classes/conf: Add eventmasks for event handlers

Now that bitbake supports masking events for event handlers, lets use
this so event handlers are only called for events they care about. This
lets us simplify the code indentation a bit at least as well as mildly
improving the event handling performance.

(From OE-Core rev: bff73743280f9eafebe4591f7368ead91a4eb74d)

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

scripts/buildenv-internal/sanity: Update to python 2.7.3 as a minimum

We're finding bugs in python 2.6 and starting to require unittest
functionality in python 2.7.x. Its time to bump the minimum version
requirement. Anyone without python 2.7.x can use the buildtools-tarball
out to install a standalone python/git/tar setup which will work
with the system.

(From OE-Core rev: 8b90f1becd40a7f857d2fbe30eaffe218a976419)

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

sanity: Drop pseudo test, now unneeded

(From OE-Core rev: 075278db6f7a4783c952c72eebddf51a12f76075)

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

scripts/bitbake: Drop horrible double execution mess

There is no longer anything of value in the bitbake wrapper script since pseudo
is handled by bitbake internally. We can therefore drop it (yay).

(From OE-Core rev: d716d095751086e72fd789721005f0dc6d632997)

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

sanity.conf: Update minimum bitbake version requirement to 1.19.1 for bitbake-worker

This is due to the need to rely on bitbake-worker being present going
forwards.

(From OE-Core rev: ccf4fe860096f4b2a56b51b612b2a1dcb3525e59)

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

scripts/oe-buildenv-internal bitbake: Migrate python version checks

In preparation for removal of the bitbake wrapper script, move the
python version checks to the environment script. There are also
checks within bitbake itself but these may not always function
correctly on every version of python so this is really insurance.

(From OE-Core rev: 07792e4a83ca4f1c8152c228813c7f795fa6a545)

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

scripts/bitbake sanity.bbclass: Migrate tests for git and tar versions

Migrate tests for correct git and tar versions from the wrapper script
to the sanity class.

This sets the scene to allow us to remove the bitbake wrapper script.

(From OE-Core rev: 7b370e23594da5dcb53cd5507ec289c3ef2d9fb5)

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

bitbake.conf: Add extra fakeroot parameters

In order to correctly handle fakeroot at the bitbake level we need some extra
information which we provide with these new variables.

(From OE-Core rev: 003ea0fd1017dde50ced710179d0dc2e835d5185)

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

bitbake: bitbake: Add event mask flag support

Add a flag to event handlers which lists the events a given handler wishes to
process. By default event handlers recieve all events but this means
we can stop running code in many cases if we know it doesn't want the event.

This is part of the fix for YOCTO #3812, but implements filtering only
for class event handlers; the other part (events filter for UIs) will be
the subject of a different patch.

(Bitbake rev: 074003a4e7530a72863b9c685fc5c31b0f08c039)

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

bitbake: bitbake: Update version to 1.19.1

(Bitbake rev: 6acd444a557bba977ae1772371fdadf5f510e3b6)

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

bitbake: bitbake: Bump minimum python version to 2.7.3

(Bitbake rev: a01d153a8baac7136c2797c95357e74dd8872026)

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

bitbake: prserv: Adapt autostart to bitbake-worker

With the change to bitbake-worker we need to ensure the workers know
how to contact the PR service, the magic 0 port and singleton is
no longer enough.

(Bitbake rev: c761751e259bb8e940552a28794b45887b5a72d9)

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

bitbake: runqueue: Spawn a separate worker for fakeroot tasks

(Bitbake rev: 860ec42b220b7ed3f3bbe52c3546bba66644eac8)

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

bitbake: runqueue: Abstract the start and teardown worker functions

We're going to need a fakeroot/pseudo version of the worker so
abstract the code to start the worker process.

(Bitbake rev: b5d0f12f9df3ab211700473ed145ee6fbd9ca8e9)

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

bitbake: runqueue: Move the bitbake-worker execution to a higher level

The worker was being executed by each execution queue so would get
constructed twice for each build. This is wasteful so move execution
to the main runqueue so we only have to start the worker once.

