Commit Graph

2789 Commits

Author SHA1 Message Date
Florian Müllner
018786dab8 output-xrandr: Don't treat 0 as invalid backlight value
Whether a value is in range depends on the backlight-min/max values,
and I didn't spot anything that excludes 0 as a valid lower limit
outright.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2947>
2023-04-04 19:30:06 +00:00
Florian Müllner
aa1eaa8120 output-xrandr: Consistently return -1 on error
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>
2023-04-04 19:30:05 +00:00
Chris Mayo
28a599631d backend: Fix build with -Dremote_desktop=false
../mutter-44.0/src/backends/meta-backend.c: In function ‘meta_backend_real_post_init’:
../mutter-44.0/src/backends/meta-backend.c:560:7: error: ‘MetaBackendPrivate’ {aka ‘struct _MetaBackendPrivate’} has no member named ‘remote_access_controller’
  560 |   priv->remote_access_controller =
      |       ^~

Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2655
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2935>
2023-04-03 21:34:54 +00:00
Jonas Ådahl
8ca94f4d3f backends: Use meta_gamma_lut_new_sized() in a few more places
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2944>
2023-03-30 11:21:02 +00:00
Daniel van Vugt
61801a713a onscreen/native: Avoid freezing the frame clock on failed cursor commits
Since 73fb64cbb6 we have preferred to handle failed updates via callback
but the cursor-only update path was forgotten and so wasn't getting frame
notifications. And so the frame clock would freeze.

Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2691
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2933>
2023-03-29 10:43:31 +00:00
Jonas Ådahl
83a6a011d7 backend/nested: Ignore setting pointer constraint
This avoids a crash when pointer constraints are enabled by Wayland
clients.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2932>
2023-03-22 13:46:21 +00:00
Sebastian Wick
3988f5a47f backends: Fall back to the default and not the unknown color space
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>
2023-03-20 10:00:36 +00:00
Daniel van Vugt
1be2f635e9 kms/impl-device/simple: Avoid destroying a NULL hash table
Which would trigger:
```
g_hash_table_destroy: assertion 'hash_table != NULL' failed
```
on non-KMS systems like with `nvidia-drm.modeset=0`.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2904>
2023-03-20 09:20:16 +00:00
Jonas Ådahl
176418d0e7 kms/device: Avoid leaking some fields
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2926>
2023-03-19 22:32:43 +00:00
Sebastian Keller
7fbef2cb6e kms/impl-device: Fix result listener list leak
queue_result_feedback() takes ownership of the result listeners list via
meta_kms_update_take_result_listeners(), but does not free it.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2922>
2023-03-19 22:05:00 +00:00
Sebastian Wick
0ed2503140 backends/native: Report correct Gamma support on the KMS properties path
The gamma property blob can be zero to indicate passthrough but Gamma is
still supported in that case.

Fixes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2686
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2916>
2023-03-18 17:52:27 +00:00
Jonas Ådahl
214a7d393b monitor-manager: Apply switch-config in idle callback
Just as with restoring the previous monitor configuration in case the
user clicked "revert" in GNOME Shell's monitor configuration
confirmation dialog, we need to do switch configs in an idle callback as
well.

Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/2694
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2912>
2023-03-18 16:20:49 +00:00
Jonas Ådahl
d31b781efb monitor-manager: Restore old config in idle callback when unconfirmed
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>
2023-03-18 13:52:10 +00:00
Jonas Ådahl
dbe8141a67 monitor-manager: Use g_clear_handle() to clean up D-Bus name owning
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2901>
2023-03-18 13:52:10 +00:00
Jonas Ådahl
e6a53ef113 monitor-manager: Use guint for handle IDs
To follow convention.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2901>
2023-03-18 13:52:10 +00:00
Sebastian Wick
950fdc6c04 backends/native: Convert MetaOutputColorspace to DRM Colorspace
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2890>
2023-03-04 19:53:30 +00:00
Sebastian Wick
f729a2023e tests: Add tests for HDR Static Metadata encoding/decoding and equality
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>
2023-03-04 09:30:41 +00:00
Sebastian Wick
af714edafb output: Check the EDID for HDR Static Metadata support
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>
2023-03-04 09:30:41 +00:00
Sebastian Wick
7ecdfaa78d edid: Fix MetaEdidStaticMetadataType bitmask
Support for HDR Static Metadata Type 1 is announced in bit 0.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
2023-03-04 09:30:41 +00:00
Sebastian Wick
a8d73e669a backends/native: Wire up color space and HDR metadata
Announce support for color spaces and HDR metadata in OutputKms and
prepare KMS update in OnscreenNative.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
2023-03-04 09:30:41 +00:00
Sebastian Wick
650d30c7bf backends/native: Process color space and HDR md updates
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
2023-03-04 09:30:41 +00:00
Sebastian Wick
f092b6c78c backends/native: Add color space and HDR metadata updates
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>
2023-03-04 09:30:41 +00:00
Sebastian Wick
5984aa8137 backends/native: Let updates require ALLOW_MODESET
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
2023-03-04 09:30:41 +00:00
Sebastian Wick
76e69f2375 backends/native: Read color space and HDR metadata connector state
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>
2023-03-04 09:30:41 +00:00
Sebastian Wick
2e69eafa06 backends/native: Add color space and HDR metadata KMS properties
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>
2023-03-04 09:30:41 +00:00
Sebastian Wick
48e759c443 monitor-manager: Make color space and HDR metadata accessible from lg
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>
2023-03-04 09:30:41 +00:00
Sebastian Wick
fad0efa687 output: Add color space and HDR metadata
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>
2023-03-04 09:30:40 +00:00
Sebastian Wick
47fcb52435 backend/native: Make is-privacy-screen-enabled not CONSTRUCT_ONLY
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2879>
2023-03-04 09:30:40 +00:00
Jonas Ådahl
995f9f9db8 onscreen/native: Hold output and CRTC refs until dispose
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>
2023-03-03 22:36:30 +00:00
Jonas Ådahl
08593ea872 onscreen/native: Hold ref to the output and CRTC until detached
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>
2023-03-03 17:32:04 +00:00
Jonas Ådahl
d1785a3917 onscreen/native: Detach onscreens when disposing
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>
2023-03-03 17:32:04 +00:00
Simon McVittie
ed12df1099 color-device: Don't close lcms profile on error from cd_icc_load_handle
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>
2023-03-03 12:07:02 +00:00
Simon McVittie
e2be2d271b color-device: Make sure lcms_context is not NULL
lcms interprets a NULL context as using a default, non-thread-safe
context, which is unsuitable for mutter's use. Make sure we're always
using a non-trivial context.

