Age | Commit message (Collapse) | Author |
|
We can never hit the case where we try to light up an output without
a buffer. output_ensure_buffer() will catch this for now, and when that's
removed, output_basic_test() will catch this case.
|
|
drm_connect_state_init() will set primary_fd to null if no CRTC is active
for the connector and can crash later if the code expects a CRTC (like
when lighting up an output).
|
|
init_plane tries to free a single plane on error, but this is an element
in a calloc'd array by the caller that we should not touch.
|
|
|
|
On startup, we fetch the previous MODE_ID blob ID so that
compositors can keep using the previous mode if they want to.
However, that blob doesn't belong to us, it belongs to the
previous DRM master. As a result, we get an error when trying to
destroy it.
Fix this by tracking whether the blob belongs to us or not.
Closes: https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3811
|
|
Stop trying to maintain a per-file _POSIX_C_SOURCE. Instead,
require POSIX.1-2008 globally. A lot of core source files depend
on that already.
Some care must be taken on a few select files where we need a bit
more than POSIX. Some files need XSI extensions (_XOPEN_SOURCE) and
some files need BSD extensions (_DEFAULT_SOURCE). In both cases,
these feature test macros imply _POSIX_C_SOURCE. Make sure to not
define both these macros and _POSIX_C_SOURCE explicitly to avoid
POSIX requirement conflicts (e.g. _POSIX_C_SOURCE says POSIX.1-2001
but _XOPEN_SOURCE says POSIX.1-2008).
Additionally, there is one special case in render/vulkan/vulkan.c.
That file needs major()/minor(), and these are system-specific.
On FreeBSD, _POSIX_C_SOURCE hides system-specific symbols so we need
to make sure it's not defined for this file. On Linux, we can
explicitly include <sys/sysmacros.h> and ensure that apart from
symbols defined there the file only uses POSIX toys.
|
|
If all connectors and planes already have the right CRTC set, or
are disabled, we can skip the device-wide reset after a VT switch.
I've contemplated using a more fine-grained logic to only reset
the connectors, CRTCs and planes that need to be migrated. However,
writing a correct algorithm for this would be quite involved, and it
doesn't seem worth the trouble anyways.
Closes: https://github.com/swaywm/sway/issues/7956
|
|
|
|
We'll need to use a bunch of internal functions from that file in
the next commits.
|
|
This function fetches property IDs. These don't change for the
lifetime of the connector. Instead of refreshing the property IDs
on hotplug (and leaving property IDs unset for disconnected
connectors), only fetch the property IDs when we create the
connector.
|
|
|
|
We only need base POSIX here.
|
|
This also fixes a memleak.
|
|
The kernel patches [1] have been merged.
[1]: https://patchwork.freedesktop.org/series/120103/
|
|
We can grab the event loop from the wlr_session instead.
|
|
Leave it up to the compositor to decide what to do when a backend
becomes unavailable.
|
|
wlr_output.refresh is populated by core wlr_output, and thus will
be zero for a custom mode with an unset refresh rate.
Save the refresh rate from the drmModeModeInfo in wlr_drm_connector
instead.
Closes: https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3791
|
|
This function touches queued_fb/current_fb, which the renderer has
nothing to do with.
|
|
|
|
Many files used to require wlr_matrix but no longer do.
|
|
We don't need the whole wl_display here anymore. The wl_event_loop
is enough.
|
|
These aren't really tied to wlr_output.
|
|
Instead of having this condition checked in multiple places,
centralize it so that they don't go out-of-sync.
|
|
This chunk of code was moved by mistake.
Fixes: 3b53d1cbf199 ("backend/drm: introduce page-flip tracking struct")
|
|
When we destroy the backend, page flips will no longer be invoked meaning
those won't clean up the page flips for us.
|
|
|
|
Introduce a per-page-flip tracking struct passed to the kernel
when we request a page-flip event for an atomic commit. The kernel
will pass us back this pointer when delivering the event.
This eliminates any risk of mixing up events together. In particular,
if two events are pending, or if the CRTC of a connector is swapped,
we no longer blow up in the page-flip event handler.
Closes: https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3753
|
|
Right now this is done "by chance" because we disable all CRTCs
on shutdown. However, we'll stop doing this. Plus, if disabling
a CRTC fails, we don't cleanup properly.
|
|
We can just assume CLOCK_MONOTONIC everywhere.
Simplifies the backend API, and fixes clock mismatches when multiple
backends are used together with different clocks.
|
|
The name "allow_artifacts" and associated description is very vague, and
theoretically allow for tearing behavior. Clarify that we only intend to
mean artifacts related to output configuration (e.g., modesets).
References: https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3740
|
|
DRM_MODE_PAGE_FLIP_ASYNC doesn't allow user-space to submit new
buffers before waiting for the uevent: the kernel will return EBUSY
in that case.
Fixes: c2aa7fd965cb ("backend/drm: Add async page flip support to legacy")
|
|
Atomic doesn't support such flags yet.
|
|
|
|
Since e5fc8cd4c734 ("output: trigger frame/present events on all
commits on enabled output"), any commit on an enabled output is
supposed to trigger frame/present events.
The DRM backend was skipping the commit completely for no-op
commits. Stop doing so.
Closes: https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3719
|
|
Up until now, frame/present events were only triggered when the
user submitted a buffer. Change the wlr_output API so that these
events are triggered when any commit is applied on an enabled
output.
Closes: https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3708
|
|
|
|
https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3664#note_2020392
we're gonna have to trust Simon on this one 🤞
|
|
|
|
|
|
|
|
wlr_client_buffer is a bad indicator for direct scan-out.
Compositors might use the underlying wlr_buffer instead.
|
|
This slightly improves type safety.
The culprits were found with:
git grep -E '\([a-z0-9_ ]+ \*\)\W?[a-z]'
|
|
This is a bit more type-safe.
|
|
This fixes an issue where the panel_orientation parameter was set but wlroots was ignoring it and
returning OUTPUT_TRANSFORM_NORMAL.
Fixes 2e12de96
|
|
Removes duplication across all the backends to finally apply the mode
to the output.
|
|
Nobody remembers why this is done and it isn't that great if people
use a lot of custom modes.
|
|
wlr_output_update_needs_frame() is called unconditionally at the
end of the function already.
|
|
This changes the semantics of wlr_output_state. Instead of having
fields with uninitialized memory when missing from the committed
bitflag, all fields are always initialized (and maybe NULL/empty),
just like we do in wlr_surface_state. This reduces the chances of
footguns when reading a field, and removes the need to check for
the committed bitfield everywhere.
A new wlr_output_state_init() function takes care of initializing
the Pixman region.
|
|
Destroying the liftoff_device invalidates all of the liftoff_planes.
Destroy the liftoff_planes before the liftoff_device to fix this.
|
|
|