libgsmd - remove

This project has been unmaintained for some time, and even the OpenMoko
project is not using it any more (in favour of FSO). Since we have ofono
in OE-Core which replaces and surpasses its functionality, we can remove
libgsmd.

(From OE-Core rev: 4cd880c61e9d74dbf1a747f3654239cadadf45ce)

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

linux-yocto/3.2: perf scripting updates

Updating the 3.2 SRCREVs to reflect the perf scripting fixes by TomZ

standard/default/base:
  0ec416e perf: change --root to --prefix for python install
  18ad076 perf: add 'libperl not found' warning
  e0f641d perf: use pkg-config instead of python-config
  3195098 Revert "perf: hard-code NO_LIBPERL/NO_LIBPYTHON"

meta:
  07ee09b features/ftrace: enable FTRACE_SYSCALLS

(From OE-Core rev: c8f60299424dd912c6402847773f89c0490a1857)

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

bitbake: bitbake/cooker: Print which pkgs would be built in -g output

This is for giving the user a clear list to show which pkg would be
built, we have the "bitbake -g", but it is not easy to read for people,
it is for "dot". Improve the "bitbake -g" to also save a pn-buildlist:

$ bitbake -g core-image-sato
...
NOTE: PN build list saved to 'pn-buildlist'
[snip]

The contents of pn-buildlist:

busybox
shadow-native
pth
sysfsutils
qemu-helper-native
curl-native
ncurses-native
gdbm
xserver-xorg
linux-libc-headers
[snip]

[YOCTO #2404]

(Bitbake rev: 18aff925aece774d0172894e25584353519ca03f)

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

linux-yocto/3.4: perf scripting & unionfs fixes

Updating the SRCREVs to pickup the following changes:

standard/base:
 95c79fb Unionfs: update ->show_options prototype
 49f1599 perf: change --root to --prefix for python install
 8ff1ab6 perf: add 'libperl not found' warning
 b4bc865 perf: use pkg-config instead of python-config
 a2257bf Revert "perf: hard-code NO_LIBPERL/NO_LIBPYTHON"

meta:
 a8cf770 features/ftrace: enable FTRACE_SYSCALLS

(From OE-Core rev: 8d8b9e77aa3403e880cbbdfefc7f24a993b3161f)

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

psplash - Update to latest psplash version. LIC_CHKSUM Tweak

* Update psplash to the latest version.
* Change the license checksum to use the lines in the psplash.h that contains
  license information instead of doing a checksum on the entire file.

(From OE-Core rev: e15dd9dede74c8906e118647ab93941c1fcb35b1)

Signed-off-by: Franklin S Cooper Jr <fcooper@ti.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

kernel.bbclass: add non-santized kernel provides

If the kernel version string uses characters or symbols that
need to be santized for the package name, we can end up with a
mismatch between module requirements and what the kernel
provides.

The kernel version is pulled from utsrelease.h, which contains
the exact string that was passed to the kernel build, not
one that is santized, this can result in:

 echo "CONFIG_LOCALVERSION="\"MYVER+snapshot_standard\" >> ${B}/.config

 <build>

 % rpm -qp kernel-module-uvesafb-3.4-r0.qemux86.rpm --requires
update-modules
kernel-3.4.3-MYVER+snapshot_standard
 % rpm -qp kernel-3.4.3-myver+snapshot-standard-3.4-r0.qemux86.rpm --provides
kernel-3.4.3-myver+snapshot-standard = 3.4-r0

At rootfs assembly time, we'll have a dependency issue with the kernel
providing the santizied string and the modules requiring the utsrelease.h
string.

To not break existing use cases, we can add a second provides to the
kernel packaging with the unsantized version string, and allowing the
kernel module packaging to be unchanged.

   RPROVIDES_kernel-base += "kernel-${KERNEL_VERSION}"

 % rpm -qp kernel-3.4.3-myver+snapshot-standard-3.4-r0.qemux86.rpm --provides
kernel-3.4.3-MYVER+snapshot_standard
kernel-3.4.3-myver+snapshot-standard = 3.4-r0

(From OE-Core rev: 7be043178f5b9d16d9a06696bc0b96689f202a8d)

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

connman: ship the empty plugins directory in the connman package

Otherwise we get a QA warning that /usr/lib/connman/packages isn't shipped.