(Bitbake rev: 8117f8480125b121b2b5ac0afc31b108d9e670ae)

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

bitbake: runqueue: Split runqueue to use bitbake-worker

This is a pretty fundamental change to the way bitbake operates. It
splits out the task execution part of runqueue into a completely
separately exec'd process called bitbake-worker.

This means that the separate process has to build its own datastore and
that configuration needs to be passed from the cooker over to the
bitbake worker process.

Known issues:

* Hob is broken with this patch since it writes to the configuration
  and that configuration isn't preserved in bitbake-worker.
* We create a worker for setscene, then a new worker for the main task
  execution. This is wasteful but shouldn't be hard to fix.
* We probably send too much data over to bitbake-worker, need to
  see if we can streamline it.

These are issues which will be followed up in subsequent patches.

This patch sets the groundwork for the removal of the double bitbake
execution for psuedo which will be in a follow on patch.

(Bitbake rev: b2e26f1db28d74f2dd9df8ab4ed3b472503b9a5c)

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

bitbake: fetch2: Fix AUTOINC handling

AUTOINC was meant to appear once at the start of the version string.
The list of names may not be sorted meaning it could get inserted in
the middle. This patch simplifies the code and ensures it appears at
the start.

Include cache version bump to ensure the cache picks up these changes.

(Bitbake rev: ad8bf10d873abb94d987860a3f6d06b134fb8a99)

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

buildtools-tarball: Add nativesdk-ncurses-terminfo-base

We've been seeing issues where some TERM values don't work well with the
tarball. These show up as escape codes coming from simple python calls such
as printing the python version. Adding the terminfo goes part way to addressing
the problem, the remainder is ensuring the relocation code correctly searches
the right path for it (which will be in a follow on patch).

(From OE-Core rev: 697e64c98affc4e0d8aa4828e74331c918b23a0e)

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

python: -Wformat is needed by gcc 4.8

This fixes errors in packages using python( build with gcc 4.8)

|
/home/jenkins/oe/shr-core-branches/shr-core/tmp-eglibc/sysroots/qemuarm/usr/include/python2.7/modsupport.h:27:1:
error: 'PyArg_ParseTuple' is an unrecognized format function type
[-Werror=format=]
|  PyAPI_FUNC(int) PyArg_ParseTuple(PyObject *, const char *, ...)
Py_FORMAT_PARSETUPLE(PyArg_ParseTuple, 2, 3);
|  ^
| cc1: all warnings being treated as errors
| cc1: all warnings being treated as errors

(From OE-Core rev: 5745a482a85c064a1eec960aff104cf8ce588e30)

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

mesa: introduce gallium-llvmpipe PACKAGECONFIG

(From OE-Core rev: d9f840725f3ea67fbd28d0903c99b729df5b1377)

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

mesa: use PACKAGESPLITFUNCS

(From OE-Core rev: ce127b8d20c9e812f2486405b080ab73d163e8de)

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

mesa: introduce openvg PACKAGECONFIG

(From OE-Core rev: 33d7c8a2e5087f84fb365e57e2555edb96f5fba5)

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

mesa: introduce gallium PACKAGECONFIG

(From OE-Core rev: 26276a215d7e09856d5e418edacab63197508050)

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

mesa: remove more .la files

(From OE-Core rev: 5eea21b86ca2ca09d4eb03aa9fcdb4564bdccb5d)

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

mesa: merge mesa-git.inc to mesa_git.bb

(From OE-Core rev: fb595d8a1181f30daa1c80938a335f9d249de913)

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

mesa: move LIC_FILES_CHKSUM to .inc

* we can also include .inc as first statement in recipe like most recipes

(From OE-Core rev: add5a0c17647a6b44461dd8181a412dac2ef263a)

