Commit Graph

376 Commits

Author SHA1 Message Date
Simon Ser 5b1b43c68c backend/drm: make wlr_drm_plane.{pending,queued,current}_fb pointers
This will be useful once we start re-using wlr_drm_fb.
2021-01-10 11:24:35 +01:00
Simon Ser adfb7cd35a backend/drm: stash pending page-flip CRTC
wlr_drm_connector.crtc may be updated by the DRM backend while a
page-flip is pending. In this case, the page-flip handler won't be able
to find the right wlr_drm_connector from the CRTC ID.

Save the CRTC when performing a page-flip to ensure we always find the
right connector when we get the event.
2021-01-03 19:35:02 +01:00
Simon Ser bec1e6b149 backend/drm: remove special linear case for cursor plane
We now properly mark the cursor plane's formats as linear-only, and we
now have a version of wlr_drm_format_intersect that handles the case of
linear-only formats and implicit modifiers.

We can remove the special drm_plane_init_surface flag we had for cursor
planes. This also allows us to use a non-linear layout for cursor planes
on drivers that support it.

Tested on amdgpu GFX9.
2020-12-30 11:17:43 +01:00
Simon Ser 7ea0e9f277 backend/drm: force LINEAR for cursor plane formats
If the kernel driver doesn't support modifiers, it still expects cursor
FBs to have a LINEAR layout. See [1] for expectations for framebuffers
attached to the cursor plane.

[1]: https://patchwork.freedesktop.org/patch/408512/
2020-12-30 11:17:43 +01:00
Simon Ser b9460ab724
Stop using wlr_texture_get_size
Just use wlr_texture.{width,height} directly.
2020-12-25 12:21:29 +01:00
Simon Ser a7a230ebef
backend/drm: make drmModePlane arg in add_plane const
Make it clearer it'll be free'd after add_plane returns.
2020-12-24 12:34:13 +01:00
Simon Ser c011a0e2ed
backend/drm: add special case for cursor in add_plane
The code calling add_plane now makes sure not to add a cursor plane to a
CRTC which already has one.
2020-12-24 12:31:20 +01:00
Simon Ser 5548406667
backend/drm: log whether ADDFB2 modifiers are supported 2020-12-24 12:29:30 +01:00
Simon Ser 64a2ca4dba
backend/drm: don't retry page-flip when modifiers are unsupported
Parse WLR_DRM_NO_MODIFIERS at startup. Don't parse IN_FORMATS when
WLR_DRM_NO_MODIFIERS is set, so that the legacy behaviour is better
reproduced.

When modifiers aren't supported, try the initial page-flip once only.
2020-12-23 19:49:27 +01:00
Simon Ser 54ec17ff64
backend/drm: use plural form for possible_crtcs
Make it clearer that this is a bitmask of CRTC indices.
2020-12-23 12:14:36 +01:00
Simon Ser 2de400a541
backend/drm: assert connector state after wlr_output_destroy 2020-12-22 19:53:33 +01:00
Simon Ser 85cf4b235d
backend/drm: destroy DRM connectors that disappear
This was lost during the refactoring. We were previously calling
wlr_output_destroy, which destroyed the connector as well.

Fixes: 248c7787c7 ("backend/drm: refactor wlr_output destruction")
2020-12-22 19:38:29 +01:00
Simon Ser d3bcd63a40 backend/drm: remove drm_fb_acquire
Instead, import the buffer into GBM and KMS in drm_fb_import. Also move
the multi-GPU copy there if necessary.
2020-12-22 15:53:10 +01:00
Simon Ser dabd2e7207 backend/drm: grab DMA-BUF from wlr_buffer instead of gbm_bo
Get the DMA-BUF directly out of the wlr_buffer instead of relying on the
gbm_bo. This eliminates a roundtrip through GBM.
2020-12-22 15:53:10 +01:00
Ronan Pigott 917ecca58e backend/drm: avoid gcc stringop-truncation warning 2020-12-18 22:19:16 +01:00
Simon Ser 352fdd1bb0
backend/drm: remove unused wlr_drm_plane.drm_format 2020-12-18 11:44:50 +01:00
Simon Ser 248c7787c7 backend/drm: refactor wlr_output destruction
Instead of hand-rolling our own manual wlr_output cleanup function, rely
on wlr_output_destroy to remove an output from the compositor's state.
2020-12-18 10:48:44 +01:00
Simon Ser 019ffe8a5b backend/drm: introduce wlr_drm_connector.name
The DRM backend is a little special when it comes to wlr_outputs: the
wlr_drm_connectors are long-lived and are created even when no screen is
connected.

A wlr_drm_connector only advertises a wlr_output to the compositor when
a screen is connected. As such, most of wlr_output's state is invalid
when the connector is disconnected.

