license.bbclass: Added get_boot_dependencies function

This function gets the dependencies from the classes that
create a boot image, this is required because sometimes
the bootloader dependecy is in these classes. The current
classes covered are bootimg and bootdirectdisk because
these are the only clases that add dependencies.

[YOCTO #6772]

(From OE-Core rev: b165c2d3349115546c9db2f084eb6def4913b82d)

Signed-off-by: Mariano Lopez <mariano.lopez@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

license.bbclass: Split license create manifest

This changes moves the writing of the licenses to a
separated function that could be called for other packages.

With these change it will be easier to reuse the writing of
the license for the packages deployed but not installed in
the rootfs.

[YOCTO #6772]

(From OE-Core rev: 71fb5da7a3a8e61b15be4efc05b968c0c9920d2f)

Signed-off-by: Mariano Lopez <mariano.lopez@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

license.bbclass: Write recipeinfo file in license folder

Currently there is no way to get the recipe version when
creating the rootfs. It is needed because the manifest
file for the image has to contain this important piece
of information.

This change writes a new file in the license folder for
every recipe. This file is called "recipeinfo" and have
the information used to write the manifest file for the
recipes deployed next to the image.

[YOCTO #6772]

(From OE-Core rev: 2d97d82fb51281fe130e8d135f3cf9a78423dd94)

Signed-off-by: Mariano Lopez <mariano.lopez@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

populate_sdk_ext.bbclass: Be more permissive on the name of the buildtools

We want to support different names for the buildtools tarball.  The
name may not always be of the default oe-core format.

For instance, at Wind River we define the built-tools name to be:

${SDK_ARCH}-buildtools-nativesdk-standalone-${DISTRO_VERSION}

because thes standard SDK_NAME has additional information that is not
relevant to the builtools tarball.

(From OE-Core rev: b49c6f179b06a8b97106aa4c95f2cdb3c4dc0920)

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

populate_sdk_base: Add sysroot symlink check

Add optional check to do_populate_sdk() that verifies SDK sysroots don't
contain dangling or escaping symlinks before attempting to tar an archive.
Such links may fail a `tar -h` operation (-h => follow symlinks) or
archive the build system's files.

Set CHECK_SDK_SYSROOTS = "1" to enable this check.

Use case: The -h option may be set via SDKTAROPTS in some configurations
to create symlink-less SDK archives for Windows file systems.

(From OE-Core rev: 2658200fa2b3df08880ee937a3de5cb2866f8a50)

Signed-off-by: Haris Okanovic <haris.okanovic@ni.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

classes/populate_sdk_ext: fail if SDK_ARCH != BUILD_ARCH

The extensible SDK relies upon uninative, and with the way that
uninative works, the build system architecture must be the same as the
SDK architecture or the extensible SDK won't be usable. At some point in
future hopefully we can remove this limitation, but until then it's
disingenuous to allow this to build, so add a check to ensure
SDK_ARCH == BUILD_ARCH and fail if it isn't.

(From OE-Core rev: 9e30e849eda3b0a0c54d3f7ed0102760fdaef06c)

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

classes/populate_sdk_ext: tweak reporting of workspace exclusion

If you have a local workspace layer enabled when building the
extensible SDK, we explicitly exclude that from the SDK (mostly because
the SDK has its own for the user to use). Adjust the message we print
notifying the user of this so it's clear that we're excluding it from
the SDK, and scale it back from a warning to a note printed with
bb.plain().

(From OE-Core rev: 90f46f74a088a7b965d2205eceb9eff6f276dd38)

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

classes/populate_sdk_ext: make it clear when SDK installation has failed

When SDK preparation fails:

* Insert an ERROR: in front of the error message
* Add an error message to the environment setup script

Hopefully this should make it more obvious when this happens.

Fixes [YOCTO #8658].

(From OE-Core rev: 105df569b3b1982005c2edb37f4690f9ba6bde35)

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

classes/populate_sdk_ext: tidy up preparation log file writing

Use a variable for the log file which includes the full path; this is
not only neater but avoids us writing the first part (the output of
oe-init-build-env) to a file in another directory since we are
changing directory as part of this subshell.

