Since the PACKAGECONFIG parameters were changed in the libva.inc file
for the newer version of the libva recipe, this older version of libva
recipe started throwing following QA warning.
WARNING: QA Issue: libva: configure was passed unrecognised options:
--enable-wayland --enable-x11
Fix the warning by nullifying PACKAGECONFIG variables coming from
libva.inc.
Fixes Bug:
[YOCTO #5764]
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
* Disable the dumper since it needs swig and python3;
* refresh the install-fitter.patch;
Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
With ${B} != ${S} the build fails due to files being accessed incorrectly.
Disable the functionlaity for now until the software can be fixed.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When building GL apps for non-x86 machines (e.g. raspberrypi)
emgd-driver-bin is being dragged in as a valid provider. To avoid
build breakage fix it at the source by limiting emgd-driver-bin to x86
architectures.
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Acked-By: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
Support the intel-common package arch and BSPs on linux-yocto-3.10 as
well.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
(cherry picked from commit e75ea81899beb885d892e5f5bf995c445981da3f)
The intel-core* BSPs now have their own BSP description in linux-yocto.
Update the recipe accordingly.
All the existing BSPs (with the exception of the emgd variants) are now
supported across these two BSPs.
For machines that derive from the intel-common-pkgarch, allow the
common linux-yocto-dev recipe to build for them by specifying $MACHINE
for the common machine overrides (core2-32-intel-common and
corei7-64-intel-common).
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Ensure package strip logic stays the
same after oe-core package.bbclass changes.
See http://patches.openembedded.org/patch/65311/
Signed-off-by: Koen Kooi <koen.kooi@linaro.org>
Tested-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Acked-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
Add a linux-yocto-dev bbappend to the common area for use by the two
upcoming new core2-32 and corei7-64 common BSPs.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
The top level LICENSE file in OE-Core changed so we need to update the checksum
references that refer to this to match the new file.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
GLES2/gl2.h is missing the GLchar typedef, so at least gstreamer1.0-plugins-bad
fails to compile. Khronos say the typedef should be there so patch it in.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
gstreamer-vaapi 0.5.x supports GStreamer 0.10/1.0/1.2, so add seperate recipes
for 0.10 and 1.0 builds.
This adds a new recipe name so keep the old gstreamer-vaapi_0.4.3 for now until
all users of that have migrated away.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
Remove the .bb/.inc split as we're only supporting one version.
Add PACKAGECONFIG for X11 and Wayland features. Add patch to fix compile error
in builds with Wayland but without X11.
Drop the workaround patch, isn't relevant for released hardware.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
Use tarballs instead of git tags (upstream git means using submodules, which are
tricky).
Also use PACKAGECONFIG to respect x11/wayland DISTRO_FEATURES.
Delete glib-includes.patch, as it wasn't being used.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
Use PACKAGECONFIG to respect x11 and wayland DISTRO_FEATURES.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
There isn't any need to PROVIDE these as the ony user is in emgd-driver-bin's
RDEPENDS and bitbake looks at PACKAGES when resolving that.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
The dummy driver is only useful for advanted testing with additional tooling, so
don't build it.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
It's not recommended to split -dev and -dbg packages up this finely, so merge
them back together into single -dev and -dbg packages.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
EMGD's user space driver contains MIT code. The current Intel
user space license contains MIT text and other licenses that
are not within the actual binary. We should strip out the
MIT text (and others) from the Intel user space license (next commit) and
correct the LICENSE field to reflect MIT license.
Signed-off-by: Elizabeth Flanagan <elizabeth.flanagan@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
Remove openssl-qat-module and zlib-qat-module recipes and any
references to them from meta-intel.
This is to remove recipes for obsolete and unmaintained software
ingredients related to QAT.
Signed-off-by Ong Boon Leong <boon.leong.ong@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
These components can come from either mesa recipe of some other
custom gl providers such as emgd recipe.
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Both emgd and mesa recipes provides opengl components. With this
change recipies which depend on opengl components, can specify
what they want, and BSP config selects the right provider recipe.
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
One of the drivers links to libva so we need to list it as an RDEPENDS
so that when we package, we can ensure any library renaming gets applied
correctly. The bug can be seen if you "bitbake emgd-driver-bin", then
build an image using that library.
