oprofile: fix cross compile on powerpc and add libpfm4 dep

Fixes [YOCTO #3717]

(From OE-Core rev: 88959e42411a40b15fe8907da00a97a7732a9cc1)

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

libpfm4_4.3.0.bb: add libpfm4 recipe needed by oprofile for ppc

(From OE-Core rev: 52c94978992d9084d685d4999064a8cee35220d1)

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

quota_4.01.bb: add e2fsprogs in DEPENDS

Fixes:

| /local/home/mattsm/git/poky/build-master/tmp/sysroots/x86_64-linux/usr/libexec/ppce300c3-poky-linux/gcc/powerpc-poky-linux/4.7.2/ld: cannot find -lext2fs

(From OE-Core rev: 514c479816da79bb2f71602c3bf3c4805a4416e6)

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

glib: disable selinux for native builds

This improves reusabiliy of sstate-cache across different hosts

(From OE-Core rev: 4c223e2b2ba552b832b51c9071f003de67493c27)

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

pulseaudio: do not postpone postinstall

Since populate-volatile.sh will run everytime the device boots, no
need to postpone the postinstall when the rootfs is created.

[YOCTO #3840]

(From OE-Core rev: c237103096530a06fd0991b4335936e509dea76e)

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_ipk, rootfs_ipk: remove the "set -x"

Comment the "set -x" lines since these will add a lot of extra, not
always necessary, debug messages in the log.do_rootfs.

[YOCTO #2599]

(From OE-Core rev: 5d8119f3d749073e355351e0e15f2703fad738f2)

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>

liberation-fonts: use the new fontcache.bbclass

This will add the proper postinst/postrm scriptlets.

[YOCTO #2599]

(From OE-Core rev: 94b53e5f996a09b9f1f73d0f90b65261a76c0cbe)

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>

fontconfig: add sysroot option to fc-cache and fc-cat

This is needed in order to be able to generate the cache on host.

Additionally, remove the volatile config file, as /var/cache was moved
out of tmpfs.

[YOCTO #2599]

(From OE-Core rev: b675e9917b0a1e774c95ee7a946f515c5a996b59)

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: add a proper error message if hook script fails

(From OE-Core rev: 5e737d3c6e6546c1368e804f4c45ab25d8791ea3)

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>

qemu.bbclass: return qemuwrapper instead of qemu-allarch

When qemu bbclass is inherited from a recipe that is not architecture
dependent, qemu_run_binary will return "qemu-allarch". However this
binary does not exist. Instead, return "qemuwrapper" which will, in
turn, execute the right binary for the target the image was built for.

[YOCTO #2599]

(From OE-Core rev: 149a564bba7d3e1c2054ae6d908835ebd95b9084)

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>

add qemuwrapper-cross recipe

This will just install a wrapper script in STAGING_BINDIR_CROSS that
will execute the proper qemu user binary for the current target.

[YOCTO #2599]

(From OE-Core rev: faaa5e7fd4353b73289f163d9f601cf0869698f3)

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>

add fontcache.bbclass

All font packages should inherit this class in order to generate the
proper postinst/postrm scriptlets.

The scriptlets will actually create a host intercept hook that will be
executed at the end, at do_rootfs time, after all packages have been
installed. This is good when there are many font packages.

[YOCTO #2923]

(From OE-Core rev: 0c12f7fb3c2c42e5b633682bb1277b943ac19ea6)

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>

tcp-wrappers: remove size_t.patch

1. it introduces bug in 64bit big endian process with __GLIBC__, At that
condition, size_t is 8byte, and the third parameter of getpeername is
socklen_t which is 4 byte. As a result, getpeername sees third parameter
is always 0, and can not return right value.

The similar program is below, the output is 0, not 9 on PPC64 cpu
main()
{
long aa=9;
printf("%d \n", *((int *)&aa));
}

2. The correct fix is to change getpeername/getsockopt/recvfrom.. last
parameter type from int to socklen_t, but to simplify, we can remove
size_t.patch, since the size of int is same as socklen_t in 32bit/64bit
cpu. and size_t.patch only change three places, there are other places
which uses int, and work well.

2. Fedora, redhat el4 do not use this patch, but Debian uses it, does not
find why this patch is written, maybe it is gcc legency issue which does
not exist.

(From OE-Core rev: 6c418ec278335c93692c2e19ec0b7b84b471e2b9)

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>

rpcbind: replace obsolete automake macros with working ones

Add obsolete_automake_macros.patch that replaces automake macros
no longer supported by automake-1.13 with modern constructs.

(From OE-Core rev: 0f8583e6fe729f8cacc8fec8c66a5c7f7c944947)

Signed-off-by: Marko Lindqvist <cazfi74@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

matchbox-wm: update to git master head

The one new commit from matchbox-wm git this gets fixes
build with automake-1.13

(From OE-Core rev: 50d7135c3f5530e0852294183cdba60fae67e040)

Signed-off-by: Marko Lindqvist <cazfi74@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

opkg: replace obsolete automake macros with working ones

Add obsolete_automake_macros.patch that replaces automake macros
no longer supported by automake-1.13 with modern constructs.

