generate-manifest-3.3.py: Add importlib

importlib isn't currently included in any of the python packages, so
create a new one for recipes that require it.

Cc: Paul Eggleton <paul.eggleton@linux.intel.com>
(From OE-Core rev: b17d2e1838f1f1c3310926a4f3eed375898c60f3)

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

gstreamer1.0-rtsp-server: Add recipe to support gst-rtsp-server

A Gstreamer library which provides an API to create an RTSP server (e.g. to stream RTP to VLC clients and similar).

Tested, works with GStreamer 1.2.3 + videotestsrc based pipeline on RPi.

(From OE-Core rev: 228736f2ffba6c2e06e72042a1fdf3fc0807f9b7)

Signed-off-by: Alex J Lennon <ajlennon@dynamicdevices.co.uk>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

Added new test cases for bitbake modes (-e -n -p -r -R -c -k)

(From OE-Core rev: eec4976e3b22efe73e823ad4876d78ad933113f4)

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

systemd: update a uclibc specific patch to avoid segment fault error

The alloca() function allocates space in the stack frame of the caller,
so using alloca(new_size - old_size) would possibly crash the stack,
causing a segment fault error.

This patch fixes the above problem by avoiding using this function in
journal-file.c.

[YOCTO #6201]

(From OE-Core rev: c69816d2bf84369ba578bf9d92e01c9d91351a64)

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>

apr: upgrade to 1.5.1

The LICENSE has no change, except a newline added.

Fix do_compile failed:
| tools/gen_test_char > include/private/apr_escape_test_char.h
| /bin/sh: tools/gen_test_char: No such file or directory
| make[1]: *** [include/private/apr_escape_test_char.h] Error 127

The tools/gen_test_char was invoked at build time, and it didn't
work for the cross compiling, so we built it with $BUILD_CC.

To make sure there was no side effect on the change, not adding this
patch, I did the test:

1. Built the native tools/gen_test_char on x86-64 host, and invoke
   it on that host;

2. Built the target tools/gen_test_char which arch was arm, and invoke
   it on the qemuarm;

3. Comparing the results, they were total the same;
...
/* this file is automatically generated by gen_test_char, do not edit. "make include/private/apr_escape_test_char.h" to regenerate. */

static const unsigned char test_char_table[256] = {
    32,30,30,30,30,30,30,30,30,30,31,30,30,30,30,30,30,30,30,30,
    30,30,30,30,30,30,30,30,30,30,30,30,6,16,63,22,17,22,49,17,
    17,17,1,16,16,0,0,18,0,0,0,0,0,0,0,0,0,0,16,23,
    55,16,55,23,16,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
    0,0,0,0,0,0,0,0,0,0,0,23,31,23,23,0,23,0,0,0,
    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
    0,0,0,23,23,23,17,30,30,30,30,30,30,30,30,30,30,30,30,30,
    30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,
    30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,
    30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,
    30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,
    30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,
    30,30,30,30,30,30,30,30,30,30,30,30,30,30,30,30
};
...

Fix do_install failed:
| DEBUG: Executing shell function do_install
| NOTE: make -j 16 DESTDIR=tmp/work/x86_64-linux/apr-native/1.5.1-r0/image
 install
| tmp/work/x86_64-linux/apr-native/1.5.1-r0/apr-1.5.1/build/mkdir.sh tools
| gcc  -isystem/home/pokybuild/yocto-autobuilder/yocto-slave/build-appliance/
build/build/tmp/sysroots/x86_64-linux/usr/include -O2 -pipe
tools/gen_test_char.c -o tools/gen_test_char
| make[1]: Entering directory `tmp/work/x86_64-linux/apr-native/1.5.1-r0/
apr-1.5.1'
| tmp/work/x86_64-linux/apr-native/1.5.1-r0/apr-1.5.1/build/mkdir.sh tools
| gcc  -isystem/home/pokybuild/yocto-autobuilder/yocto-slave/build-appliance/
build/build/tmp/sysroots/x86_64-linux/usr/include -O2 -pipe
tools/gen_test_char.c -o tools/gen_test_char
| tmp/work/x86_64-linux/apr-native/1.5.1-r0/apr-1.5.1/build/mkdir.sh
include/private
| tools/gen_test_char > include/private/apr_escape_test_char.h
| /bin/sh: tools/gen_test_char: Permission denied

Remove the 'tools' dir creation, it always existed. And it caused
gen_test_char unexpected rebuilt at do_install time.