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

mesa: use PACKAGECONFIG instead of DISTRO_FEATURES to define EGL_PLATFORMS

* it makes it easier to change PACKAGECONFIG in .bbappend and get
  consistent EGL_PLATFORMS

(From OE-Core rev: 50dcfeab2b18422f3ab067893a0b68b17706d6e2)

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

mesa: introduce PACKAGECONFIG for dri

* move DEPENDS which were originally applied only for mesa-dri to
  dri PACKAGECONFIG

(From OE-Core rev: f780ec28cbaade55361390a04ac70574bde426de)

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

mesa: merge mesa-common.inc to mesa.inc

(From OE-Core rev: 72afb46920202588948e256738ee80d585519974)

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

cogl: make cogl be able to build on ARM arch

1. Fix asm() register constraints in cogl when building for ARM.
2. Fix cogl to handle Thumb builds.

(From OE-Core rev: cdfea71ff1c4f80ff3a0ade1d7514cbf3c22abde)

Signed-off-by: Roy.Li <rongqing.li@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemd: update install-quotaon-once.patch

This patch has been merged upstream now, so refresh the patch with a backport
(the patch was changed before being merged).

(From OE-Core rev: cdad18589d08fdf592a50685f3b5fed32f4f4c76)

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

classes/qmake_base: allow parallel make

Remove MAKEFLAGS= from EXTRA_OEMAKE to allow parallel make.

(From OE-Core rev: 1aeecaeee9bb1eee779973fce7f15cc7fad269a0)

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>

qt: update qmake2 class to export qconfig.pri mkspec

This is necessary when Qt applications want to know variables such as
QT_MAJOR_VERSION, QT_VERSION, QT_MINOR_VERSION, GCC version, etc.

(From OE-Core rev: dcaa40e65d986587fa7c44f4285bbd0883478cc3)

Signed-off-by: Felipe F. Tonello <eu@felipetonello.com>
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>

gnome-doc-utils: drop the usage of create_wrapper

create_wrapper was previously abused to invoke the Python interpretter directly
instead of relying on #! parsing, which is read into a 128 byte buffer so breaks
with deep build directories.

However now that we already install a "nativepython" binary into the native
sysroot, use sed to change the #! line to /usr/bin/env nativepython.

(From OE-Core rev: 470a67d9047ccabe3b624238f31c0381793404be)

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>

createrepo: drop the usage of create_wrapper

create_wrapper was previously abused to invoke the Python interpretter directly
instead of relying on #! parsing, which is read into a 128 byte buffer so breaks
with deep build directories.

However now that we already install a "nativepython" binary into the native
sysroot, use sed to change the #! line to /usr/bin/env nativepython.

(From OE-Core rev: 7d91b44e00d6c50b7b5fbef17f2bae8e3a983a4c)

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>

python-native: add nativepython symlink

Add a symlink to the native Python so that scripts can just invoke
"nativepython" and get the right one without needing absolute paths (these often
end up too long for the #! parser in the kernel as the buffer is 128 bytes
long).

The name "nativepython" was chosen to match the existing "nativeperl" which
serves the same purpose.

(From OE-Core rev: d99dac66e3e121e96c4d8a29aee846d4b8f38622)

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>

dpkg: drop the usage of create_wrapper

create_wrapper was previously abused to invoke the Perl interpretter directly
instead of relying on #! parsing, which is read into a 128 byte buffer so breaks
with deep build directories.

However we already install a "nativeperl" binary into the native sysroot, so
use sed to change the #! line to /usr/bin/env nativeperl.

(From OE-Core rev: c4f987239fe78783d732df673a5e592bb05a877d)

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>

scripts/contrib/build-perf-test.sh: fix passing arguments

Don't pass arguments to bitbake as a single one,
because this will break when the bitbake double-exec
is removed.