(From OE-Core rev: 495bea3911be164225c91b696389fc16dab356fd)

Signed-off-by: Marko Lindqvist <cazfi74@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

send-pull-request: add extra CC argument

* useful e.g. when sending pull-request to release branch with extra CC
  for release maintainer

(From OE-Core rev: 52bc47756eb8a81ea07ef4bc06345ef335b30ceb)

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

ipk: use OPKGLIBDIR in all places

* it's not recommended to change this value, because it breaks upgrade path on target
  (all old u-a alternatives are forgot in old OPKGLIBDIR value
* but make it consistent, so if someone really want to change that, then
  setting OPKGLIBDIR_distro in distro.conf would be enough
* without this there were at least 4 places to change:
  rootfs_ipk: opkglibdir variable (notice that I've removed /opkg from it to correspond
              with EXTRA_OECONF option used in opkg recipes
  package_ipk: ${target_rootfs}${localstatedir}/lib/opkg/ hardcoded in
               package_install_internal_ipk
  opkg-collateral: value in lists file
  opkg: EXTRA_OECONF for all 3 classes, FILES_libopkg, do_install
* validated with buildhistory that without OPKGLIBDIR explicitly set the
  output is the same and that after setting
  OPKGLIBDIR_forcevariable := "${libdir}"
  everything including empty directory from package_ipk is moved to
  libdir

(From OE-Core rev: cf0aa9c4fdae8855803e96b1922d54a2431795d3)

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

net-tools: upgrade to v1.60-24.2

(From OE-Core rev: 8abcf2c3541d7fe96d7717cb161812784dca6c66)

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>

qmmp: update to 0.6.6 and fix

* Update to 0.6.6
* Fix compilation failure due to unconditional call to
  QApplication::commitData()
* Disable /usr/local host paths to avoid compilation warnings
* Force use of the cmake configure instead of qmake as the latter is not
  dynamic with Qmmp and doesn't let you disable certain dependencies
  (e.g. enca)
* Add libsndfile1, libsamplerate0 and curl to DEPENDS since their use
  will be non-deterministic otherwise
* Explicitly disable library-requiring options that we don't have
  available in OE-Core so that they don't get used if they happen to be
  in the sysroot
* Update LIC_FILES_CHKSUM since there were cosmetic changes to the
  license file (mostly reformatting and change of references to LGPL
  from "Library" to "Lesser".)

Fixes [YOCTO #3822].

(From OE-Core rev: 12484dca3bf09dd9a03442a223885deb7472a6cd)

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>

nativesdk-qt4-tools: fix DEPENDS, as nativesdk is now prefixed

(From OE-Core rev: 871549996cfcaf860083d867eb6a9522f4407cd4)

Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake: fetch2: Print the complete SRCREV variable name when INVALID

If a particular SRCREV (say for a particular branch) is missing, the
fetcher will currently just report an obtuse error about the "SRCREV"
being invalid. If there is more information is to be had (say from name,
i.e. branch, and pn) then display that as well.

The new error looks something like this:

ERROR: ExpansionError during parsing /home/dvhart/source/poky/meta/recipes-kernel/linux/linux-yocto_3.4.bb: Failure expanding variable do_patch: ExpansionError: Failure expanding variable SRCPV, expression was ${@bb.fetch2.get_srcrev(d)} which triggered exception FetchError: Fetcher failure for URL: 'git://otcgit.jf.intel.com/dvhart/linux-yocto-minnow-3.4.git;protocol=git;nocheckout=1;branch=standard/minnow,meta,emgd-1.14;name=machine,meta,emgd'. Please set SRCREV_emgd_pn-linux-yocto to a valid value

Note the variable listed as invalid is
"SRCREV_emgd_pn-linux-yocto", making it explicit what is wrong.

(Bitbake rev: 63774f5b4edb999300bddd891233f6050f4af877)

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

crosstap: handle hyphenated x86_64 target arch

systemtap_target_arch() should also translate x86-64 (hyphenated) into
x86_64 for the -a param.  Failing to do that causes systemtap to see
an architecture mismatch and create a cloned session with a bogusly
synthesized build directory path, and fails to compile the probe.

 Fixes [YOCTO #3756]

(From OE-Core rev: 98cae0544884cb5700d42409ec4a9584a17dc9a4)

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libgcc: Disable multilib install for nativesdk

Fixes errors when building SDK for multilibbed enabled
architectures.