Helps: 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>
2023-03-03 12:07:02 +00:00
Jonas Ådahl
d03dce4378 monitor: Use fractional scale result to decide the non-fractional scale
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>
2023-03-03 08:46:02 +00:00
Jonas Ådahl
f3f9200f9a tests: Port headless start test case to VKMS
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>
2023-03-02 23:49:15 +00:00
Jonas Ådahl
9abce2fca7 gpu/kms: Report that we can have outputs if we have connectors
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>
2023-03-02 23:49:15 +00:00
Jonas Ådahl
ffd1e8106c kms/update: Merge connector updates too
This includes privacy screen updates, underscanning etc; this is needed
when merging mode set updates that happen to change these too, which is
rather likely.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2881>
2023-03-02 22:09:16 +01:00
Jonas Ådahl
f27ca241f9 renderer/native: Move per frame KMS update to MetaFrameNative
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>
2023-03-02 01:40:16 +01:00
Jonas Ådahl
90ae14e792 kms/device: Move result queueing to impl device
This means there will be no entry points to the dispatch machinery that
doesn't report about the results.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2855>
2023-03-02 01:39:16 +01:00
Jonas Ådahl
25172c21c9 kms/impl-device: Move fd hold to update when processing
This will let us move the update ownership to the backend, allowing it
to handle it with less restrictions.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2855>
2023-03-02 01:39:16 +01:00
Jonas Ådahl
6d043e6a8f kms/impl-device: Let process() always take ownership of the update
This allows more freedom how to handle the update in the backends.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2855>
2023-03-02 01:39:16 +01:00
Jonas Ådahl
135ed27d27 clutter/frame: Allow setting a backend release function
The one set by the native backend is empty, but it will be used later.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2855>
2023-03-02 01:39:16 +01:00
Jonas Ådahl
7b634df379 renderer-view/native: Allocate backend specific ClutterFrame
This will carry an on-demand created MetaKmsUpdate.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2855>
2023-03-02 01:39:16 +01:00
Dor Askayo
1abd930c56 backends/native: Introduce MetaRendererViewNative
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>
2023-03-02 01:39:16 +01:00
Jonas Ådahl
2dfbbf0068 tests/kms/render: Add test for scanout failure handling
Make sure we repaint correctly if direct scanout updates fail.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
2023-03-01 23:57:32 +00:00
Jonas Ådahl
73fb64cbb6 kms: Always handle KMS update result via callback
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>
2023-03-01 23:57:32 +00:00
Jonas Ådahl
a2e442ab9d kms/result-listener: Let listeners own a feedback ref
Dispatch results via a helper, and let each result hold a reference to
the feedback. This allows for more asynchronous feedback management.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
2023-03-01 23:57:32 +00:00
Jonas Ådahl
5dbd9cc1b8 onscreen/native: Process TEST_ONLY updates directly on the device
No point in going via a thin meta_kms_* wrapper.

Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/2854>
2023-03-01 23:57:32 +00:00
Jonas Ådahl
b1b9a9aac1 kms/update: Change lock() to seal()
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>
2023-03-01 23:57:32 +00:00
Jonas Ådahl
02b1cfe08f onscreen/native: Handle unexpected scanout failures async
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>
2023-03-01 23:57:32 +00:00