bitbake: fetch2/perforce: Ensure command has a default

Ensure there is default command available if FETCHCMD_p4 isn't set.

Also clean up the getVar references whilst here.

(Bitbake rev: d150226d11d5f041f78c8c3ce4abc5465dbc81d8)

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

bitbake: parse: Improve file inheritance logging

The file inheritance logging has been here since the dawn of time. It
duplicates output many times over and logs to debug level 2. When running
with the debug option, its understandable the user may want to see the
paths of files included in the build.

These changes remove pointless/duplicate output and print the
include paths clearly at debug level one in a form which
users should be able to more easily parse.

(Bitbake rev: 633a56987df639295fe9128418da5634050063e5)

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

bitbake: knotty: Fix output printed to the user at debug log levels

If a user runs with the -v or -D options, its understandable they'd
expect to see log output from the workers yet right now a bug in the
log handling does not show this.

Fix the conditional to ensure such log output is shown on the terminal
when it has been requested. Ideally this data should always flow to
the logfiles but that is for another patch.

This also fixes the code to do what was always intended in the comments,
i.e. if the user specifies -D or -v, output is shown from the tasks,
otherwise notes are suppressed.

(Bitbake rev: 20a3c93d8572969e76563f29bff89400b93ffae7)

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

gstreamer1.0-plugins-good: add libxfixes and libxdamange to x11 depends

* fixes following issue in test-dependencies report:
  gstreamer1.0-plugins-good/gstreamer1.0-plugins-good-ximagesrc/latest lost dependency on  libxdamage

(From OE-Core rev: 23521fccb122b64f3469d46176f73fe476b0dfc5)

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

gst-plugins-good: add PACKAGECONFIG for x11

* fixes following issues from test-dependencies report:
  gst-plugins-good/gst-plugins-good-ximagesrc/latest lost dependency on libxdamage

(From OE-Core rev: 8da4462b04bdce101efd989ed6b41d88d213baa0)

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

gst-plugins-bad: add PACKAGECONFIG for schro, dc1394 and faac

* resolve following issues from test-dependencies issues:
  gst-plugins-bad/gst-plugins-bad-meta/latest lost dependency on
    gst-plugins-bad-dc1394 gst-plugins-bad-faac gst-plugins-bad-schro
  Found errors:
  gst-plugins-bad/gst-plugins-bad-dc1394/latest doesn't exist
  gst-plugins-bad/gst-plugins-bad-dc1394-dev/latest doesn't exist
  gst-plugins-bad/gst-plugins-bad-faac/latest doesn't exist
  gst-plugins-bad/gst-plugins-bad-faac-dev/latest doesn't exist
  gst-plugins-bad/gst-plugins-bad-schro/latest doesn't exist
  gst-plugins-bad/gst-plugins-bad-schro-dev/latest doesn't exist

(From OE-Core rev: 6e12dcd3a98417099f961eb9574d0d6f261afa42)

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

bitbake: parse/ast: Show append logging at lower log level

It was reported that bitbake -D made no mention of which append files it
was using. bitbake -DD does but it makes sense to increase the log level
of this piece of debug information.

[YOCTO #6262]

(Bitbake rev: 5824bf9c6feea05567d155911f4ab2e371911d34)

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

bitbake: fix cloning of mercurial repository with username and password specified in url

(Bitbake rev: 9a328262e047562a6edf7b2ec7b8a1949e287cbe)

Signed-off-by: Volker Vogelhuber <v.vogelhuber@digitalendoscopy.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

mesa-demos: Specify the lib dir to locate glut libraries

(From OE-Core rev: 55ecbebdf13ef46f0fb4d87ef11651fe692be33d)

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

classes/lib/oe: Fix cross/crosssdk references

With the renaming of the cross packages, its no longer possible to use
endswith("-cross") and similar to detect cross packages. Replace these
references with other techniques.

This resolves certain build from sstate failures which were due to the
system believing cross packages were target packages and therefore
dependency handling was altered.