(From OE-Core rev: 001af71752a9e9aab460cbd49ed049e1eb726295)

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

boot-directdisk.bbclass: remove HDDIMG before create

Fixed when rebuild:
mkdosfs: file /path/to/hdd.image already exists

(From OE-Core rev: 9abcd309c098558360cde2bff65be840ead25f83)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sstate: Ensure siginfo and sig files are also touched

Maintaining an sstate mirror is complicated at the moment as only the main
sstate tgz files have their timestamp updated when they're used.

This causes problems as the siginfo files can be removed from the server
due to inactivity but the sstate fetch code tries to fetch them leading
to confusing warnings and reduced debug capability.

This change ensures we touch any sig/siginfo files present and should
help ageing of the objects on sstate mirrors. It doesn't handle the intermediate
task siginfo files but those are never fetched by the current sstate code
so are an order of magnitude less problematic.

(From OE-Core rev: 80b3974081c4a8c604e23982a6db8fb32c616058)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

weston: Add PACKAGECONFIG option for colord CMS

(From OE-Core rev: e98d01c1ca060ad984d7a14a3f56fe049812df44)

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

opkg: add cache filename length fixes

(From OE-Core rev: 8e53500a7c05204fc63759f456639545a022e82b)

Signed-off-by: Alejandro del Castillo <alejandro.delcastillo@ni.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

openjade-native: statically link local libs

Statically link local libs to avoid gold link issue. This is clearly
a workaround, but does get us past the failures with systems using gold by
default until we find a better solution.

[YOCTO #2972]

(From OE-Core rev: 415287be0cce596ea0d33ded0f3e6ffa9f26b775)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sysklogd: inhibit updatercd for non-sysvinit

This recipe doesn't inherit systemd, so we need to take care of it ourselves.

(From OE-Core rev: 4b382b79a0acce0e2704d841288cef7dad660690)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

connman: depend on readline

When the client is enabled, which we do by default, readline is required.
Currently it gets pulled in by other dependencies of connman, indirectly,
but we should be explicit rather than relying on this.

(From OE-Core rev: 8d10de63e09754ed600ad3130b685bb1624eba5e)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

latencytop: obey LDFLAGS

Patch update courtesy Abdur Rehman <abdur_rehman@mentor.com>.

(From OE-Core rev: 448f211ba8a2a05380276de648176422c0b6eb37)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

tcf-agent: obey LDFLAGS

This silences a GNU_HASH warning when using external toolchains. The patch is
courtesy Abdur Rehman <abdur_rehman@mentor.com>.

(From OE-Core rev: 909c77ad13b139478bd1a9eee288604a75d59ee5)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

blkspace: fix ldflags for iowatcher

This quiets a GNU_HASH warning.

(From OE-Core rev: ad8a2ca46075a7ea236d6db9da963dc0040d2206)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bluez5: enable sysvinit support

This is from Shrikant Bobade <Shrikant_Bobade@mentor.com>.

(From OE-Core rev: f4c16f53af27ad459e77ae7f43e7bf1bad70645a)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sysprof: use packageconfig for the gui

This makes the gtk dependencies optional.

(From OE-Core rev: 191a1817c9dce90f3094bc98d6e0f5ab02f8a5c7)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

mc: upgrade to 4.8.15

4.8.14 -> 4.8.15

(From OE-Core rev: 35c5ab8f34400000129078e1a5246232b20fd265)

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

packagegroup-core-directfb: Don't depend on pango-modules

Recent pango upgrade means there is no pango-modules.

(From OE-Core rev: 2aea260b99d38bf506c6a0ebcf0f299b2c675da7)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xkeyboard-config: Upgrade 2.15 -> 2.16

(From OE-Core rev: 8983e571a6e2491a6253995414f35e05661cb9f0)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xkbcomp: Upgrade 1.3.0 -> 1.3.1

(From OE-Core rev: 4605abb8d56f902e3dd278d890d4293caa95fa40)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xinput: Upgrade 1.6.1 -> 1.6.2

(From OE-Core rev: 649963e0d4c3061b5794823809ecaa363fade6e9)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xf86-video-omap: Upgrade 0.4.3 -> 0.4.4

Remove backported patch.

