bitbake: fetch2/svn.py: use log instead of info to retrieve revision

We have faced a corner case situation where the 'last changed
revision' returned from svn info is wrong. It happens when the last
revision is a directory move. e.g. if we assume that the svn
repository at revA has root/x/y/z/foo/bar and it is moved to
root/a/b/c/foo/bar in revB, then svn info 'last change revision' will
return revA. As such when using AUTOREV, we are going to attempt to
retrieve root/a/b/c/foo/bar (as per SRC_URI) but at revA when it did
not exist.

So this patch changes how we retrieve the latest revision and uses
'svn log --limit 1' which gives correct result in all tested cases.

(Bitbake rev: 17d8ef0b813a05c231e3dbe6e8bc82a4a9b1d2f8)

Signed-off-by: Nicolas Dechesne <nicolas.dechesne@linaro.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sgml-common: make postinst run successfully at rootfs time

Take $D into consideration and make postinst run successfully at
rootfs time.

(From OE-Core rev: 0636093711547957a8f5b25322bd3e0da367cfc4)

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

groff: fix pkg_postinst and remove unneeded do_install_prepend

Fix to take $D into consideration in pkg_postinst so that the postinst
script could run at rootfs time.

The existence of ${D} is ensured by the in base.bbclass.
    do_install[dirs] = "${D} ${S} ${B}"
So there's no need for this do_install_prepend.

(From OE-Core rev: 4facf4df0bc0d71d11ba53737faff6f2c6e5863b)

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

jpeg: enable postinst to run at rootfs time successfully

(From OE-Core rev: 41eda2f5a256f495b788e1ced22e6d19c007e34e)

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

busybox: fix sed auto insert newline testcase

backport the patch from busybox upstream to fix the
auto insert newline issue.

busybox defect:
https://bugs.busybox.net/show_bug.cgi?id=6584

(From OE-Core rev: db00fa405c025c61844df2e9589fe635fc5df0e2)

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

udev: fix typo in local.rules

(From OE-Core rev: b8e6b73d4ef6aa957bfcae299f754c2dcb702b84)

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

grub-efi.bbclass: Fix startup.nsh to work on more EFI revs

Some revs of the EFI firmware + shell do not automatically setup the
path in a such a way as to execute a binary without an absolute
reference like "FS0:\EFI\BOOT\bootx64.efi".  All the versions that I
have tested work properly by simply calling the binary which is in the
EFI\BOOT directory by name like "bootx64.efi".

The error you see on the console looks like the following:

startup.nsh> EFI\BOOT\bootx64.efi
'EFI\BOOT\bootx64.efi' is not recognized as an internal or external command, operable program, or batch file
Shell>

This patch simply drops the EFI\BOOT for greater compatibility.

(From OE-Core rev: 754b52ea7a3cdf8e7e939a314525d16c4dfb52cb)

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

update-rc.d.bbclass: Cleanup package scripts

* Add proper preinst script
* Use -n "$D"/-z "$D" checks throughout the scripts
* Only set OPT in cases where its used
* Don't pass unnecessary -f to update-rc.d in postrm script
* Remove outdated comment

(From OE-Core rev: 2d9114aaf130179cb13c20c8b045c2694eca6fd9)

Signed-off-by: Jacob Kroon <jacob.kroon@mikrodidakt.se>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

rxvt-unicode: upgrade to 9.19

(From OE-Core rev: a896ac91109ab3f2a6b8c7253c6da3e3e60203b6)

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

grub-efi.bbclass: Fixes GRUB_IMAGE when using boot-directdisk class

When boot-directdisk class is used and EFI boot is set the
grub-efi-${TRANSLATED_TARGET_ARCH}-native need to be dependent.
Allowing GRUB_IMAGE to be created and bootia32.efi got from the
image directory.

(From OE-Core rev: b9778975db410b8cd01ef6854c7cd3ea22a0b5b7)

Signed-off-by: Joao Henrique Ferreira de Freitas <joaohf@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

Mesa: upgrade to 9.2.2

* License is still MIT
* removed patches were already merged or
  another solution was backported.

