linux-yocto has been carrying this patch for years. It allows the kernel
to link with a non-host ncurses.
This patch is needed for all linux-intel kernels, and that area of code
is rarely changed, so maintenance of the patch should be minimal.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
RMC is confusing as a default because it is only supported by legacy
(iso, hddimg) image types. Its also not being actively maintained,
causing it to lag behind in updates (currently against systemd-boot
v232 instead of v234).
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Including a non-existant .inc doesn't cause errors, so just move the
systemd-boot.inc file into the main bbappend. This avoids rebuilding the
package without our changes should grub-efi be the EFI_PROVIDER and we
build systemd-boot, while keeping the RMC changes when rmc-boot is the
EFI_PROVIDER.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
systemd-boot shouldn't be stuck on a certain commit just because RMC is.
This effectively upgrades our systemd-boot version from v232 to v234. We
also get to drop a patch that was accepted upstream.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Updated version with patches to ensure thermald with musl and x32.
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Fixes the following error seen in 32-bit builds:
| ../../intel-gpu-tools-1.20/lib/igt_x86.c:96:10: error: redefinition of 'igt_x86_features'
| unsigned igt_x86_features(void)
| ^~~~~~~~~~~~~~~~
| In file included from ../../intel-gpu-tools-1.20/lib/igt_x86.c:38:0:
| ../../intel-gpu-tools-1.20/lib/igt_x86.h:47:24: note: previous definition of 'igt_x86_features' was here
| static inline unsigned igt_x86_features(void)
| ^~~~~~~~~~~~~~~~
| ../../intel-gpu-tools-1.20/lib/igt_x86.c:148:7: error: redefinition of 'igt_x86_features_to_string'
| char *igt_x86_features_to_string(unsigned features, char *line)
| ^~~~~~~~~~~~~~~~~~~~~~~~~~
| In file included from ../../intel-gpu-tools-1.20/lib/igt_x86.c:38:0:
| ../../intel-gpu-tools-1.20/lib/igt_x86.h:51:21: note: previous definition of 'igt_x86_features_to_string' was here
| static inline char *igt_x86_features_to_string(unsigned features, char *line)
I unfortunately missed this before merging the upgrade.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
This driver was added for a platform that is no longer supported.
This recipe is also maintained in meta-oe in case it's required outside
of meta-intel.
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Part of Intel graphics stack 2017Q3 release.
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Anuj gets video related pieces.
Stephano gets RMC and poky-tiny related pieces.
Cal picks up other small pieces.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
These new dependencies are needed for some configurations.
elfutils-native is required for CONFIG_UNWINDER_ORC (disabled by default
for now due to out-of-tree module complications).
openssl-native and util-linux-native are required for some certificate
processing options.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
It wasn't appropriate to use the 4.9 cache for 4.14. This also brings in
the following relevant commits:
* b6b5ac8d x86-64: disable orc by default
* 82743ab8 common-pc*.scc: Add igb to common-pc drivers
* 6bcc7298 features/i915/i915.cfg: compile i915 as a module
This adds Intel gigabit Ethernet drivers, fixes the i915 firmware load
issue, and fixes an lttng-modules compilation failure with the new
kernel due to the new CONFIG_UNWINDER_ORC=y default.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
This brings in the changes up to v4.9.76 from linux-stable.
Also update the cache, bringing in the following relevant config
changes:
* 3e7fedea common-pc*.scc: Add igb to common-pc drivers
* 4341ae3a features/i915/i915.cfg: compile i915 as a module
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
This is the newest LTS kernel, and will be the preferred kernel going
for this release.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Perf fails to compile for 32-bit without this patch. I have requested
for Bruce to submit it upstream.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Trusty and USB fixes in source.
Compile i915 as module in cache. This makes sure that the video drivers
and firmware are available at the same time, fixing firmware load
issues.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
As the Quark machine has been EOL'ed at the end of 2017, remove this
machine type from the 2018 planned release of meta-intel
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Since Quark is getting depercated remove reference and WKS file
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
This is an early step to removing the meta-quark machine type, as
the Quark processor has been EOL'ed by Intel, this release will
remove it from the kernel recipes
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
This commit adds the 4.14.0-LTS kernel, but does not change the
PREFERRED_VERSION. This will be updated later, to use this kernel
set PREFERRED_VERSION in your local.conf or distro.conf files
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
This has a SRCREV update to bring linux-intel to HEAD with some
audio and pnp patches.
The refactor moves the selection of branch to the versioned recipe
so we can have a shared linux-intel-lts repo with multiple LTS version
branches.
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Non-Intel* machines did not get COMPATIBLE_MACHINE set at all, allowing
the package to be built when it probably shouldn't have been. For
example, it would be built when the MACHINE was set to qemux86. This
caused lttng-modules to fail because lttng-modules only gets patched
for our kernel when an Intel MACHINE was being used in order to maintain
Yocto Project Compatibility status.
By setting the default it fails when an invalid MACHINE is used. Note
that while qemux86 is not a compatible MACHINE, the kernel itself does
have the necessary components to run under qemu.
Also change the COMPATIBLE_MACHINE overrides to use the x86-intel-common
override instead of machine-specific overrides since they all have the
same value.
[YOCTO #12278].
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Remove the concept of the common directory and move all the recipes-* dirs
to the top level as a normal layer would be. layer.conf is updated appropriately
Signed-off-by: Saul Wold <sgw@linux.intel.com>
do_uefiapp_deploy copies files into the /boot directory of the rootfs
and thus, like do_rootfs, should run under pseudo so that the files
end up being owned by root.
This did not trigger the uid leak check, perhaps because the task runs
later.
For some (still unknown reason), sometimes the installed files did end
up being owned by root, which tripped up a swupd test because the
system update randomly included unexpected changes to bootx64.efi or
rmc.db. With this change, they are owned by root consistently.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
meta-tlk gets moved to a standalone layer since it's can't be
properly made Yocto Project Compatible, move it out of meta-intel
to allow meta-intel to be fully Compatible.
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Adding the special override limits the impact of the append to
just when we are building for meta-intel machines
Signed-off-by: Saul Wold <sgw@linux.intel.com>
This reverts commit 075b81ae1b.
The audio backports in "c27cfdc linux-intel/4.9: Update kernel SRCREVs"
fixed the broken snd_soc_skl module, fixing the issue this was working
around.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
We never hit this because we're usually only building core-image-rt with
the -rt kernel.
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Depmod during do_install is irrelevant when cross-compiling.
Remove the depmod steps during do_install.
Signed-off-by: sweeaun <swee.aun.khor@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Depmod during do_install is irrelevant when cross-compiling.
Remove the depmod steps during do_install.
Signed-off-by: sweeaun <swee.aun.khor@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
oe-selftest leaves python bytecode around for the selftests it
has found making git status dirty.
Add .gitignore that hides those files from git. While we're at it,
ignore some other common patterns too.
Signed-off-by: Mikko Ylinen <mikko.ylinen@linux.intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