We want to stop using wlr_output state on disconnected connectors.
Introduce wlr_drm_connector.name which is always valid regardless of the
connector status to avoid reading wlr_output.name when disconnected.
2020-12-18 10:48:44 +01:00
Simon Ser c89b131f29 backend/drm: introduce wlr_drm_conn_log
Simplify and unify connector-specific logging with a new
wlr_drm_conn_log macro. This makes it easier to understand which
connector a failure is about, without having to explicitly integrate the
connector name in each log message.
2020-12-18 10:48:44 +01:00
Simon Ser 253f447329
backend/drm: print DRM device name when scanning connectors 2020-12-17 20:50:19 +01:00
Simon Ser 60001a75a2
backend/drm: remove nouveau workaround
The workaround is broken because drm_fb_acquire doesn't leave the EGL
context current anymore. We'll need to re-introduce it.

References: https://github.com/swaywm/wlroots/issues/2525
2020-12-17 20:38:02 +01:00
Simon Ser da2a216934
backend/drm: add wlr_drm_connector.backend
This allows the DRM code to have direct access to the wlr_drm_backend
without having to go through an upcast via get_drm_backend_from_backend.
2020-12-15 22:56:14 +01:00
Simon Ser 93cd3a79b2 backend/drm: stop using GBM flags
gbm_bo_create_with_modifiers doesn't take GBM flags, so our
wlr_gbm_allocator interface doesn't either. We were still internally
using GBM flags in the DRM backend, leading to awkward back-and-forth
conversions.

The only flag passed to drm_plane_init_surface was GBM_BO_USE_LINEAR, so
turn that into a bool to make sure other flags can't be passed in.

Move the "force linear" logic out of init_drm_surface, because the
supplied wlr_drm_format should already contain that information.
2020-12-13 13:20:39 +01:00
Simon Ser 4c363a564f backend/drm: remove workaround for amdgpu DP MST
Closes: https://github.com/swaywm/wlroots/issues/2533
2020-12-11 23:27:13 +01:00
Simon Ser 6ff478632a backend/drm: remove EGL config
Since we're using wlr_swapchain, we don't need to provide an EGL config.
2020-12-09 14:25:41 +01:00
Stephane Chauveau b790e5ea34 backend/drm: don't assume possible_crtcs has only one bit set
This isn't necessarily the case [1].

This should fix an assertion failure on Raspberry Pi 4 dual screen.

[1]: https://lists.freedesktop.org/archives/dri-devel/2020-August/275142.html

Closes: https://github.com/swaywm/wlroots/issues/1943
Co-authored-by: Simon Ser <contact@emersion.fr>
2020-12-08 18:38:42 +01:00
Simon Ser 863acb26c0 backend/drm: stop tracking overlay planes
We don't do anything with them. Once we do, we can easily add this back.
2020-12-08 18:38:42 +01:00
Ilia Bozhinov 54b7ca56c0 drm: do not unset make/model before emitting destroy event 2020-12-07 10:39:29 +01:00
Simon Ser 1336ad2a23
backend/drm: remove unused if in drm_connector_move_cursor
We return early if we don't have a plane.
2020-12-04 19:34:35 +01:00
Simon Ser 8bc5a92a98 Revert "backend/drm: stop force-probing connectors"
This reverts commit 713c1661b7.

It turns out we do need to force-probe on startup and on hotplug [1].
This is unfortunate, but that's just how the uAPI works. Fixing this
would require patching the kernel.

[1]: https://lists.freedesktop.org/archives/dri-devel/2020-November/289506.html

Closes: https://github.com/swaywm/wlroots/issues/2499
2020-12-01 11:31:03 +01:00
Simon Ser 713c1661b7 backend/drm: stop force-probing connectors
After discussing with Pekka and Daniel on #dri-devel, we concluded [1]
that user-space shouldn't need to force-probe connectors. Force-probing
can take some time, so using drmModeGetConnectorCurrent can result in
faster start-up.

Users can manually trigger a force-probe if necessary:

    echo detect | sudo tee /sys/class/drm/card0-DP-1/status

Or just by running a tool like drm_info.

A similar change has been submitted to Weston [2].

[1]: https://lists.freedesktop.org/archives/dri-devel/2020-November/287728.html
[2]: https://gitlab.freedesktop.org/wayland/weston/-/issues/437
2020-11-24 15:00:46 +01:00
Kenny Levinsen fb3bea8014 backend/drm: Use legacy gamma size for legacy backend
We would always return the GAMMA_LUT_SIZE property if available, and
only fall back to legacy gamma size otherwise. This leads to issues if
both are available in differs in size while we use the legacy backend.

Ensure that we only return the legacy size if we're using the legacy
backend.

