bitbake: tinfoil: fix for changes to cooker config structure

Fix the code here for recent changes to the initialisation of
configuration objects for cooker.

(Bitbake rev: 9d3ca9aa73a448b0594f03ac8e8317403ec0dc8d)

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

bitbake: bitbake-layers: fix for move of calc_bbfile_priority within cooker

calc_bbfile_priority is now in CookerCollectFiles which can be accessed
on the collection attribute of a cooker instance.

Fixes [YOCTO #4513].

(Bitbake rev: 5d941631ad7198737d9a5c5a920a9062fa0431f8)

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

bitbake: command: Fix function module reference after code rearrangement

(Bitbake rev: ff9f62fd5f76892ad41a5329b75472501e17e712)

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

bitbake: runqueue.py: Ensure export flag is set for fakeroot environment variables

This means the variables show up in the shell execution "run" files since
its useful to know what the fakeroot environment is and how to set it up
manually.

(Bitbake rev: bdf437747b664479acde6deaa9096e2a6bcdf483)

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

bitbake: bitbake/cookerdata: Explicitly specify cooker configuration options

As the code stands today its hard to know which configuration variables
are used by which parts of the system. Some are used by the UIs, some
by bin/bitbake itself, some by cooker.

This patch changes the configuration to just contain the variables cooker
uses, and changes bin/bitbake to access the variables it needs directly
which hopefully lets us start to untangle this mess.

(Bitbake rev: e57497a24b6157c92519a34accd66035a39ad1f8)

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

bitbake: cooker: Split configuration parsing code into cookerdata

In order to have a memory resident bitbake and to allow task execution, we need
to be able to rebuild the base configuration without a cooker. This moves the
code into its own class so it can be built independently.

The interface is less than ideal here but I didn't want to add parsing methods
a subclassed DataSmart, at least until we've experimented further with this code
and are certain that makes sense. At the very least, the methods are ugly and need
cleaning up.  Spliting the code out seems to be the right thing to do though and
should unblock various activities on BitBake so I believe this code is a step in
the right direction.

Based on a patch from Alexandru Damian <alexandru.damian@intel.com>

(Bitbake rev: 22a0b3cf73d2689db0c118b37aa7492632f8b0a7)

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

bitbake: cooker: Move commandline parsing back into the UI/cookerdata

Building up a set of actions for the server is tricky since we depend upon the
commandline but fall back to values from the datastore. We should be able to build
a datastore without a commandline and vice versa. Ultimately the UI should send
the commands to the server.

This patch amounts to code rearranging, moving the heavy lifting to the UI, though
a helper in the configuration option. This will need further cleanup/tweaking but
this should be the only update needed to the UIs. The code now queries the server
for any missing data should it need to.

This code allows various knowledge of configuration variables to move to the UI side
only, partcularly pkgs_to_build but also all the command specifiers. It should also
be possible to move cmd eventually, I'm just unsure if any callers call the commands
expecting this to default to something sane right now.

(Bitbake rev: 2dbbb1d51dafd4451fef8fe16f095bcd4b8f1177)

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

bitbake: bitbake: Create cookerdata splitting config from cooker and bin/bitbake

Currently the UI and server configuration is one big incestuous mess. To
start to untangle this we creater cookerdata, a new module which contains
various confiuration modules and the code for building the base datastore.

To start with we add a ConfigParameters() class which contains information
about both the commandline configuration and the original environment.

The CookerConfiguration class is created to contain the cooker.configuration
options. This means we can transfer new paramters to the server over something
like XMLRPC and then build a new configuration from these on the server.

Based on a patch from Alexandru Damian <alexandru.damian@intel.com>

(Bitbake rev: 35bd5997e8d8e74bc36019030cc10c560a8134f9)

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

bitbake: bin/bitbake: Improve exception handling

Due to the internal event processing, this excepting handler usually raises an
Empty error, masking the underlying failure. Ensure the original exception is
raised.

(Bitbake rev: 7d548568a55adfe84a976f2a549995e42da1afef)

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

bitbake: fetch2: Shorten long srcrevs

The long srcrevs are mainly used or the workdir construction as well as
the package version. The long entries are hashes generated by the git fetcher
and other scms using a similar revision mechanism.