(From OE-Core rev: db13f10d233873148156880ab709ec76f8d3c329)

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>

scripts/contrib/build-perf-test.sh: add branch name and sizes to results

Be more descriptive about the revision we are running on
in the global results file: add branch:commit and git describe fields.
Also add the sizes for tmp dir not only times. (previously these were
only available in the output.log)

(From OE-Core rev: 769a2c8ce797ee3afa39ab0fe9d9206a60cc4ba1)

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>

site: add more alignment values for at-spi2-core

x86 and x86_64 values were added in 8c46ec. The x86-64 values were missing an
entry, add MIPS and PowerPC values from myself in qemu, and ARM values from
Martin Jansa.

(From OE-Core rev: a6a12ef5cad0dbb2d773bdccc340f1f767c5a782)

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>

runqemu: when tunctl can't be found, say what package builds it

If runqemu is used without actually building any qemu images (i.e. you
downloaded the images) it's likely that qemu-helper-native hasn't been built.
Instead of just saying what command can't be found, tell the user how to solve
their problem.

(From OE-Core rev: 1498c431a161e8b3ddebefb5f03f4f11d5796c1d)

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>

nspr: Update to 4.10

Source directory name changed from mozilla/nsprpub -> npsr

(From OE-Core rev: e4b206f28032916e95cc21a893af687409efe83f)

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

libusb: Update tp 0.1.5

Removed obsoleted patches

(From OE-Core rev: ed72e1867d517b99e60be66cb9f2318919c36056)

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

libxkbcommon: Update to 0.3.1

(From OE-Core rev: 6c0080817476a5a4c6f2a05676d323926318c89a)

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

sysstat: Update to 10.1.6

(From OE-Core rev: 2f27ca61d4f5bd3042ea9aefc2d2824e554daa69)

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

lsbinitscripts: Update to 9.47

(From OE-Core rev: 15644f3fae1525c8847a17deed0e681df8d92b8d)

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

libidn: Update to 1.27

COPYING, idna.h and ida.c:
  Year change only 2012 -> 2013

(From OE-Core rev: ca5f7c2c977e18c4e07555f9a2f8c5f9cbe75060)

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

cups: Update to 1.6.2

(From OE-Core rev: dc6d641ed00ef2426d6ca9f3966c4c5974aea053)

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

cracklib: Update to 2.9.0

(From OE-Core rev: d42e34eb497b4ef72492b469ce0b035bb3dacb43)

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

help2man: Update to 1.43.2

(From OE-Core rev: 6d5d89deed0dca25220e5bc20de005d427b5e144)

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

cmake: Update to 2.8.11.1

(From OE-Core rev: 56f4e7101d2bbb0692607a54b3392f2703c11a32)

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

socat: Update to 1.7.2.2

COPYING changed formating and address, no license changes

(From OE-Core rev: e83bc6b5f68b12c2a989001f2cdac02b46bb3c9b)

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

qemu: update to 1.5.0

Patches removed:

0001-doc-Fix-texinfo-table-markup-in-qemu-options.hx.patch
 - applied
0002-docs-Fix-generating-qemu-doc.html-with-texinfo-5.patch
 - fixed different upstream
Revert-vmware_vga-Add-back-some-info-in-local-state-.patch
 - fixed different upstream
arm-bgr.patch
 - this should be also fixed upstream, if still issue please resend.
fallback-to-safe-mmap_min_addr.patch
 - this should be also fixed upstream, if still issue please resend.
linker-flags.patch
 - unneeded with makefile changes

from configure the obsolete --audio-cards-list option has been removed.

(From OE-Core rev: 408f933f3b5523fc26ee818a16bb4162c4343192)

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

gcc: Upgrade to 4.8.1

(From OE-Core rev: 1584e587288d355eef29f4f447a75388ed86f314)

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>

ofono: Add run time dependency for ofono test scripts

Some ofono test scripts i.e. enable-modem use python-dbus module
and this must be installed along with ofono-tests package.