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

connman: remove some obviously wrong (circa early-GNOME 2?) directories from FILES

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

ncurses: fix packaging issue in multilib build

ERROR: QA Issue: lib32-ncurses: Files/directories were installed but not shipped
  /usr/bin/clear.ncurses
  /usr/bin/reset.ncurses

(From OE-Core rev: e3a29405a59df45ff72f088c602181e1435d8abe)

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

bluez4: fix packaging issue after update

WARNING: QA Issue: bluez4: Files/directories were installed but not shipped
  /usr/share/dbus-1
  /usr/share/dbus-1/system-services
  /usr/share/dbus-1/system-services/org.bluez.service

(From OE-Core rev: 8635372d8e1886d3f0c066341f6909e5015d81f6)

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

packagedata.py: Fix get_subpkgedata_fn for multilib

This happens when tryng to add libgcc-dev to as a multilib package
(e.g. IMAGE_INSTALL_append = " lib32-libgcc-dev")

| Processing task-core-boot...
| Processing fman-ucode...
| Processing dosfstools...
| Processing lib32-libgcc-dev...
| Unable to find package lib32-libgcc-dev (libgcc-dev)!
NOTE: package fsl-image-full-1.0-r1.1.3.6: task do_rootfs: Failed

RPM (or bitbake?) is looking in the tmp/pkgdata, however some of these file
paths are mungned for the multilib scenario:

$ find tmp/pkgdata/ | grep libgcc-dev$
tmp/pkgdata/ppce5500-fsl-linux/runtime/lib32-libgcc-dev
tmp/pkgdata/ppc64e5500-fsl-linux/runtime/libgcc-dev

This patch fixes where we look for these files so they can be found and
properly installed for the multilib root file system

(From OE-Core rev: 4e65ddda5d351bd9f2172e77e656903f61a52f34)

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

qemu-0.15.1: add patch to fix compilatation problems on powerpc

ERROR: Function failed: do_compile (see /opt/yocto/cache-build/p5020ds-64b/build_p5020ds-64b_release/tmp/work/ppc64e5500-fsl-linux/qemu-0.15.1-r6/temp/log.do_compile.28447 for further information)
ERROR: Logfile of failure stored in: /opt/yocto/cache-build/p5020ds-64b/build_p5020ds-64b_release/tmp/work/ppc64e5500-fsl-linux/qemu-0.15.1-r6/temp/log.do_compile.28447
Log data follows:
| DEBUG: SITE files ['endian-big', 'bit-64', 'powerpc-common', 'common-linux', 'common-glibc', 'powerpc-linux', 'powerpc64-linux', 'common']
| ERROR: Function failed: do_compile (see /opt/yocto/cache-build/p5020ds-64b/build_p5020ds-64b_release/tmp/work/ppc64e5500-fsl-linux/qemu-0.15.1-r6/temp/log.do_compile.28447 for further information)
| NOTE: make -j 24
|   LINK  ppc-linux-user/qemu-ppc
| /opt/yocto/cache-build/p5020ds-64b/build_p5020ds-64b_release/tmp/sysroots/x86_64-linux/usr/libexec/ppc64e5500-fsl-linux/gcc/powerpc64-fsl-linux/4.6.4/ld:/opt/yocto/cache-build/p5020ds-64b/build_p5020ds-64b_release/tmp/work/ppc64e5500-fsl-linux/qemu-0.15.1-r6/qemu-0.15.1/ppc64.ld:84: syntax error
| collect2: ld returned 1 exit status
| make[1]: *** [qemu-ppc] Error 1
| make: *** [subdir-ppc-linux-user] Error 2
| make: *** Waiting for unfinished jobs....
| ERROR: oe_runmake failed

(From OE-Core rev: a9207aad5b163a071cd8298517d61514c587e0ed)

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

libxml-parser-perl_2.41.bb: fix MakeMaker issues with using wrong CC/LD/etc

MakeMaker has a bug where it does not propagate CC/LD/etc information
down to subproject it generates Makefiles for... this recipe has has an
Expat subproject which has issues building if we are using sstate-cache
and it will reference the old sysroots and be unable to build properly.
There is an upstream MakeMaker bug for this issue but we can work around
it by fixing up the Makefiles for now

See:
https://rt.cpan.org/Public/Bug/Display.html?id=28632

(From OE-Core rev: 2c3629a698e537c69c71039bc6172b3c0e617ff1)

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

utils.bbclass: add helper function to add all multilib variants of a specific package

This is useful for the scenario where we want to add 'gcc' to
the root file system for all multilib variants

(From OE-Core rev: 7cd54693363882854cd026ebe071e9f2e03d364e)

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

pseudo: Update to 1.3.1 (fixing chroot crash)

Yocto bug #2639.  If a chroot path was long, expanding absolute
paths within the chroot path could overrun a buffer.

(From OE-Core rev: 282cbf7002112f6b58ca1ee070c0b1285b838db7)

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

grub-efi: Do not use help2man

Fixes [YOCTO #2527]

Modify configure.ac and the generated configure script to avoid using
help2man during the compilation process. For grub-efi we are only
deploying the EFI payload and are not installing grub on the target
root filesystem. Therefor, we do not need the man pages.

Cleanup the SRC_URI whitespace while we add a line to it.