(From OE-Core rev: 91edf4cac223298e50a4b8e59dd19f1b272e3418)

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

systemd: Upgrade past 212 release

Remove unconditional enable-xz, its specified by PACKAGECONFIG anyway

(From OE-Core rev: b9a44cec782e833d4f29f6ceb64908552d5cdd7b)

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>

update-rc.d: Handle multilib case for initscripts

When building multilib, we also have to add the multlib prefix otherwise
we get a WARNING:
WARNING: Multilib QA Issue: lib32-dbus package lib32-dbus - suspicious values 'initscripts-functions' in RDEPENDS

[YOCTO #6164]

(From OE-Core rev: 2e14dbb7f85532220f9aec293ddd4143fae8407b)

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

rt-tests: bump version 0.87 => 0.89

cyclictest was updated to correct some problems with CPU affinity selection
and display of help and/or execution details.
These patches were accepted by the maintainer of the cyclictest utility.
The changes were reflected in an updated release of rt-tests - v0.89
The cyclictest changes were successfully tested on both an ARM v7 and x86_64
host environment.
The OE recipe was updated locally to incorporate the newer rt-tests release
and the resulting builds were verified for the same ARM v7 target
and for qemux86.
Patch files obsoleted by the rt-tests 0.89 release were removed from the recipe
deleted:    meta/recipes-rt/rt-tests/files/0001-rt-tests-Allow-for-user-specified-PYLIB.patch
deleted:    meta/recipes-rt/rt-tests/files/0002-rt-tests-Break-out-install_hwlatdetect.patch
deleted:    meta/recipes-rt/rt-tests/files/added-missing-dependencies.patch

(From OE-Core rev: 7996ca20c24157aee259bfa88b6d7743511d5eff)

Signed-off-by: Anders Roxell <anders.roxell@linaro.org>
Signed-off-by: Gary S. Robertson <gary.robertson@linaro.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

nettle: disable openssl

Do not include openssl glue in the benchmark program.
This should fix a compile issue on multilib.

(From OE-Core rev: 22c90592347582854410b95db8b7765c85c1568b)

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

libpng12: upgrade to 1.2.51

License is the same.

(From OE-Core rev: 303a152ddd38fbb11c7d296a6fbacb33b2749695)

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

libsoup: upgrade to 2.46.0

Removed merged patches:
fix-compiling-failed-while-mips-gcc-optimization-enabled.patch

(From OE-Core rev: 5ae63515ccef232ab93446f7ce09cd8593a0733f)

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

libtasn1: upgrade to 3.5

(From OE-Core rev: 08efbacd016072c3345dc115e8b87519e7d36a89)

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

icu: upgrade to 53.1

Removed patches:
(*) add_buffer_length_check_to_UTF_16_or_32_detector.patch
    - not needed anymore

License is the same. BSD-like copyright inserted for lao-dictionary.

(From OE-Core rev: cce150502fada2068ef1ab92b88a08ed0b64d766)

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

chrpath: properly handle rootdir with '..' in path

When there is a '..' in the rootdir path, rootdir will not be a substring of
fpath.  This causes an incorrect rpath of the difference between the workdir
and the sysroot to be computed, which is incorrect.  Normalizing basedir
fixes this issue.

(From OE-Core rev: 753cfcadd8cc683e69b6707b823dc49dfb34ab0b)

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

libxcb: disable check to avoid floating dependencies

The "check" module is checked for in configure without any way to explicitly
enable or disable it, but it isn't a build dependency of libxcb.  If it's found
libxslt is also checked for, which is a build dependency but will only be used
if check is also present.

As the libxcb unit test suite is minimal at present, forcibly disable the test
for check and remove the libxslt-native build dependency.

(From OE-Core rev: 510e64d90d3ef2f9757bfa148e8cc09e4834051d)

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

librsvg: upgrade to 2.40.2

Also add Upstream-Status to gtk-option.patch

(From OE-Core rev: aa983d9117d2ddc0906aafdb7a9e584097475275)

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

util-linux: fix parallel build issure for PATHFILES

When compiling util-linux, it's possible to encounter the following error.

     /bin/sh: line 2:: misc-utils/uuidd.8.tmp: No such file or directory

This is because that the misc-utils directory doesn't exist when trying to
write to misc-utils/uuidd.8.tmp.

When generating misc-utils/uuidd.8 (or anything in PATHFILES), its directory
may not have been created yet. So we need to ensure the existence of the
directory to avoid the compilation error.

[YOCTO #6292]

(From OE-Core rev: b8c0a4f6a319a7cf84530b891707a5887f3caaf4)

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

mirrors.bbclass: Add mirror site for savannah

* The SRC_URI is not accessible.
  So need to add mirror site referred by the original site.
* The problem is that
  http://download.savannah.gnu.org/releases redirects to closest mirror
  and few mirrors (e.g. .jp) weren't working correctly while
  http://download-mirror.savannah.gnu.org/releases/ seems to be reliable.
* Add SAVANNAH_GNU_MIRROR and SAVANNAH_NONGNU_MIRROR variable in bitbake.conf.
* Change the SRC_URI using the new variable.

