This version emits a different internal name so it passes a check
added to egl-gbm at version 1.1.1. This does not alter functionality
in any way.
Signed-off-by: Kurt Kiefer <kekiefer@gmail.com>
The nvidia-drm kernel module and associated load config are needed
by libdrm but were only getting picked up by weston-init.
There are valid configurations (headless, egl-only) that need
libdrm but do not pull in weston, so the runtime recommendation
is moved to the libdrm bbappend.
Signed-off-by: Kurt Kiefer <kekiefer@gmail.com>
* Updates for wayland-related changes
* Drop NV12 blocklinear support in weston-dmabuf-formats due to missing
source file
Signed-off-by: Matt Madison <matt@madison.systems>
The tegra-drm + nvidia-drm setup on r36 is tenuous with respect
to egl-gbm (and particularly calls to eglGetPlatformDisplayEXT with
EGL_PLATFORM_GBM_KHR). It was only working somewhat by chance due
to the order that devices are probed and then made available to
readdir().
This change adds a patch to egl-gbm to provide a fallback in the
case where eglGetPlatformDisplayEXT(EGL_PLATFORM_GBM_KHR) might have
otherwise failed for no good reason, thus eliminating the need for
workarounds to enforce ordering of the modules loading.
Signed-off-by: Kurt Kiefer <kekiefer@gmail.com>
OE-core has dropped the version number from these recipe names
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Matt Madison <matt@madison.systems>
Weston removed support for the wl_shell protocol in version 11.0. Clients
are expected to use the xdg-shell extension instead, so convert
the nvgldemo wayland adapter to do that.
Signed-off-by: Matt Madison <matt@madison.systems>
weston-init is already a machine-specific package, so it's a good place to
mention our tegra-specific runtime dependencies (well, recommendations)
for running Weston.
Signed-off-by: Matt Madison <matt@madison.systems>
* No more 'S = "${WORKDIR}"', which is not permitted
* All ${WORKDIR} references for local files updated to ${UNPACKDIR}
* Custom do_unpack for the recipes that are in the public_sources
archive in the BSP adjusted for the new unpacking approach
* Some SRC_URIs updated with 'destsuffix=' parameters where needed
Signed-off-by: Matt Madison <matt@madison.systems>
cairo has dropped its experimental GL backends, so we can
drop the bbappend we had for configuring it correctly for the
tegra platforms.
Signed-off-by: Matt Madison <matt@madison.systems>
While NVIDIA's Vulkan support really does require OpenGL
and X11 support, modifying the builds of these packages for
the tegra platforms to turn off wayland support causes
other build failures (e.g., gtk4).
Adding bbappends for every other recipe that might try calling
on Vulkan's Wayland APIs isn't practical, so folks will just
have to discover the incompatibility at runtime.
Signed-off-by: Matt Madison <matt@madison.systems>
Support for enabling glvnd support in mesa landed in
OE-Core master a while back, so we can take advantage
of that here.
Signed-off-by: Matt Madison <matt@madison.systems>
This patch was required to fix explicity buffer synchronization,
such as that used by weston-simple-dmabuf-egl. The standard sync
methods now work properly, so we don't need this patch anymore.
Note that in order to run weston-simple-dmabuf-egl, the default drm
render node can't be used. The following will work (where card0 is
the drm display device corresponding with the integrated graphics):
weston-simple-dmabuf-egl -d /dev/dri/card0
Signed-off-by: Kurt Kiefer <kekiefer@gmail.com>
Builds with the recently-introduced make 4.4 are broken, with a race
between building and using the internal libraries (particularly
libnvtexfont2.a). The makefile rules don't look completely race-proof,
which is probably the root cause, although it's possible there's a bug
in make. For now, we'll just disable parallel builds of this recipe.
Signed-off-by: Matt Madison <matt@madison.systems>
A recipe for libglvnd has landed in meta-oe with a version
newer than our copy here. We need to handle builds with and
without meta-oe in the layer set, so update to work like we
do with libv4l2/v4l-utils:
* Rename our recipe to libglvnd-tegra, restricted to tegra
machines only. Add appropriate PROVIDES/RPROVIDES so users
can still reference just 'libglvnd', update it to 1.5.0 to
match the meta-oe recipe, and fix a typo.
* Add a bbappend for libglvnd under external/openembedded-layer.
The bbappend applies the tegra-specific build settings we need
and makes the packages built tegra-specific.
Signed-off-by: Matt Madison <matt@madison.systems>
* Update sha256 checksums
* Switch to using functions from the l4t_* bbclass for forming
package names/versions
Signed-off-by: Matt Madison <matt@madison.systems>
egl-gbm is required to use gbm along with egl on tegra devices.
This was an rdep on weston, but other generic gbm clients (like
kmscube) that want to use egl will need it as well. It's not
strictly required to make gbm itself work, but not being able to
initialize egl without it would be a common and unexpected surprise.
If a tegra platform is pulling in mesa to get libgbm, include this.
Signed-off-by: Kurt Kiefer <kurt.kiefer@arthrex.com>
With updated egl in L4T 35.1.0, we no longer have to workaround
missing support for EGL_NV_stream_consumer_eglimage and broken
eglQueryStreamConsumerEventNV.
Signed-off-by: Kurt Kiefer <kekiefer@gmail.com>
Both the tegra-udrm-gbm and tegra-libraries-gbm-backend work to
provide buffers for libgbm (they are gbm backends).
This change adds a provider so that either implementation can be
chosen, preferring NVIDIA's official binary version.
Signed-off-by: Kurt Kiefer <kekiefer@gmail.com>