(From OE-Core rev: 035bc3f5ee99ed3bd74219d9717239299e4bc765)

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
CC: Paul Eggleton <paul.eggleton@linux.intel.com>
CC: Radu Moisan <radu.moisan@intel.com>
grub-efi-native whitespace cleanup (INC)

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

init-install: Clean up partition alignment

The current partitioning scheme leaves a 1MB gap between all the
generated partitions by adding a 1 to the end of the last partition to
use as the start of the next. parted is smart enough to not overlap
start and end positions of the same value. This avoids the 1 MB gaps.

Rather than pad the disk with 1MB in the beginning and cut it off at the
MB boundary on the end, we can use 0% and 100% to allow parted to do the
required math and use as much of the disk as possible.

(From OE-Core rev: 8aac6ecc5194c734dfd3d677017ab3ea045b2339)

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

init-install: Correct ext2->ext3 typo in logging

We create both the boot and root partitions as ext3 now, update the
logging accordingly.

(From OE-Core rev: 4436639eed57d818992596d6f0f7b53d3bbd4800)

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

init-install: Use swap_ratio in the calulation of swap_size

swap_size currently uses a hard coded percentage and ignores the
swap_ratio variable. Fortunately they are the same value currently. Make
the calculation use the variable to avoid problems in the future.

(From OE-Core rev: 2678ce668499af0e90994b9da8c518e85de56651)

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

classes/license: fix manifest to work with deb

Prepend the license manifest creation call to ROOTFS_POSTPROCESS_COMMAND
instead of appending to ROOTFS_POSTINSTALL_COMMAND. The latter is not
implemented for the deb backend (and probably ought to just be removed
completely), and by using _prepend we can still ensure it occurs before
package info is removed (and before buildhistory in case it is needed
there in future).

(From OE-Core rev: 56fafe083fdb63f156a081344fb1d836e3182bf0)

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

runqemu: fix support for ext4 rootfs images

(From OE-Core rev: c9479ae46d0b891eda8f0db89bc66fdf08c3f7c2)

Signed-off-by: Scott Garman <scott.a.garman@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

runqemu: fix usage() help for MACHINE setting

(From OE-Core rev: 36482a0064993b047829631d063beadbc03f2cbf)

Signed-off-by: Scott Garman <scott.a.garman@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sanity.bbclass: the tmpdir can't be longer than 410

There will be errors when the length of the tmpdir is longer than 410:

1) Longer than 420:
Can't exec "/bin/sh": Argument list too long at /usr/lib/perl/5.10/IO/File.pm line 66.

This error happens on both Ubuntu 10.04 and 10.10 when the pkg needs run
"autoreconf", this is because it passes many files with absolute path to
aclocal, aclocal passes them to perl, this is a limitation of the perl
on Ubuntu 10.04 and 10.10, and the perl-native is not ready at this very
early stage.

2) Longer than 490:
bitbake/lib/bb/persist_data.py", line 197, in connect(database=...)
    >    return sqlite3.connect(database, timeout=5, isolation_level=None)
OperationalError: unable to open database file

This error happens on Ubuntu 10.04, 10.10 and Fedora 17. This is because
the length of the database in sqlite3 module (host's) can't be longer
than 490 (or little smaller). The python-native is not ready at this
very early stage.

The 2 errors are host related, I think that limit the length of the
TMPDIR to 410 is OK for most of the build, rarely build sets TMPDIR's
longer than 410.

[YOCTO #2434]

(From OE-Core rev: ebcf949853ff667478a1ea1d3f1f8f41d643e708)

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

apt 0.7.14: runtime error: Method file has died unexpectedly

When the length of the tmpdir is longer than 400, there is an error
when run "apt-get update":

Method file has died unexpectedly!

This is because the "char S[1024]" is not enough for long URI, S[2048]
would be enough.

[YOCTO #2689]

(From OE-Core rev: 3ed08bd24cef39a85c528159a494e8f0dd001739)

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

apt 0.7.14: runtime error: filename too long (tmpdir length)

when the tmpdir dir is longer than 220, there is no files saved in
tmp/sysroots/x86_64-linux/var/lib/apt/lists/ after run apt-get update,
this is because apt-get uses the path as the file name, but the file
name can't be longer than 255 according to /usr/include/linux/limits.h.

[YOCTO #2688]

(From OE-Core rev: 9a0c0393871eda4bbcecfdd4b595f0c1b8e42edf)

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

curl: upgrade to 7.26.0

Config system changed from 7.24.0 and the noldlibpatch
is no longer needed, thus deleted.

(From OE-Core rev: 0d2d59420b5924491ccd5c091c823b9c277a6721)

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

gcc-common: Don't use "is" for comparing strings, use "=="

Needed because the equality check was failing here even though upon
printing the LHS and RHS were the same.
As per http://stackoverflow.com/a/2987975/64537, using "is" compares the
memory addresses of the two objects which is not what we want here. We
just want to compare the values.