(From OE-Core rev: 263e654e5d28fa7f6b0c8cc23f57a31448e88a40)

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

irda-utils: restart irda daemon correctly

irattach init script restart faulty logic prevents irda daemon
from restart correctly.

root@qemu0:~# /etc/init.d/irattach restart
Restarting IrDA: Terminated
root@qemu0:~# ps aux | grep irattach
root       541  0.0  0.2   2400   612 ttyS0    S+   09:05   0:00 grep irattach

As above shows, irattach not started after executing restart command.
This commit changed the restart command logic: firstly stop, then
start.
Prompt telling user irattach start successfully or failure also
added.

(From OE-Core rev: 39f266138b972b550979909b235a5779828d7d89)

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

lttng-tools: make ptest able to work on target

The paths of executable files in test cases are the source code path,
need to fix to make them work on target

(From OE-Core rev: 48067c2093e397255e41b0222da3b7612f1458a7)

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>

elfutils-native: Update the patch to include the missing pieces needed for tests

In the last patch we missed similar update for test/line2addr.c
which meant that the build still failed.

(From OE-Core rev: cb966afe86a5856d846fb0d4fb627cdce2ae0670)

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>

syslinux.bbclass: Fix default serial port string

The default value of SYSLINUX_SERIAL_TTY is not correct.

It should be console=ttyS0,115200 else the boot string generated in
the syslinux menus for the serial choice is not correct.  The kernel
boot parameters will get set to:

/vmlinuz initrd=/initrd LABEL=boot root=/dev/ram0 ttyS0,115200

Note that the above is missing the "console="

The default value will now work the same as the value found in
grub-efi.bbclass.

(From OE-Core rev: fbc864241933c6f40814f47e7a85dd71ce255393)

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

cross-canadian: Improve comment

The previous cross-canadian change was missing some tweaks
to the comments. This clarifies them slightly.

(From OE-Core rev: 154ecc40c289b15fe9cbb33befb20dd10112e788)

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

telepathy-glib: upgrade to 0.23.0

(From OE-Core rev: f8c52949aba97d1b4d8c0f5c9f6d40dc5a857e03)

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>

telepathy-mission-control: upgrade to 5.16.0

(From OE-Core rev: 6bae01a04a375906731555b997766d41f887b496)

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>

iproute2: upgrade to 3.11.0

(From OE-Core rev: 6c2f95c1cce38de2905a69b79f132eb28f0a64f6)

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>

libuser: upgrade to 0.60

(From OE-Core rev: 2a89bcf6f604b14ccc9aa98c674b6a4632015c52)

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>

lsb: add begin function to lsb_log_message

Some init scripts call a "begin" action to log early init phase
messages, e.g. openvswitch-controller.  Add the "begin" function to
lsb_log_message.

(From OE-Core rev: c956290902afd0e4f6d8f545dfe0a55796deec5e)

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

libx11: upgrade to 1.6.2

Adapted X18NCMSstubs.diff to correctly apply.

Tested on core-image-sato (for both libx11 and
libx11-diet)

(From OE-Core rev: 1179a6dc039fc4ce7a540450de55f5a39681370e)

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

shared-mime-info: upgrade to 1.2

Adapted parallelmake.patch to the new Makefile.am.

(From OE-Core rev: 30a3dba92f38b8877954e2c7da0c9c18be420d8a)

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

ncurses-terminfo: Remove bashism from basic terminfo installation

The vtX terminfo files aren't being copied on systems where bash isn't
the default shell (debian, etc.).  I removed the bash specific syntax
so the files are properly copied on these systems.

(From OE-Core rev: edd7d53c6149b27d5636a458db91650c8c400612)

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

e2fsprogs: Escape filenames in populate-extfs.sh

Without this patch filenames containing spaces do not get into the final
ext2/3/4 filsystem.

[YOCTO #5401]

(From OE-Core rev: 1350b461ed0c9d4afa1ab909a5b1ff60fb160c97)

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

coreutils: support command arch

The "arch" is removed in util-linux 2.23, but some old
software and devs look for arch:
http://lists.gnu.org/archive/html/bug-coreutils/2007-06/msg00010.html

Use coreutils version.
[YOCTO #5404]

(From OE-Core rev: 9df8591a6e18afa592673cb454dc699a1341461d)

Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
[sgw: fixed ordering to be alphabetized]

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

vala.bbclass: add dependency on vala

This class points the inheritor, if it is a target,
to directories in the target sysroot, so we want to
be sure the .vapi files are there.