(From OE-Core rev: b751144ba467b1358af2db8a4c30b8e0644cffa5)

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

apr-util: upgrade to 1.5.3

(From OE-Core rev: 512ec9fb677d920c813d084b85238f3be86f376f)

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

readline-5.2: exclude autoheader

readline hand-maintains config.h.in so exclude autoheader and re-use a patch
from readline-6.3 to fix a typo in variable names.

See oe-core 8c37d32d6133c6ad2b9142e7a42775e7a979b570 against readline-6.3 for
further rationale.

(From OE-Core rev: 8281133c6dcb2f31666d76e282d02bafe65e15d7)

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>

readline-5.2: use upstream's aclocal.m4 as acinclude.m4

Instead of shipping a fork of the upstream aclocal.m4, simply rename it to
acinclude.m4 at configure time.  We don't need the fork now that autoheader is
excluded.

(From OE-Core rev: 7b98575797c2e8822e6a26d95cb7e727d6efef4b)

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>

grep: upgrade 2.18 to 2.19

Upgrade grep from 2.18 to 2.19.

(From OE-Core rev: 61deb74c37236adef4a72a0930150752993cf89d)

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>

clutter: update to 1.18.2

* Updated Ross's install-examples.patch to current
  ** AM_CPPFLAGS changes no longer needed
  ** example_DATA = ... redhand.png no longer needed
  ** minor line # changes

(From OE-Core rev: 7a15b1855021eec9dd292bb76c0543a886e9c2ad)

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>

cogl: update to 1.18.0

Announcement:
http://lists.freedesktop.org/archives/cogl/2014-March/001636.html

* License changed from LGPL to MIT
* kms-include-stddef.h-before-drm.h.patch fixed upstream

(From OE-Core rev: 0e3b38baddf6341465288323dfdc0440c47579d7)

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>

mesa-demos: Use DEMOS_DATA_DIR.

Modify mesa demo apps to respect the DEMOS_DATA_DIR
configuration parameter for locating data files.

(From OE-Core rev: a6a6cd089352157d35a8730e37153ce9df62ec01)

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>

mesa-demos: Add missing data files.

Add some data files that are present in the git repository:
   http://cgit.freedesktop.org/mesa/demos/tree/?id=mesa-demos-8.1.0
but not in the release tarball
   ftp://ftp.freedesktop.org/pub/mesa/demos/8.1.0/mesa-demos-8.1.0.tar.bz2

(From OE-Core rev: 347b25be7136ec639926ed7b8cfa6da55f00f17c)

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>

mesa-demos: Build with system data files

Use the "--with-system-data-files" when configuring mesa-demos
so that the data files in /usr/share/mesa-demos/ will be properly
located at run time.

Copy some additional data files into the target filesystem.

(From OE-Core rev: d602264267f49bf72704b6c5ab1247b4027313bb)

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>

Update tzdata to 2014d

moved build data to an inc file since it changes little.

(From OE-Core rev: d5a95dc8985a42bb7e50bc4e7dc6b012d711ff08)

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

Update tzcode to 2014d for zic fix

Changes affecting code

    zic no longer generates files containing time stamps well before
    the Big Bang.
    This works around GNOME bug 730332
    <https://bugzilla.gnome.org/show_bug.cgi?id=730332>.
    (Thanks to Leonardo Chiquitto for reporting the bug, and to
    Arthur David Olson and James Cloos for suggesting improvements to
    the fix.)

Changes affecting documentation

    tz-link.htm now mentions GNOME.

(From OE-Core rev: 5ffc10fcea78d1ad54dbcf854e16b0e2f19116fc)

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

image_types: Fix ubi filesystem return codes

If the first command returns an error, it will not cause the image generation
step to fail. Simply split up the statement into multiple lines to avoid
this issue, they no longer need to be one line expressions.

