy path
when driving DisplayLink screens. This is due to a very slow CPU access
to the zero-copy texture. Instead they fall back on primary GPU doing a
copy of the texture for fast CPU access. This commit accelerates texture
copy by working through damage regions only.
Tests on a 4K screen with windowed applications show significant
reduction of GPU utilisation.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2033>
If one would end up with an actor attached to mapped actor, where the
attached actor doesn't itself have an up to date stage view list while
listening on the stage for updating, when clearing the stage views of
the list, anything that would query the stage views list at this time
would end up accessing freed memory.
This could happen if
1) An actor was added to a newly created container actor attached to
the stage
2) The actor got a timeline attached to it
3) The actor was moved to a container that already was mapped
4) A hotplug happened
After (1) both the container and actor would not have any stage views.
After (2) the timeline would listen on the stage for stage views
updates. After (3) the actor would still listen on the stage for stage
views updates. When (4) happened, the actor would be signalled when the
stage got its stage view cleared, at which point it would traverse up
its actor's tree finding an appropriate stage view to base its animation
on. The problem here would be that it'd query the already mapped
container and its yet-to-be-cleared stage view list, resulting in
use-after free, resulting in for example the following backtrace:
0) g_type_check_instance_cast ()
1) CLUTTER_STAGE_VIEW ()
2) clutter_actor_pick_frame_clock ()
3) clutter_actor_pick_frame_clock ()
4) update_frame_clock ()
5) on_frame_clock_actor_stage_views_changed ()
6) g_closure_invoke ()
7) signal_emit_unlocked_R ()
8) g_signal_emit_valist ()
9) g_signal_emit ()
10) clear_stage_views_cb ()
11) _clutter_actor_traverse_depth ()
12) _clutter_actor_traverse ()
13) clutter_actor_clear_stage_views_recursive ()
14) clutter_stage_clear_stage_views ()
...
Avoid this issue by making sure that we don't emit 'stage-views-changed'
signals while the actor tree is in an invalid state. While we now end up
traversing tree twice, it doesn't change the Big-O notation. It has not
been measured whether this has any noticible performance impact.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/1950
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2025>
Certains keys (such as ~ and |) are in the keyboard map behind the
second shift level. This means in order for them to be input, the
shift key needs to be held down by the user.
The GNOME Shell on-screen keyboard presents these keys separately on
a page of keys that has no shift key. Instead, it relies on mutter
to set a shift latch before the key event is emitted. A shift latch
is a virtual press of the shift key that automatically gets released
after the next key press (in our case the ~ or | key).
The problem is using a shift latch doesn't work very well in the face
of key repeat. The latch is automatically released after the first
press, and subsequent repeats of that press no longer have shift
latched to them.
This commit fixes the problem by using a shift lock instead of a shift
latch. A shift lock is never implicitly released, so it remains
in place for the duration of key repeat.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2045>
The `guess_candidates()` function scores each display that an input
device could be mapped to and then uses the `sort_by_score()` comparator
to find the best option. The function expects the list to be sorted from
best to worst, but the comparator currently sorts them in the opposite
order. This causes the function to end up returning the _worst_ match
rather than the the best. This commit reverses the sort order of the
comparator so that the best display can be returned as intended.
Closes: #1889
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1934>
This does two things to frown upon:
- Modifies ClutterEvent structs, while the effort is to have those
completely opaque, and readonly after creation from the input
thread side.
- Stores state in the ClutterInputDevice struct, event though those
are also considered static after creation, managed by the input
thread, etc.
Stop doing that. This makes all events just forwarded as-is in
the ClutterStage/clutter-main.c code.
Handling of click count sounds like material for a ClutterGestureAction
(or perhaps ClutterClickAction), all of both callers now do it in place
at the moment, while gestures lack a better state tracking and management.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2024>
This will not try the captured-event shenanigans to emulate grab
behavior, instead relying on event delivery being influenced by
other grab mechanisms.
While at it, improve handling of additional touchpoints by
cancelling the click action right away, as the differences in
event handling make this unwanted behavior surface.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2024>
By default, the pan action performs matrix translations on the
child widget. Nobody wants that (or, nobody wants *just* that).
It's cleaner not to mix mechanism and effect in ClutterGestureAction
subclasses, so drop this base implementation, and change the signal
accumulator so it's more similar to event signals (not that it's
used any longer, anyway).
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2024>
Mutter already calculates and tracks the damage rectangles to redraw
only areas of the screen that change since the last time a buffer was
used.
This patch extends this by using the EGL_KHR_partial_update extension to
inform the GPU in advance that only those areas will be changed, which
may allow for further optimization.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2023>
Add support for a cogl function to set the damage_region on an onscreen
framebuffer.
The goal of this is to enable using the EGL_KHR_partial_update extension
which can potentially reduce memory bandwidth usage by some GPUs,
particularly on embedded GPUs that operate on a tiling rendering model.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2023>
Defining valid values makes
1. changing settings less error prone
2. sure they are discoverable.