(From OE-Core rev: 2da8bbd47686f54efeec521d521f176f6aeb8d39)

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

cross-canadian: Handle powerpc linux verses linux-gnuspe

PowerPC toolchains can use the OS "linux" or "linux-gnuspe". This
patch links them together so the one cross-canadian toolchain can support
both.

GCC_FOR_TARGET is set for the GCC recipe as otherwise configure
can pick up an incorrect value.

[YOCTO #5354]

(From OE-Core rev: a1d6331238982b0c5d39b0a18794f6654b00d46a)

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

systemd-compat-units: run-postinsts fix script link

in

commit fe039170236080291c0220476a5809774f82ee5c
Author: Muhammad Shakeel <muhammad_shakeel@mentor.com>
Date:   Wed Oct 2 10:55:32 2013 +0000

    systemd-compat-units: Use correct run-postinsts script link

    OE-Core commit 75a14923da1ba91eddde47f0907345c19c82d6f0 has moved
    run-postinsts script execution from S98 to S99 in rcS.d. run-postinsts.service
    should check for this script and run it on first boot rather than
    S98run-postinsts, which is for opkg/dpkg.

the link was corrected but the mentioned commit is not available. Instead of
reverting, we use the same variable as opkg for init script ordering and drop
a note in case somebody wants to change default.

(From OE-Core rev: 7aabc9408fb382f0ae39f9932b6d9ac391528b76)

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>

nfs-utils: modify nfsserver init script indent

Using sysvinit testing service status, nfsserver status
allways display as [?] unknown.

This is because sysvinit package check whether service's
init script supporting status function or not by:
grep -qs "\Wstatus)" "$SERVICE"

So, this commit modified the indent for status etc, as
most service's init script does.

(From OE-Core rev: a6b02fe439fa13c8482383fba2bfdcb0e9742141)

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

classes/ptest: do not set EXCLUDE_FROM_WORLD

Various normal recipes that we expect to build during a world build now
inherit this class; it cannot therefore set EXCLUDE_FROM_WORLD.

(From OE-Core rev: 80f29b04b1d0a9357f193085224ccf2bf8f530fa)

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

classes/ptest: fix quoting

BitBake currently allows using the same quotes outside and inside the
value, but it isn't really right, looks odd and might stop working in
future.

(From OE-Core rev: 0af9cf31851896276a219170001047406f45de50)

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

core-image-basic.bb: Allow user extensions

Allow the user to provide additional packages to this image.
This lets core-image-basic behave like all other core-image*
recipes (which do support CORE_IMAGE_EXTRA_INSTALL), as well
as match the documentation which suggests this as the mode to
extend any core-image* image.

v2 - drop redundant setting of CORE_IMAGE_EXTRA_INSTALL

(From OE-Core rev: 5faabf398819d40b55c46bc83ae03942d115024b)

Signed-off-by: Gary Thomas <gary@mlbassoc.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

recipes: Remove PR = r0 from all recipes

Remove all PR = "r0" from all .bb files in oe-core. This was done
with the command sed -e '/^PR.*=.*r0\"/d' recipes*/*/*.bb -i

We've switching to the PR server, PR bumps are no longer needed and
this saves people either accidentally bumping them or forgetting to
remove the lines (r0 is the default anyway).

(From OE-Core rev: 58ae94f1b06d0e6234413dbf9869bde85f154c85)

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

openssl: create package for openssl configuration file

* Add the openssl-conf package to the list of packages to
  be created.  This package contains the openssl.cnf file
  which is used by both the openssl executable in the
  openssl package and the libcrypto library.

* This is to avoid messages like:
    WARNING: can't open config file: /usr/lib/ssl/openssl.cnf

* When running "openssl req" to request and generate a certificate
  the command will fail without the openssl.cnf file being
  installed on the target system.

* Made this package an RRECOMMENDS for libcrypto since:
* libcrypto is a RDEPENDS for the openssl package
* Users can specify a configuration file at another
      location so it is not stricly required and many
      commands will work without it (with warnings)

(From OE-Core rev: 5c3ec044838e23539f9fe4cc74da4db2e5b59166)

Signed-off-by: Chase Maupin <Chase.Maupin@ti.com>
Signed-off-by: Qiang Chen <qiang.chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

dbus: no messages of status command print

/etc/init.d/dbus-1 use "set -e" to let the script exit when any command failes.
This will cause "/etc/init.d/dbus-1 status" command can't display messages when dbus is stopped.