[YOCTO #6391]

(From OE-Core rev: f8125a1e9b6893a12355d55d4df584a8d97f0bff)

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

resolvconf: Update to 1.75

Archive compression changed to xz

(From OE-Core rev: 1a6ee66aa48e024fc5d112e940488157b74b5fe0)

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

cups: Update to 1.7.3

(From OE-Core rev: bb80dd49b5631772276772f9551ca2b72b9e72d4)

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

man-pages: Update to 3.68

README updated with general text regarding installation location

(From OE-Core rev: dd4d713a9db3b6e1fe411cd581a068bf60d76873)

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

libffi: Update to 3.1

LICENSE file has updated year

(From OE-Core rev: 14b777301371280ec7ad6cf528ebb597c025c988)

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

libcheck: Update to 0.9.13

(From OE-Core rev: eb4452b214c5dd81dab5ee82d9d5048d29c33613)

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

libxkbcommon: Update to 0.4.2

(From OE-Core rev: 541a0f1afd066d73e956c347190a65cca6f2504a)

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

x264: Update SRCREV to match commit in upstream git repo

It seems that 585324fee380109acd9986388f857f413a60b896 is no
longer there in git and it has been rewritten to
ffc3ad4945da69f3caa2b40e4eed715a9a8d9526

Change-Id: I9ffe8bd9bcef0d2dc5e6f6d3a6e4317bada8f4be
(From OE-Core rev: b193c7f251542aa76cb5a4d6dcb71d15b27005eb)

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>

grub_git: fix build warning->error

This patch fixes a build warning which is promoted to an error via -Werror.

changes from v1:
 * put "Upstream-Status" in patch itself, not in OE patch header
 * fix commit wording

(From OE-Core rev: 4b41f4199d77144a629897173cff3d1efcfa85f9)

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

qt4: add patch for GIF denial-of-service vulnerability

For further details, see:
https://bugreports.qt-project.org/browse/QTBUG-38367

(From OE-Core rev: c322f67808bb36c5fea3fbabd30aa242e408fc50)

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>

qt4: upgrade to version 4.8.6

* Remove patches merged upstream
* Add a qtscript translation package for an additional .qm file added in
  this release

(From OE-Core rev: 32a9c7101c7212c69d40893ee597eaf34955b641)

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>

elfutils: upgrade to 0.158

Add 'm4-biarch.m4-tweak-AC_RUN_IFELSE-for-cross-compiling.patch' to fix cross compiling failure;

Rebase 'elf_additions.diff' for 0.158;

Drop obsolete patches:
- nm-Fix-size-passed-to-snprintf-for-invalid-sh_name-case.patch
- elfutils-ar-c-fix-num-passed-to-memset.patch
- fix-build-gcc-4.8.patch

Pick patches from debian:
http://ftp.de.debian.org/debian/pool/main/e/elfutils/elfutils_0.158-2.debian.tar.xz

We could not directly add elfutils_0.158-2.debian.tar.xz to SRC_URI, because it
contains other souce codes which are not pathces.

(From OE-Core rev: d9c7a02240ce37d5b2569d9177e8ba534b9295ce)

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

sanity.bbclass: Add libsdl-native check

If libsdl-native is in ASSUME_PROVIDED, check for it in the sanity tests.
This warns the user if they've said its being provided but it isn't and
prevents silent build issues.

(From OE-Core rev: d9d7b0515fcf47c4cf7533a12915ea92298ce834)

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

oprofileui-server/oprofileui/pcmanfm: Add missing intltool-native DEPENDS

These recipes all require intltool-native to build but were missing
a dependency on it.

(From OE-Core rev: 3c5fdd10c3fe70d650469556b501055ebaecd628)

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

libcroco/libfm/menu-cache: Add missing gtk-doc class dependencies

These recipes all use gtk-doc m4 macros but were missing dependencies
on the class.

(From OE-Core rev: 500d48ba760f50c780377fb1ebed257a6e8b1887)

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

gsettings-desktop-schemas: Add missing DEPENDS on gobject-introspection-stub-native

It uses gobject-introspection macros but was missing a dependency
upon something which provides them.

(From OE-Core rev: a3ed1ec9ff9f963925bd7ed336c32d5e428bd15a)

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

lame: Add missing DEPENDS on gettext-native

Without this, configure will fail due to a missing AM_ICONV macro.

(From OE-Core rev: 36ace50aea4e47a0ddf9365a6dec1e0db0b31107)

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

recipes: Add missing pkgconfig class inherits

These recipes all use pkg-config in some way but were missing
dependencies on the tool, this patch adds them.