We need these to change when the package changes however collisions are
unlikely to happen within the domains we care about. The long revisions
have generated negative user feedback due to the use in path and file
names.

This patch therefore truncates the revisions to 10 characters maximum.

This should be safe in the contexts where these revisions are used as
the chances of spatially close collisions is very low (distant
collisions are not a major issue in the way we use these).

(Bitbake rev: 43a8319cda7fae37862dae323eeb24cb39ca21b7)

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

bitbake: fetch2/git: Clean up sortable_revision

Now we no longer try and provide increasing values from the fetcher,
we can simplify the function structure for the sortable_revision
pieces and move the AUTOINC handling directly into the function
which needs it, simplifying the code.

(Bitbake rev: fb068bee47bb1a06f02447daf16c2b2a79c03288)

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

bitbake: data: Ensure dependencies of subfunctions are accounted for

Currently we account of the top level function's vardeps but not
those of any subfunction. This would imply we'd have to manually
write the dependencies of all parent functions which would be crazy.

This patch adds the dependencies to fix the issue.

(Bitbake rev: 658008cc6a8acd5c1f26877c9444c96002db01e7)

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

packagegroup-core-clutter: use clutter-1.0 packages

(From OE-Core rev: f45aa6e07318f22ede4614b28ea15cd48f5e4095)

Signed-off-by: Tomas Frydrych <tomas@sleepfive.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

mx: Update to improved naming scheme

The package has been renamed to mx-1.0, keeping up with the upstream versioning
policy. The 1.14.7 tag does not build with clutter 1.14, so git revision
9b1db6b8060bd00b121a692f942404a24ae2960f from the 1.14 branch is used instead.

(From OE-Core rev: e49f3e23d7ae7d105d9c32a33bd28590f5c300cd)

Signed-off-by: Tomas Frydrych <tomas@sleepfive.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

clutter-gtk: Replace with 1.4.2 version

(From OE-Core rev: 095581d3b888ba702ba0b1b7e9e3ef8fe69dc164)

Signed-off-by: Tomas Frydrych <tomas@sleepfive.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

clutter-gst: Replace with 1.6.0 version

(From OE-Core rev: 4476e789ac41f44ad74385d85a16e505d96bc100)

Signed-off-by: Tomas Frydrych <tomas@sleepfive.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

clutter: Replace with 1.14.4 version

The package has been renamed to clutter-1.0 instead of clutter-vmajor.vminor,
keeping up with the upstream versioning policy (all 1.x packages install
clutter-1.0 pc file and headers and are backward compatible), and to simplify
dependency management (worth noting that since the 1.x development files are
not parall installable, it is not possible to use two versions of clutter 1.x
at the same time anyway).

Package configuration is provided via PACKAGECONFIG options as follows:

  'x11'  : enable X11 backend,
  'glx'  : enable GLX backend,
  'egl'  : enable EGL backend,
  'evdev': enable evdev input backend

Default configuration is 'glx'; typical configuration for embedded HW using
'native' EGL would be 'egl evdev'.

(From OE-Core rev: 4fbae4449fbb3f5eba499418b374c12bccfb4243)

Signed-off-by: Tomas Frydrych <tomas@sleepfive.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

cogl: Replace with 1.14.0 version

The package has been renamed to cogl-1.0 instead of cogl-vmajor.vminor, keeping
up with the upstream versioning policy (e.g., all 1.x packages install cogl-1.0
pc file and headers and are backward compatible), and to simplify dependency
management (worth noting that since the 1.x development files are not parallel
installable, it is not possible to use two versions of cogl 1.x at the same
time anyway).

Package configuration is provided via PACKAGECONFIG options as follows:

GL flavour:   'gl' for big GL or 'gles2' for GLES2
              (GLES1 is availabe in cogl, but not supporeted here at present.)

EGL platform: 'egl-null' -- PVR-style null platform
              'egl-kms'  -- kms platform provide by Mesa
              'egl-x11'  -- egl over xlib platform
              (Additional EGL platforms, e.g., Wayland are supported by cogl,
              but not supported here at present.)

GLX:          'glx' for the GLX extension support (implies 'gl')

Default configuration is 'glx'; typical configuration providing 'native' egl
on embedded HW would be 'gles2 egl-null'.