(From OE-Core rev: 4ca6274aff4c7c489f148f811267634a760c5b77)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xf86-input-synaptics: Upgrade 1.8.2 -> 1.8.3

(From OE-Core rev: 15b4e5d2ab0ac117979cb367f07f444b8e123b64)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xf86-input-evdev: Upgrade 2.9.2 -> 2.10.0

This version installs the 10-evdev.conf configuration file
that is no longer included in xserver-xorg.

(From OE-Core rev: 52114d07cedd49f7db08520ce67d2635fff8f688)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xorg-driver-input: add xorg configuration to FILES

It's common for input drivers to to add files into
${datadir}/X11/xorg.conf.d/.

(From OE-Core rev: 757fb421ec498b7c268944ba5d2942d2b286bed8)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xserver-xorg: Upgrade 1.17.2 -> 1.18.0

* Remove glibc-guard patch: xserver no longer uses termio.h on linux
  in any case.
* Remove "--enable-glx-tls": this hasn't done anything since glapi
  was separated from glx
* Note that xserver no longer installs evdev configuration
  (evdev and/or libinput should do it instead)
* Add patch to check for libsystemd correctly
* Add PACKAGECONFIG for libsystemd

(From OE-Core rev: 9057678e75ec09af081b27d5935907ec955c3664)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libxcb: Remove unused git-version of the recipe

(From OE-Core rev: 39c6dae9286f649ff02acb56d0214d9d2a3cadf7)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libxcb: Upgrade 1.11 -> 1.11.1

(From OE-Core rev: 3a3f65ffafeba5cb73877bcb2be13c02ae6ef034)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pixman: Upgrade 0.32.6 -> 0.32.8

Update Upstream-Status on a patch.

(From OE-Core rev: 43f67da6bfd5870424ebb32e5d25c6c193b6d41f)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libxi: Upgrade 1.7.4 -> 1.7.5

(From OE-Core rev: 473a73e77b53b059ac99cc938831faa1c7cd72c0)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gtk-icon-utils-native: Upgrade 3.16.6 -> 3.18.2

(From OE-Core rev: f989fc4699a4e616b1f56515466d1f62379f0eb5)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package_regex.inc: Add gtk-icon-utils-native

(From OE-Core rev: dad769b0d62f44e1dcd5cca305b7ce1dfd879390)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gtk+3: Upgrade 3.16.6 -> 3.18.2

* Package new development tool gtk-builder-tool.
* Disable colord support (for explicitness, no functional change)

(From OE-Core rev: 59498a8b4fa69a359e72225d14c554aae133571a)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

adwaita-icon-theme: Upgrade 3.16.2.1 -> 3.18.0

Remove a backported patch.

(From OE-Core rev: 136c62b5dcb7d01278b18217d136d55d4ecee040)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

librsvg: Upgrade 2.40.10 -> 2.40.11

(From OE-Core rev: a74b956b74fefb060bb8dd4bd503b7d53b9b4c4e)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pango: add RPROVIDES for removed packages

Now that Pango has merged the external modules and engines the packages don't
exist.  For convenience have the Pango library RPROVIDE these names.

(From OE-Core rev: 9e4c511ed155e1af50b69b65cf43a6320750ca8c)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pango: Upgrade 1.36.8 -> 1.38.1

Pango is now a simple a simple library and no longer has
 - modules
 - engines
 - configuration
Dropping all related things from the recipe.

(From OE-Core rev: 9a86d2a876f5cce9c2825f8fd5bf5dc9aaf9d7a1)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gdk-pixbuf: Upgrade 2.30.8 -> 2.32.1

Remove a backported patch.

(From OE-Core rev: d19d221ca77ff5cfbe92cf8234b2f4f35dd07329)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libsoup-2.4: Upgrade 2.50.0 -> 2.52.1

Don't build the now-included vala bindings: This requires gobject-introspection.