(From OE-Core rev: af00b6544f60e4d7581f9d9767f9d3f574392359)

Signed-off-by: Changhyeok Bae <changhyeok.bae@lge.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: modify SQL data type constraints

This patch modifies the data type constraints for the SQL
fields.

The original constraints have been based on guess worked, and
they weren't strictly enforced for sqlite. On MySQL, the
constraints are strictly enforced, and need to match the real
data that we have.

After measuring data on sample builds, the new constraints
have been validated to work with MySQL, and this patch updates
the database schema with the new field types / constraints.

(Bitbake rev: aaefbc0b4e7484d6f4854ce49602821a6e4adc4b)

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

bitbake: toaster: fix migration code for MySQL

This is a patch that fixes the 0004, 0005 migrations of the toaster
model to get them to properly work with MySQL.

These migrations had a conflict around Build.timespent field.
The sqlite3 constraits were not enough to detect the conflict, and
the migrations worked as expected.

MySQL objected to adding the field twice, so I did regenerate
the migrations with the correct model listing. The net effect
is the same, so the migrations work the same, but now we
can use these two migration on the MySQL and other more advanced
SQL engines.

(Bitbake rev: 29afc29154b948d270ce52978a1eed4cd8887f92)

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

bitbake: toaster: fix span2 headings for packages

[YOCTO 6147]

Fix bug in views.py that put 'span2' class id in a 'clclass' key
instead of the 'dclass' key.  This fix exposed another problem in
target.html and the target view function that used 'package_size'
instead of 'size' as the 'clclass' which is inconsistent with the
model field name, causing table data columns to be shown when
the header is hidden.

(Bitbake rev: 436865ac8520c1d75401dde7df373dba9a386c9d)

Signed-off-by: Dave Lerner <dave.lerner@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Make sure layer branch only shows when not empty

In the recipe details and package details pages, the layer branch
definition list elements should only show when the branch field in
the database is populated. If the branch field is empty, we don't
show them.

The patch also removes all unnecessary data-toggle and
data-original-title attributes.

[YP #6152]

(Bitbake rev: 72adcadb931519f803dad7488544767241561fb7)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: sort outcome and sstate columns by string value

Sort the 'outcome' and 'sstate_result' columns by their
description string values and not their numerical values.

[YOCTO #5921]

(Bitbake rev: eee0abe70e4d85a0753f0b122518d7a9f06181bc)

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

bitbake: Fix refetch of local files when they are read only

(Bitbake rev: 42b7be1a6c5ff559f4a213eed04b370ca955460d)

Signed-off-by: Nick Lewis <nick.lewis@usa.g4s.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

attr: Fix uclibc builds

attr needs libintl headers and libs. Add in the missing dependency and
ensure the linker flag gets passed in multilib builds by replacing the
PN == BPN check with a class-target override instead.

(From OE-Core rev: c19fb913006eb53025e5db1574d973c8d7ceec76)

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

gcc: Handle uclibc linker relocation for multilib support

We need to handle the UCLIBC_* linker variables in the same way
as we do the GLIBC_* ones to allow uclibc multilib to work properly.

(From OE-Core rev: 025ec5958b7e1fd71caa0079ec3c573126b30886)

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

openssh: add /var/log/lastlog to volatile list

The /var/log/messages reports /var/log/lastlog as missing, since openssh
needs this file, create it as a volatile.

[YOCTO #6172]

(From OE-Core rev: a29af8c20187a65fbdbbedd0b7158c07d3e713cf)

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

separatebuilddir.inc: Drop, now obsolete

(From OE-Core rev: 0e9d2a0abc9764b6d94581ff0d9e8c8e5f548f6f)

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

sudo: upgrade from 1.8.9p5 to 1.8.10p2

Upgrade sudo from 1.8.9p5 to the newest stable release 1.8.10p2.

The license checksums are modified as required, because the doc/LICENSE
file now declares that compat/inet_pton.c bears the ISC license.

As /var/run/sudo is the default directory for sudo's time stamp
files, this patch adds a configuration file to manage this directory.

(From OE-Core rev: af625d2f78a24dac96d11f159569f6465a534dfd)

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>

libxml2: fix CVE-2014-0191

It was discovered that libxml2, a library providing support to read,
modify and write XML files, incorrectly performs entity substituton in
the doctype prolog, even if the application using libxml2 disabled any
entity substitution. A remote attacker could provide a
specially-crafted XML file that, when processed, would lead to the
exhaustion of CPU and memory resources or file descriptors.