This does reset the values once on update, but fortunately does *not*
change the way to set the values. Thus e.g. enabling fractional scaling
via terminal command still works as before and internet guides stay valid.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1961>
This will make clients immediately aware of the output disappearing,
while still allowing for a grace period of 10 seconds for attempting to
bind to it before it turning into a protocol error. This API added as
part of wayland 1.18.
This requires us to not add the output resource to the output resource
list, if the output was made inert. This effectively makes the resource
useless, but that is harmless, since shortly after, the client will
clean it up anyway.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1796>
This will be crucial when we start to remove the global directly when an
output is removed, as that means Xwayland might have removed the output
before we managed to get our queries in.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1796>
We setup Xwayland in an early phase of the X11 display, before we had a
MetaX11Display, and teared down in a couple of places happening when
tearing down the Xwayland integration if the X server died or
terminated. It was a bit hard to follow what happened and when it
happened. Attempt to clean this up a bit, with things being structured
as follows:
* Early during X11 display connection setup, only setup the rudimentary
X11 hooks, being the libX11 error callbacks, and adding the local
user to XHost.
* Move "initialize Xwayland component" code to a new
'x11-display-setup' signal handler. Things setup here are cleaned up
in the 'x11-display-closing' handler.
* Connect to 'x11-display-setup' and 'x11-display-closing' up front,
and stay connected to these two.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1796>
This old handling of session files looked on ~/.mutter, which has
been unused and unsupported for a long time. It also had paths were
the GError was leaked. Fix both by dropping the legacy code, and
falling back to the common error paths.
CID: #1502682
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2061>
This warning is actually dead code, since should_be_mapped and
must_be_realized are always set to the same value, so it does not
make sense to check for "a && !b".
Turn this into an assert so we avoid the dead branch, but do not
remove the variable duplication so the more aptly named variable
is used where it belongs, for clarity.
CID: #1506254
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2061>
When grabbing the devices, there's no error paths that would quit
late enough that both pointer and keyboard would need ungrabbing,
so the keyboard checks were dead code.
Fix this by dropping the boolean variable checks, and adding goto
labels to unroll the operation properly at every stage.
CID: #1418254
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2061>
The monitor orientation tests do a lot of things in sequence. Replace
some of the comments with g_test_message() so that the log from a failed
test gives us a better idea of how far we got.
Signed-off-by: Simon McVittie <smcv@debian.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2049>
Previously, we were waiting up to 300ms for the signal, then proceeding
anyway. However, 300ms is not necessarily long enough to wait on an
autobuilder that might be heavily loaded, particularly if it's a non-x86
with different performance characteristics.
Conversely, if mutter responds to the D-Bus signal from the mock sensor
before we have connected to the signal, then we cannot expect to receive
the signal - it was already emitted, but we missed it. In this case, we
need to avoid waiting.
One remaining use of wait_for_orientation_changes() that would previously
always have timed out was in
meta_test_orientation_manager_has_accelerometer(), which does not
actually expect to see an orientation-changed signal. Make this wait
for the accelerometer to be detected instead.
Resolves: https://gitlab.gnome.org/GNOME/mutter/-/issues/1967
Bug-Debian: https://bugs.debian.org/995929
Signed-off-by: Simon McVittie <smcv@debian.org>
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2049>
When we use gbm together with the NVIDIA driver, we want the EGL/Vulkan
clients to do the same, instead of using the EGLStream paths. To achieve
that, make sure to only initialize the EGLStream controller when we
didn't end up using gbm as the renderer backend.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2052>
This switches the order of what renderer mode is tried first, so that
the gbm renderer mode is preferred on an NVIDIA driver where it is
supported.
We fall back to still try the EGLDevice renderer mode if the created gbm
renderer is not hardware accelerated.
The last fallback is still to use the gbm renderer, even if it is not
hardware accelerated, as this is needed when hardware acceleration isn't
available at all. The original reason for the old order was due to the
fact that a gbm renderer without hardware acceleration would succeed
even on NVIDIA driver that didn't support gbm.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2051>
This replaces functionality that MetaRenderDevice and friends has
learned, e.g. buffer allocation, EGLDisplay creation, with the usage of
those helper objects. The main objective is to shrink
meta-renderer-native.c and by extension meta-onscreen-native.c, moving
its functionality into more isolated objects.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1854>
All render devices that have a device file backing them might be able to
allocate dumb buffers, so add a helper for doing that. Will indirectly
result in an error up front on a surfaceless render device due to lack
of a device file.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1854>
It might not be needed by the user of the buffer, so don't always
require it up front. Instead make sure that any user that needs it first
calls "meta_drm_buffer_ensure_fb_id()" to create the ID.