(From OE-Core rev: 7fd82cc90a48302ed42b6bfa962bb0de2c652b45)

Signed-off-by: Kartik Mohta <kartikmohta@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

recipes.txt: Add entries for "recipes-rt" and "recipes-support"

(From OE-Core rev: 9705ccc10892b15fa3269ed071bb41d45987694c)

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pixman: upgrade to 0.26.2

Also, disable Loongson MMI for now because Pixman's build system will
autodetect a Loongson2f architecture revision and the build will fail.

(From OE-Core rev: 7dded4da52bc73677583da740620b0c81e594d62)

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

xinput: upgrade to 1.6.0

Licence file changed, an entire part has been removed.

See diff below:

<<
 TORTIOUS  ACTION, ARISING    OUT OF OR   IN  CONNECTION  WITH THE USE OR
 PERFORMANCE OF THIS SOFTWARE.

-
-Copyright 2007 Peter Hutterer
-
-Permission to use, copy, modify, distribute, and sell this software and its
-documentation for any purpose is hereby granted without fee, provided that
-the above copyright notice appear in all copies and that both that
-copyright notice and this permission notice appear in supporting
-documentation.
-
-The above copyright notice and this permission notice shall be included
-in all copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
-OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
-MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
-IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY CLAIM, DAMAGES OR
-OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
-ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
-OTHER DEALINGS IN THE SOFTWARE.
-
-Except as contained in this notice, the name of the author shall
-not be used in advertising or otherwise to promote the sale, use or
-other dealings in this Software without prior written authorization
-from the author.
-
-
+Copyright © 2007 Peter Hutterer
Copyright © 2009 Red Hat, Inc.

Permission is hereby granted, free of charge, to any person obtaining a
>>

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

libxi: upgrade to 1.6.1

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

libx11-diet: upgrade to 1.5.0

The X18NCMSstubs.diff was modified to please the -Werror=return-type
compiler option. Otherwise, it will throw an error.

(From OE-Core rev: 708785266d1d671f5d2a8e8af81a77e34c70662f)

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

libx11-trim: upgrade to 1.5.0

(From OE-Core rev: 817742fa1a5261b3b2c697de3c14e602d55713e2)

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

libx11: upgrade to 1.5.0

Removed one backported patch.

(From OE-Core rev: 52d4ff7b35602faa52360a3e7a800891ef60c739)

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

directfb-examples: upgrade to 1.6.0

(From OE-Core rev: 37a1e3c0ed5347c77ace9f3ce262f34d9ac53528)

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

directfb: upgrade to 1.6.1

Removed two backported patches that already exist in the current
version and fixed a compilation issue when zlib is used.

(From OE-Core rev: 47800f9bff7a799489851156c341151ee9cf364f)

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

multilib: Enable multilib remapping for SDK generation

Enable the remapping for SDK generation, this is required to be able to
create an SDK that targets an alternative multilib.  Note, this work does
not finish SDK/multilib support, but it is one more step toward making
it work properly.

(From OE-Core rev: d66d2bdf6214939e8b131f47db2f35d3c64e7dd2)

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package_rpm: Avoid duplicate package generation failures

When constructing a multilib based image, it's possible to end up with
the same package listed in multiple steps of the install.  During resolution
we use the --replacepkgs option to avoid errors, as the resulting install
solution will be as expected.  (We do not enable the replacepkgs option
for the final install step, only the generation of the install solution.)

(From OE-Core rev: 68ea7d827710ae7c5cf402887d4c12319038e064)

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

populate_sdk_rpm: Sync multilib configuration with rootfs_rpm

An anonymous python chunk configures some multilib settings for both
populate_sdk_rpm and rootfs_rpm.  The two classes should contain identical
versions to ensure that the generated multilib configuration is the same
for the SDK and the rootfs.

(From OE-Core rev: 0f60722ef460fa826e4f02e7457004e64ca41d76)

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

rpm: Fix PACKAGECONFIG dependencies

neon requires expat to be built first

lua similarly requires expat if enabled

selinux package name should be "libselinux" (from meta-selinux).

(From OE-Core rev: 6b3c9802b659ca1d3b7638ff80e6492cc0e16d6f)

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gnutls: make sure native is patched for gettext version

(From OE-Core rev: 92ede8a994b9a24387d3d7dc79e80705f6be4524)

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

lsbinitscripts: increase ALTERNATIVE_PRIORITY

[YOCTO #2133]

Increase ALTERNATIVE_PRIORITY of file functions to be higher than
the value in package initscript. Then when lsbinitscripts installed,
file functions provided by lsbinitscripts will be used.