Reference: https://access.redhat.com/security/cve/CVE-2014-0191

(From OE-Core rev: 674bd59d5e357a4aba18c472ac21712a660a84af)

Signed-off-by: Maxin B. John <maxin.john@enea.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

kmod: update Upstream-Status for one patch

I will mark Inappropriate because the patch replaces some functions
unavailable in older versions of glibc.

(From OE-Core rev: fe19e4d13a645efccd778cfd3a7b55fe74498d60)

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

openssl: update upstream status for a patch

The patch is not included in 1.0.1g, but it is included on 1.0.2
branch.

(From OE-Core rev: f99ca886da274fafa212e354f9e4871eb7e59e87)

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

perl: update ptest package dependency list

Add "sed" in the ptest runtime package dependency list

[YOCTO #6260]

(From OE-Core rev: 5506333737f7e3c98d165b1025ef6c5301b16d50)

Signed-off-by: Maxin B. John <maxin.john@enea.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package_manager: Fix NoneType Object on do_populate_sdk

PACKAGE_EXCLUDE can be not defined or empty, leading to a build error.

File: '/var/lib/jenkins/jobs/qt5022-cesium/workspace/repo/yocto/meta/lib/oe/package_manager.py', lineno: 1649, function: _create_configs
     1645:                    "Pin-Priority: %d\n\n" % (arch, priority))
     1646:
     1647:                priority += 5
     1648:
 *** 1649:            for pkg in self.d.getVar('PACKAGE_EXCLUDE', True).split():
     1650:                prefs_file.write(
     1651:                    "Package: %s\n"
     1652:                    "Pin: release *\n"
     1653:                    "Pin-Priority: -1\n\n" % pkg)
Exception: AttributeError: 'NoneType' object has no attribute 'split'

ERROR: Function failed: do_populate_sdk
ERROR: Logfile of failure stored in: /var/lib/jenkins/jobs/qt5022-cesium/workspace/build/tmp/work/bobcat_64-poky-linux/meta-toolchain/1.0-r7/temp/log.do_populate_sdk.21363
NOTE: recipe meta-toolchain-1.0-r7: task do_populate_sdk: Failed
(From OE-Core rev: 26314886c3712f980ccc589b014a8f1802193b56)

Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package_manager: Fix Argument list too long

Function buildhistory_list_installed_image fails with error "Argument
list too long". This patch uses a temporal file to pass the package list
to opkg-query-helper.py

File: '/var/lib/jenkins/jobs/qt5022-cesium/workspace/repo/yocto/meta/lib/oe/package_manager.py', lineno: 421, function: list
     0417:            try:
     0418:                output = subprocess.check_output("echo -e '%s' | %s" %
     0419:                                                 (output, opkg_query_cmd),
     0420:                                                 stderr=subprocess.STDOUT,
 *** 0421:                                                 shell=True)
     0422:            except subprocess.CalledProcessError as e:
     0423:                bb.fatal("Cannot compute packages dependencies. Command '%s' "
     0424:                         "returned %d:\n%s" % (e.cmd, e.returncode, e.output))
     0425:
Exception: OSError: [Errno 7] Argument list too long

ERROR: Function failed: buildhistory_list_installed_image
ERROR: Logfile of failure stored in: /var/lib/jenkins/jobs/qt5022-cesium/workspace/build/tmp/work/qt5022-poky-linux/qimage-dev/1.0-r0/temp/log.do_rootfs.16747
NOTE: recipe qimage-dev-1.0-r0: task do_rootfs: Failed
ERROR: Task 7 (/var/lib/jenkins/jobs/qt5022-cesium/workspace/repo/yocto/../qtec/meta-qt5022/recipes-core/images/qimage-dev.bb, do_rootfs) failed with exit code '1'
NOTE: Tasks Summary: Attempted 4999 tasks of which 30 didn't need to be rerun and 1 failed.
NOTE: Writing buildhistory
Auto packing the repository for optimum performance.

Summary: 1 task failed:
  /var/lib/jenkins/jobs/qt5022-cesium/workspace/repo/yocto/../qtec/meta-qt5022/recipes-core/images/qimage-dev.bb, do_rootfs
Summary: There were 74 WARNING messages shown.
Summary: There were 2 ERROR messages shown, returning a non-zero exit code.

(From OE-Core rev: 36cba6e00d76462e4ae314dd2af0b47472835538)

Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

x-load: remove recipe

x-load has been superseded by u-boot SPL for all cortex based TI SoCs.
For legacy boards meta-ti layer (that contains a maintained recipe)
should be used instead.