(From OE-Core rev: 2543b14dd0ca13005be0df027543431fc8e882ae)

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

libfm: Ensure m4 directory exists

configure.ac references the m4 directory so things like gtkdocize
will try and place m4 files there. We need to ensure the directory
exists or the builds can fail.

(From OE-Core rev: 0ced649c055dd25c500d0fd34e4ea94504d98580)

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

toaster.bbclass: exclude variable dependency on toaster postfunc

This patch fixes an issue where the toaster postfuncs were
added to the task signature, making impossible the reuse
of sstate caches from builds outside toaster control.

Now the signatures do not differ between toaster and toaster-less
builds.

(From OE-Core rev: 552c5daceb9f51d2b7331a12dfa033f1ca3d7468)

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

qemu: upgrade to 2.0.0

- switched to ${BP} variable.
- updated download link;
- fxrstorssefix.patch no longer needed, superseded;

(From OE-Core rev: b6f6e389b68468266926856bd318c245696ea932)

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

v86d: fix for SeaBIOS VGA BIOS compatibility

This fixes a compatibility issue between v86d
and SeaBIOS VGA BIOS where the leal instruction
is not decoded properly. Read the
http://thread.gmane.org/gmane.comp.emulators.qemu/271806
thread for more details.
Patch taken from upstream.

(From OE-Core rev: 3e890dd00ed27cc30800caefa85827cb16da5a3a)

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

diffstat: patches have now been submitted upstream

(From OE-Core rev: 6433f5d68eba1ce7306fbfb19265ea6786715d7c)

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

diffstat: fix autoheader warning due to missing AC_DEFINE template

(From OE-Core rev: 9c9da40b6de2cd29378e56fef643305872a52f62)

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

socat: upgrade to 1.7.2.4

socat manually maintains config.h.in instead of using autoheader, so exclude
autoheader from autoreconf and remove all patches that are intended to enable
use of autoheader.

The license checksum changed, because the company address changed.

Based on a upgrade by Hongxu Jia <hongxu.jia@windriver.com>, with cleanup after
noticing that config.h.in was hand-maintained.

(From OE-Core rev: ea3ec30c19af23f6b62ce3d4d9d42c1fcb23a215)

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

lrzsz: fix autoheader warning due to missing AC_DEFINE template

(From OE-Core rev: 5eb90ef508b041dcae5a16b9696d3688659b8f39)

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

bitbake: siggen: Print warning about tainted tasks

The big warning printed when people use -f is easily ignored/forgotten.
To raise user awareness, print a warning any time we include a tainted
stamp file into a build instead.

(Bitbake rev: 18f9bcbad059608e22fca20309314e1c399acec7)

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

bitbake: build/utils: Fix broken exception handling

Checking for explicit exception names is bad, we also want to be able top
rely on inheritance. Fix these checks to be part of the real except clauses
so SkipPackage is recognised as being inherited from SkipRecipe.

(Bitbake rev: b131229145e1f2c372d6230a7b554e436c13c3f9)

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

bitbake: bb.fatal: Raise a BBHandledException instead of exiting

With new bitbake UIs having the cooker exit at 'random' points
in the codebase is problematic. This patch raises an exception
which matches the situation instead.

(Bitbake rev: 181a9735d02ebd517378558e909efc8b1b118973)

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

bitbake: event: Add SkipRecipe event to replace SkipPackage

In the depths of time we were rather confused about naming. bb files
are recipes, the event to skip parsing them should be SkipRecipe,
not SkipPackage. This changes bitbake to use the better name but
leaves the other around for now. We can therefore start removing
references to it from the metadata.

(Bitbake rev: 98d9e6e0f514a7cb7da1d99bf4bd5602b89426d6)

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

bitbake: event.py: Allow passthrough of BBHandledException events

We need BBHandledException events to be passed through to the higher
layers, they don't need addition of any traceback since they've already
been reported to the user.

(Bitbake rev: f63b61f8997862439519f474cc09f3e05e95288c)

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

bitbake: cookerdata: Improve error handling

If we see errors during parsing, firstly its bad to show a traceback
for an expansion error so lets suppress this.

Secondly, raise a BBHandledException instead of a SystemExit to show
we've informed the user about the condition (printing a traceback in the
default unknown case).

