Re: [PATCH RFC v7 07/10] drm/atomic: Loosen FB atomic checks

2023-12-05 Thread Maxime Ripard
Hi,

On Fri, Oct 27, 2023 at 03:32:57PM -0700, Jessica Zhang wrote:
> Loosen the requirements for atomic and legacy commit so that, in cases
> where pixel_source != FB, the commit can still go through.
> 
> This includes adding framebuffer NULL checks in other areas to account for
> FB being NULL when non-FB pixel sources are enabled.
> 
> To disable a plane, the pixel_source must be NONE or the FB must be NULL
> if pixel_source == FB.
> 
> Signed-off-by: Jessica Zhang 

This breaks some plane kunit tests we have:

See 
https://lore.kernel.org/dri-devel/20231204173536.51003-1-naresh.kamb...@linaro.org/

Maxime


signature.asc
Description: PGP signature


Issues with cursors (cursor-shape-v1)

2023-12-05 Thread Campbell Barton
Hi, I've been looking to support mouse cursors under wayland (via
cursor-shape-v1 but also more generally).

Attempting to use cursor-shape-v1, I ran into some mouse cursors not
being available, so I had to fallback to loading the "default" theme,
this might be OK however the cursor size used by cursor-shape-v1 isn't
known, so it can result in the cursor changing sizes within the same
application.

More generally not being able to know the preferred cursor size in
Wayland causes odd glitches, as the cursor size also changes between
applications, when LIBDECOR is used, the application and it's own
window decorations can use different cursor sizes too.

The same issue exists with themes, although I find it more noticeable
when the cursor size changes.

While accessing the preferred theme & size is possible it relies on
reading the settings for desktop environments (DBUS to read the GNOME
setting for e.g.).

It would be good if there was a way to access the cursor size & theme
via generic Wayland API's,
the size particularly so applications that use theme-cursors or define
their own cursor bitmaps can match the size of system cursors.

This could be implemented as part of cursor-shape-v1 although I'm not
sure if this is the appropriate level to expose the API.

-- 
- Campbell


Re: Issues with cursors (cursor-shape-v1)

2023-12-05 Thread Simon Ser
For wlroots-based compositors this is passed down via XCURSOR_THEME and
XCURSOR_SIZE just like on X11 although env vars aren't a good fit for
configuration.

There was an earlier xcursor-configuration protocol with per-seat live
XCursor config which ended up being abandoned. I wouldn't recommend this
approach, it's too complicated and roundabout.

What exactly are the missing cursor shapes? I'm surprised it's not enough.


[ANNOUNCE] libei 1.2.0

2023-12-05 Thread Peter Hutterer
libei 1.2.0 is now available.

libei is a library to send Emulated Input (EI) to a matching Emulated Input
Server (EIS) which can receive those events with libeis, also part of this
project.

libei uses GitLab releases, for tarballs please see:
https://gitlab.freedesktop.org/libinput/libei/-/releases/1.2.0

Notable fixes in this release include better handling of various protocol
errors and less log spam:
- if the write buffer is full, outgoing messages are now cached
  instead of returning EAGAIN.
- if a message send fails, stop trying to send more messages
  which will do nothing but spam the log with more errors
- if we get disconnected by the server, stop trying to send things to the
  server (and spam the log on error)
- use EINTR checks instead of blocking SIGALRM
- keep a cache of defunct objects to avoid spamming the log when there
  are in-flight messages arriving after destroy

No change in the API

Alban Browaeys (1):
  Fix duplicate negation in log message

Peter Hutterer (33):
  ei: improve a debug message, the number we print is the serial
  util: fix clobbered errno in sink_add_source
  util: allow sources to have write notifications
  ei: queue unsent messages for later delivery if our buffer is full
  Add a static assertion that EAGAIN == EWOULDBLOCK
  tools: add a --interval option to the demo server and client
  util: block SIGALRM during posix_fallocate
  CI: add a separate job for the documentation build
  doc/protocol: fix some hugo warnings
  doc: use a specific revision of the hugo template
  CI: add missing dependencies line
  ei: if sending a message fails, don't send any more
  ei: remove the source if we get disconnected
  tools: add an --iterations argument to ei-demo-client
  util: replace most of SIGALRM blocks with EINTR checks
  util: don't check for EINTR on close()
  util: retry posix_fallocate() if it fails
  util: add helper macros for the ELF section handling
  util: add a global setup option to our munit wrapper
  test: set up a timer to trigger SIGALRM
  doc: drop the git depth from the hugo theme
  test: expand log messages to accommodate for 4-digit line numbers
  test: make peck_dispatch_until_stable() print more useful debug info
  test: fix a test missing frame events
  eis: only queue a seat bind event if the caps change
  eis: fix a typo
  test: hack a peck_debug() function
  test: make bug logs fatal by default
  test: add the ability to add offsets to ei_now()
  ei: keep a cache of defunct objects to avoid spamming the log
  ei: improve debug messages for keymap failures
  pre-commit: autoupdate and add two more useful checkers
  libei 1.2.0



signature.asc
Description: PGP signature