(From OE-Core rev: 3a675df3dd608b6e7250a3b0bfc9362d2f4e8aa9)

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

uclibc: Set MULTILIB_DIR based on baselib

baselib contains the current "lib" value so pass this into
uclibc's configration to allow multilib builds to work correctly.

(From OE-Core rev: 665b03b338d8d710250bd92d6b9a227255f808f8)

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

libiconv: Fix B != S with uclibc builds

Without this, uclibc builds fail with libtool version mismatches. The issue
is that we need to remove the files in ${S}, not ${B} which is now
the default after the B != S change.

(From OE-Core rev: 054151c77dae51e2826eeac9b9ae2de3a08f0bb6)

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

ptest: Work from PTEST_ENABLED, not DISTRO_FEATURES

Some classes of recipe disable ptest even though its in DISTRO_FEATURES
(e.g. nativesdk). We shouldn't attempt to build ptest packages when
its disabled. This replaces some DISTRO_FEATURE checks with PTEST_ENABLED
checks instead.

(From OE-Core rev: 8b14fcc62f31bbbb231790136cdb984db96d9ba9)

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

scripts/runqemu: avoid pipe with sed

Replace:
cat <file> | sed -e xxx
By:
sed -e xxx <file>

(From OE-Core rev: 173769374b1b37244d64767144398337001f1c6f)

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

initscripts: avoid pipe with sed

Small optimisation in /etc/init.d/populate-volatile.sh.

Replace:
cat <file> | sed -e xxx
By:
sed -e xxx <file>

(From OE-Core rev: c91739cd08ed5451a0927586a14db54c4c328ad7)

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

sysvinit-inittab: avoid pipe with grep

Replace:
cat <file> | grep xxx
By:
grep xxx <file>

(From OE-Core rev: 350c7c63c05a977158da91b4e06f84a6fb84a6ee)

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

dropbear: avoid pipe with sed

Replace:
cat <file> | sed -e xxx
By:
sed -e xxx <file>

+ fix indentation

(From OE-Core rev: e6a25f3377f75d995c996374e155e10ced519dce)

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

apmd: avoid pipe with sed

Replace:
cat <file> | sed -e xxx
By:
sed -e xxx <file>

(From OE-Core rev: 3bfaea36a4484f1db0340cd67f7783ccec23c738)

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

acpid: avoid pipe with sed

Replace:
cat <file> | sed -e xxx
By:
sed -e xxx <file>

(From OE-Core rev: 49345d83f280da13b7993971a336c6176587961d)

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

*.bbclass (shell): avoid pipe with sed

Replace:
cat <file> | sed -e xxx
By:
sed -e xxx <file>

(From OE-Core rev: e2026f5d32ac05396615224ac9ec927439e7e6b4)

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

tcl: cleanup

Changes:
- add missing licence file: tcl
- use binconfig to patch tclConfig.sh paths
- avoid subshell in do_configure()
- use ${PV}

(From OE-Core rev: 6ef8ac2ff97a408d4311ebebbafcfaf0a84be4d9)

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

rxvt-unicode: upgrade to 9.20

(From OE-Core rev: 17ea13e0133d77426fabc8d7a87e20c245692faf)

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

babeltrace: upgrade to 1.2.1

The release for 1.2.1 is on branch stable-1.2.
Backported a patch to fix out of tree build.

(From OE-Core rev: 699d6c02140a37262d8e5daf29143db22a5a5d0b)

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

mmc-utils: fix compilation failure for mips64 target.

This patch fixes mmc-utils compilation failure for qemumips64.
Remove the 'include <asm-generic/int-ll64.h>' line from mmc.h,
because this file is automatically included if _MIPS_SZLONG
is not 64, otherwise, <asm-generic/int-l64.h> is included.

Expicitly including <asm-generic/int-ll64.h> will cause the
compilation failure for mips64 target.

[YOCTO #6267]

(From OE-Core rev: 5477cc86015be24cc5b232ac1f8c67cd8b65a740)

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>

ltp: enable PM and RT tests

(From OE-Core rev: 370d8954dddd90addd0e955bbd4592316af113b8)

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

perl: fix missing Module::Metadata

Module::Build depends on Module::Metadata (since perl 5.13.9).
Module::Metainfo is only a wrapper for Module::Metadata.