(Bitbake rev: e01988d9a1b7c40e31161c6ce7b85c4405671068)

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

bitbake: event: Fix event handlers to raise SkipPackage

If an event handler triggers a SkipPackage event, we really want that
event to be received and processed by the higher code levels. Currently
it was getting caught and ignored which was leading to recipes
being present when they clearly shouldn't have been.

In general this exception catching looks to be doing the wrong thing. It
was introduced in
http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/bitbake/lib/bb/event.py?id=37cb4cc02b2e2b6c338c5943747e0a1ef15176b3
but we likely want exceptions to pass through to the higher layers.

This patch therefore removes that code.

(Bitbake rev: 79211b3434855942f9fad4a1db69ce7be911327c)

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

util-linux: add sulogin to alternatives list

sulogin is provided by both busybox in /sbin/sulogin and util-linux provides one
in /usr/sbin/sulogin, so move util-linux's to sbin and setup ALTERNATIVE_LINK.

[YOCTO #6384]

(From OE-Core rev: a827ca74d27b916df8d40ec8155bc0340f8a0487)

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

systemd: Upgrade 212 -> 213

boots on all qemu machines

(From OE-Core rev: 29022a6a3a44089a14048e685c57d65382709094)

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

libogg: upgrade to 1.3.2

- Switched to BP variable.

(From OE-Core rev: 0697cf110a847ebb14809e92c7b98698026da8f7)

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

weston: make lcms explicitly configurable

Per Ross's comment, make --enable-lcms deterministic.
Follow upstream style to make it more likely to be merged upstream.

(From OE-Core rev: e2f45c6284f0a26cc858f9ae5887a4c1ef844d96)

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

weston: fix make-lcms-configurable.patch

In AS_IF([test "x$enable_lcms" != "no"],
"no" should have been "xno"

(From OE-Core rev: 1942c417d94e0b4e39613b15a632482b557a51d0)

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

weston: drop superfluous --disable-lcms

Thank you to Henning Heinold for the catch.
"Is this needed when you use PACKAGECONFIG later on?"

(From OE-Core rev: 815341fd1232e7739650497f94d851af41f6af79)

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

gzip: Avoid host contamination during build

On systems with /bin and /usr/bin merged the gzip recipe may find grep
in /usr/bin, while OE puts it in /bin. Force the recipe to find grep in
the correct place.

(From OE-Core rev: 8ae714bb59c67433ded1efe6a8750e0e9c126dab)

Signed-off-by: Dan McGregor <dan.mcgregor@usask.ca>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake.conf: Set a dafault value for TUNE_PKGARCH

If we don't do this, we see an exception:

ERROR: Failure expanding variable MACHINE_ARCH, expression was ${@[d.getVar('TUNE_PKGARCH', True),
d.getVar('MACHINE', True)][bool(d.getVar('MACHINE', True))].replace('-', '_')} which triggered
exception AttributeError: 'NoneType' object has no attribute 'replace'

Setting a default value avoids this error and allows the sanity checker
to trigger instead.

(From OE-Core rev: 106e9a3f594658b6a207f1f29bd4007616cc31d6)

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

gcc: Clean up configure_prepend and fix for mingw

The do_configure_prepend was duplicated in gcc-4.X.inc and
gcc-configure-common.inc leading to confusion when reading the resulting
do_configure task where the file was processed twice.

The only difference was the removal of the include line for gcc 4.8/4.9.

On mingw were were seeing two issues, firstly that the if statements meant
the values we wanted weren't being set, the second that the include
paths were still wrong as there was no header path set.

To fix the first issue, the #ifdef conditionals were removed, we want
to set these things unconditionally. The second issue is addressed by
setting the NATIVE_SYSTEM_HEADER_DIR variable here (it was already
set in t-oe).

(From OE-Core rev: db44be06c75f2ac17a55dd1764471e869e872b8b)

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

classextend: Fix crosssdk remapping for multilib

Multilib builds only require one crosssdk toolchain. We therefore shouldn't
be remapping crosssdk names. This resolves build failures looking for
weird multilib crosssdk toolchains.

(From OE-Core rev: aa8b93e2db06866529d20939452f81fb9e18aaab)

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

opkg-utils: Update SRCREV

Recent changes in opkg-utils allow package files to be stored in a different
directory to the package index if desired.