(From OE-Core rev: 9844b5e2a544b2c2f76aac497c3a2cdfcc46577c)

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

attr: upgrade to 2.4.47

Remove the patch that has been integrated upstream

(From OE-Core rev: 39c556756d1af274d7868710ef80663ae8042f97)

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

gnupg: upgrade to 2.0.22

(From OE-Core rev: 87800b8941cd96265d2c3dd25c25d0b4fd9a3572)

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

cups: upgrade to 1.7.0

Rename and consolidate cups16.inc to cups.inc use PACKAGECONFIG
for pam.

(From OE-Core rev: b749bed86efcdc063d720fa095ffa7a7984cd53c)

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

udev: Update touchscreen rule for ID_INPUT_TOUCHSCREEN

Since udev 174 udev has been running input_id as a built-in command and
setting this value in the environment for touchscreens.  Use this logic
to detect when to make a touchscreen0 symlink.

(From OE-Core rev: 5abcfcd4380aacafc45d776f557738fb18089113)

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

connman: upgrade to 1.19

Improvements:
- On startup keep interfaces already configured
with static IPv4 addresses if there is a
service configuration file containing
the very same static values. If not,
take the interface down and flush
the old configuration as before (Jukka Rissanen).
- Handle FallbackNameservers properly;
use them when no other nameservers have been
configured for a service (Patrik Flykt).

(From OE-Core rev: 2d0840780525e9a911e567f45b764850419d49f3)

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>

libcap: fix CAP_LAST_CAP

(From OE-Core rev: 9032c10cc882a96acdfd0739f090d121ab625a18)

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>

glib-2.0: upgrade to 2.38.1

(From OE-Core rev: a738580b108ae11519176e9302913dbd06f0742d)

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>

bluez5: upgrade to 5.10

(From OE-Core rev: d0ebb8382a6799774634252cd615de0787bc6409)

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>

pixbufcache.bbclass: gdk-pixbuf-query-loaders depends on libz

ldd sysroots/x86_64-linux/usr/bin/gdk-pixbuf-query-loaders.real
<snip>
libz.so.1 => /sysroots/x86_64-linux/usr/bin/../../usr/lib/libz.so.1 (0x00007fab55393000)

If zlib-native has not been unpacked, host libz is used which can fail.

(From OE-Core rev: 8422c759ae674856aaaee176eab5a395a620443c)

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

sudo: upgrade to 1.8.8

License checksum updated to reflect additional Copyright for
file getopt_long.c, owned by The NetBSD Foundation, Inc

(From OE-Core rev: e78a5cf3af993d87300d6a527fa14ec6ef9ed79a)

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

curl: upgrade to 7.33.0

(From OE-Core rev: 5b7ae85d1de82de0e78f7aaec01c5510445ceb21)

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

at: upgrade to 3.1.14

Removed 2 patches because they were merged upstream

(From OE-Core rev: ff1707449d851a365e686e6d2cbe822631001147)

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

opkg: putting the service files into PN

(From OE-Core rev: f0ec7f81c1951211f049c342fd6bd1cad424564a)

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>

qt4: add upstream QTBUG-34218/QTBUG-34234 misaligned selection patch

(From OE-Core rev: 3af8f2e0697a9523d3b505ba4c48eca35f6de3a9)

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

wic: Remove selinux_check()

This seems to be an obsolete check - we don't have any problems with
image creation under selinux, so remove it.

(From OE-Core rev: 12e81eceab9e0a483765566ad3791b14718195b5)

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>

strace: remove += assignment of a OVERRIDES variable

In some cases, it's unfit to use "+=" in a conditional appending, we would
end up with the variable being set rather than being appended, which is not
it mean to.

(From OE-Core rev: 7eae843aaa4eaa534a2b825949563ca11bd3680c)

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

pigz: Add pigz to buildtools tarball

When using the tar executable in the buildtools, tar will execute
gzip. If this happens before zlib-native is built, then the gzip
on the host will be used and can fail if the libz in the buildtools
is not compatible. Adding pigz to the build tools avoids this host
contamination.