(From OE-Core rev: b508fdd2b19ca30da8d09caf646897dc4cf195c8)

Signed-off-by: Tomas Frydrych <tomas@sleepfive.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

clutter.bbclass: helper class for clutter and friends

(From OE-Core rev: 19b7e58b71d31071361f19fd82666d40653569c4)

Signed-off-by: Tomas Frydrych <tomas@sleepfive.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

zlib: put shared libraries in base_libdir

We put the shared libraries in base_libdir because at least
one library under base_libdir, pam_cracklib.so, needs them
and will cause a qa warning when it is built.

(From OE-Core rev: ebb8382af892bef8e11fb590292506e1124276c5)

Signed-off-by: Joe Slater <jslater@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

cdrtools-native : upgrade to 3.00

Upgrade from 2.01 -> 3.00

- Updated md5 of the license file because new information
was added by the owner.
- Removed glibc-conflict-rename.patch because it is not
required anymore.
- Updated no_usr_src.patch because it didn't apply.

(From OE-Core rev: 951a8e422be93a3d06d0149f9b070eddcdb37afb)

Signed-off-by: Andrei Dinu <andrei.adrianx.dinu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bind: use /var/run/named instead of /var/run/bind/run

Change /var/run/bind/run to /var/run/named to avoid the following error message.

  chmod: cannot access '/var/run/bind/run': No such file or directory

[YOCTO #4429]

(From OE-Core rev: a32c05f691ef5620516b2f84452fb5129e16bb14)

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

opkg: remove unnecessary extra leading slash in run-postinsts

(From OE-Core rev: 047826f034f00c2402c0a078b5c918019fcf36a4)

Signed-off-by: Jonathan Liu <net147@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

nfs-utils : upgrade to 1.2.8

upgrade from 1.2.7 -> 1.2.8

Removed patches that are already in the upstream.

Added a make clean routine needed because the package comes with
precompiled 64-bit objects that break the build.

(From OE-Core rev: f505e913eae5e91d494234ee38a38ac961583b12)

Signed-off-by: Andrei Dinu <andrei.adrianx.dinu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

qt: fix warning -Wunused-local-typedefs introduced by gcc 4.8

in meta-oe maliit-framework failed with

| In file included from /home/andreas/tmp/oe-core-eglibc/sysroots/overo/usr/include/qt4/QtCore/QtCore:53:0,
|                  from ../utils/core-utils.cpp:18:
| /home/andreas/tmp/oe-core-eglibc/sysroots/overo/usr/include/qt4/QtCore/qtconcurrentfilter.h: In function 'QtConcurrent::ThreadEngineStarter<void> QtConcurrent::filterInternal(Sequence&, KeepFunctor, Redu
| /home/andreas/tmp/oe-core-eglibc/sysroots/overo/usr/include/qt4/QtCore/qtconcurrentfilter.h:108:47: error: typedef 'Iterator' locally defined but not used [-Werror=unused-local-typedefs]
|      typedef typename Sequence::const_iterator Iterator;
|                                                ^
| cc1plus: all warnings being treated as errors

(From OE-Core rev: 4cb93a42bae3dfa53880a9475a92cc046dde5767)

Signed-off-by: Andreas Müller <schnitzeltony@googlemail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

image_types.bbclass: remove the old images

Remove the old image before the new one generated to save disk
space when RM_OLD_IMAGE is set to 1, this is an easy way to keep
the DEPLOY_DIR_IMAGE clean.

[YOCTO #4391]

(From OE-Core rev: b3e2e405c53d63bc71872d41f455507be833e7eb)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: test/fetch: Allow the conditional network tests to work under python 2.6

The skipIf decorator is unavailable under python 2.6 and finding a syntax
works there is hard. This patch does allow network tests under 2.6.

(Bitbake rev: 565b3e31e0226c6e193ee0b031bd5e7bef25591e)

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

bitbake: test/fetch: Split the network tests from the rest of the fetcher tests

(Bitbake rev: f32265bcc6b4fe5566aca16f6b05555ec6a96d01)

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

bitbake: fetch2/utils: Clean up imports

Move various random imports to the start of the modules as cleanup
and avoid an import issue with bb.process on python 2.6.