[YOCTO #3832]

(From OE-Core rev: 09934f38df057e12af7d14791f7ab752e81093db)

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

dropbear: fix RPROVIDES

(From OE-Core rev: 3b1beb8b15b8e3e397b3aa8320490103d4f29bac)

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

openssh: fix RPROVIDES

(From OE-Core rev: 43ddc955727361458fec92731775a1fc7f29917c)

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

sip.bbclass: improve RDEPENDS to add python-sip only for PN

(From OE-Core rev: 9aeb36d452154a766c0ada6d4594dcabc226752f)

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

v86d: Use BP instead of P

P is expanded to it's multi and other prefix / suffix name,
so use BP instead which is the BaseName and Version.

(From OE-Core rev: 4e475a66c23cd2e4e109ff1fcfa2975d595537c5)

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

blktool: fix SRC_URI typo

(From OE-Core rev: 393c673964493f9dcc15dd6dc394b0e0f6994afd)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gcc: target: fix libiberty removal

- use ${libdir} instead of ${exec_prefix}/lib

Fix the following warning:
WARNING: QA Issue: gcc: Files/directories were installed but not shipped
  /usr/lib64/libiberty.a

(From OE-Core rev: b3643415ad91dc77880cc5b95e9ad8cd9aef5c44)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

consolekit: add arch independent unpackaged files

Fix the following warning:

WARNING: QA Issue: consolekit: Files/directories were installed but not shipped
  /usr/lib
  /usr/lib/ConsoleKit
  /usr/lib/ConsoleKit/run-seat.d
  /usr/lib/ConsoleKit/scripts
  /usr/lib/ConsoleKit/run-session.d
  /usr/lib/ConsoleKit/scripts/ck-system-restart
  /usr/lib/ConsoleKit/scripts/ck-system-stop

(From OE-Core rev: 98ef35b319320ac9530e31926ad18d5f5ecd87db)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

run-postinsts: use BPN instead of PN

Fix the following warning:
WARNING: QA Issue: lib32-run-postinsts: Files/directories were installed but not shipped
  /usr/share/lib32-run-postinsts
  /usr/share/lib32-run-postinsts/run-postinsts.awk

[YOCTO #3438]

(From OE-Core rev: fb42fae0f6eb9821b5f1fedfaebf4307dc6590fe)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

taglib: use LIB_SUFFIX in order to determine the correct baselib

WARNING: QA Issue: taglib: Files/directories were installed but not shipped
  /usr/lib
  /usr/lib/libtag.so
  /usr/lib/libtag.so.1.12.0
  /usr/lib/libtag.so.1
  /usr/lib/libtag_c.so.0.0.0
  /usr/lib/libtag_c.so
  /usr/lib/libtag_c.so.0
  ...

(From OE-Core rev: ebc3e1f43a558165d16f663be796d731ca4b2a74)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

foomatic-filters: fix CUPS directory config

- use CUPS_SERVERBIN to configure the CUPS directory
- the removed variables are not used anymore

Fix the following warning:
WARNING: QA Issue: foomatic-filters: Files/directories were installed but not shipped
  /usr/lib64
  /usr/lib64/cups
  /usr/lib64/cups/filter
  /usr/lib64/cups/backend
  /usr/lib64/cups/filter/foomatic-rip
  /usr/lib64/cups/backend/beh

(From OE-Core rev: 141621d956ea66a026b3571c3e0d30f1ab43961e)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

syslinux: use BPN instead of PN

Fix the following warning:
WARNING: QA Issue: lib32-syslinux: Files/directories were installed but not shipped
  /usr/lib
  /usr/share/syslinux
  /usr/share/syslinux/com32
  /usr/share/syslinux/com32/libcom32gpl.a
  /usr/share/syslinux/com32/libcom32.a
  ...

[YOCTO #3438]

(From OE-Core rev: 0d014d0f42de4af76226799b04c8a2daa52f787e)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sgml-common: use ${PN} instead of sgml-common

Fix the following warning:
WARNING: QA Issue: lib32-sgml-common: Files/directories were installed but not shipped
  /etc
  /etc/sgml
  /etc/sgml/sgml.conf
  /usr/bin
  /usr/bin/install-catalog
  /usr/bin/sgmlwhich

[YOCTO #3438]

(From OE-Core rev: 1a43fba81749618f9f1c18b99cb74ae1399bdc35)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

console-tools: fix linking when using the multilib compiler

fix-libconsole-link.patch: add patch for fixing libconsole
linking problems when using multilib gcc

(From OE-Core rev: f70371a7c2da892a480a73d8571497dd7b367c7b)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

enchant: use BPN instead of PN

- use BPN (PN includes mlprefix) to package all files
when building with multilib options

[YOCTO #3438]

(From OE-Core rev: 6039e0048d324569620868ec774cd88aa191eeaf)

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

cpan-base.bbclass: Improve RDEPENDS to be package specific

(From OE-Core rev: 3300d5fdcc60b14e184e0c6c40ba25d8c4d5ed46)

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

initramfsframework: Improve RDEPENDS to be package specific

(From OE-Core rev: 8c9a604d3ef33d47a48000d6c38159a64204e81e)

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

dbus-ptest: Improve RDEPENDS to be package specific

(From OE-Core rev: 962dc38c78ce539c74f90811a022995d15d76ffc)

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

hwlatdetect: Improve RDEPENDS to be package specific

(From OE-Core rev: 1207bb402adfbe6a0600e4540fc35a53282a857c)

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

libxml-sax-base-perl: Improve RDEPENDS to be package specific

(From OE-Core rev: b72f277a8952649f23a3e476a10bf1e1706918be)

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

initramfs-live-boot: Set RDEPENDS on the specific package that needs it

Set the RDEPENDS on the specific package that has the dependency and stop it
being applied to for example ${PN}-doc (and others).