(From OE-Core rev: 33f9abed7658bd5eae39e987b210651d202e1f4e)

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

initscripts: use update-alternative to handle file functions

lsb need a more abundant /etc/init.d/functions file to handle lsb test.
Use update-alternative to install file funtions in package initscripts,
when package lsbinitscripts installed the file functions of
lsbinitscripts will be used.

(From OE-Core rev: 478f821462665a9dc6f1755059cdbf642aebbf4d)

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

matchbox-panel-2: bump PR because libxcb-util soname was changed

(From OE-Core rev: 31b8fe8d1a9921e4efc64b6ab00c26e4bd09ed2c)

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

startup-notification: bump PR because libxcb-util soname was changed

(From OE-Core rev: 1dc4100dd35c050b38e4733cb98752c1323ef7b1)

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

bitbake: usermanual: Fix missing markup

(Bitbake rev: 68af2b09315ba35eae24933a599014a662789c2c)

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

bootimg: Use STAGING_KERNEL_DIR

bootimg.bbclass using STAGING_DIR_HOST/kernel instead of
STAGING_KERNEL_DIR, resulting in build failure of live images.

| install: cannot stat `/usr/local/dev/yocto/fishriver-test/build/tmp/sysroots/fishriver/kernel/bzImage': No such file or directory

Replace it with STAGING_KERNEL_DIR.

(From OE-Core rev: 8f16811a8d51982a8b3d70e6087aef4a41926840)

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

perf: remove unconditional python-ext install

This shouldn't be unconditional - a later patch made it so, but that's
not yet pulled in.  In the meantime, to fix build failures remove the
unconditional install.

(From OE-Core rev: dcc43c34bfa9304233ecfd3f518a84ddfdc2fd90)

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

dpkg 1.15.8.7: bad interpreter when long tmpdir

When use PACKAGE_CLASSES = "package_deb", and the length of tmpdir is
longer than 177, the error occurs during the image generation:

/very/long/path/totmp/sysroots/x86_64-linux/usr/bin/dpkg-scanpackages:
bad interpreter: No such file or directory

The interpreter is perl, and it does exist, this is because the first
line:

  #!/very/long/path/to/perl

in the script can't be very long.

Create a wrapper for it would fix the problem.

[YOCTO #2640]

(From OE-Core rev: 7d3281be18567207de8ce2cabce03957bf2cbca8)

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

netbase: update to ver. 5.0

(From OE-Core rev: 5beb7176d361d4bfa279bbfab5a17b8fe4fa3b73)

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

bluez4: update to ver. 4.101

(From OE-Core rev: 1b2f6a21dde6045c4604ca5463e6c4db3499ee06)

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

ofono: update to ver. 1.8

(From OE-Core rev: 80f05ff2afed7ef24d17209d39a17cf8139855bb)

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

gst-plugin-bluetooth: update to ver. 4.101

(From OE-Core rev: 5d0e81b370a4c6b0473efd7989b6881a16e02f29)

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

telepathy-glib: update to ver. 0.19.2

(From OE-Core rev: 2cc263bf5402e21429a5057c183a10918491da48)

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

u-boot: make FILESDIR a shared setting via FILESPATH

The setting is the same in all recipes, so move it to
the shared settings in u-boot.inc

Since FILESDIR is also being phased out, use the FILESPATH
setting as suggested by Richard Purdie.

Cc: Richard Purdie <richard.purdie@linuxfoundation.org>
(From OE-Core rev: 1b78904b115b50172cf82948d8a68dc3a98005ce)

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-firware: bump PE because git SRC_URI was changed resulting in LOCALCOUNT reset

(From OE-Core rev: 088b59e7888ae670710be19c868d0fe90ca44b78)

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

bitbake: data_smart: Fix multiple override interaction with append and prepend operators

Variables which used multiple overrides and the append/prepend operators were
not functioning correctly. This change fixes that.

This fixes the testcase:

OVERRIDES = "linux:x86"
TESTVAR = "original"
TESTVAR_append_x86 = " x86"
TESTVAR_append_x86_linux = " x86+linux"
TESTVAR_append_linux_x86 = " linux+x86"

[YOCTO #2672]

(Bitbake rev: dc35a2e506e15fb7ddbf74c3b3280e9e83ab33bb)

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

lib/oe/sstatesig.py: Exclude subversion-native from sstate checksums

If we don't do this, target and cross recipes end up with different sstate
checksums for shared work directory tasks which is bad in the case of gcc.
It leads to multiple fetch/unpack tasks against the shared directory
which ends up with build failures/races.

(From OE-Core rev: 793ce6cd9aa632e0f13789c8293770a86085d28d)

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

gcc: fix collect2 host contamination problem properly

We added the autoconf cache line a while back to ensure that configure doesn't
poke into some hardcoded host paths looking for things it shouldn't. Applying
it as part of do_configure wasn't getting it to the do_compile tasks
where much of the configure scripts are run by gcc. This changes it to a simple
export to ensure it reaches the places it needs to and truly gets rid of
the cross compile badness messages from the logs.