See discussion in http://patches.openembedded.org/patch/66233/

(From OE-Core rev: 3babe5704fbda62920ec17a910d0f3fe9f468229)

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

openssl: fix CVE-2014-0198

A null pointer dereference bug was discovered in do_ssl3_write().
An attacker could possibly use this to cause OpenSSL to crash, resulting
in a denial of service.

https://access.redhat.com/security/cve/CVE-2014-0198

(From OE-Core rev: 580033721abbbb4302bc803ebc70c90e331e4587)

Signed-off-by: Maxin B. John <maxin.john@enea.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

python3: Revert python-config to distutils.sysconfig

The newer sysconfig module shares some code with distutils.sysconfig,
but the same modifications as in

12-distutils-prefix-is-inside-staging-area.patch

makes distutils.sysconfig affect the native runtime as well as cross
building.  Use the old, patched implementation which returns paths in
the staging directory and for the target, as appropriate.

This change reverts this upstream patch
http://hg.python.org/cpython/diff/712970b019f7/Misc/python-config.in

(From OE-Core rev: 7b2ffd68ae8235dcc3ddff9cbe8525e61f3b3d28)

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

python3: Substitute correct python version in shebang

If python2 and python3 are both available, scripts that are subject to
this substitution can possibly run with the wrong python version.
python3-config is one such script.

(From OE-Core rev: 23849347d0fe60a01578efdd6c6e23ebb444dcd6)

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

distutils.bbclass: only modify *.py file if it contains path to be removed

Currently sed command touches every single *.py file. This modifies the
timestamp of the file. All *.pyo files will be recompiled during the first
boot, because timestamp will not match. This should be only necessary if
sed command changes the file.

(From OE-Core rev: 2d01c5a4989dcf03a202c27730a2a8f334e0c37a)

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

python3/gcc/autoconf: Fix Upstream-Status in some patches I authored

(From OE-Core rev: 337798fa5c0a1d1e745a143f6a9f398b07f0628f)

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

base.bbclass: Fix multilib PREFERRED_VERSION handling

Now that gcc-cross, gcc-crosssdk and others have expanded variables in
their names, the code that previous just applied to cross-canadian needs
expanding to cover the other cases. Improve the conditional and also
fix a bug where the multilib prefix wasn't being added into the generated
versions.

(From OE-Core rev: d2b18d7692229d4b87b38becf173b1f8b6e90025)

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

bitbake: data_smart: Ensure all possible overrides are cached including those with '_' in the name

Unfortunately we've been neglecting to pay the correct tributes to the
cookie monster and hence the datastore is malfunctioning.

Currently tributes are only paid on the last part of a variable after
the last "_" character. We need to split by *all* "_" characters since
an override may contain the character.

This fixes the code so the correct number of tributes are made. Paradoxically
parsing appears to be faster after this change.

(Bitbake rev: d1c712fd3a59fa804e6fd451612c30487671f3a2)

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

tcmode-default: Fix up sdk related PREFERRED_VERSION variables after toolchain changes

Also drop an obsolete gcc-cross-intermediate reference.

(From OE-Core rev: d807e87708d5e4dc9267aa1611ba7ad7beec1d40)

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

tcmode-default.inc: Adjust for new toolchain component changes

Recently we renames the cross compilers as well as spun out
libgcc-initial from gcc-initial out. We need to also express
it in default tcmode otherwise defining GCCVERSION and/or BINUVERSION
outside OE-Core does not cover all cases

(From OE-Core rev: 82813d27640549255dc81a15d7fb63688431f747)

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

linux-firmware: upgrade to commit dec41bce44e0dff6a2c3358a958fadf22bf58858

- Updated copyright years for radeon firmware;
- Changed copyright terms (privilege of use) for qla2xxx firmware;
- Changed Broadcom firmware name for SDIO;
- Added TI wl18xx firmware package.

(From OE-Core rev: 4a9595dc704012b21f7bb272b82a8a4d548e347f)

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>

taglib: upgrade to 1.9.1

- Updated website;
- Updated tarball location;
- Switched to BP variable;
- Improved source dir definition;

(From OE-Core rev: 6c06a67c1a94ebc81f9bbbd71bbb512539eacdaf)

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>

linux-yocto/3.10: mei.cfg: enable Intel chipsets

Importing the mei meta data change from the yocto 3.14 kernel tree.

(From OE-Core rev: 2a39c6c11d9438a6d9cab9255e69c52779709e3e)

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>

linux-yocto/3.14: update to v3.14.2

Updating to the korg 3.14.2 -stable release.

(From OE-Core rev: 34afc38d86d169f0c0c5f2427f644b0dcc3bf9a1)

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>

linux-yocto/3.10: update to v3.10.38

Updating to the korg -stable release.