(From OE-Core rev: 7437a864f03ff56a4fba9d8ce9baf845b945ed9e)

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

gdb-cross-canadian: Set RDEPENDS on the specific package that needs it

Set the RDEPENDS on the specific package that has the dependency and stop it
being applied to for example ${PN}-doc (and others).

(From OE-Core rev: 51257c8665282e2b7f647adb4bdf8d07e2b40e1c)

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

qemu: Set RDEPENDS on the specific package that needs it

Set the RDEPENDS on the specific package that has the dependency and stop it
being applied to for example ${PN}-doc (and others).

(From OE-Core rev: 3c57a755ff1aec3806770443b73dc899d981c678)

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

v86d: Update SRC_URI to point to new file location.

The existing SRC_URI pointed to a Gentoo developer's private web
space.  It appears that that developer has retired and that his web
space is no longer active.  I've updated the SRC_URI to point to a
location where the file can now be found.

(From OE-Core rev: 2046c2a0922c4d67408578d4bf8549435fec8cd6)

Signed-off-by: Andy Dalton <a.spam.filter@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

dpkg: Add missing pkgconfig dependency

configure touches pkg-config for various tests so we need the DEPENDS
which we can gain from the class inherit

(From OE-Core rev: 2602575108a39723f9975391e83290573cbd2ec9)

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

scripts/wipe-sysroot: add script to safely wipe the sysroots

Add a script to wipe the sysroots and all of the relevant stamps, so that it
will be correctly re-populated.

(From OE-Core rev: ef98ff5ba562eb710b5a6fbd181fb1c4380010b2)

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>

package.bbclass: Allow per-package SKIP_FILEDEPS

The existing check for SKIP_FILEDEPS can be overridden per recipe
using SKIP_FILEDEPS_pn-${PN}. However, there's no mechanism for
letting a single package within a recipe use SKIP_FILEDEPS.

This patch adds SKIP_FILEDEPS_<pkg>, by analogy to FILES_<pkg>.
Note that it only works one way; if the recipe has SKIP_FILEDEPS = 1,
the checks for individual packages will never be reached.

(From OE-Core rev: 94557b500ad38a49aec40629015ed0b24e167f76)

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

kernel-yocto/linux-yocto-custom: support low bandwidth options

To support configurations where active development is not being done within
the oe/bitbake build environment and restricted bandwidth situations, this
commit allows the SRC_URI to point to a kernel tgz instead of a full git
repository.

Outside of the upstream tgz instead of a kernel git repository, the
restrictions, config and patch process is the same as any linux-yocto-custom
recipe.

An example linux-yocto-custom based recipe would have a configuration like
this to build the 3.7 kernel, using an externally supplied config, from the
3.7 tgz:

  SRC_URI = "http://kernel.org/pub/linux/kernel/v3.0/linux-3.7.tar.bz2"
  PV = "3.7"
  S = "${WORKDIR}/linux-3.7"
  SRC_URI[md5sum] = "5323f3faadd051e83af605a63be5ea2e"
  SRC_URI[sha256sum] = "dc08d87a579fe2918362e6666e503a95a76296419195cb499aa9dd4dbe171a9e"

[YOCTO #2686]

(From OE-Core rev: 08b3a282ce75a9972694f0c4379179505b9ec91f)

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: allow multiple / shared kernel feature directories

To promote the reuse and sharing of configuration fragments this change
allows any kernel-yocto based recipe to have multiple alternate git repositories
which provide kernel feature directory trees listed on the SRC_URI.

These feature directories are in addition to any in-tree kernel meta data branches
that may be available (described via the KMETA variable in linux-yocto recipes).

Features found within these directories can be used from recipes via the
KERNEL_FEATURES variable. Features found within a feature directory are free
to include any other features that are available in any directories. In both
cases the path to a feature description (a .scc file) is relative to the
root of a given feature directory (which is how existing .scc files work)

The search order for features is determined by the order that repositories
appear on the SRC_URI.

Normal SRC_URI rules apply to any repository that is added as a kernel
feature container. A SRCREV must be supplied and it must be unpacked to
a unique directory, which is controlled via the "destsuffic" url parameter.