(From OE-Core rev: e5422ed7f3e4b1ee8554ffe3a98006477fb52c4d)

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

gzip : upgrade to 1.6

tcmode-default: Update gzip PREFFERED_VERSION to 1.6

(From OE-Core rev: 0a3dff5bc0b98df54f8e0f4f198999e00a5f2ce7)

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

libxt: upgrade to 1.1.4

The following fragment has been added at the end of the licence file:

Copyright (c) 1993, 1994  X Consortium

Permission is hereby granted, free of charge, to any person obtaining a
copy of this software and associated documentation files (the "Software"),
to deal in the Software without restriction, including without limitation
the rights to use, copy, modify, merge, publish, distribute, sublicense,
and/or sell copies of the Software, and to permit persons to whom the
Software furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
THE X CONSORTIUM BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF
OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

Except as contained in this notice, the name of the X Consortium shall not
be used in advertising or otherwise to promote the sale, use or other
dealing in this Software without prior written authorization from the
X Consortium.

(From OE-Core rev: a7d4370efa4f51d79c2f37e92fe978659a9a2d6d)

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>

libdrm: upgrade to 2.4.45

(From OE-Core rev: 3b097ef359d884cb077bafad5f47fa9737f6886c)

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>

xwininfo: upgrade to 1.1.3

(From OE-Core rev: a89702a8e416864d40a7aa76f5f9b475d618f9f1)

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 2.21.9

(From OE-Core rev: a5b4b8a52c1523dad24e2a3ef7c6a73a9b286581)

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>

xdpyinfo: upgrade to 1.3.1

(From OE-Core rev: 58fcded3b5657d6120b97fdc605506608eb2b0a5)

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>

xkeyboard-config: upgrade to 2.9

(From OE-Core rev: 8b04e2757c07a6f583025ef52c4c9770f673ef7d)

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 1.7.1

(From OE-Core rev: 38992002d6fcf8147b2f2a0172d8acefbdcbfe8d)

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>

libnewt: split python module into libnewt-python

This is useful as it avoids pulling python into a build just to build
chkconfig. The python recipe uses the libnewt installed in the sysroot by the
main package, so there's little loss of build time, other than the additional
unpack/patch.

(From OE-Core rev: b255ebbf8283568d732827086bf0bd2b8db1d905)

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

latencytop: Deprecate tracing_enabled for tracing_on

tracing_enabled should not be used, it is heavy weight and does not
do much in helping lower the overhead. see commmit (Deprecate
tracing_enabled for tracing_on 6752ab4a9c30d5411b2dfdb251a3f1cb18aae48)
in kernel

(From OE-Core rev: ee95048cf0c79dc52857dbc0e5da9363f2ad8b9b)

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

util-linux: update to 2.23.1

Removed patches integrated upstream.

(From OE-Core rev: 4f7c228c841ce1244e92bceeef363e29748adadd)

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>

wpa-supplicant: Enable EXTRA_CFLAGS

Even if we define EXTRA_CFLAGS, but it never work, since the source codes
donot refer it, and CFLAGS is given a fixed value.

(From OE-Core rev: 4a7b5963a9d18924ae0564119f4edeefaca6b415)

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

tune-cortexa*.inc: fix tunings for cortex a5, a7, a8, a9, a15 machines.

Using CORTEX_ID variable reference in the tuning overrides did not work.
This reverts those changes, and adds a tuning file for the cortex-a5.

Revert "tune-cortexa5.inc: Add tune file for cortex-a5"
Revert "tune-cortexa.inc: create a common include for cortex-a armv7a tuning"

(From OE-Core rev: 74158c2e99c6d8631800ae80025d1cc9f19336d2)

Signed-off-by: Andy Voltz <andy.voltz@timesys.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: Add missing file that went out of sync

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

buildtools-tarball: add difflib and pprint python modules

Needed for the unittest part of bitbake-selftest can work correctly.

(From OE-Core rev: fe744381e8844ad17dcc6e036650a442dbc9575d)

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

sanity.bbclass: non-zero status means that we need -march.

A non-zero status from the march test for gcc means that the "march" flag
is needed. Correct the logic to return True in this case.