(Bitbake rev: aed4adfbe3a591ca4f8e41fb763c9f961bf2e6d5)

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

bitbake: tests/fetch: Allow to work on python 2.6

Currently the fetch tests fail on python 2.6 since check_output
doesn't exist on this version of python. Use bb.process instead
to avoid this problem.

(Bitbake rev: a3637155e829e06bbfe50fb8753de3de7ee8c22e)

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

python-pygobject-native: fix build dependency on python-native

The class-native override is undoing the dependency that distutils-base
added which we require. This patch adds in the missing dependency manually
to ensure the build functions correctly.

Fixes [YOCTO #4502].

(From OE-Core rev: ae28ee3f7a060b9e0d13154a84f2444a98490b5b)

Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

Revert "initramfs-live-install*: fix the "install" boot option"

This was not the correct fix for this issues, it turns out that
base-files package was getting installed un-intentionally when
rpm-postinsts was split out. The base-files recipe lays down the
link that caused the cat failure.

[YOCTO #4504]

This reverts commit 45e460d0846f0f660128dc06064b597ce40282b3.

(From OE-Core rev: 7d50133dd5c955d97193cc26458ebbf84dbfe0a4)

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

rpm-postinst: remove un-needed RDEPENDS on base-files

This was transferred un-intentionally when we split the rpm code,
the base-files then layed down additional files (specificaly /etc/mtab)
in the initramfs image and caused problems for the installer code.

Removing the RDEPENDS will fix the image generation issue and thus
the installer, so we will need to revert a previous patch that attempted
to fix the earlier issue #4229.

[YOCTO #4504]
[YOCTO #4229]

(From OE-Core rev: 6861eb34e903c5ddf491eca04011b2219fe14267)

Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libcroco: allow native builds

Use BBCLASSEXTEND to get native builds, and inherit gnomebase instead of gnome.
libcroco uses the class just for the SRC_URI helpers, and the full class was
pulling in many unrequired dependencies including hicolor-icon-theme-native,
which doesn't exist.

(From OE-Core rev: f2c438ddfeb17bbff384e612bb247f3652d85a74)

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

gtk+3: disable Xinerama support

GTK+ has enabled by default support for Xinerama, but it's not a build
dependency so this can be enabled or disabled depending if xinerama has been
built previously.

This can cause problems with sstate, and result in situations where
PKG_CHECK_MODULES() for GTK+ can appear successful but the CFLAGS are not set.

(From OE-Core rev: c6862c8162bd79c45961a4b9f8f570a3d5346148)

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

linux-yocto/3.8: set qemumips64 SRCREVs

When qemumips64 was enabled, its machine specific SRCREV was missed.
Without it, qemumips64 builds from the wrong branch and we miss important
oprofile/ftrace fixes that are required for boot.

(From OE-Core rev: 6c94a53f7c1b674927486fdbf81613b2832bae40)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

qemuimage-testlib: change default directory for target tests

/tmp is a better location, and it allows copying files
on read only fs images

(From OE-Core rev: e3561c1cae467a4fb79723f83dea54d9d62adf7d)

Signed-off-by: Stefan Stanacar <stefanx.stanacar@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

qemuimage-testlib: pass optional timeout to ssh function

Sometimes we need to change the timeout used by the function for
certain kinds of tests.

(From OE-Core rev: 21950ff5eb032fefc4753bd68af57f655d0c61f2)

Signed-off-by: Stefan Stanacar <stefanx.stanacar@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

qemuimage-testlib: add support for extra arguments passed to runqemu

Some tests might want to pass extra arguments to runqemu.
I can think of "kvm" or qemuparams="-m 1024" when we want extra muscle.

(From OE-Core rev: 1a5446ca73736753d172c06dcb48858887c7a896)

Signed-off-by: Stefan Stanacar <stefanx.stanacar@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

tcf-agent: Use common two-digit sequence numbers for init script links

The common init script links use two-digit sequence numbers and
systemd interprets everything behind the two-digits as service name.
This leads to the following failure during boot:
  Cannot add dependency job for unit 9tcf-agent.service

(From OE-Core rev: e29a13bf047ce90e9e1aae953044b9cab85f9aab)

Signed-off-by: Stefan Herbrechtsmeier <stefan@herbrechtsmeier.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

glib-2.0: drop dependency on python-argparse

Since 2.20 the use of argparse was removed, so we don't need these dependencies
anymore.