In addition to these standard requirements, any kernel feature repository
reference should identify itself via the "type=kmeta" url parameter. If
type=kmeta is not supplied, the repository will not be processed for
kernel features.

As an example, the following in a linux-yocto bbappend makes two additional
feature directories available to KERNEL_FEATURES and fragments.

 SRC_URI += "git://git.yoctoproject.org/yocto-kernel-cache;protocol=git;branch=master;type=kmeta;name=feat1;destsuffix=kernel-cache/"
 SRC_URI += "git://${KSRC_linux_yocto_3_4};protocol=file;branch=meta;name=feat2;type=kmeta;destsuffix=kernel-features-experimental/"

 SRCREV_feat1 = "${AUTOREV}"
 SRCREV_feat2 = "${AUTOREV}"

(From OE-Core rev: 02ad603a104b70ab74548c8018e738bfbb3c59db)

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: fix .scc and .cfg matching

SRC_URIs that contained git repositories or other constructs that resulted
in an extension of "." or a substring of "scc" or "cfg" were matching the
tests for patches and configs. This was due to a python tuple being used
instead of an array. Switching to an array makes the match exact and the
behaviour we want.

(From OE-Core rev: 22aa5d040604b37ba984bae9e800e56ba6e4956d)

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: make configuration audit details debug only

The details of the kernel configuration audit are typically a
debug action, so should be moved to bb.debug(). But in order
to maintain visibility of the results, a reference to the log
file is provided in the standard message.

(From OE-Core rev: 9ab80ad88d34622a81670cdc45cc3275fc3ebabe)

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>

guilt: remove GIT_EXEC_BASE export

git-core provides the following routines (among others): git-sh-setup,
cd_to_toplevel, die, and more.  But it is not always in the same location
relative to the guilt binary if git is not part of a 'host tools'
sysroot. Modern git versions don't need this, so commenting it out (until it
breaks again) is the solution to the problem of it not existing, and the
library routines will sourced and provided by git itself.

Since bitbake.conf has git-native as ASSUME_PROVIDED, this means that
when the system git binary is used, errors such as this can be seen in
failed patch logs:

| /opt/poky/build/tmp/sysroots/x86_64-linux/usr/bin/guilt: line 29: /opt/poky/build/tmp/sysroots/x86_64-linux/usr/bin/../libexec/git-core/git-sh-setup: No such file or directory
| /opt/poky/build/tmp/sysroots/x86_64-linux/usr/bin/guilt: line 717: cd_to_toplevel: command not found
| error: patch failed: Makefile:2
| error: Makefile: patch does not apply
| /opt/poky/build/tmp/sysroots/x86_64-linux/usr/bin/guilt-push: line 137: die: command not found
| [ERROR] unable to complete push
| pending patches are:
| /opt/poky/build/tmp/sysroots/x86_64-linux/usr/bin/guilt: line 29: /opt/poky/build/tmp/sysroots/x86_64-linux/usr/bin/../libexec/git-core/git-sh-setup: No such file or directory
| /opt/poky/build/tmp/sysroots/x86_64-linux/usr/bin/guilt: line 29: /opt/poky/build/tmp/sysroots/x86_64-linux/usr/bin/../libexec/git-core/git-sh-setup: No such file or directory
| links/linux-yocto-custom/0001-linux-version-tweak.patch
| ERROR. could not update git tree

With the export removed, we have a normal "clean" failure message when
the sysroot or system git is used.

(From OE-Core rev: 9ca697a7f83957205bc418acfd7f45fe4cbddbee)

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>

polkit: Enable systemd support

By default the polkit will autodetect if systemd is available
and setup the configuraton accordingly, potentially
contaminating sstate.
In this patch systemd is explicitly enabled/disabled.
When enabled, polkit will use systemd for session tarcking,
otherwise it will use ConsoleKit.

(From OE-Core rev: 67c13656343d55ae9d1c79dcace5c79004c4c186)

Signed-off-by: Radu Moisan <radu.moisan@intel.com>
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>

matchbox-stroke: remove

matchbox-stroke was never more than a proof of concept, so remove it.

(From OE-Core rev: f9491f2ed276d591b7a2c6eff2a7241f627db50c)

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>

clutter-box2d: remove

This is a specialised physics engine for Clutter, and isn't suitable for
oe-core.

(From OE-Core rev: 9eec14d658c12525cb81d0ef3227b55e65d1c69b)

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>

packagegroup-core-tools-testapps: remove clutter-box2d

(From OE-Core rev: 175af7282fc423f1f473b686b43a0d413adff5a3)

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>

pax-utils : upgrade to 0.6

 - changed the archive extension because it changed
 on the repository.

(From OE-Core rev: 7eb31a99925af3cc4ca9c322d1c988124cc812db)

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

orinoco-conf: remove

This is only used by old hardware, anything that still needs it can put it in
their BSP layer directly.