(From OE-Core rev: 217fd857df78c66eae853f935e9cdafcbeb3bc31)

Signed-off-by: Randy MacLeod <Randy.MacLeod@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

packagegroup-base: s/dosfsprogs/dosfstools/

(From OE-Core rev: 15e3c72bdcabf0a21efff60333323ac04908efb0)

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

systemd: don't install quotaon.service twice

If the same file is installed twice there's a race which can lead to "make
install" failing.  Remove the redundant installation of quotaon.service to
eliminate the race.

(From OE-Core rev: c6c5c2285e1b6cb3598ccba08aee3dfd090e7129)

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

copyleft_compliance: set 'dirs' flag on task

Without this, due to bitbake defaulting to ${B}, it's possible to get ${S}
inadvertantly created before do_unpack, breaking the unpack of git sources,
which expect the destination to not exist.

(From OE-Core rev: 694a15382fe9f3749ffdc0168a6dc2da8a29ae2d)

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

sstate.bbclass: fetch archive+siginfo in a single pstaging_fetch

This avoids unnecessary duplication of setup. The only visible change in
behavior will be the case if siginfo exists and the archive does not, in which
case it'll redownload both, but this doesn't seem unresasonable to me,
particularly since the archive is downloaded first, making this case
particularly unlikely.

(From OE-Core rev: aa4991c307d4bbdd06c3cbf8448240b74c5e01c4)

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

sstate.bbclass: kill MIRRORS and FILESPATH to speed up fetching

The default FILESPATH isn't really of use, as we don't expect to find sstate
archives buried in layer recipe directories, and the default MIRRORS is
intended for use for fetching SRC_URI, not sstate.

(From OE-Core rev: 46402b2f5b69004751f6663d435bedae0ad9dab1)

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

copyleft_compliance: handle localpaths ending with '/'

(From OE-Core rev: 1f30d2eb2ef45550c71969c735275db1448b6ac6)

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

gconf: inherit gettext

(From OE-Core rev: b35411d0ed069d135c3052ae457a488503636221)

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

avahi: inherit python classes, use PACKAGECONFIG

Without using our python classes and having appropriate dependencies, the
build is nondeterministic, and whether a python-avahi package is produced will
vary depending on the host environment, yet avahi-discover is always produced,
and it depends on python-avahi.

(From OE-Core rev: 4599ef630c13224506671bf84569bfc240cd3032)

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

license.bbclass: include all licenses in the manifest

When we don't have a generic license file for the license in question, we can
warn, but we should still include it in the manifest, otherwise the manifest
doesn't reflect reality. Failing to include a license listed in the recipe in
the manifest can't be allowed.

(From OE-Core rev: e87232828b761d56f1ce6a27e4009d350d68209c)

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

packagegroup-core-tools-profile: don't exclude mips

Now that liburcu supports mips, this packagegroup can pull in lttng recipes
(lttng-ust, lttng-tools, lttng-modules, babeltrace) on that architecture as it
does for the others.

(From OE-Core rev: c0f435cde195e1d037af32c69aa1732797362cfa)

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

insane.bbclass: quiet package errors not in *_QA

Currently, if a package QA type from package.bbclass is not listed in
ERROR_QA, it gets shown, regardless of whether it's in WARN_QA. This differs
from the behavior of the rest of the QA handling (which doesn't call
package_qa_handle_error at all if it's in neither _QA variable), and is
nonintuitive. Change this to use bb.note() if it's listed neither in ERROR_QA
nor WARN_QA, so it ends up in the task logs, but doesn't clutter up the user's
output.

(From OE-Core rev: 8f5675e6d3eb8473f367de72d2d3d8a702011a47)

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

packagegroup-base-vfat: include dosfsprogs

The ext2 group includes e2fsprogs, so this is consistent and should be
appropriate, I think.