(From OE-Core rev: 25991b97ce027d1ff9b803f4ab57350659465fd8)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

at-spi2-atk: Upgrade 2.16.0 -> 2.18.1

(From OE-Core rev: c6a14343bf58329c69a2c4c9cb780647ae5e9ded)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

atk-spi2-core: Upgrade 2.16.0 -> 2.18.1

(From OE-Core rev: ba8016f0e01178030aa9075fc3a48245f630d49a)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

atk: Upgrade 2.16.0 -> 2.18.0

(From OE-Core rev: bf804ba484c08ed3531e639777a9be3b7858b1e3)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

glib-networking: Upgrade 2.44.0 -> 2.46.1

(From OE-Core rev: 07fd82d5909b2227d6b163356468cf8a73df8358)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

glib-2.0: build dependency cleanup

New GLib doesn't require glib-2.0-native to build itself so remove it from the
DEPENDS.  However it is needed for the test suite so respect PTEST_ENABLED and
add it to DEPENDS if required.

(From OE-Core rev: 216d534258f4fd7aa3c46093aba6fa2967ae5283)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

glib-2.0: Enable more tests while cross-compiling

(From OE-Core rev: e6e9260b97d873f14991a1143f1bda5b1b624955)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

glib-2.0: Upgrade 2.44.1 -> 2.46.1

Fix cross-compilation build issue in gio tests.

(From OE-Core rev: 17bbd1957720d6faea04530211a76655541fe71e)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

qemu: Backport malloc-trace disabling

Allocation tracing has been removed from GLib 2.46 and trying to use
it results in an ugly warning: Backport patch to not use it in Qemu.

(From OE-Core rev: 5e6105a90acb86bf7e2c0d5e7fe51e6112080916)

Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

logrotate: do not move binary logrotate to /usr/bin

In oe-core commit a46d3646a3e1781be4423b508ea63996b3cfca8a
...
Author: Fahad Usman <fahad_usman@mentor.com>
Date:   Tue Aug 26 13:16:48 2014 +0500

    logrotate: obey our flags

    Needed to quiet GNU_HASH warnings, and some minor fixes.
...
it explicitly move logrotate to /usr/bin without any reason,
which is against the original Linux location /usr/sbin.

So partly revert the above commit which let logrotate be
kept in the original place /usr/sbin.

(From OE-Core rev: 0007436b486fd0bea9e6ef60bf57603e7cfce54b)

Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemd: drop unneeded $D check in prerm

(From OE-Core rev: 2bfc772425893ed5674eef76174be6e4b0b5cfd7)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemd: chown hwdb.bin to root:root for do_rootfs

This is created by qemu for the do_rootfs case, which bypasses pseudo, so we
need to correct the ownership. This fixes a warning issued by
rootfs_check_host_user_contaminated.

(From OE-Core rev: 4ff6b8cadec10e17dbf884a873a227e29944f5d1)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemd: for valgrind, define VALGRIND=1

Per the systemd README, this should be defined to run systemd under valgrind,
otherwise false positives will be triggered.

(From OE-Core rev: c643fcf5152c50de130b16b567e2a9bad99546a1)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemd: make coredump a PACKAGECONFIG

(From OE-Core rev: 70ddba0a5798362923c4c437541ab9992b5020fb)

Signed-off-by: Daniel McGregor <daniel.mcgregor@vecima.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemd: add machine-id to conffiles

If / is mounted read-write machine-id will be overwritten on first boot.
This change ensures that the machine-id file persists across package
upgrades to systemd.

(From OE-Core rev: a25335967ac4aa62d77e862b4b80fed293eba0ad)

Signed-off-by: Dan McGregor <dan.mcgregor@usask.ca>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemd: ignore .so filenames in systemd-doc

systemd names two manual pages for .so files ${foo}.so.2.8,
the library being named ${foo}.so.2. This hits the libdir
sanity checker:
WARNING: QA Issue: systemd-doc: found library in wrong location:
/usr/share/man/man8/libnss_mymachines.so.2.8 [libdir]

Disable the libdir sanity check for systemd-doc.

(From OE-Core rev: 2c10bc96257a934d9431bef0c39006c7e100924d)

Signed-off-by: Dan McGregor <dan.mcgregor@usask.ca>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemd: fix Upstream-Status tag

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

mdadm: fix CFLAGS and ptest issues

* Pass global CFLAGS to build:

The CFLAGS does not pass to build at all since it was redefined by
mdadm Makefile:

  CFLAGS = $(CWFLAGS) $(CXFLAGS) ...

This could be done by setting 'CXFLAGS="${CFLAGS}"'.