(From OE-Core rev: 2a6067f6e69c8f2d04b8cf7e4a97e5085f758654)

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

connman: replace hardcoded path in init script and systemd service

The connman init script sources a setup file from /usr/lib/connman,
so we end up with no network in qemu multilib enabled images.

The init script it's installed by connman and because wired-setup
it's installed by another package (connman-conf) we can't use
libexecdir here and now (in the init script and systemd service file).

Once libexecdir changes from ${libdir}/${bpn} to something else like
/usr/libexec we could use that instead of ${libdir}/connman.

Changed in v2: - better commit message

[YOCTO #4493]

(From OE-Core rev: fca3a884e9cae13a521d840838eee3c01f0b6acf)

Signed-off-by: Stefan Stanacar <stefanx.stanacar@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemd: make udev-hwdb postinstall run on host

Run udevadm inside qemu in order to create the HW database index on
host.

The alternative would be to build a native version
of udevadm which would imply several things: split out systemd recipe,
create a common inc file and create a new recipe for udevadm native.

However, this latter solution might also add up to the build time
(the native recipe would need to run configure, make udevadm, install),
besides the time spent in the actual postinstall...

So, having the postinstall run through qemu is a good trade-off.

[YOCTO #4341]

(From OE-Core rev: 0f6c0ed7888603c8d026a671f2acb1515ce799bf)

Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

busybox-mdev: Startup at level S03 instead of S06

The udev hotplug deamon starts at 03, so mdev should start at the same
point.

This fixes the bug that when modutils (runs at 04) initializes devices
that the mdev hotplug will often not pick them up, resulting in missing
device nodes and similar problems.

(From OE-Core rev: 18d3e5eb9102a984d00837f8fec16fda522c511a)

Signed-off-by: Mike Looijmans <mike.looijmans@topic.nl>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

toolchain-scripts.bbclass:Add CROSS_COMPILE variable to environment-setup file

when compiling the linux kernel manually. users need to set CROSS_COMPILE separately.
adding the CROSS_COMPILE variable will be nice for using.

(From OE-Core rev: b66ec45e5dd9418a6568c04ef30854531a4b66f3)

Signed-off-by: Zongchun Yu <b40527@freescale.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

connman: Add PPTP support

(From OE-Core rev: 8f1ecefc0c3d86ee6c467a9cc7119be237f74083)

Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

connman: Add L2TP support

(From OE-Core rev: 53cb0d8502d895c47c87130bc9f3e9452cbd04b5)

Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

connman: Add vpnc support

(From OE-Core rev: 6c23a3f66cad6f66b5ca599773d7cc3925fd777d)

Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

connman: Add OpenVPN support

(From OE-Core rev: 428c15a7b1e9c1ffa90e83a276c367f3f84268d8)

Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

connman: Add VPN support

One needs to add following statement into local.conf or distro config

PACKAGECONFIG_append_pn-connman = " openvpn vpnc l2tp pptp"

in order to activate support for these VPN technogies in ConnMan.

(From OE-Core rev: 92da847ed6cea6342bdc86de121534259332a2c3)

Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gpgme: Updated from 1.4.0 to 1.4.1

(From OE-Core rev: 775529a3891d3f3317ff79b0b1bd2f196f7fa2e8)

Signed-off-by: Ionut Radu <ionutx.radu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pixman: Update from 0.29.2 to 0.29.4

(From OE-Core rev: b5d834f4320579fd2fb688d0bfb36b01e8b95511)

Signed-off-by: Ionut Radu <ionutx.radu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

ethtool: Updated from 3.8 to 3.9

(From OE-Core rev: deb79dc43a13621490f700d0269cedc126a32ac3)

Signed-off-by: Ionut Radu <ionutx.radu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libfontenc: Updated from 1.1.1 to 1.1.2

(From OE-Core rev: bfcae5426c6fd48da513c55b14dc90416b83f130)