(From OE-Core rev: 237b9700d449de03a3f5dc524c15709f46941cf9)

Signed-off-by: Paul Barker <paul@paulbarker.me.uk>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

combo-layer: fix revlist taking into account file_filter

If file_filter is set, git format-patch takes account but git rev-list
does not. So revlist is going to get with wrong revisions. And last_revision
will be updated with wrong revision. The next time that user run
combo-layer it complain about applying patches.

So ensure that 'git rev-list' are using file_filter as 'git format-patch'.

(From OE-Core rev: c22e40900d8e33c2c884e714c11ddb771b86923f)

Signed-off-by: João Henrique Ferreira de Freitas <joaohf@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

update-rc.d: fix to handle priority numbers correctly

The update-rc.d treated priority numbers begin with '0' as octal
numbers. This led to problems of update-rc.d being unable to handle
priorities like '08' or '09' correctly.

This patch fixes the above problem.

[YOCTO #6376]

(From OE-Core rev: 675fd834b9ed696cd87809830d57d3da083580d3)

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

eglibc: Fix configure when re-building with gold

(From OE-Core rev: 2a36a02c2444551579b8498468debfaff0faa6d8)

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

gcc, uclibc: Add/Fix Upstream-Status in patches

(From OE-Core rev: 68a0e34260f884f6fb39aae2d0bad035b2b1d177)

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

layer.conf: Add musl to safe recipes for exclusion

(From OE-Core rev: ad8ccccf442e29a0b733753c1951f402baa330a9)

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

gnu-config: Let it recognise *-*-musl* triplets

This will help autotools based packages to recognise
musl

(From OE-Core rev: a51f790bd657011d871aab603d1695937bfa2033)

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

cross-canadian.bbclass: Recognise musl

(From OE-Core rev: 66fd622058f690dbb291a648ec1583191bf44df5)

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

image.bbclass: Remove dependency on ldconfig-native for musl

it does not grok glibc ldconfig format

(From OE-Core rev: 9c85aef3ce25f6eb1d370a1a94e3fe16d59ec627)

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

multilib_header: Do not install mutlilib headers for musl

musl is not multilib and this creates trouble. eg. when
util-linux probes for ncurses it does not find it because
ncurses has installed the multilibbed header and this
header includes bits/wordsize.h and this header does not
exist on musl systems. If and when musl adds multilib
support we will revisit it.

(From OE-Core rev: dad1c2746326912db41a3ff180679cdfe0e844f9)

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

tclibc-musl.inc: Add musl specific distro policy file

With this we could use TCLIBC=musl to switch to images
based on musl

(From OE-Core rev: 797ef28c55a30f1b465ce512fffa4e06c7f1c658)

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

siteinfo, insane: Recognize musl specific triplets

We will use '-musl' to identify musl based systems
this patch lays the foundation for recognising those
and map them to internal variable representations

(From OE-Core rev: 9cd77aed67373e33dc69158ab02b94d7045c1119)

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

bitbake: utils: avoid printing traceback on ExpansionError during parsing

If an ExpansionError occurs during better_exec() we should just raise it
instead of printing the traceback, so that recipe errors (such as broken
URLs in SRC_URI) are more easily comprehensible.

(Bitbake rev: 5b0da8932c318813138c113d2bb20498145dbd42)

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

bitbake: fetch2: improve handling of two classes of URL parameter mistakes

Handle the following situations in a URL (e.g. in SRC_URI):

* Trailing semicolon in a URL - this is now ignored.
* Parameter specified with no value (no equals sign). This still
  produces an error, but at least it is MalformedUrl with a proper
  message rather than "ValueError: need more than 1 value to unpack".

(Bitbake rev: bfd13dfbc4c9f1dd8315002271791b1d9e274989)

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

bitbake: Revert "toaster: toaster oe-selftest support"

This reverts commit bb5b1d6b139b886e54bfdc0c17f2b556db6a7fde.

Applied to incorrect repo.

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

toaster: toaster oe-selftest support

This patch adds toaster tests using the oe-selftest infrastructure.
You need to have builds done - the tests will verify data integrity
after the toaster collection phase.

