color-management: add color-mangagement cap for backends; rework how enabling color management and shadow fb works

makes sure that both framebuffer fetch + shader blending
and shadow framebuffer + ff blending work

Harish Krupo authored May 24, 2019 and Sebastian Wick committed Aug 02, 2019

The upcoming patches introduce 2 pass rendering using a shadow fbo which
is used for linear light blending for HDR. This 2 pass rendering,
unsurprisingly, affects performance especially when the full output is
damaged (fullscreen video playback). To remedy this, a new configuration
option called "use-shadow-fbo" is introduced in the output section of
weston.ini which controls the use of a shadow fbo for an output. The
default value is false for the config option.

Signed-off-by: Harish Krupo <harishkrupo@gmail.com>

Harish Krupo authored Apr 18, 2019 and Sebastian Wick committed Aug 02, 2019

Linear light blending is required when the presented surfaces are
of different colorspaces. The surfaces are linearized, colorspace
convertion is applied (if required) and then blended together. All
this is done to an intermediate texture so that it can be blitted
on to the framebuffer while applying the required non linear curve.

v2: Use /* */ instead of // (Pekka)
    Rename fbo and tex to shadow_{fbo,tex} (Pekka)
    Check for OpenGLES capabilities before creating
    shadow_{tex,fbo} (Pekka)

Signed-off-by: Harish Krupo <harishkrupo@gmail.com>

gl-renderer: only create color space conversions if needed
gl-renderer: fragment debug now shows which parts are color
	corrected by the compositor

Signed-off-by: Sebastian Wick <sebastian@sebastianwick.net>

Since version 7 clients must use MAP_PRIVATE to map the keymap fd so we
can use memfd_create in os_ro_anonymous_file_get_ref, for older version
support_shared is set to allow compatibility with MAP_SHARED.

Signed-off-by: Sebastian Wick <sebastian@sebastianwick.net>

The ro_anonymous_file is a abstraction around anonymous read-only files.
os_create_ro_anonymous_file creates such a file with the contents passed
in data. Subsequent calls to os_ro_anonymous_file_get_ref get a
reference to a fd that's ready to be send over the socket. If
support_shared is not set the socket is only guaranteed to be mmap-able
readonly with MAP_PRIVATE but does not require duplicating the file for
each resource when memfd_create is available.

See also:
weston commit 76829fc4
wayland commit 905c0a341ddf0a885811d19e2b79c65a3f1d210c

Signed-off-by: Sebastian Wick <sebastian@sebastianwick.net>

gather uniform values in struct gl_shader_uniforms, update uniforms and bind all resources in shader_uniforms

Harish Krupo authored Apr 18, 2019 and Sebastian Wick committed Jul 12, 2019

This patch modifies the shader generation code so that the shaders are
stitched together based on the requirement instead of creating them
during initialization. This is necessary for HDR use cases where each
surface would have different properties based on which different
de-gamma or tone mapping or gamma shaders are stitched together.

v2: Use /* */ instead of // (Pekka)
    Move shader strings to gl-shaders.c file (Pekka)
    Remove Makefile.am changes (Pekka)
    Use a struct instead of uint32_t for storing requirements (Pekka)
    Clean up shader list on destroy (Pekka)
    Rename shader_release -> shader_destroy (Pekka)
    Move shader creation/deletion into gl-shaders.c (Pekka)
    Use create_shaders's multi string capbility instead of
    concatenating (Pekka)

v3: Add length check when adding shader string (Pekka)

Signed-off-by: Harish Krupo <harishkrupo@gmail.com>

the GL part is broken