Add a simple converter to test sampling from fp16 textures.

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>

Picking 0.47 allows us to use `feature` options, simplifying some of the code.

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>

Add meson support

See merge request !5

cube-tex: close the fd after EGLImage creation.

See merge request mesa/kmscube!2

planes[i].fd are opened in a `nplanes` loop, so they should be closed in
a same-length loop. Also, the name is a hint :P

common: free temporary log buffer

See merge request !3

add a trivial build-test ci

See merge request !4

EGLImage takes the fd ownership.

Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

Gallium drivers that do not implement resource_create_with_modifiers()
will fail gbm_surface_create_with_modifiers() so fall back to the old
way.

Fixes: 4f7cec00 Use weak functions to handle lack of gbm modifiers
Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

Fixes rendering issues with mode rgba on etnaviv. I have applied
the same change for nv12 variants but they are not supported on
etnaviv.

v1 -> v2:
 - check if EGL_EXT_image_dma_buf_import_modifiers is supported
 - use weak function trick for gbm_bo_get_modifier(..)

v2 -> v3:
 - check if modifier != DRM_FORMAT_MOD_INVALID

Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>

Emil Velikov authored Mar 30, 2018 and Christian Gmeiner committed Apr 02, 2018

Add weak function declaration and check if they're valid prior
to calling the functions.

This allows us to remove conditional compilation, yet allowing the
modifiers codepath to work if API is available.

Cc: Christian Gmeiner <christian.gmeiner@gmail.com>
Cc: Rob Clark <robdclark@gmail.com>
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Rob Clark <robdclark@gmail.com>

Emil Velikov authored Mar 30, 2018 and Christian Gmeiner committed Apr 02, 2018

Currently the default modifier is MOD_INVALID. At a later stage, if the
modifier is the same (the user may have explicit selected MOD_INVALID),
we'll fallback to MOD_LINEAR.

Simplify things to a) use MOD_LINEAR by default b) use the user modifier
- even if they request MOD_INVALID.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Rob Clark <robdclark@gmail.com>

Ilia Mirkin authored Feb 09, 2018 and Daniel Stone committed Feb 16, 2018

The GBM surface format has to match the DRM mode. Both are used in a
couple of places, so unify it so that it's only set in one place. Note
that the GBM and DRM formats are identical.

Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>

The select fds have to be reinitialized before each call to select.

Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Rob Clark <robdclark@gmail.com>

With atomic kms backend, userspace can get further ahead of the gpu.  In
the decoder, we unref the previous frame when retrieving the current
frame.  If userspace gets too far ahead, this can happen while the gpu
is still sampling from the previous frame.

Simple solution is add a fence.

Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>

Remove some boilerplate and check more specifically for all the entry-
points that different modules need.

Ok, I guess it isn't likely for a driver to expose eglCreateImageKHR
but not eglDeleteImageKHR.. but otoh kmscube is useful for bringing
up drivers.

Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>

Signed-off-by: Ben Widawsky <ben@bwidawsk.net>

Seems work better than
commit 88858d32
Author: Lucas Stach <l.stach@pengutronix.de>
Date:   Fri Apr 28 18:18:52 2017 +0200

    common: use %llx to print modifier

Signed-off-by: Ben Widawsky <ben@bwidawsk.net>

Lucas Stach authored May 02, 2017 and Daniel Stone committed May 02, 2017

With the modifier support it is possible to force a modifier which isn't
compatible with any of the GPU targets. In that case the EGL platform will
be unable to allocate the back buffers.

Provide an error message in that case, instead of crashing.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Daniel Stone <daniels@collabora.com>

The initial pattern of:
    initial_modeset();
    while (1) {
        pageflip();
    }

was relying on getting a buffer despite not having rendered anything. It
would also show undefined content. Replace this with just using the
pageflip loop for the initial modeset as well.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>

Atomic does not let us have multiple requests in flight at once; if we
don't synchronise with the request completion on the CPU side, we can
get -EBUSY from the atomic commit.

We already have everything required to do this, namely the out-fence
from the last commit. Block on that with a CPU-side wait before we call
atomic commit, to make sure we'll never attempt to queue an atomic
commit before the previous one has completed.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

eglGetProcAddress is allowed to return any old garbage for symbols it
doesn't know about. To avoid any mishaps, check for the appropriate
extension presence (split into EGL client extension, EGL display
extension, and GL extension, checks) before we look up any symbols
through it.

The walk through the extension list is taken from libepoxy.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

Sinclair Yeh authored Apr 28, 2017 and Emil Velikov committed Apr 29, 2017

The proper flag is to check is "enable_gbm_modifier"

Signed-off-by: Sinclair Yeh <syeh@vmware.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>

Carlos Rafael Giani authored Apr 26, 2017 and Rob Clark committed Apr 29, 2017

* Make EGL image attribute specification code more generic, and not
  specific to certain pixel formats, implicitely gaining support for YUY2
* Better handling of gstbuffers with multiple memory blocks
* Print out more information about the stream
* Use the GST_VIDEO_INFO_* macros instead of directly accessing the
  GstVideoInfo fields; this is what the GStreamer documentation recommends

Signed-off-by: Carlos Rafael Giani <dv@pseudoterminal.org>