(From OE-Core rev: af6424e8c2bf3a938fddabc669c0956d68964ed0)

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

debugedit: fix segment fault while file's bss offset have a large number

While ELF_C_RDWR_MMAP was used, elf_begin invoked mmap() to map file
into memory. While the file's bss Offset has a large number, elf_update
caculated file size by __elf64_updatenull_wrlock and the size was
enlarged.

In this situation, elf_update invoked ftruncate to enlarge the file,
and memory size (elf->maximum_size) also was incorrectly updated.
There was segment fault in elf_end which invoked munmap with the
length is the enlarged file size, not the mmap's length.

Before the above operations, invoke elf_begin/elf_update/elf_end
with ELF_C_RDWR and ELF_F_LAYOUT set to enlarge the above file, it
could make sure the file is safe for the following elf operations.

[YOCTO #5356]
https://bugzilla.redhat.com/show_bug.cgi?id=1019707
https://bugzilla.redhat.com/show_bug.cgi?id=1020842

(From OE-Core rev: 35c8b1ac7c3b1e4209b1e30d1dbd1a457286b97b)

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

cogl-1.0: depend on virtual/mesa

- Wayland support depends on wayland-egl, which is provided by mesa.

(From OE-Core rev: a1a379b3c9728a06b086b4c1f06f663f54d7d37d)

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

rpm: Fix FILES_${PN}

In a recent commit, the FILES expression for rpm accidently moved a
large portion of rpm-build into rpm:

http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=2e557556e03f4e86795cf331ccf06acd4c373045

Since rpm-build requires perl, this caused a number of build failures
when the rpm-build package was attempted to be used since it references
/usr/bin/perl. This commit restores the previous contents of the rpm
package.

(From OE-Core rev: a72ca9b8068f71cfed8b7cbe0f08247bd3f0fbe7)

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

scripts/oe-pkgdata-util: fix global name 'debug' is not defined

This global variable is no longer present, so pass in the value
specified via the command line.

(From OE-Core rev: fa90f92e52330a9bf5836c0832412af0927b19a9)

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

update-rc.d.bbclass: Fix host/target test in postinst

When running the postinst script I get a shell warning:

sh: argument expected

and the service is never stopped. This patch fixes the warning
message and stops the service.

Patch v2: Hans Beckérus pointed out that the patch is not correct.
          This version uses the syntax proposed by Hans. I've tested
          that the postinst script works correctly when run on the target,
          both when the init script exists and when it doesn't exist.

(From OE-Core rev: 7e23557835f756b22b95fa7a1926b5d1d21872c3)

Signed-off-by: Jacob Kroon <jacob.kroon@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pcmanfm: upgraded to 1.1.2

Removed cross-compile-fix.patch; pcmanfm doesn't
use xml-purge anymore.

Tested on core-image-sato (qemuarm & qemux86).

(From OE-Core rev: ba93c0faba5f4b5418ecc99da5e04d2ce0642875)

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

grep: upgrade to 2.15

(From OE-Core rev: b4996f3ef827cdc12c39fe58fdb50d7486bfc7f0)

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

scripts: Add ksize.py and dirsize.py

Fixes [YOCTO #5388]

These scripts can be useful when working to reduce the size of the Linux
kernel and the root filesystem.

ksize.py displays the kernel build size by the built-in.o files.

dirsize.py displays the various sizes of the components of the root
directory.

(From OE-Core rev: 26099eb8ac855aa08e5e1a307affe42fe5f43859)

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

libsoup: upgrade to 2.44.1

Tested with midori-browser on qemux86
and qemuarm.

(From OE-Core rev: 67a562d236735d1bb08e7c2f5ea01409f59741d3)

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

libxv: upgrade to 1.0.10

(From OE-Core rev: db0af1b3ca6b4de61dbcb8b264e418fe40c36fae)

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

pseudo: fix library path in FILES_${PN}

libpseudo.so is always installed into ${prefix}/lib/, not ${libdir},
so fix these paths; and skip libdir WARN_QA checking to ignore the
warning in 64bit and multilib enabled system

(From OE-Core rev: 47c7850c025994685aa1811057f4f9a5f0f2a3ae)

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

pulseaudio: Fix build break on armeb

There is no need for += when using append hence removed and added a
leading space appropriately

(From OE-Core rev: fb9cde0fc1a54b073edf5979f4cb7dc297b790fd)

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

tzcode & tzdata: update to 2013h

Update recipes to version 2013h of timezone
code and data.