Also exclude these recipes from world since they should get built when
needed through the virtual/ mechanism, building them for say a qemux86
world build just corrupts the system due the the multiple shlibs providers.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
A newer version of this recipe is in oecore. And this version
is not used anymore.
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
emgd recipe needs 'opengl' distro feature enabled. Otherwise target image
build fails without much clue. This commit is marking the 'opengl' as a
required distro feature, which helps user understand the distro
configuration issue in case the requirement of distro features is not met.
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Acked-By: Ross Burton <ross.burton@intel.com>
Update to the latest gummiboot and remove all patches from the recipe
which are now upstream in one for or anoter. Make use of the new
configuration methods for inc/lib/lds directories. Change the PV to use
the git SRCREV beyond the release tag.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
oecore is updating mesa to version 9.1.6. Sync-up the bbappend version
in the meta-intel layer accordingly.
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
This issues was already fixed for the emgd-driver-bin-1.16 recipe as
seen in the commit message below. Now reapplying the same fix to the
emgd-driver-bin-1.18 recipe.
commit edf3ef3cca
Author: Nitin A Kamble <nitin.a.kamble@intel.com>
Date: Mon Jun 10 14:20:20 2013 -0700
emgd-driver-bin: add xserver-xorg to RDEPENDS
When any package using libraries provided by the emgd-driver-bin recipe,
is included in a image (such as: tools-testapps), the rootfs generation
process encounters an error mentioned below.
The code for checking the abi version of the xorg xserver needs xorg
xserver recipe included in the image. Otherwise it gets following rootfs
generation error:
| Computing transaction...error: Can't install
emgd-driver-1.16-r0@core2: no package provides xorg-abi-video-8
Fix the issue by marking xserver-xorg as a dependency of the
emgd-driver-bin recipe.
Fixes this bug:
[YOCTO #4615]
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
This patch was updated but mistakenly deleted by the previous commit:
commit 8834d83768
Author: Darren Hart <dvhart@linux.intel.com>
Date: Fri Jul 26 10:36:06 2013 -0700
gnu-efi: Update to 3.0u
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
gummiboot is a simple UEFI boot manager. It uses configuration files to
setup the default boot entry or presents an on-screen menu.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Update to 3.0u. Remove the now upstreamed cross-compile-support patch
and update the parallel-make-archives patch for the 3.0u Makefiles.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Creating version 1.18 of the emgd recipe for the recently released EMGD
driver.
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
Needed to be compliant with 2013Q2 Intel Graphics Stack Release.
[YOCTO #4828]
Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
Needed to be compliant with 2013Q2 Intel Graphics Stack Release.
[YOCTO #4828]
Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
From similar patch for lms8 (fc9cba84):
The configure script looks at the *host* environment to decide where
to install the init script, so it's location at packaging time can
change.
Signed-off-by: Thibault Guittet <guittet.thibault[at]gmail.com>
Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
These recipes are x86 specific packages so lets make
sure they dont get built for non-x86 architectures
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
The configure script looks at the *host* environment to decide where to install
the init script, so it's location at packaging time can change.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
This package contains numerous useful tools for working with Intel GPUs. Add a
patch to install intel_panel_fitter as it's genuinely useful and not just a
debugging tool.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
The changes this bbappend makes have been merged into OE-Core so its now
unneeded.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
With the switch to the PR server we don't need to use PRINC. This
patch matches a change to OE-Core to bump the PR values of the appropriate
core recipes so PR values don't go backwards.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Some of the newer boards have AMT/MEI firmware of version 8. These BSPs
need verion 8 of the lms package, which this recipe is providing.
Also the lms binary & script is renamed to lms8 to avoid runtime clash
with lms7 packages.
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
Preparations for upcoming lms8 recipe. Rename the lms recipe as lms7.
This will allow BSPs to choose either lms7 or lms8 package to suite
the version 7 or 8 firmware available on the board.
Also the lms binary & script is renamed to lms7 to avoid runtime clash
with lms8 packages.
Simplified the dounpack2 function.
And this build time warning is muted:
WARNING: lms7: No generic license file exists for: Modified in any
provider
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
lms7 : tail conexist
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Tom Zanussi <tom.zanussi@intel.com>