(From OE-Core rev: 4ec66dc5f38913056942e1288c0b01a00ea0551f)

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

connman: use PACKAGECONFIG for WISPr support

(From OE-Core rev: a5161ed721791dbd32b90f1621b3eca92b2a9400)

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

mesa: remove unrecognized configure options

--disable-glut and --disable-glw are not recognized since Mesa 8.0.
--disable-glu is not recognized since Mesa 9.0.

(From OE-Core rev: dac1f95e03f897237e6551240c200370d5102807)

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

glib-2.0: drop Makefile.in hunks from patch

Patching a generated file is prone to conflicts, and we'll be replacing this
file anyway when configuring, so drop the Makefile.in hunks.

(From OE-Core rev: a3fa323c4d24375be9b783281c3cf31c0707ecb7)

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

handle two-word commands with tmux

Trying to make a devshell using tmux can fail because "tmux new"
expects a single command, not a series of arguments. It does, however,
split strings in a suitable way. So you can quote the command.

The failure mode is particularly arcane, in that you end up
with a message like:

ERROR: Unable to spawn terminal auto: \
Execution of 'pseudo /bin/bash' failed with exit code 1:
usage: new-session [-d] [-n window-name] [-s session-name] \
[-t target-session] [command]

which is confusing because there's no "new-session" anywhere in
sight (that's actually "tmux new"), and because what failed to execute
wasn't either pseudo or bash.

(From OE-Core rev: f8ed7446755eeb88191e16749350efa1e7e6197c)

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

librsvg: configure needs a native gdk-pixbuf binary

Build failure was seen on the autobuilder, so add gdk-pixbuf-native to depends list

(From OE-Core rev: 6ea293a8696843779cb23d6176b54efcde0462ec)

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

rpm: replace rpm-postinsts dependency with run-postinsts

run-postinsts is generic and will handle rpm delayed postinsts just
fine. No need to use another recipe.

(From OE-Core rev: 1195a188f366fbbb04b9379d7ea3c7c02130cc98)

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>

package_rpm.bbclass: handle pre/post remove scriptlets

Currently the scriptlet_wrapper is designed to handle only pre/post
install scriptlets. This patch will slightly change the wrapper script
to handle also pre/post remove scriptlets when we want to remove
packages from the final image.

(From OE-Core rev: aea47c77d69407b2e62f151cabba35293d179f0c)

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>

rootfs_rpm.bbclass: switch to using run-postinsts

Since run-postinsts was made generic, use it for RPM too.

(From OE-Core rev: 1ba44302f522a42535c9ad7d2fcccc06ba4c20b3)

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>

run-postinsts: make it generic

This recipe is used only when the package manager is not deployed on
target and there still are delayed postinstalls that need to be run on
target. At the moment, run-postinsts is targeted to dpkg/opkg. RPM uses another
recipe, called rpm-postinsts which, in turn, is used even when the rpm
package is deployed.

This patch intends to make run-postinsts generic and deprecate
rpm-postinsts. Here's why:
 * when opkg/dpkg are not deplyed, the meta-data files are useless.
   Hence, the awk script is not needed. The postinstall files are already
   saved in /etc/(deb|ipk)-postinsts directory and we can use the same
   procedure as in rpm's case;

Also, the recipe is moved out from meta/recipes-devel/dpkg to
meta/recipes-devel/run-postinsts.

(From OE-Core rev: 0222e6740ee4670bbf42d246ab3ecad415ffcdee)

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>

core-image-minimal: do not remove packaging files in the rootfs postprocess

Since the removal of PM's meta-data was moved in image.bbclass, after
rootfs_*_do_rootfs() runs, there's no need to remove the files in rootfs
postprocess stage. Also, if the meta-data is removed then there's no way
we can save the postinstalls to /etc/(deb|ipk)-postinsts later.

For RPM, it doesn't really matter since the postinstalls are saved to
/etc/rpm-postinsts as soon as the postinstall fails.