(From OE-Core rev: b8dc4fc76e3a179d2712ce76e19b664236260724)

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>

packagegroup-core-x11-sato: remove matchbox-stroke

matchbox-stroke was never more than a proof of concept, so let's not pretend
it's a serious gesture input method and remove it.

(From OE-Core rev: 465958f2fc6a633d697277aa06dd1c67d485f0d0)

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>

distcc: updated to version 3.1

Added two new flags to the configure script:

--disable-Werror: don't treat all warnings all errors (which breaks compilation).

PYTHON=/dev/null: this prevents distcc from detecting the host Python and trying
to build its include server using the host Python. This disables the include
server completely. If the include server is needed, that should be the object of
another patch (and would introduce a dependency on python for distcc).

The 'distcc-avahi' and 'makefile-param-order' patches are not needed anymore, as
they were merged upstream.

(From OE-Core rev: 3f33a6ecd9f1703381e175d688bdfce291ffdc8a)

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

initramds-framework: add shutdown hook for udev

In cases where other initramfs modules need to rely on
udev running (ie in my case I have to load firmware on
a device that is slow to start) there needs to be a way
to keep it running during the lifecycle of the initramfs
but still be shut down before swith_root is called.  I
added a  module_pre_hook that will shut down udev before
the finish module is called.

(From OE-Core rev: ce690659ef797bd26dc2be59167aa01744841510)

Signed-off-by: Ian Reinhart Geiser <igeiser@devonit.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pulseaudio_2.1: Remove the dependence of fftw.

(From OE-Core rev: ddbd713293929d361a7ec7f8f1eb2986b7adc729)

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

cmake: reset B from autotools, as this class doesnt like it

(From OE-Core rev: f8eae815186976f6111f30ae88ac33e723863982)

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

gdk-pixbuf: use correct build path when installing

The native install was invoking a binary in $S, but it's been built into $B.

(From OE-Core rev: b6ffded282d2b4aeb924db443169a9263aade014)

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

libpcap: fix relative path references

do_configure was using relative paths in do_configure with the assumption that
$S is the same as $B.  This isn't always true, so explicitly use $S.

(From OE-Core rev: a01c103f54d2f588d78756d19f879eb82d0e5ae4)

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

gtk+: add missing $S references in do_install

When $B != $S the relative paths are wrong, so add an explict $S.

(From OE-Core rev: 68838f1d61f91229ee489889ad2c7f4b61a1f95c)

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

avahi: fix relative path in do_configure

do_configure was assuming that $B is $S, so the relative path was wrong. Add a
missing $S.

(From OE-Core rev: 9c947cf76235624f3661341154575688cf528bed)

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

update-modules: remove recipe from oe-core

update-modules is not used anymore. Any references to it have been
removed, some time ago, from all recipes/bbclasses. It stayed in oe-core
in case anybody still wanted to use it. Time for it to go.

(From OE-Core rev: b8c5f1facd9f9878e137f803b45e99d4e7214d20)

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

qt4: Add space for _appends

With out the space we get -x11-no-neon concatanted incorrectly

(From OE-Core rev: 82602d2c55f5d233bd8351e95a84a65b792786e5)

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

gtk-icon-cache: Fix rdepends construction

The rdepends getVar is returning NoneType and the string constuction fails
this occurs with the hicolor-icon-theme recipe.

(From OE-Core rev: a718cf3179540e049fd0d750ce11a97f84addf40)

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

qemux86-64: Disable paravirt guest, causes test failure on older kernels

The 2.6.37 kernel on the AB occasionally fails dmesg and shutdown
tests with general protection faults.

(From OE-Core rev: 64749308fadabb4aa7c39f360c6395827bc5eb3a)

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

e2fsprogs : upgrade to 1.42.6

 - removed the usage of the patches already contained in the new version
 - adapted patch remove.ldconfig.call.patch so that it applies on new version

(From OE-Core rev: 7ceba1eb102b3f855f561764833f8a7a407b7785)

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

icu : upgrade to 50.1.2

(From OE-Core rev: 1ce0a0b56e37eb4807f5d7e44e0afaba8a05c278)

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

eglibc: don't list the same path twice in RTLDLIST

Before modifying RTLDLIST in ldd, make sure that it doesn't already
contain the right path, thus avoiding duplicate entries in RTLDLIST.

[YOCTO #2655]

(From OE-Core rev: 3cef117439aea2d724e92dcb0f862f1cc8e8bfa5)

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

gcc: enable multilib for target gcc

- add a task to setup multilib configuration for target gcc
- this commit adapts Nitin Kamble's work to gcc 4.7
- use a hash for storing arch-dependent multilib options
- patch gcc in order to use the multilib config files from the
build directory

Tests:
root@qemux86-64:~# gcc -m64 t.c -o t
root@qemux86-64:~# file t
t: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.16, not stripped
root@qemux86-64:~# ./t
Hello World !
root@qemux86-64:~# gcc -m32 t.c -o t
root@qemux86-64:~# file t
t: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.16, not stripped
root@qemux86-64:~# ./t
Hello World !