(From OE-Core rev: af58d5c3a5c5e632aa20b1d007d47dbbfb4ed6e6)

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

bitbake: runqueue.py: Fix recursive task pruning to only prune self referencing tasks

(Bitbake rev: 4962a59793504b26b06cf058dda600a07fbbd951)

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

prexport.bbclass: Fix typo in warning: "AUROPR" -> "AUTOPR".

(From OE-Core rev: 0b79c730f3e5c3c5329e439dffd137f2605b6cba)

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

lsbinitiscripts: Add glib-2.0 to DEPENDS list

ppp-watch requires it.

x86_64-poky-linux-gcc    -m64 --sysroot=/intel/poky2/builds/fetch/tmp/sysroots/qemux86-64 -O2 -pipe -g -feliminate-unused-debug-types  -Wall -D_GNU_SOURCE `pkg-config glib-2.0 --cflags` -c ppp-watch.c -o ppp-watch.o
Package glib-2.0 was not found in the pkg-config search path.
Perhaps you should add the directory containing `glib-2.0.pc'
to the PKG_CONFIG_PATH environment variable
No package 'glib-2.0' found
ppp-watch.c:78:18: fatal error: glib.h: No such file or directory
compilation terminated.
make[1]: *** [ppp-watch.o] Error 1
make[1]: Leaving directory `/intel/poky2/builds/fetch/tmp/work/x86_64-poky-linux/lsbinitscripts-9.03-r0/initscripts-9.03/src'

(From OE-Core rev: 488610f4a5b7591a989ff1a6137f276a85fd7a85)

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

qemumachines: make MACHINE_FEATURES append follow qemu.inc include

qemu.inc does a straight assign to MACHINE_FEATURES so overwriting the
preceding append to MACHINE_FEATURES, so the MACHINE_FEATURES append
needs to be moved after the include.

This situation came about as a result of commit 71a4bf386:

    qemumachines: Enable xserver-xorg as default xserver

    For qemux86 and qemux86-64 include qemu.inc after defining XSERVER

which missed this side-effect (and maybe others).

(From OE-Core rev: 4f336e5f416df382fdd2b405314741164d537b22)

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

perf: add libexec/perf-core and contents

libexec/perf-core contains all the pre-canned scripts and modules
needed by both the Perl and Python bindigs.  Add libexec/perf-core
along with all the pre-defined perf scripts underneath it.

(From OE-Core rev: 00f9e8aed9f111d351c2dc71cd8605ccbd11f68f)

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

perf: enable Perl binding

Add support to enable the perf Perl binding.

The build depends on perl-native to retrieve the configuration
settings needed for the binding.  cpan-base adds some useful functions
like is_target() and get_perl_version() that we need for
PERLCONFIGTARGET and related settings, which allow us to use the
target's Config_heavy.pl settings for ExtUtils:Embed when building for
the target.

Also adds the perl-modules dependency to give the target the perl
modules that scripts using the binding need.

(From OE-Core rev: 73a8eed010a821fc4f5c0e71ef749938ed780b44)

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

perf: enable Python bindings

Add support to enable the perf python bindings.  The combination of
these changes and the changes in the python-config sections in the
kernel Makefile enable all the python bindings currently available in
perf.

(From OE-Core rev: 487c3a72f0ac7121b79d2f78b81f1d4732bc258a)

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

perl: use @STAGINGDIR@ in config.sh

archlibexp in the snapshotted config.sh needs to be relative to
staging dir - it gets generated into Config_heavy.pl as one of the
settings and is used by ExtUtils::Embed during the target build.

The substituted setting doesn't make it into the target package
because it's removed by perl_package_preprocess() before packaging.

(From OE-Core rev: b647d2ab84ff0eb94f4b6a1b731d74aaf8c60b9d)

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

perl: add @STAGINGDIR@ for config.sh substitions

For some config.sh substitions, we'd like to be able to use
STAGING_DIR_HOST; add @STAGINGDIR@ to allow for that.

(From OE-Core rev: cc594be0877b2a6616453c402db0a58ef265bc8d)

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

perl: keep original libperl location

Things like ExtTools expect to find libperl at the original
(${libdir}/perl/${PV}/CORE/libperl.so) location, so keep it there but
symlink to it for the original (?) reason.

(From OE-Core rev: deb5d5ededfc404f4c54b18f3dcf8192e1c835e0)

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

mkelfimage: Add stable git build (initial recipe)

(From OE-Core rev: c04136393325967cbd491636b582d3e0f394de16)

Signed-off-by: Raymond Danks <ray.danks@se-eng.com>
Acked-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

image_types: Add elf image type

On x86, an ELF image file may be stored as a coreboot payload.
The image file is constructed, using the mkelfimage utility,
from a kernel and an initrd.