[YOCTO #4484]

(From OE-Core rev: aac8a59216819759d6b97bc30a7c717dbee7060a)

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>

image.bbclass: remove unneeded files from the image

This patch creates a new function, rootfs_remove_unneeded(), that will
be called after rootfs_*_do_rootfs is called and which will handle
delayed postinstalls in a generic fashion, for all backends in the same
way.

(From OE-Core rev: d3d810faf8e93ce2d3faaa70da6deeb25b1cec26)

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>

rootfs_*.bbclass: add some helper functions

This patch adds the following helper functions:
 * delayed_postinsts - outputs the list of delayed postinstalls;
 * save_postinsts - this will save the delayed postinstalls for ipk/deb
   in /etc/(ipk|deb)_postinsts;
 * rootfs_remove_packages - removes packages from an image;

Additionaly, this patch will remove a piece of code in
rootfs_ipk_do_rootfs which will be moved to image.bbclass and used for
all backends;

[YOCTO #4484]

(From OE-Core rev: 928df79cd3964f775c4c6e4283ef84b8882f9328)

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>

qt4: fix QMAKE_QT_CONFIG being overwritten with empty value

QMAKE_QT_CONFIG is set by the linux-oe-g++ mkspec but it is overwritten
with an empty value read from .qmake.cache. Avoid this by first checking
if the value from .qmake.cache is not empty before assigning it to
QMAKE_QT_CONFIG.

This allows variables from qconfig.pri such as QT_ARCH, QT_CONFIG,
QT_VERSION, etc. to be queried by qmake projects.

(From OE-Core rev: 5330905749828ea8befdee559c180754ac506b75)

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>

qt4-native: update configure paths to match target

qmake2 was being built with paths different to the target Qt.

The incorrect paths can be seen by running qmake2 -query:
QT_INSTALL_DOCS:${STAGING_DIR_NATIVE}/usr/doc
  should be /usr/share/doc/qt4
QT_INSTALL_PLUGINS:${STAGING_DIR_NATIVE}/usr/plugins
  should be /usr/lib/qt4/plugins
QT_INSTALL_IMPORTS:${STAGING_DIR_NATIVE}/usr/imports
  should be /usr/lib/qt4/imports
QT_INSTALL_TRANSLATIONS:${STAGING_DIR_NATIVE}/usr/translations
  should be /usr/share/qt4/translations
QT_INSTALL_EXAMPLES:${STAGING_DIR_NATIVE}/usr/examples
  should be /usr/bin/qt4/examples
QT_INSTALL_DEMOS:${STAGING_DIR_NATIVE}/usr/demos
  should be /usr/bin/qt4/demos
QMAKE_MKSPECS:${STAGING_DIR_NATIVE}/usr/mkspecs
  should be /usr/share/qt4/mkspecs

Use the same paths as target Qt for consistency. As the mkspecs are
installed to /usr/share/qt4/mkspecs not /usr/mkspecs, qmake2 was
unable to detect and load the webkit module (it is detected by loading
modules/qt_*.pri inside the mkspecs directory) so webkit was missing
from QT_CONFIG if a qmake project reads the QT_CONFIG variable.

(From OE-Core rev: 37b57b44ff83c718021da4f717fc66a28204c96c)

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>

classes/qmake_base: update .pro hack

This adds support for lconvert and xmlpatterns tools and tweaks the
sed expressions to handle more complex assignments in .pro files like:
LRELEASE = $$targetPath($$[QT_INSTALL_BINS]/lrelease)

(From OE-Core rev: 542fefa61f0ac1dcdacac2630fd44950ac11cdd9)

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>

classes/qmake2: export OE_QMAKE_LCONVERT and OE_QMAKE_XMLPATTERNS

These will be used by qmake_base.bbclass.

(From OE-Core rev: 405d1325ebbd31676d88a3e157f55d40d04cc3f4)

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>

qt4-native: build lconvert and xmlpatterns

These tools are needed by some Qt projects (e.g. Qt Creator).

(From OE-Core rev: e2f3232fdc2836685549e59dc5c71363d3bbab5f)

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>