[YOCTO #1369]

(From OE-Core rev: b26819c85881e82ee1b5c68840011e78c321f18e)

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

module-base.bbclass: Call make instead of oe_runmake in do_make_scripts

This makes the external module compilation a bit more robust for
cases where external module recipes may be passing extra params
to make via EXTRA_OEMAKE, and more than often one needs to pass
M=$(PWD) when building external modules and if we use EXTRA_OEMAKE
that would mean that it would appear in do_make_scripts as well
and since we are only changing the reference kernel src tree here
it will not run the make scripts in desired directory.

It is also well explained in top makefile in kernel tree around
line# 1335 in 3.4

|else # KBUILD_EXTMOD
|
|###
|# External module support.
|# When building external modules the kernel used as basis is considered
|# read-only, and no consistency checks are made and the make
|# system is not used on the basis kernel. If updates are required
|# in the basis kernel ordinary make commands (without M=...) must
|# be used.

Therefore passing  M=... will not do the updates in the basis
kernel as expected with 'make scripts' so we have to bypass EXTRA_OEMAKE

[Yocto #3787]

(From OE-Core rev: 0e0a5ee405bab478f35690e95219a1e5f2ac7aa6)

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>

libfm: add dependency on libexif

* it's autodetected from sysroot

(From OE-Core rev: 945953ed40dafb88d0b559f056fc49f63a84d6de)

(From OE-Core rev: 58a6a61b076b7ad4c2c3aba0336ae91bc7869394)

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

module.bbclass: Allow for modules to be packaged seperate from ${PN}

This patch will allow recipes that provide kernel modules to package
the module or modules in specific packages. That list is contained in
MODULE_PACKAGES, this defaults to  to preserve the current behavior.
The package can also define MODULE_FILES to specify files.

[YOCTO #3803]

(From OE-Core rev: c1ff0467bf03a3342846f0d9dde74e34b740798f)

(From OE-Core rev: 977aee43868499ab87a098f3798e90d6978836b9)

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

gst-plugins-bad: remove musicbrainz build-dependency

This plugin only works with an old version of the musicbrainz API, so it hasn't
been built for some time now.

So, remove the libmusicbrainz dependency as it isn't used.

(From OE-Core rev: 993b4fd2a63939359ea1833374912c15fccf6120)

(From OE-Core rev: 775621d65b5b72169913c817ed9de7973b2d5dcf)

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

git: updated to 1.8.1.2

(From OE-Core rev: 81942e454e4c5aadf640dd3b2431f5726330c52d)

(From OE-Core rev: f07be36a5f699d08d0c018fb7ec327d6a509d76d)

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

dhcp: Fix unshipped WARNINGS

This fixes:
WARNING: QA Issue: dhcp: Files/directories were installed but not shipped
  /etc/dhcpd.conf.example
  /etc/dhclient.conf.example

(From OE-Core rev: 209c7a31ff62291a1aa21dfa4062c7fd3d1a6155)

(From OE-Core rev: d3c6125001516955faf8fa52032cb18681238916)

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

license: fix LICENSE_CREATE_PACKAGE to stay disabled by default

* as reported by Enrico on #oe
  11:06:50 < ensc|w> JaMa: might this be caused by dc78ef91a2bf01efb8028c9afbe69e506e016265
  which checks for 'd.getVar('LICENSE_CREATE_PACKAGE', True)' evaluating to 'True' for every
  string (including the default 0)

(From OE-Core rev: bc42585ad9ca3a9891459ec3234893dff420b95b)

(From OE-Core rev: 0ab56ad52937823dc66b541dd4eda09fb6a34407)

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>

multilib.conf: add TARGET_ARCH & TARGET_SYS to MULTILIB_SAVE_VARNAME

(From OE-Core rev: 81f3dd49693a7ad2e98c182896c98fc71b01592f)

(From OE-Core rev: 92e74eb2f501b10a513e043d3d62007d8ab5f468)

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

autotools.bbclass: add EXTRACONFFUNCS variable

- add EXTRACONFFUNCS variable in order to make it possible
to inject tasks after autotools_preconfigure

(From OE-Core rev: eafaeee58ab7a8f0613f54b8411f41ccefdf94c3)

(From OE-Core rev: d7950a25eda96a271aa6d541e03177d47e533f3f)

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

systemd: systemd-analyze remove python runtime dependencies

systemd-analyze was rewritten in C

(From OE-Core rev: d0682242fb44042497764ecc821b19c3f89054a0)

(From OE-Core rev: 83b0af0aa50e1b0259ae13ea723289a3edb7d9e2)

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

systemd: systemd-analyse has moved to base_bindir

(From OE-Core rev: e24a5a72b73017c59fa0503b9745cb58444ea121)