Only the plain gbm implementation creates the ID lazilly, the other
still does it on construction due to the objects used to create them
only existing during construction.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1854>
Mostly calls into gbm_bo_* API, or something somewhat similar when on
dumb buffers. Added API are:
* get offset for plane
* get bpp (bits per pixel)
* get modifier
This will allow users of MetaDrmBuffer to avoid having to "extract" the
gbm_bo to get these metadata.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1854>
The purpose of MetaRenderDevice is to contain the logics related to a
render device; i.e. e.g. a gbm_device, or an EGLDevice. It's meant to
help abstract away unrelated details from where it's eventually used,
which will be by MetaRendererNative and the MetaOnscreenNative
instances.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1854>
Meant for MetaRenderer and everything related that deals with turning
composited frames, or client buffers, into mode set updates. This is
slightly related to the debug topic 'kms' is meant for the KMS details.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1854>
It was a feature relevant for when Clutter was an application toolkit
that wanted the application window to communicate a minimum size to the
windowing system.
Now, clutter is part of the windowing system component, so this feature
doesn't make any sense, so remove it.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2002>
This feature was configured depending on whether the Cogl backend
reported COGL_WINSYS_FEATURE_MULTIPLE_ONSCREEN or not. All cogl backends
do report this, so any code handled the 'static' case were never used.
While we only ever use one stage, it's arguable more correct to
consilidate on the single stage case, but multiple stages is something
that might be desirable for e.g. a remote lock screen, so lets keep this
logic intact.
This has the side effect of completely removing backend features, as
this was the only left-over feature detection that they handled.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2002>
This changes the setup phase of clutter to not be result of calling an
init function that sets up a few global singletons, via global singleton
setup vfuncs.
The way it worked was that mutter first did some initial setup
(connecting to the X11 server), then set a "custom backend" setup vfunc
global, before calling clutter_init().
During the clutter_init() call, the context and backend was setup by
calling the global singleton getters, which implicitly created the
backend and context on-demand.
This has now changed to mutter explicitly creating a `ClutterContext`
(which is actually a `ClutterMainContext`, but with the name shortened to
be consistent with `CoglContext` and `MetaContext`), calling it with a
backend constructor vfunc and user data pointer.
This function now explicitly creates the backend, without having to go
via the previously set global vfunc.
This changes the behavior of some "get_default()" like functions, which
will now fail if called after mutter has shut down, as when it does so,
it now destroys the backends and contexts, not only its own, but the
clutter ones too.
The "ownership" of the clutter backend is also moved to
`ClutterContext`, and MetaBackend is changed to fetch it via the clutter
context.
This also removed the unused option parsing that existed in clutter.
In some places, NULL checks for fetching the clutter context, or
backend, and fetching the cogl context from the clutter backend, had to
be added.
The reason for this is that some code that handles EGL contexts attempts
to restore the cogl EGL context tracking so that the right EGL context
is used by cogl the next time. This makes no sense to do before Cogl and
Clutter are even initialized, which was the case. It wasn't noticed
because the relevant singletons were initialized on demand via their
"getters".
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2002>
In various places we retrieved the default seat from the ClutterBackend.
All the clutter backends implement this by calling
meta_backend_get_default_seat() which will then return
MetaBackendPrivate::default_seat.
Lets avoid this by fetching the default seat directly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2002>
Rename all instances of `MetaClutterBackendX11` so they are called
`clutter_backend_x11`. This is because `MetaBackendX11` will start to be
used for some things, and having both be named `backend_x11` would be
confusing.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2002>
This one is a trivial wrapper around clutter_actor_get_children(), so just
use that in the two places where clutter_container_get_children() is used,
and remove clutter_container_get_children().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2057>
Right now we damage the stage even if an actor is not mapped, for
example in the overview.
Stop doing so, reducing over-paint significantly in some situations.
Clones will still do stage damage on their own.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2035>
Quoting the spec for `wl_data_device::drop`:
> If the resulting action is "ask", the action will not be considered
> final. The drag-and-drop destination is expected to perform one last
> wl_data_offer.set_actions request, or wl_data_offer.destroy in order
> to cancel the operation.
We did not respect the action choosen by the drop destination when
it called `wl_data_offer::set_actions` after `wl_data_device::drop`
if a user override was still active. This eventually resulted in
a protocol error in `wl_data_offer::finish`, as the current action
could still be `ask`.
Fix this by only allowing a user override to `ask` before `drop` is
called, thus making sure the final `set_actions` preference is
honored.
Closes https://gitlab.gnome.org/GNOME/mutter/-/issues/1952
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2043>
Since the CI pipeline now runs `meson dist` which includes a test
for an up-to-date NEWS entry, post-branch/release version bumps
have become unwieldy.
It's still useful to increase the API version right after branching
though, so return to bumping it manually and do so now.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2053>
2021-10-14 18:37:45 +02:00
2218 changed files with 237632 additions and 161807 deletions
print("\nIssue the following command in your local tree to apply the suggested changes (needs uncrustify installed):\n\n $ git rebase origin/main --exec \"./check-style.py -r\"\n")
print(f"""
Issue the following commands in your local tree to apply the suggested changes:
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