(From OE-Core rev: 93e9de4e27919f59b6783d53f314df6e49ccb436)

Signed-off-by: Raymond Danks <ray.danks@se-eng.com>
Acked-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package.bbclass: enable the use of package_qa_handle_error

This will allow the reporting of these errors as either WARNINGs (default)
or ERRORs if installed_vs_shipped is added to the ERROR_QA of the policy
file (such as a <distro_name>.conf file.

V2: found the code I had intended to send instead of that other junk,
    was just not watching what I pushed on that one, sorry. (this is edit in
    no in the actual commit message)

(From OE-Core rev: 20d4205a4e408b6a99db392d4df458156113106a)

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

oprofile: Change /root to /home/root

OE uses /home/root instead of /root for the root user's directory.  Update
oprofile to match.

Upstream-status: Inappropriate [OE Specific]

(From OE-Core rev: d50c44585e72057532b8e0d9221ebf058e514672)

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

speex: use lib_package bbclass

This is a clean-up following the recent addition of PN-bin
to bitbake.conf and lib_package

(From OE-Core rev: 4c41954227144606d2d22fbdf07231f4c7207ac5)

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

gnutls: use lib_package bbclass

This is a clean-up following the recent addition of PN-bin
to bitbake.conf and lib_package

(From OE-Core rev: b9553b5fe8e06032770fd34e26ab97e6834292ab)

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

db: use lib_package bbclass

This is a clean-up following the recent addition of PN-bin
to bitbake.conf and lib_package

(From OE-Core rev: b3bfeec4db7d8fa6b8cc4bb2153c7f94cbe1dcc3)

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

pulseaudio: use lib_package bbclass

This is a clean-up following the recent addition of PN-bin
to bitbake.conf and lib_package

(From OE-Core rev: 42dfedc01beb90adfa222a0c671ca8876849f938)

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

libacpi: use lib_package bbclass

This is a clean-up following the recent addition of PN-bin
to bitbake.conf and lib_package

(From OE-Core rev: 6597f7c77e1ccb21d73f8e3cec25bb27a901b500)

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

lib_package: use PACKAGE_BEFORE_PN

Since this check effective is not a change becase the package
order that was in this file has not really changed no PR Bumps
are needed.

(From OE-Core rev: d807d5dd5bd0e161057ee115256dc050dbe8609b)

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

bitbake.conf: Add PACKAGE_BEFORE_PN to PACKAGES

By doing this we can easily add addtional packages that can be
greedy before PN, but after some of the other standard packages.

This will also allow us to simplify the lib_package class to set
this variable.

(From OE-Core rev: b41b2bc584756956a48c9c5f91e7337e821105b2)

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

kernel: Add kernel headers to kernel-dev package

[YOCTO #1614]

Add the kernel headers to the kernel-dev package. This packages what was
already built and kept in sysroots for building modules with bitbake.
Making this available on the target requires removing some additional
host binaries.

Move the location to /usr/src/kernel

Before use on the target, the user will need to:

    # cd /usr/src/kernel
    # make scripts

This renders the kernel-misc recipe empty, so remove it.

As we use /usr/src/kernel in several places (and I missed one in the
previous version), add a KERNEL_SRC_DIR variable and use that throughout
the class to avoid update errors in the future.

Now that we package the kernel headers, drop the
kernel_package_preprocess function which removed them from PKGD.

All *-sdk image recipes include dev-pkgs, so the kernel-dev package will
be installed by default on all such images.

(From OE-Core rev: 6125ea40d4483965f793bd847b3ce14b668a5b1e)

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
CC: Bruce Ashfield <bruce.ashfield@windriver.com>
CC: Tom Zanussi <tom.zanussi@intel.com>
CC: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-yocto-tiny: add KMETA specification

With recent kern tools changes, having a meta branch is opt-in. The
other linux-yocto recipes were updated, but linux-yocto-tiny was
missed.

Without specifying a meta branch, the kernel configuration audit
won't return any useful results, and throws a warning message.

 WARNING: Can't find any BSP hardware or required configuration fragments.
 WARNING: Looked at //cfg///hdw_frags.txt and //cfg///required_frags.txt in directory: //cfg//
 NOTE: package linux-yocto-tiny-3.2.18+git1+ee78519365bdb25287703bbc31c06b193263c654_1+27b68a93eb791e830da8d3a2c0fc99780897ad89-r3.0: task do_kernel

(From OE-Core rev: 29dbe7d7ba66c66e9e5c8dc438f89d9b46c509a0)

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

linux-yocto/3.4: update preempt-rt to 3.4.4-rt13

Updating preempt-rt in the 3.4 kernel to the latest upstream version
3.4.4-rt13.