Signed-off-by: Ionut Radu <ionutx.radu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

json-glib: updated from 0.15.2 to 0.16.0

(From OE-Core rev: 21907426ff25966f0bf1ac6854a4d2742659167b)

Signed-off-by: Ionut Radu <ionutx.radu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libxi: updated from 1.7 to 1.7.1

(From OE-Core rev: fa3aec02b58d30b9c78a5e44627bfafc4f2ef221)

Signed-off-by: Ionut Radu <ionutx.radu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bbclass: bb.fatal() clean up

The bb.fatal() is defined as:

def fatal(*args):
    logger.critical(''.join(args))
    sys.exit(1)

So anything after bb.fatal() in the same code block doesn't have any
effect, e.g.:

    bb.fatal("%s_%s: %s" % (var, pkg, e))
    raise e

The "raise e" should be removed.

I searched all the files which use bb.fatal(), only the following 3
classes have this issues:

  insane.bbclass
  package.bbclass
  package_rpm.bbclass

[YOCTO #4461]

(From OE-Core rev: 4c47525c5a171b1282615c9fbc7d84addef85f92)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xf86-video-modesetting: upgrade to 0.7.0

(From OE-Core rev: d6fb741dfd828d40317175a624065918b3e5bfec)

Signed-off-by: Jonathan Liu <net147@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

machine/qemumips64: Add machine definition

(From OE-Core rev: 3f16954c6fae2de78fa9c5ece69354be7ec25cf2)

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

binutils: Add broadcom XLP support

(From OE-Core rev: f72a8257388f3df1d9b1f7ae516295afe227d3e3)

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-yocto: add qemumips64 compatibility

The 3.8 and linux-yocto-dev kernel trees carry qemumips64 support. Adding
the board to the compatibility directly in the recipes is the first step
to adding mips64 support to the broader set of packages.

(From OE-Core rev: 0b4187fac16f78a1f9ba1c696edbc8c2aa189cb1)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

runqemu: Replace use of ifconfig with ip

ifconfig and its ilk (net-tools package) is deprecated in favour of iproute2 package
and is now removed by many distro's e.g. Archlinux. So we replace ifconfig with ip utility

(From OE-Core rev: c19e5d19ae8e6e6eb9b37549d80765b8315f79a4)

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

kern-tools: refresh and make dash compliant

The separately packaged merge_config.sh in the kern-tools package was
missing upstream fixes, and in particular a change that ensures it is
dash compatible.

By grabbing that upstream commit and rebasing the existing patches on
top of the new baseline, we are up to date and working on systems
where /bin/sh is dash.

[YOCTO #4473]

(From OE-Core rev: 681bcd2783e100dd2882273f28f16ef118161e89)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pkgconfig: Update to 0.28

Remove older patches that have been fixed upstream, added new patch to
use the correct alias-libtool instead of ./libtool.

pkgconfig provides an internal glib library, so use it to avoid ciruclar
dependecny issues.

(From OE-Core rev: d91a43228c5ba35335a28de278194df23dcdf978)

Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

boot-directdisk: Scope HDDDIR and HDDIMG variables to avoid conflicts

These variables should not be shared with other image classes.
The bootimg class also has an HDDDIR variable that could be overwritten
if executing concurrently in the same image recipe.

(From OE-Core rev: 94c83fa26af104d28aa4e6951bcb109bb0c82265)

Signed-off-by: Jonathan Liu <net147@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

syslinux: Update to 4.06

Patches are now part of upstream or no longer needed.
Added new util-linux dependency for isohybrid.
Paths updated to reflect directory structure changes.
Add CFLAGS and LDFLAGS overrides.

[YOCTO #4438]

(From OE-Core rev: 17e7ac0c5e75245d17a90e5cc49ade3d18a168ba)

Signed-off-by: Jonathan Liu <net147@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake.conf: define empty EXTRA_IMAGE_FEATURES before using them

* fixes build without EXTRA_IMAGE_FEATURES defined:
  ERROR: '${EXTRA_IMAGE_FEATURES}' in IMAGE_FEATURES is not a valid
  image feature. Valid features: dbg-pkgs ....