(From OE-Core rev: 95c9355c35d24b60f44857a8c2e3c9860a91d23b)

Signed-off-by: Joe Slater <jslater@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

meta-toolchain-qt: put QT_CONF_PATH in environment script

This will allow apps using QLibraryInfo class to find qt.conf.

[YOCTO #5339]

(From OE-Core rev: fffa4c37c49b169f663d28612b9251819cef9577)

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

nativesdk-qt4-tools: create qt.conf file

When installing the SDK to another location than the default one, qmake
will look for libraries, headers, etc. in the default location. That's
because the paths are hard-coded in the binary itself. Luckily, QT
allows to override this using a qt.conf file installed in the same
directory with the application executable. However, we already have a
patch that allows for the installation of qt.conf in another place and
read the location from QT_CONF_PATH environment variable.

Hence, install qt.conf in ${sysconfdir}. This will allow other apps, that
use QLibraryInfo class, to find it.

[YOCTO #5339]

(From OE-Core rev: 23f88695683a8e428375a8ccb6be935347a8768c)

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

gnome: set UNKNOWN_CONFIGURE_WHITELIST

This class disable introspection support using --enable-introspection=no but not
all GNOME modules support introspection.

This can cause unknown-configure-options QA warnings, so clarify the option by
using --disable-introspection and add it to the sanity test whitelist.

(From OE-Core rev: a97600330d626eb0ca75178a9011f1a63c8a29f3)

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

xorg-proto-common: set UNKNOWN_CONFIGURE_WHITELIST

This include adds configure options globally but not all X libraries support
them, so to avoid unknown-configure-option warnings add them to the whitelist.

Also change the options from --enable-foo=no to --disable-foo as they're clearer
and the whitelist would have to contain --enable-foo.

(From OE-Core rev: 2544dc51fdebed77ff6eddf27d0ee897c5f5c257)

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

openssh: upgrade to 6.3p1

Removed the following backported patch(es):
 * mac.patch

(From OE-Core rev: ce04ba6a5385ad2d021d472cf9236787c6ea7357)

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

ethtool: upgrade to 3.11

(From OE-Core rev: ba5a3e7ed4e58865c770e9480e322a4c4a939c02)

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

cmake: upgrade to 2.8.12

(From OE-Core rev: 5d63ad569d6beab5beea82aa51c1625d0e1e8231)

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

libgcrypt: remove --without-pth

This isn't recognised by configure, and the random number daemon that requires
Pth isn't enabled.

(From OE-Core rev: 6553c807e19042d3b6add4e7c4b3999fb641abe6)

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

gnome-common: inhert gnomebase not gnome

This recipe doesn't need any of the helpers that come with gnome (GConf, icons,
MIME, etc), so just inherit gnomebase.

(From OE-Core rev: eadd68010f923edc12b4962074850ea724477a64)

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

xorg-lib-common: set UNKNOWN_CONFIGURE_WHITELIST

This include adds configure options globally but not all X libraries support
them, so to avoid unknown-configure-option warnings add them to the whitelist.

(From OE-Core rev: 1f77801e6837b23d9123097dba77ca1374fbbea4)

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

scripts/oe-pkgdata-util: add ability to find a recipe from a target package

Add a "lookup-recipe" command to show which recipe produced a particular
package.

(From OE-Core rev: 4ab561ac3df105b4b6487271b6ccc29445518d52)

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

scripts/oe-pkgdata-util: add ability to look up runtime package names

Add a "lookup-pkg" command to oe-pkgdata-util that can be used to find
the runtime name of a package (after e.g. Debian library package
renaming).

(From OE-Core rev: d923846d91ae307372f1e48483e86807feeeb09d)

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

scripts/oe-pkgdata-util: add ability to search for a target path

Add ability to search for a target path in produced packages, in order
to find which package provides a specific file.

(From OE-Core rev: 0824f2f5cf4e05f82b6986ce6fb22fa1392b7776)

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

scripts/oe-pkgdata-util: check path arguments to ensure they exist

Show an error if the specified paths don't exist.