Once you have your toaster builds done, to run the automated backend
tests via oe-selftest do the followings:

1. Update builddir/conf/bblayers.conf to contain the meta-selftest
layer
2. From the builddir run:

'oe-selftest toaster'

or if you just want to run a single test:

'oe-selftest toaster.Toaster_DB_Tests.testname'

This first part adds the meta/lib/oeqa toaster file.

(From OE-Core rev: 762d425ed6f6d9046d3e3230c44b42ea6173b447)

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

Update tzcode to 2014c

(From OE-Core rev: 3fa9508521d27e17bfe1a0aeb15d7fc2377218cd)

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

Update tzdata to 2014c

Removed solar-time experiment  as per
http://mm.icann.org/pipermail/tz/2013-November/020488.html

(From OE-Core rev: 57af3fb9662106f0a65a1b4edf83e2398be0a8f1)

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

ethtool: use serial-tests config needed by ptest.

buildtest-TESTS and runtest-TESTS targets are required by ptest.
In order to have those targets in automake 1.13.4 serial-tests
should be specified since parallel-tests is assumed by default
and serial-tests is optional.

ptest results:
PASS: test-cmdline
PASS: test-features
==================
All 2 tests passed
==================

(From OE-Core rev: 15bdef1f25ef567caf2f2e270de899e35da7cca9)

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

sstatesig: Move saferecipedeps handling to be earlier

We want to use the saferecipedeps handling code to allow gcc-cross-* to
work on multiple different tunes. Its currently in target only code
so it needs to be earlier to allow it to work on native-> target
dependencies.

This change has no effect on existing uses but makes gcc-cross become
shared as desired.

(From OE-Core rev: 9e03db2dfab0b534b86fd48c9190b2d7d0d21238)

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>

openssl: add openssl-CVE-2010-5298.patch SRC_URI

make openssl-CVE-2010-5298.patch truely work

(From OE-Core rev: eab33442480cc27a5cd00b3f46984fea74b7c0f9)

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>

readline: use upstream's aclocal.m4 as acinclude.m4

Instead of shipping a fork of the upstream aclocal.m4, simply rename it to
acinclude.m4 at configure time.  We don't need the fork now that autoheader is
excluded.

(From OE-Core rev: e531923c4c17becb2f1a8a89adfeff0a82961a4a)

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

readline: exclude autoheader from autoreconf

readline ships a hand-maintained config.h, instead of letting autoheader
generate one from configure.ac.  The required arguments to AC_DEFINE are not in
configure.ac so autoheader will produce warnings and the generated code will not
behave as expected.

Solve this by excluding autoheader from autoreconf, so the upstream config.h.in
is used.

(From OE-Core rev: 8c37d32d6133c6ad2b9142e7a42775e7a979b570)

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

readline: add missing STRUCT_DIRENT_D_* symbols to config.h.in

readline maintains config.h.in by hand but several symbols are incorrect.  Fix
these so that the test results are reflected in config.h.

(From OE-Core rev: bc0d0c71eca48be05490209261b88b1f92bcf847)

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

tcmode-default.inc: Default to using gcc 4.9

(From OE-Core rev: 050dbf916b7da792be0f9ca2ee7895ceb397fbce)

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

gcc: add patch to fix errors with Decimal64 type

[OE-core bug #6270] - https://bugzilla.yoctoproject.org/show_bug.cgi?id=6270

(From OE-Core rev: 8f8ef80131d4aa62a4b106d365a5e7b6273c766d)

Signed-off-by: Alexandru-Cezar Sardan <alexandru.sardan-KZfg59tc24xl57MIdRCFDg@public.gmane.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

i2c-tools: Add i2c-tools to the core

i2c-tools has been sitting outside of oe-core for long enough now. It is
a required tool for board validation, and many people are pulling it
into their builds and their own layers. Let's add it to the core.

This patch includes the i2c-tools recipe from meta-oe as of:

  commit 9df13b4140e8c6bfa0e4fb89107a6146981d2cdc
  Author: Khem Raj <raj.khem@gmail.com>
  Date:   2014-04-26

      i2c-tools: Fix build when S != B

(From OE-Core rev: 32ac58819580d359e22161be1abf62215d202250)

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Cc: Richard Purdie <richard.purdie@linuxfoundation.org>
Cc: Khem Raj <raj.khem@gmail.com>
Cc: Martin Jansa <Martin.Jansa@gmail.com>
Cc: Matthieu Crapet <Matthieu.Crapet@ingenico.com>
Cc: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-yocto-dev: Dynamic SRCREV update

The current implementation would result in the default SRCREVs being
used by the fetcher, even though the anonymous python would update them
to AUTOREV. This appears to be something to do with early parsing
bitbake black magic.