* Also fix ptest build errors caused by global CFLAGS:

  raid6check.c:352:2: error: ignoring return value of posix_memalign, \
  declared with attribute warn_unused_result [-Werror=unused-result]

  raid6check.c:315:8: error: 'stripe_buf' may be used uninitialized \
  in this function [-Werror=maybe-uninitialized]

(From OE-Core rev: 60f71fa4da86ca4c7c37115c343db194a3b7b47b)

Signed-off-by: Wenzong Fan <wenzong.fan@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gcc-4.9: Fix various _FOR_BUILD and related variables

This patch is based on the patch for gcc-5.2 (41cbfd7af60f93a4bd496b7b6bf477215a286950)

When doing a FOR_BUILD thing, you have to override CFLAGS with
CFLAGS_FOR_BUILD. And if you use C++, you also have to override
CXXFLAGS with CXXFLAGS_FOR_BUILD.
Without this, when building for mingw, you end up trying to use
the mingw headers for a host build.

The same goes for other variables as well, such as CPPFLAGS,
CPP, and GMPINC.

(From OE-Core rev: e6dc4b2ac0cc6417c0e0ddcdcbe3f334581af8fc)

Signed-off-by: Juro Bystricky <juro.bystricky@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

devtool: add sync command

The sync command is similar to the extract command, except it
fetches the sync'ed and patched branch to an existing git repository.

This enables users to keep track the upstream development while
maintaining their own local git repository at the same time.

(From OE-Core rev: e2fe4c99f1020a20b579832c4feafcd0e9bc2f75)

Signed-off-by: Tzu-Jung Lee <roylee17@currantlabs.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

boost.inc: remove unused parameter from get_boost_parallel_make()

The bb parameter seems to have been accidentally left behind after:

  http://git.openembedded.org/openembedded-core/commit/?id=1ff36aaec25a7ee89514366fe484345e8d1d7b64

(From OE-Core rev: 871d3279fd3360628c8fd9a37134eca541b10636)

Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

wireless-tools: remove unused files

Remove wireless-tools.if-pre-up and zzz-wireless.if-pre-up.
As per the comment, these files were not in use from 20/8/08.

(From OE-Core rev: 85ed9f581184d4028d407cd69f1ae6c58836aa77)

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gstreamer1.0: fix install race

The install hook needs to be a data hook not an exec hook (because automake),
the explicit dependency upstream added to fix the ordering results in
install-helpersPROGRAMS executing twice and potentially racing.

(From OE-Core rev: 0fde7c96bd81f16a70ad7193a91c3c2021c4aca4)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gcc-multilib-config: make aarch64 support multilib

Fixed:
MACHINE = qemuarm64
require conf/multilib.conf
MULTILIBS = "multilib:lib32"
DEFAULTTUNE_virtclass-multilib-lib32 = "armv7at-neon"

$ bitbake core-image-minimal -cpopulate_sdk

WARNING: gcc multilib setup is not supported for TARGET_ARCH=aarch64
WARNING: gcc multilib setup is not supported for TARGET_ARCH=aarch64

[YOCTO #8638]

(From OE-Core rev: 9e0cad83eefae4c1a5e5e0334cc1cfdfc1f51057)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libxml2: fix CVE-2015-7942 and CVE-2015-8035

CVE-2015-7942 libxml2: heap-based buffer overflow in xmlParseConditionalSections()
CVE-2015-8035 libxml2: DoS when parsing specially crafted XML document if XZ support is enabled

[YOCTO #8641]

(From OE-Core rev: 27de51f4ad21d9b896e7d48041e7cdf20c564a38)

Signed-off-by: Armin Kuster <akuster@mvista.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

terminal: Open a new window instead of split on older tmux versions (<1.9)

If an old version is detected (<1.9), create a new window instead of split:
the reason is that there is no easy way to get the active pane height if no
nested formats are supported.

(From OE-Core rev: 457bd6297ae99627c5f596c3c09086d787d5a5ab)

Signed-off-by: Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

flex: fix test-bison-yylval and test-bison-yylloc failed

The tests in flex test-bison-yylval and test-bison-yylloc
failed, so we backport a patch from upstream to fix it.

(From OE-Core rev: 04810bf36720240cf0e1b8ba2cb1bba16b2ccac8)

Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gdbm 1.8.3: install libgdbm_compat

It has been built but not installed by default, while 1.11 builds and
installs it.