(From OE-Core rev: eff2690f7976664e6257c1c19c89feea9152eb9e)

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

scripts/oe-pkgdata-util: improve help text and command line parsing

* Use optparse to parse command line
* Make help text actually helpful by describing what each command does
* Drop comment at the top listing the commands which is now superfluous

(From OE-Core rev: feb317513fff638ad7abdba8ab34b8413f0ab055)

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

scripts/oe-pkgdata-util: remove remnants of former pkgdata structure

OE-Core commit 1b8e4abd2d9c0901d38d89d0f944fe1ffd019379 removed the
vendor-os argument from the command line, and the code using the package
architectures, so clean these items up.

(From OE-Core rev: 8b9ee57afbbcf633cba66e4b6e8ab7339ad6d391)

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

pigz: bump to 2.3.1

Update pigz to latest release - 2.3.1
Drop ldflags.patch as it has been merged upstream

(From OE-Core rev: 8081dcb03f54efd551d1c8fe8a0484f8270053e0)

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

harfbuzz: upgrade to upstream version 0.9.22

Code that was under ICU license has been removed.

(From OE-Core rev: 12e2e82a7cfcbcb989b5254b2148320db69a820d)

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

cairo: upgrade to upstream version 1.12.16

png.patch dropped as it's part of upstream now

(From OE-Core rev: c645545d245e77a83a6ce2feb7211fbea0242d46)

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

gtk+: upgrade gtk+ to upstream version 2.24.22

(From OE-Core rev: f7c43757a0a2ad3b24c9f2ce3ad13059d929ba39)

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

cracklib: cracklib-native should not depend on zlib

(From OE-Core rev: 89d7d46947d9bb8c7bf568c65e52d5bbe159027f)

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

nfs-utils: nfsserver restart should kill and recreate nfsd kernel threads

nfsserver restart without killing kernel threads worked when portmap
was the rpc publishing process and portmap was restarted.
When rpcbind replaces portmap, nfsserver restart in this way does not
work after an rpcbind restart.

Steps to reproduce:
1). Make ext3 filesystem image on local host.
cd /root
dd if=/dev/zero of=test bs=1024K count=50
mkfs.ext3 -F test

2). runqemu qemux86-64
mkdir /mnt/wrtest
mount -t ext3 -o loop test /mnt/wrtest
echo "/mnt/wrtest *(sync,rw,no_subtree_check,no_root_squash)" > /etc/exports
/etc/init.d/rpcbind restart
/etc/init.d/nfsserver restart
showmount -e localhost
mkdir wrtest
mount -t nfs localhost:/mnt/wrtest wrtest

mount: mounting localhost:/mnt/wrtest on wrtest failed: Connection refused

Modifying the nfsserver script to kill and restart kernel threads on
restart makes the problem go away and is consistent with current
RHEL/SUSE and Ubuntu/Debian mechanisms of handling the nfs server.

(From OE-Core rev: 1a96b8d7dfc490fc61bbd470a8b09065750cd563)

Signed-off-by: Rich Dubielzig <rich.dubielzig@windriver.com>
Signed-off-by: Qiang Chen <qiang.chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libfm: allow build with automake-1.14

Do not consider automake warnings about future incompatibility
errors.

(From OE-Core rev: f7281ebf1fd593805f2cc10828ecb9723e1554d0)

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

libfm: upgrade to upstream version 1.1.2.2

(From OE-Core rev: 47f803a9cc6a7b43f10775172e6bc89632af322e)

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

libcheck: allow build with automake-1.14

Do not consider automake warnings about future incompatibility
errors.