(From OE-Core rev: f71e40ff6ae98644800e15f8f472be6022a32f1e)

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>

linux-yocto/3.4: update to v3.4.88

Updating the 3.4 SRCREVs to import the korg v3.4.88 -stable release.

(From OE-Core rev: faf8ccedf8956590cfcf01825ad3d421fb7f7a0e)

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>

linux-yocto-custom: provide defconfig guidance

As reported by Andrea, defconfigs created with savedefconfig should specify
their expansion mode as "--alldefconfig' for custom yocto recipes. To ensure
that this is documented, we can add a comment in the skeleton recipe.

Reported-by: Andrea Adami <andrea.adami@gmail.com>
(From OE-Core rev: 2c41a28c9a9772531dcfc7079604ff2441c71f8f)

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>

kernel-yocto: quote kconfig mode check

We allow inheriting recipes to control the kconfig mode used by merge_config.sh
via the KCONFIG_MODE variable. An error crept into the variable reference, and
since it is not quoted, the true condition always runs.

The result is that operations without an explicit kconfig mode cannot trigger
allnoconfig for defconfig builds, which can result in some options being
dropped from the final .config.

Quoting the reference allows it to evaluate properly.

(From OE-Core rev: aad19e4381a8a09c354e5899885997c5b4cd115b)

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>

qt4: Explicitly disable sqlite2

* notice that this disables only sqlite2 (sqlite dependency),
  sqlite (sqlite3 dependency) stays enabled

* fixes these errors detected by test-dependencies.sh:
  qt4-embedded/qt4-embedded-plugin-sqldriver-sqlite2/latest doesn't exist
  qt4-embedded/qt4-embedded-plugin-sqldriver-sqlite2-dbg/latest doesn't exist
  qt4-x11-free/qt4-plugin-sqldriver-sqlite2/latest doesn't exist
  qt4-x11-free/qt4-plugin-sqldriver-sqlite2-dbg/latest doesn't exist

(From OE-Core rev: 279c221d9dfce43254ed1f03cf84ea454cc7b485)

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

ldconfig-native: also default to lib32 and lib64 directories

make ldconfig default to /lib+/usr/lib, /lib32+/usr/lib32 and
/lib64+/usr/lib64 on bi-ABI architectures.

(From OE-Core rev: c90bb98d4ac14562ef4882691daed3aaa9d08504)

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

serial-getty service: Add xterm as default TERM

In poky with systemd enabled, vt102 is selected for getty
causing user to experience a very crappy terminal. Default
TERM to xterm.

Reviewed-by: Jonathan Liu <net147@gmail.com>
(From OE-Core rev: 473ff65c2f69de4ece3204fadfae7c5cb992149a)

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

gdbm: enable ptest support

Install gdbm test-suite and run it as ptest

(From OE-Core rev: aaba69d598763831e8b42996154ed87d0db10714)

Signed-off-by: Maxin B. John <maxin.john@enea.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

defaultsetup: enable blacklist by default

(From OE-Core rev: 74e377eeb6de606b99ed0ff840922261f4a27abc)

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

systemtap: update to 2.5 release

(From OE-Core rev: 6c30c43af893d3d68651501c0ffb01d46c5f4d13)

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

grub git: update to latest git (grub-2.02-beta2-81-g4b8b913)

(From OE-Core rev: ec6158c0a41a929bac6df9d4ab62afa5dc39544e)

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

wpa-supplicant: add libgcrypt as a dependencie

gnutls doesn't depend on libgcrypt anymore but
wpa-supplicant does. So add it as a dependencie.

(From OE-Core rev: b5e0e0589dba0e3eb6fa070594c904fec6e6c3a8)

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

gnutls: update to 3.2.13

Removes not needed or already merged patches.

Removes unused configure flags.
Tells gnutls to use the included libopts.

Removes libextra (not needed since 3.0.5).

(From OE-Core rev: 4e360033579b9501449798b2d118f786816a557f)

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

nettle: add nettle to oe-core

Newer versions of gnutls depends on nettle.

(From OE-Core rev: e0e5eaee72e49c01e76d56cd03fbfb3e20febbb0)

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

libgcrypt: update to 1.6.1

(From OE-Core rev: fd2848df50bec43248cab6ccdea2358be4719473)

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

xinput-calibrator: fix XDG launch

In the move from meta-oe to OE-core XDG based launched was dropped
without noting it in the commit message, so fix that regression.

Gnome-session will now launch the calibrator again.