(From OE-Core rev: b2cc92595b30d96a79f33ea7a7217834c8b6bff7)

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

cachedpath.py: global name 'error' is not defined

The error is not global since we don't use "from os import *", so it
should be os.error.

[YOCTO #4489]

(From OE-Core rev: ac0e2781c2723257bd380cac170d4c8b97e36747)

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

bitbake: bin/bitbake: Fix queue import for pyhton 2 and 3

(Bitbake rev: 782bc0e797802224f06cb6f9098a50d658c86523)

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

image.bbclass: Add package-management image feature to validitems

Some images dont use core-image bbclass and add this feature directly

Fixes parsing errors like

/builds1/angstrom/cache/prserv.sqlite3, IP: 127.0.0.1, PORT: 34734, PID:
24551 Parsing recipes...ERROR: Unable to parse
/builds1/angstrom/sources/meta-kde/recipes-images/angstrom-kde-desktop-imag
e.bb: Exited with "1" ERROR: 'package-management' in IMAGE_FEATURES is
not a
valid image feature. Valid features: dbg-pkgs debug-tweaks dev-pkgs
doc-pkgs ptest-pkgs read-only-rootfs splash staticdev-pkgs ERROR:
'package-management' in IMAGE_FEATURES is not a valid image feature.
Valid
features: dbg-pkgs debug-tweaks dev-pkgs doc-pkgs ptest-pkgs
read-only-rootfs splash staticdev-pkgs ERROR: Command execution failed:
Exited with 1

(From OE-Core rev: f9f1397d688cc4eb8d736066b1280bee40a8916a)

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: hob: fix for XMLRPC mode

Contains a small fix to the builder.py code
(a missing "self" in a variable assignment).
With this changes, "hob" can run with the XMLRPC backend:

$ hob -t xmlrpc

(Bitbake rev: 3497478f803986af32099ddd124c47df8e89f6da)

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

scripts/hob: enable additional arguments

Allow "hob" to receive other arguments in the command line (for example
the server type and the address of the remote end if running remotely).

(From OE-Core rev: 1bd6fa9c81dea90f66641835a4c2ed6f2b7a239a)

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package.bbclass: Fix sources contents

http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=c376f1f49cea182a2887945840ab97a20970a373
fixed a valid issue where the sources file was accumulating information
and subsequent task runs of do_packacge were not cleaning it. The
fix is wrong however since we're removing the file within a loop.
This fix removes the file outside the loop ensuring it is not truncated
and contains the correct information.

(From OE-Core rev: a015881f2207aded601459ba3eebbefb0002b3c5)

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

sanity.bbclass: Attach the missing value to a format string.

The tuning changes to sanity.bbclass were almost right, but one of
the messages had a %s with no % operator.

(From OE-Core rev: cf5e40598ae9a83f22cabedc7b72000beb62703c)

Signed-off-by: Peter Seebach <peter.seebach@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: cooker/bitbake-layers/tinfoil: Fix recipecache typo

(Bitbake rev: 0f5eee689992f84d263cb817dc2ce755a9a075f7)

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

bitbake: bitbake-layers/tinfoil: Catch up with status -> recpiecache rename

(Bitbake rev: 0a9cbe7a6a17c5df38cd442ee8650097d6bbf502)

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

image-vmdk.bbclass: disable the NOISO = "1"

The vmdk needed hddimg which was generated by do_bootimg in the past,
but do_bootimg generated both iso and hddimg, so we set NOISO = "1",
it now uses do_bootdirectdisk, doesn't use do_bootimg anymore, so we
don't need NOISO = "1", and need remove it, otherwise:

IMAGE_FSTYPES += "live vmdk"

Would not generate the iso.

I'd like to remove it rather than comment it, but we had comment the
do_bootimg before.

(From OE-Core rev: c5c1517726aa103a3cdb60abda95e28997cac7c4)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libdrm: upgrade to 2.4.44

Removed part of the installtests.patch since tests can now be installed
giving an option to configure script.

(From OE-Core rev: 563edb321594b007cae2294e807235ad22fd27da)

Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xproto: upgrade to 7.0.24

(From OE-Core rev: e2efcff4c4a3d6d5b4cecac57b952f8fb8680a22)

Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xf86-input-evdev: upgrade to 2.8.0

(From OE-Core rev: 6a68d5a1bfa7e29ceaee488f7455b70ca7a24d26)

Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xf86-input-synaptics: upgrade to 1.7.0

Licence checksum changed because RedHat adjusted the copyright years.