(From OE-Core rev: c7ead34c51e77067cb5dd95cf0aa1bc35567a0f3)

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

libxcb: fix build with automake-1.14

(From OE-Core rev: f8c9a88082f5acf7835647a98a394d5a44fc556f)

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

nfs-utils: Stop rpc.statd correctly

An incorrect process name in the nfsserver initscript prevented
rpc.statd from being shut down.

root@qemux86-64:~# /etc/init.d/nfsserver start
creating NFS state directory: done
starting 8 nfsd kernel threads: done
starting mountd: done
starting statd: done

root@qemux86-64:~# ps | grep rpc.statd
  650 root     10532 S    /usr/sbin/rpc.statd
  654 root      4720 S    grep rpc.statd

root@qemux86-64:~# /etc/init.d/nfsserver stop
stopping statd: done
stopping mountd: done
stopping nfsd: done

root@qemux86-64:~# ps | grep rpc.statd
  650 root     10532 S    /usr/sbin/rpc.statd
  662 root      4720 S    grep rpc.statd

As this daemon drops a pid file,simply use that instead.
Also add some initialization checks so the daemons are not
left partially started in the absence of kernel nfsd support.

(From OE-Core rev: 37e70a28e9cfc773bd70f09d7129295ce891ae18)

Signed-off-by: Andy Ross <andy.ross@windriver.com>
Signed-off-by: Qiang Chen <qiang.chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

perf: flag __SANE_USERSPACE_TYPES__ to include int-ll64.h for mips64

As the same reason to powerpc64, mips64 also need the flag.

(From OE-Core rev: d6f3cb0d71c3b6739365f085b6d5a5e20f329fa5)

Signed-off-by: Wenzong Fan <wenzong.fan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

screen: dynamically add or remove a entry to/from /etc/shells

1 Add base-files to RDEPENDS.
2 Add pkg_postinst and pkg_postrm to add and remove a entry in
/etc/shells.

(From OE-Core rev: 95618d6d5c4f8e5f0aec632d35e02ad5a33a1f75)

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

bash: dynamically add or remove a entry to/from /etc/shells

1 Add base-files to RDEPENDS.
2 Use ${base_bindir} in regex to match bash path.
3 Add pkg_postrm to remove the entry from /etc/shells that added by
pkg_postinst.

(From OE-Core rev: c3f93357e2d3ece910ff0e2d18eba3fb94fb5c3c)

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

base-files: remove invalid entries from /etc/shells

It's more reasonable and secure to keep /etc/shells a minimal file, and
then entries for valid shells be added dynamically to the system, only if
the packages that provide them are supported.

(From OE-Core rev: a1d6b55bfa5daa9ba5fa9a7a99dd8872284b6ad8)

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

relocate_sdk.py: Allow script to work with Python 2.4 and 3.

Python 2.4 does not support the 'b' string literal or the
keyword 'as' in exception handling. Python 3 does not accept
the old method of exception handling and defaults to unicode.
The b() function converts strings to bytes on Python 3 and
using sys.exc_info() avoids the exception handling syntax.

(From OE-Core rev: 1e2ec5f576f167673d7980737826987fefdc74a9)

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

rpm: Add nativesdk to BBCLASSEXTEND

Modified FILES-paths since nativesdk can't handle hardcoded paths,
Also added *.real binaries to packaging since this is not done when built
as native.

As far as /var having to be hardcoded, I have a hard time seeing someone
modifying bitbake.conf to place localstatedir somewhere else than /var.

If there exists a target/nativesdk portable way of hardcoding FILES,
please let me know, and we'll do it that way.

Cleanup of FEATURES, since it was the same for native & target.

(From OE-Core rev: 39383e9bf1aa5e29d33d8af7f8e690d2238fd14f)

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

runqemu-extract-sdk: add --numeric-owner option to tar command

If the same username exists on both target and the build host, but
the uids differ, and we start target via NFS, then the uid for the
user will be incorrect on target.

For example, if postfix's uid on host is 119 and on target is 1024,
then if we start target via NFS, the uid for postfix will be 119.

The root cause is that when we use runqemu-extract-sdk to generate
the NFS rootfs for later use, the tar command will respect the username
instead of uid. So if PSEUDO_PASSWD environment is not set correctly,
the host /etc/passwd will be used, resulting in wrong uids.

The situation for gid is completely analogous to that of uid.

It's almost impossible for the runqemu-extract-sdk to guess the correct
location of the needed password file merely based on the target tarball
name.

This patch solves this problem by adding the '--numeric-owner' option
to the tar command so that the uid/gid will be used when extracting the
tarball using runqemu-extract-sdk. In this situation, we'll always get
the correct uid/gid after extracting the tarball.

[YOCTO #5364]

(From OE-Core rev: acce6ff1a77cfd29e3868faa89b120becb58bbbf)

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