(From OE-Core rev: 64e0c8a7585a36964988f2f00ac89eda514d5f7b)

Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xdg-utils: reinstall xdg-terminal

It's being used by e.g. gnome-session when Terminal=true is set in
.desktop files, like xinput-calibrator does.

(From OE-Core rev: 9a821d80b732a52574ac5c20b323966811828852)

Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

procps: Do not detect 32/64 bit compiler options

It gets its wrong in detecting 64bit options
and OE we already pass the right options as compiler
defaults

(From OE-Core rev: ec3add1752a410fada04b4788c142b91ef63ffa3)

(From OE-Core rev: 32b826720626ec2d426ff7dbfe5607bc0dbd8082)

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

lttng-modules_2.3.3.bb: Fix build on ARM/gcc-4.9

Update the structure to work with immutable bio_vecs

(From OE-Core rev: 09d3d8f97d6c60707b8dfb3fa5b04fa2efe6f6f6)

(From OE-Core rev: b32d57e5a807a634c411628af1688aec960ac6e7)

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

mdadm: Fix build on gcc 4.9

Fix

error: right-hand operand of comma expression has no effect
[-Werror=unused-value]

(From OE-Core rev: 9e1aac485175d29951f7ad23cd7515c8f7580271)

(From OE-Core rev: 58da3355522736b93bb6d164739c795068d2a1e8)

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

oprofile: Update to 0.9.9

Backport patches to fix with gcc 4.9
delete applied patches and obsoleted ones

(From OE-Core rev: 9ef0ed510aff5e975fb6d8be9330a10172a5b5d8)

(From OE-Core rev: 5b5016ed87bcc916d29fb7e1e3d23a600847c882)

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

systemd: Fix build with gcc 4.9

We need to use gcc-nm and gcc-ar to deal with slim objects
which are generated when using -flto

(From OE-Core rev: e6d84c9f4fef201217ada60711ecfb94bc5dc2e0)

(From OE-Core rev: 56dc32f128b3029c38381c7cc81f244fbd77416e)

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

binutils: Fix building nativesdk binutils with gcc 4.9

Patches explain the issue in detail but this is exposed
with gcc 4.9 in binutils 2.24

(From OE-Core rev: fc5c467b680fc5aef4b0f689e6988e17a9322ae0)

(From OE-Core rev: 4dfb8847ebf8aab90ad8888933468e2899c96998)

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

cairo: Fix build with gcc 4.9

Fix upstream bug
https://bugs.freedesktop.org/show_bug.cgi?id=77060

gcc 4.9 now generates slim objects when LTO is used
These slim objects only contain intermediate language
representation for LTO. Use -ffat-lto-objects to create files which
contain additionally the object code. When linking some .so file, the option
'-flto' is missed, so the error happens.
We add '-ffat-lto-objects' flag to make gcc 4.9 behaves the same as gcc 4.8.

(From OE-Core rev: c4e65e14fbdf957aa34e074337c5649a0b0888b7)

(From OE-Core rev: 858f01ab9613c2b122718c350da763f8fe981d9f)

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

gcc: Add 4.9 recipes

(From OE-Core rev: f051216ea373f166016b15bbd2a2a6f136430372)

(From OE-Core rev: d4573cb750bfde488682244d30266dfe675bac06)

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

cleanup-workdir: Fix crosssdk reference after addition of PN

(From OE-Core rev: 2f922357287129f7d216393531886ed1f3ce8185)

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

sstatetests: Use TUNE_ARCH instead of TARGET_ARCH

The scripts extact variable values from bitbake -e. Unfortunately
TARGET_ARCH is unset from that environment so we use TUNE_ARCH instead
which will have the value we need.

(From OE-Core rev: b77494025b3d47fd5130ae6c445ac84cabb1f91e)

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

sstatetests: Update after binutils-cross name change with PN addition

(From OE-Core rev: c5230b459fb4ef486f657f5633b67ca925981f03)

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

curl: Backport a fix for a build issue

mkhelp: generate code for --disable-manual as well

This allows configure --disable-manual to run and build without having
to regenerate the src/tool_hugehelp.c file which otherwise is necessary
since we ship tarballs with that file present.

(From OE-Core rev: 544a96255203a6779d1f0022d003c6680f330511)

Signed-off-by: Tudor Florea <tudor.florea@enea.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: prserv/db: Avoid fsync() calls

If the power were to fail, it doesn't matter to us much if the data
makes it to disk or not, we'd have other problems. However an fsync()
call on a multi build autobuilder is painful so lets avoid them.

(Bitbake rev: 4eb2dc8048e2722d64d589f453df1ce6262c71b8)

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