(From OE-Core rev: e09796f917fd684e1402da5953d34cd9c10ed800)

Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xf86-input-mouse: upgrade to 1.9.0

(From OE-Core rev: 7c5498e42f6152e6179781dc0e96e255af3f87dd)

Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xf86-input-keyboard: upgrade to 1.7.0

(From OE-Core rev: 993b641ee56dc3dbcdb0ce73d6e875058c262c89)

Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xf86-video-intel: upgrade to 2.21.6

(From OE-Core rev: b3a0c5d13ce23e64aaa683f533dea4d399ad367c)

Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xf86-video-vmware: upgrade to 13.0.1

Removed the backported patch.

(From OE-Core rev: 01b74d26ba9f96122105987581660ee2b26521b6)

Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

Upstream-Status: Correct Capitalization

(From OE-Core rev: 43b2a2f375e2201be7a9bb6a9c5c0a9fc61f3361)

Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

classes/core-image: drop apps-console-core IMAGE_FEATURES support

Remove the legacy support for the apps-console-core IMAGE_FEATURES item;
we've kept this for a while but it's time for it to go.

(From OE-Core rev: bcd5ed8dd0e80cb76415cd2be9686d4e8023421d)

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

classes/image: show an error on invalid IMAGE_FEATURES

If the user specifies an invalid feature in IMAGE_FEATURES, show an
error during parsing. Valid IMAGE_FEATURES are drawn from
PACKAGE_GROUP_ definitions, COMPLEMENTARY_GLOBS and a new 'validitems'
varflag on IMAGE_FEATURES (so that additional non-package group features
can be added elsewhere.)

Implements [YOCTO #3308].

(From OE-Core rev: 8d25442ab795ff0f8e072da2022108eff128e2b2)

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sbc: upgrade to 1.1

(From OE-Core rev: 89be8aa7c6d49c6372df687798d006d8cfe12943)

Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

image-vmdk: Fix error recreating symbolic link

If the vmdk image symbolic link already exists from a previous build,
overwrite it instead of returning an error.

(From OE-Core rev: bcfb5a783dc94fe38539a19f4f7de249ce2391c1)

Signed-off-by: Jonathan Liu <net147@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gmp: updated to 5.1.1

A number of patches are now part of the upstream.
Tested by compiling and running core-image-minimal/qemux86.

(From OE-Core rev: e4bc2b287bdeb0abf7c198564575f719a98a6591)

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pciutils: updated from 3.1.10 to 3.2.

(From OE-Core rev: 5d8dd6ec550764b89754812f0d06cc2baa707a1e)

Signed-off-by: Ionut Radu <ionutx.radu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

glib-2.0: upgrade to 2.36.1

(From OE-Core rev: 71e83ce29f09678ed5aa67c5edf88e0f24f754f8)

Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

tcl: upgraded to 8.6.0

The license checksum changed because of a small correction in the
license text.

(From OE-Core rev: eb2b082146cf3ba229a816faa9c87fd098c0627f)

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gdb-cross-canadian: set everything in gdb-cross-canadian.inc

This allows out-of-tree versions to include one file to have all set.

(From OE-Core rev: c1bfe7985ac95644585e8b296edd9722c2ca4f57)

Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gdb: add 7.6 version

(From OE-Core rev: ace67d28ab295451a59edcf02447bcd9095b3edb)

Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

mpfr: upgraded to 3.1.2

(From OE-Core rev: 84a13eab40e8aca73e5f2421e9a32c922e00b714)

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

byacc: upgraded to 20130304

(From OE-Core rev: c6cb877ae1185952e9954ffc5e29e032f6c448f5)

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

autogen: upgraded to 5.17.3

'guile.patch' updated for the new version.

(From OE-Core rev: 0f1b721a99ee1403ffe88dde97676316bca3004c)

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

iproute2: upgrade to 3.9.0

patches removed; already included in upstream.

(From OE-Core rev: 9a8a765c2ccbae29ecabe00f358b6bbb1c9b3b98)

Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>