(From OE-Core rev: f5ed8bc876fd24e4f08d29da06454037eec818e6)

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

linux-yocto/3.4: v3.4.4 + configuration updates

Bumping the kernel version to v3.4.4 and importing the following
meta branch commits:

 9b7c74b meta: bump kver to v3.4.4
 8231dec ck: relocate config post patches
 26b965e net_sched: remove dupicate configuration option
 df8bf19 meta: remove non applying routerstationpro patch
 292d8ea arm: import upstream versatile io.h fix
 f4824b1 net_sched: select ACT as a dependency of INGRESS
 30ae722 cgroups: delete obsolete namespace option
 51f94f0 meta: add qemumipsel mapping

Built and boot tested on qemu*

(From OE-Core rev: 3d33b8c8b3be5dcdac30c115b9fec427269db2b4)

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

linux-yocto/3.4: add qemumipsel machine mapping

Updating the meta branch with a board description and configuration
that can be found from the linux-yocto recipes without doing a
MACHINE -> KMACHINE mapping.

From the meta branch commit:

    The mti malta board description for a little endian config can't
    be found by default unless it has a matching KMACHINE, or the
    KMACHINE is set in a kernel recipe. In this case, it is easier to
    just add qemumipsel in the KMACHINE list, so the recipe doesn't
    need to do the mapping.

The little endian mips support reuses everything but the endianess
settings from the big endian mti malta board emulation.

This change adds the required mappings and SRCREVs, but does not
update the machine compatibilty, since official support is still
pending. The compatibility can be set in addon layers until that
time.

(From OE-Core rev: de2535b2bb47b80bdaeb3053b0ac0d1bbb9c5d53)

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

telepathy-glib: update to ver. 0.19.1

(From OE-Core rev: 5109954afefc48d8f44776ec1f2d8a94a21c04f6)

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

pcmcia-cs: Remove references to long-deprecated pcmcia-cs.

pcmciautils is the appropriate PCMCIA package these days.

(From OE-Core rev: 5066f7e9750253bc4678f2884e15f0333e60932b)

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

lighttpd: upgrade to 1.4.31

(From OE-Core rev: 913f4831aeb83e2c7f08daa47a31089dbf06f2be)

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

xcb-util-keysyms: upgrade to 0.3.9

(From OE-Core rev: f89834e7277a77d46b21a363a82371085f028969)

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

xcb-util-image: upgrade to 0.3.9

(From OE-Core rev: 3a88f34e9aa04f5aa988f9f77cf68189da169084)

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

xcb-util-wm: upgrade to 0.3.9

(From OE-Core rev: fba0788bb5ef1743aeceb9d874eb099f7240d118)

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

libxft: upgrade to 2.3.1

(From OE-Core rev: d1e7823c33eafbdd8ca086b8ba48f61faf391f20)

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

ethtool: upgrade to 3.4.1

(From OE-Core rev: 5550c437b50ea9124ce06fa7d17166f2be6d4ee5)

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

libdrm: upgrade to 2.4.35

(From OE-Core rev: 4b640c4737806cbc035fa49bb22b3ee39e372e6a)

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

xcb-util: upgrade to 0.3.9

(From OE-Core rev: 8c6acf7edabb95051384025d3793f6346e234b71)

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

xf86-input-synaptics: upgrade to 1.6.2

(From OE-Core rev: 1e42302426780d8603517418163f46a7eef9931b)

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

xf86-input-vmmouse: upgrade to 12.9.0

(From OE-Core rev: 886c923c1f0259204399582eeb8523fd146c2c62)

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

bitbake: fetch2/svn: Enhance to cope with subversion 1.7 upgrade

svn changed working checkout formats between 1.6 and 1.7. Its convoluted to
detect what format a given working copy is in so the simplest solution is simply
to run "svn upgrade" within the working copy.

The base svn command variable is relocated slightly to enable this new code to
work effectively.

(Bitbake rev: ebd3ecdb5f3c6d3fe1fad27cbed4d80f4277e92e)

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

bitbake: usermanual: Improve task dependency documentation to match reality

(Bitbake rev: f67c0606fc681359fb0c68be55cfc9f11d410f17)

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

bitbake: runqueue.py: Allow recrdeptasks that have self references

In some cases we want to pull in DEPENDS and RDEPENDS of recrdeptask
dependencies but we need a way to trigger or avoid this behaviour
depending on context. The logical syntax to trigger such behaviour
would be a self referencing recrdeptask:

do_a[recrdeptask] = "do_a do_b"

The dependency chains already recurse this kind of expression correctly, the
missing piece is to avoid any circular reference errors. Since the dependencies
have already been recursively resolved, simply removing any recrdeptask
references is enough to break the circular references.

This patch therefore removes any circular references using the set
difference_update() operator. There will be metadata tweaks required to
add any references needed to the extra taskname.

(Bitbake rev: a5324da9b8a0c9307a6c511ea9009f34be70c92b)

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