Fixed when build non_gplv3 + perl:
powerpc64-wrs-linux-gcc -m64 -mhard-float -mcpu=e6500
--sysroot=/mnt/docker/test_p/bitbake_build/tmp/sysroots/fsl-t2xxx
-Wl,-O1 -Wl,--hash-style=gnu -Wl,--as-needed -
fstack-protector -shared NDBM_File.o  -o
../../lib/auto/NDBM_File/NDBM_File.so   \
   -lgdbm -lgdbm_compat         \

ld: cannot find -lgdbm_compat
collect2: error: ld returned 1 exit status
Makefile:473: recipe for target '../../lib/auto/NDBM_File/NDBM_File.so' failed

(From OE-Core rev: 1a50961e63222e6196b9688f48700c7ba1596dbc)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

harfbuzz: update to 1.0.6

1.0.2 -> 1.0.6

(From OE-Core rev: 1ca357c1a1e920e260d3d61863d3c6efa01f37b4)

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

ethtool: bump version to 4.2

4.0 -> 4.2

(From OE-Core rev: c3950647a3a84986380cc37d4a1dfbacc1c0acc6)

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

openssl: fix ptest issues

* follow symbolic links while copying sources from test/*
* install required target files to remove Make errors:

  make[2]: *** No rule to make target 'xxx', needed by 'yyy'.

* fix hardcode pathes:
  /usr/lib -> ${libdir}, /usr/bin -> ${bindir}

(From OE-Core rev: 928adfc807d3c812fcd748e2cf65f392eebd852c)

Signed-off-by: Wenzong Fan <wenzong.fan@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

base-files: stage /etc/skel

Having skel in the staging area lets other recipes access these files.
For ex. recipes inheriting useradd, if passing "-m" to USERADD_PARAM,
.bashrc and .profile are put automatically in $HOME, as expected.

(From OE-Core rev: 784f4c80204a09b85b7e83c9bd91cd48f001384f)

Signed-off-by: Ioan-Adrian Ratiu <adrian.ratiu@ni.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

mktemp: raise the priority to avoid conflicting with coreutils

If mktemp and coreutils have the same priority for the 'mktemp' command,
/bin/mktemp might point to different destinations depending on which package
is installed first.

Raise the priority to 200 in mktemp recipe to avoid such problem.

(From OE-Core rev: 1c0cfffee21e99a7356c9046c86628bc0acf654c)

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libunwind: fix build for qemuarm

(From OE-Core rev: 9c370e54101c2de13cbabefb38eaa6cf584f348c)

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gma500_gfx: Avoid inserting gma500_gfx module for certain devices

The gma500_gfx driver will match certain devices on which it causes incorrect functionality,
we want to avoid inserting this module, basicallly blacklist it for specific hardware,
but still allow it to work on other hardware by default; usually when we have an already working system,
using udev rules would do the job, but since we are building it, it is impossible to blacklist
a driver when a certain udev rule matches, since rootfs isn't writeable at this point during boot time,
the solution is to use modprobe install, which runs a certain command instead of inserting a matching module,
this command needs to insert the module manually afterwards and have a flag --ignore-install
so it doesnt create an infinite loop executing itself everytime it wants to insert the module,
busybox's modprobe doesn't provide this functionality, so a small hack had to be used to avoid
the infite loop instead.

(From meta-yocto rev: 70c143767a8b63921e668a62ac463b3a6b8c6132)

Signed-off-by: Alejandro Hernandez <alejandro.hernandez@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libsndfile: fix CVE-2014-9756

Fix divide by zero bug (CVE-2014-9756)

(From OE-Core rev: f47cf07ab9d00ed7eddc8e867138481f7bd2bb7d)

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

python-pycurl: update version to 7.19.5.2

No change in License. Updated no-static-link.patch

README.rst: license checksum changed to reflect the Copyright update
COPYING-MIT: checksum changed to reflect the Copyright update

(From OE-Core rev: 8b990fd7054feaaaccce2819b5a915419c636a4a)

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

rt-tests: upgrade to version 0.96

bump version to 0.96

(From OE-Core rev: 44a14209f358abfadf7fcd9934d2a3d5cdcc63a3)

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

rpcbind: don't use '-w' for starting rpcbind

While runing:

$ systemctl restart rpcbind
$ systemctl status rpcbind

There are errors like below:
  rpcbind[1722]: Cannot open '/tmp/rpcbind.xdr' file for reading, \
    errno 2 (No such file or directory)
  rpcbind[1722]: Cannot open '/tmp/portmap.xdr' file for reading, \
    errno 2 (No such file or directory)

'-w' causes rpcbind to do a "warm start" by read a state file when
rpcbind starts up. The state file is created when rpcbind terminates.