(From OE-Core rev: 3c14a1f79097f471d5897bea11d55ab4ac17b17e)

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

systemd: Drop ppc build breakage patch for test-unit-file

its been fixed differently upstream in 197

(From OE-Core rev: 50a8bc8cf32f0cf38f0a910b0ae6bba03a2d58d7)

(From OE-Core rev: 97b8958b194add71f281651be1ab1a825029977d)

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>

systemd: bump PE to provide upgrade path from meta-systemd recipe

* 197 sorts lower then v196.*, 22 ERRORs are shown when buildhistory is
  used:
ERROR: Package version for package udev-dbg went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package udev went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package udev-consolekit went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package udev-utils went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package udev-systemd went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package systemd-gui went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package systemd-vconsole-setup went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package systemd-initramfs went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package systemd-analyze went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package systemd-dbg went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package systemd-staticdev went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package systemd-dev went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package systemd-doc went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package systemd-locale went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package systemd went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package libgudev-1.0 went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package libsystemd-daemon went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package libsystemd-id128 went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package libsystemd-journal went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package libsystemd-login went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)
ERROR: Package version for package libudev went backwards which would break package feeds from (0:v196-105-gdecd634-r20.0 to 0:197-r1.0)

(From OE-Core rev: a58238aec13dc7a0fe350c65502860a9045d151d)

(From OE-Core rev: 71ae1992ee79672556b0fb1e066d317044a214cc)

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>

xinetd: enable tcp-wrappers support

(From OE-Core rev: 5890c691510797613b8418456ef5f7b6152c3410)

(From OE-Core rev: 722caaa6783e0a5796f11b77503121d9c0c7f391)

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>

Sudo : upgrade to 1.8.6p4

(From OE-Core rev: 115b1a703c26e11c631d77901104605d64f51f82)

(From OE-Core rev: 5f93bb7e783c23a5da344a9fc2f856f9b973d54c)

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

systemd: honor ROOT_HOME

Use ${ROOT_HOME} which was added in commit
a78cd0b3a69b829327cea035321051ab11cba70f.

(From OE-Core rev: b783b0fe2ac7d6733eaaa4ba9f078ed67d0de944)

(From OE-Core rev: 6ad2b3e18358a075b5a93edc2e620ac0ce9a15c8)

Signed-off-by: Enrico Scholz <enrico.scholz@sigma-chemnitz.de>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

oprofile: remove AX_KERNEL_VERSION from acinclude.m4

The version of this macro in acinclude.m4 is preventing the correct
version in the package's kernelversion.m4 from being used.

Since the version in acinclude.m4 includes the obsolete config.h, any
test that uses it with newer (> 2.6.19) kernels fails.  In the case of
oprofile it means that perf_events support is never detected and thus
'legacy oprofile' is always built.

Fixing this allows the new perf_events 'operf' support to be built.

Fixes [YOCTO #3447]

(From OE-Core rev: 3f7f17894c097ebaa81055c8449c543b734df541)

(From OE-Core rev: 6fac31c7930e228efe5fed978964885e37a4da80)

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

oprofile: add kernel dependency

oprofile-0.9.8 adds 'operf' which depends on the kernel's perf_events
subsystem to be present when building, which adds the kernel
dependency that we specify in a new DEPENDS for oprofile-0.9.8.

(From OE-Core rev: d4dd1eaf75ef3d494d16ed2561e3ff4f5928a428)

(From OE-Core rev: 16bc366877a2e02d96adb91b90e30710777e6f09)

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

busybox: update test case du-k-works

busybox test case du-k-work fails on ext3/4 image whose size less than
512M. Add patch testsuite-du-du-k-works-fix-false-positive.patch to fix
it and it has been merged by upstream.

[Yocto 2896]

(From OE-Core rev: ffe4f8a859d5e71e3e33363d61f0470a32c48d66)

(From OE-Core rev: 19eb2971c6e8c4a2847897919ec7a2cd9b1697b1)

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

python-smartpm: multilib fixes

To fix some multilib issues, change the way the RPM backend decides
if two packages can coexist: if they have a different architecture,
automatically assume that they can coexist (which is fundamental for
multilib).

[YOCTO #3681]

(From OE-Core rev: 05fd850f09c58dba8f64f3fe1de28ed9f21890a2)

(From OE-Core rev: 03c892a02568fa8a5765d9fb569a55f17ea05f96)

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

metacity: remove, migrated to meta-gnome

(From OE-Core rev: 30b4dcdd3a4f78f3e7c7b94ffb9e0f94e56568ad)

(From OE-Core rev: d5314f90747a09967fcbcccec8d86b1e2f48980d)

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

libcanberra: remove, migrated to meta-gnome

(From OE-Core rev: 30b2d9d61965c02cc62bce8d295dc0da7e8ab71f)

(From OE-Core rev: df2bb0fd9f5921a430eb2f7435b69c901f01ac91)

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