This patch ensures the default is never assigned if we are actually
building the recipe by using a function to assign it in the first place.

The USE_DEFAULT* variables are removed as they are not necessary to
allow for overriding the SRCREVs.

The anonymous python parse check is moved closer to the top of the
recipe to be a bit more logically representative of its intended
purpose.

(From OE-Core rev: a0334b0de654a41c53df54ef80625094368113f6)

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-yocto/3.10: bump kver to v3.10.40

Integrating the latest korg releases for the 3.10 kernel.

(From OE-Core rev: 574c03bd5fd73281472f8267a31cfecb235f1c65)

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

beaglebone: enable the nowayout option for the watchdog

Bumping the meta SRCREV for the following fix:

[
  The default watchdog behaviour is to stop the timer if the process
  managing it closes the file /dev/watchdog. The system would not reboot
  if watchdog daemon crashes due to a bug in it or get killed by other
  malicious code. So we prefer to enable nowayout option for the
  watchdong. With this enabled, there is no way of disabling the watchdog
  once it has been started. This option is also enabled in the predecessor
  of this BSP (beagleboard)
]

[YOCTO: 3937]

(From OE-Core rev: 7006412c285a4a6c75d5349f60dc71b0b735ff90)

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

linux-yocto-rt/3.14: update to 3.14-rt5

Updating the the latest 3.14-rt release.

(From OE-Core rev: ca1d952c964ce25bf78d47c7a856105d59d72cac)

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

linux-yocto/3.14: update to v3.14.4

Bumping the 3.14 recipes to the latest korg -stable release.

(From OE-Core rev: 5c0088767a59c63d2197b54450a54578fa10fa07)

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

bitbake: fetch2/hg: Fix missing proto param for hg checkout with user and pw

A fix for the former patch when checking out a repository with
username and password using HG

(Bitbake rev: 0e7b594ccbceb3149f38776cea204807031ef69f)

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

bitbake: bitbake-layers: show-cross-depends: add support for RRECOMMENDS

RRECOMMENDS must be satisfied at build time, and these could cross layer
boundaries, so report these if they exist.

(Bitbake rev: 5569b3dca61e6d962494ca65c7aad09b2eb2ae63)

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

bitbake: bitbake-layers: show-cross-depends: ignore self-satisfied RDEPENDS

Overlayed recipes caused this to show false positives because the
overlaying version appeared to be satisfying the overlayed version's
RDEPENDS; but you'd never be building both at the same time.

(Bitbake rev: b94318174fe7f92b9a20eabb0bc4055066cb3d51)

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

bitbake: bitbake-layers: show-cross-depends: ignore global inherits

It's not particularly useful to show globally inherited classes here
since they do not normally represent a dependency.

(Bitbake rev: d16948bb88fcf44d861985838030be7c08697963)

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

bitbake: bitbake-layers: show-cross-depends: add option to ignore a layer

By default, show-cross-depends shows dependencies on OE-Core (i.e.
"meta") which is not particularly useful. Add an option to allow you to
hide those. For example, to hide all dependencies on OE-Core:

  bitbake-layers show-cross-depends -i meta

Multiple layers can be specified by using commas as separators (no
spaces).

(Bitbake rev: 0e9062e65acbb05c1d9b3a9145eb866c3d562309)

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

lighthttp: Use pkg-config for pcre dependency

(From OE-Core rev: d2457880e7bb08b9c2f8d60e70b1d59ed84e9da9)

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

libarchive: Use pkg-config for libxml2 dependency

(From OE-Core rev: fe277bf0a61d5d7787dba699ee1ed4d979ba5cff)

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

libksba: Use pkg-config for dependencies

Use pkg-config in the m4 macros for the package, ensure we have a host
field in the .pc file.

(From OE-Core rev: 4a971a90988435902a4a8dd9c721d440cd80c0bd)

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

gpgme: Update to ensure we pkg-config for dependencies

(From OE-Core rev: ffffc627b21a3cf8b407d16a437793b5fddf7127)

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