The state file is not always there, the patch refers to:
  https://bugs.launchpad.net/ubuntu/+source/rpcbind/+bug/835833

(From OE-Core rev: 3d32a8a1fd90ca68b0d74d86165c8f3668faedcd)

Signed-off-by: Li Wang <li.wang@windriver.com>
Signed-off-by: Wenzong Fan <wenzong.fan@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libsecret: add dependency on intltool-native

intltoolize is needed by configure.

(From OE-Core rev: dc2c5a6a0316fa8d27b1b6884fb476403cf9b3e7)

Signed-off-by: Joe Slater <jslater@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

openssl: use subdir= instead of moving files in do_configure_prepend()

For clarity and correctness of source archiving, don't move find.pl from WORKDIR
to S in do_configure_prepend but tell the fetcher to put it in the right place
when unpacking.

Also re-order the files in SRC_URI so that patches are grouped together.

(From OE-Core rev: a960b6024f1b17994b0f4683a4e70fd2a079bd90)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

openssl: sanity check that the bignum module is present

The crypto_use_bigint_in_x86-64_perl patch uses the "bigint" module to
transparently support 64-bit integers on 32-bit hosts.  Whilst bigint (part of
bignum) is a core Perl module not all distributions install it (notable Fedora
23).

As the error message when bignum isn't installed is obscure, add a task to check
that it is available and alert the user if it isn't.

[ YOCTO #8562 ]

(From OE-Core rev: 2f9a2fbc46aa435a0a7f7662bb62029ac714f25a)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libsdl2: require GLES when building Wayland support

The Wayland support requires GLES2 to be enabled as otherwise the EGL support
code in SDL2 isn't enabled.

| In file included from .../SDL2-2.0.3/src/video/wayland/SDL_waylandvideo.c:34:0:
| .../SDL2-2.0.3/src/video/wayland/SDL_waylandvideo.c: In function 'Wayland_CreateDevice':
| .../SDL2-2.0.3/src/video/wayland/SDL_waylandopengles.h:38:38: error: 'SDL_EGL_GetSwapInterval' undeclared (first use in this function)
|  #define Wayland_GLES_GetSwapInterval SDL_EGL_GetSwapInterval

Solve this by adding gles2 to the Wayland PACKAGECONFIG option.

(From OE-Core rev: 8a497ef5eb3b54ed45b826fec910ed61985e04ce)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

meta: add some missing Upstream-Status tags to patches

Where Signed-off-by statements have been added they were sourced from the
original commit where the author claims creation.

(From OE-Core rev: 3e6f57059d1a5343fe1432fba408ee3f33b4c2f7)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

weston: delete unused patch

This patch was dropped from the recipe in the 1.8.0 upgrade.

(From OE-Core rev: a5e09329b8796576c158e62d87249dd4a1cc011a)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

glibc: fix Upstream-Status tag

(From OE-Core rev: 27debfb85c74ee7a4a13657e3aa4f3d1a326f776)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-firmware: package Broadcom BCM4339 firmware

(From OE-Core rev: 3d0f829740e67e9f1116ef0b8d8715c5c26d4fb7)

Signed-off-by: Tzu-Jung Lee <roylee17@currantlabs.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libusb1: fix make install race

There's an install race in when building in parallel, remove a redundant rule to
stop it happening.