Closes: https://github.com/swaywm/wlroots/issues/2429
2020-11-19 22:46:25 +01:00
Simon Ser eef8b3dde8 backend/drm: check drm_surface_render_black_frame return value
This avoids hitting an assertion in drm_fb_lock_surface when
we failed to render a black frame.
2020-11-15 22:48:42 +01:00
Simon Ser 8058e338ea backend/drm: get rid of wlr_drm_fb_type
Since all DRM FBs are backed by a wlr_buffer, there's no need for this
anymore.
2020-11-15 22:48:42 +01:00
Simon Ser 68a8d99055 backend/drm: add support for wlr_swapchain buffer age 2020-11-15 22:48:42 +01:00
Simon Ser ef846a8839 backend/drm: use wlr_swapchain 2020-11-15 22:48:42 +01:00
Simon Ser 1328477a82 backend/drm: export pending FB in export_dmabuf, if any
This allows callers to grab the current frame right after committing it,
without having to incur a full vblank worth of latency.
2020-11-05 09:43:42 +01:00
Simon Ser ee43ef3c9d backend/drm: fix "a page-flip is already pending" errors on modeset
When performing a modeset, the DRM backend will request a page-flip
event. However frame_pending wasn't set to true, so any subsequent
wlr_output_schedule_frame calls would imemdiately trigger a synthetic
frame event, asking the compositor to submit a new frame. Committing the
new frame fails with "a page-flip is already pending" error in the DRM
backend.
2020-11-03 08:09:54 +01:00
Marten Ringwelski 85757665e6 backend/drm: Check if output is enabled before sending frame event
When an output is disabled one last pageflip will happen to disable it.
Currently this pageflip causes a frame event.
Since the output is disabled we don't want to send this frame event.
2020-10-30 10:04:54 +01:00
Devin J. Pohly aaf490d794 drm: fix uninitialized read
get_drm_prop_blob does not set path_len if it returns NULL.  Check the
return value before path_len to avoid reading uninitialized memory.

(Granted, this doesn't change the logic at all, but it does make
Valgrind a bit happier.)
2020-07-31 09:32:14 +02:00
Antonin Décimo d9bb792794 Fix incorrect format parameters 2020-07-27 10:49:19 +02:00
Simon Ser b6377b59ff backend/drm: check drm_surface_make_current return value
drm_connector_set_cursor wasn't checking the return value of the
drm_surface_make_current call. On failure, this results in a failed
assertion in wlr_renderer_begin (because no rendering context is
current).
2020-07-07 11:18:07 -06:00
Simon Ser bf93d2e67c output: rename impl->rollback to rollback_render
The output backend API is now mostly state-less thanks to the atomic
hooks (commit and test). There is one exception though: attach_render.
This function makes the rendering context current. However sometimes the
compositor might decide not to render after attach_render (e.g. when
there's nothing new to render to the back buffer). Thus
wlr_output_rollback has been introduced to revert the pending state.

Because the output backend API is mostly state-less, the only thing
wlr_output_impl.rollback needs to do is revert the current rendering
context. Rename the function to rollback_render to make this clear. Add
a check in the common wlr_output code to only call rollback_render when
attach_buffer has been previously called.

On the long term, we'll be able to remove attach_render and
rollback_render together.
2020-06-19 11:50:42 -06:00
Simon Ser e77c046cf9 backend/drm: fix stack overflow in dealloc_crtc
Call drm_crtc_commit directly instead of calling drm_connector_set_mode.
This restores the previous behaviour where conn_enable was called [1].

[1]: 0c7c562482/backend/drm/drm.c (L1093)

Closes: https://github.com/swaywm/wlroots/issues/2253
2020-06-08 10:13:00 -06:00
Simon Ser 8f28f5b2f8 backend/drm: fix black screens when enabling output
This patch fixes this failure:

    01:57:16.642 [ERROR] [backend/drm/drm.c:360] Failed to page-flip output 'eDP-1': a page-flip is already pending
    01:57:16.684 [ERROR] [backend/drm/drm.c:360] Failed to page-flip output 'eDP-1': a page-flip is already pending
    01:57:16.684 [ERROR] [backend/drm/drm.c:732] Failed to initialize renderer on connector 'eDP-1': initial page-flip failed
    01:57:16.684 [ERROR] [backend/drm/drm.c:805] Failed to initialize renderer for plane
    01:57:16.684 [sway/config/output.c:423] Failed to commit output eDP-1

References: https://github.com/swaywm/sway/issues/5101
2020-06-03 10:27:21 -06:00
Simon Ser b03eebf7d4 backend/drm: always perform a CRTC commit in drm_connector_commit
When the mode, status or buffer hasn't changed, drm_connector_commit was
a no-op. Because of this individual changes to the gamma LUT or adaptive
sync status were ignored (if committed without a buffer).

Instead, perform a commit to apply the changes.
2020-06-01 11:45:11 -06:00
Simon Ser 1a2e82e327 backend/drm: drop extra wlr_output_update_enabled call
drm_connector_set_mode already takes care of keeping this up-to-date.
2020-06-01 11:45:11 -06:00
Simon Ser 80467f6ae8 backend/drm: commit/rollback FBs in drm_crtc_commit
We need to perform the FB bookkeeping on all commits, not just on
page-flips.
2020-06-01 11:45:11 -06:00
Simon Ser 70f3a0bb27 backend/drm: make adaptive_sync atomic
Stop using drmModeObjectSetProperty, set the property in the crtc_commit
function instead.
2020-06-01 11:45:11 -06:00
Simon Ser 42126c2022 backend/drm: rollback pending CRTC state on test commit
A test commit doesn't apply the pending state.

The CRTC state will be populated again if the compositor decides to
perform a regular commit afterwards.
2020-06-01 11:45:11 -06:00