(From OE-Core rev: cbdd4099b06011f2b73743c715276c730b7bf576)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libusb1: upgrade from 1.0.19 to 1.0.20

This updates libusb1 from 1.0.19 to 1.0.20

2015-09-13: v1.0.20
* Add Haiku support
* Fix multiple memory and resource leaks (#16, #52, #76, #81)
* Fix possible deadlock when executing transfer callback
* New libusb_free_pollfds() API
* Darwin: Fix devices not being detected on OS X 10.8 (#48)
* Linux: Allow larger isochronous transfer submission (#23)
* Windows: Fix broken builds Cygwin/MinGW builds and compiler warnings
* Windows: Fix broken bus number lookup
* Windows: Improve submission of control requests for composite devices
* Examples: Add two-stage load support to fxload (#12)
* Correctly report cancellations due to timeouts
* Improve efficiency of event handling
* Improve speed of transfer submission in multi-threaded environments
* Various other bug fixes and improvements
The (#xx) numbers are libusb issue numbers, see ie:
https://github.com/libusb/libusb/issues/16

(From OE-Core rev: 641a9454fbb25f1458bb8f96cbfada3e0da98dee)

Signed-off-by: Jens Rehsack <sno@netbsd.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

perl: fix spaces in brackets while using CC version

Here is the way to reproduce the issue:
...
root@localhost:~# perl -e "use Errno qw(ENOENT);"
"ENOENT" is not exported by the Errno module
Can't continue after import errors at -e line 1.
BEGIN failed--compilation aborted at -e line 1.
...

For some distros, there was extra spaces in the
brackets while using CC version:

For Windriver:
$CC --version
x86_64-wrs-linux-gcc (Wind River Linux 5.2.0-8.0-intel-x86-64) 5.2.0

For Ubuntu:
$ gcc --version
gcc (Ubuntu 4.8.4-2ubuntu1~14.04) 4.8.4

So we replace the contects between brackets with semicolon
and then use space to split.

[YOCTO #8367]

(From OE-Core rev: 115bf201a775410121d2f9769a4a5bb909cac5fd)

Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

u-boot: Update to 2015.10 release

The U-Boot 2015.10 has been released at October 20th 2015. This also
removes the GCC workaround, for the inline behavior, as this version
properlu supports the GCC 5.2 as compiler.

(From OE-Core rev: aebd1a87bbf960e69301bd29137aabb5dfab05fc)

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake-prserv-tool: check file name

Fixed:
$ bitbake-prserv-tool import /tmp/1
  File "/path/to/lib/bb/parse/__init__.py", line 114, in handle(fn='/tmp/1', data=<bb.data_smart.DataSmart object at 0x2369bd0>, include=True):
                     return h['handle'](fn, data, include)
    >    raise ParseError("not a BitBake file", fn)

ParseError: ParseError in /tmp/1: not a BitBake file

But 1.conf or 1.inc works well, check the filename and print proper
error message.

(From OE-Core rev: 273eee7a3614caea17c5bc93b720353641293cf7)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

recipetool.append: don't choke on a trailing ; in a url

Apparently bb.fetch.URI can't handle this at the moment.

(From OE-Core rev: d3e0a300810251f34932f46daf5263a23846fedd)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

yocto-bsp: Set SRCREV meta/machine revisions to AUTOREV

By default, checkout to latest revision from the machine branch specified by
the user.

(From meta-yocto rev: f79a43406b5b323587415380ecffc87527c64653)

Signed-off-by: Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

yocto-bsp: Set KTYPE to user selected base branch

Fixes the hardcode branch name set to KTYPE, where its value is used as a base branch
when user decides to create a new branch. Tested on x86_64 architecture.

[YOCTO #8630]

(From meta-yocto rev: ab895be90a0cae7dfa77a8aab3b19e5571e7e7bc)

Signed-off-by: Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

yocto-bsp: Typo on the file extension

By mistake, the file initially had a wrong extension name, so changing to the
correct one.

(From meta-yocto rev: 32c2278b8fe93429d4cfa097eefccd20157cd3b8)

Signed-off-by: Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

yocto-bsp: Avoid duplication of user patches ({{=machine}}-user-patches.scc)

On linux-yocto-dev or linux-yocto_X.YY bbappend files, the SRC_URI includes
{{=machine}}-standard.scc, which in turn includes {{=machine}}-user-parches.scc,
thus there is no need to include it again on the corresponding bbappend file.

[YOCTO #8486]

(From meta-yocto rev: 11c93b5dd8c651df478d4810e1b6ff6ad9fa57e8)

Signed-off-by: Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>