zypper: Fix build on uclibc

uclibc does not have rpmatch() so we define
it if zypper is being compiled for uclibc

(From OE-Core rev: 1ea9b31f304e54ce4b7c2d38582d1775a46cd8c6)

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

rpm: pass lrt and lpthread to link step explicitly for uclibc

These need to appear in DT_NEEDED for librpmio.so
for uclibc somehow it does not get added to link
cmdline so we do it explicitly.

(From OE-Core rev: e5d141fff4a96fbdac63586638fb286f263af216)

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

sat-solver: Fix build on uclibc

futimes is not available on uclibc so use utimes
qsort is also not as expected by sat-solver therefore
for uclibc we resort to using internal version of
qsort

(From OE-Core rev: e0800be814c8f8adf8d6682aacea1d3ce9b14c9e)

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

rpm: Use link time check for libssp

-fstack-protector needs libssp to link with
so when checking for this option support we
need to find if libssp is staged in root file
system

(From OE-Core rev: 492297fce0b80547d3dac9f611f463c6cc95160b)

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

rpm: Fix compilation on uclibc

rpmatch is not there in uclibc therefore add logic to configure
to detect it.

x* wrappers need to be defined for uclibc as well.

(From OE-Core rev: 0cd3a4f3cfd0eb70915e3b59be8d04ee2d92749e)

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

site/x86_64-linux-uclibc: Cache definitions of guint32/guint64

Without this glib 2.32.x wont build.

(From OE-Core rev: 6cb8bb9a338097dd63be63732b8e67fc87eb1a5e)

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

uclibc_0.9.33.bb: Upgrade to latest git

This is needed otherwise glib-2.0 wont run
correctly

(From OE-Core rev: c495e31f278e49e15fb48285f3b68794f6cf888e)

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

debian.bbclass: invoke target objdump

Using the host objdump can lead to errors like:

objdump: library.so.1.0.0: File format is ambiguous
objdump: Matching formats: elf32-littlearm elf32-littlearm-symbian elf32-littlearm-vxworks

with certain configurations of binutils.

(From OE-Core rev: 6d761ae17ddbd3d936e7fe985b40825ad62b2418)

Signed-off-by: James Limbouris <james@digitalmatter.com.au>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

tzdata, matchbox-session, mini-x-session: replace RCONFLICTS with RCONFLICTS_

(From OE-Core rev: 8f4e7dfd8b60a10411261666b4da2b4aff1c4231)

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

recipes: replace CONFLICTS with RCONFLICTS_${PN}

(From OE-Core rev: fbcda6eff010e0a39cc0145a514e7de99f475145)

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

qt4-embedded: fix QT_ARCH usage in QT_CONFIG_FLAGS

After the change to shell style functions (from python style), the
ability to use oe_filter_out on QT_CONFIG_FLAGS got broken.

This patch solves that by referring to QT_ARCH in a more correct way.

(From OE-Core rev: abd5d5a712686a379a1a1da29d78a0f374f13e33)

Signed-off-by: Anders Darander <anders@chargestorm.se>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

mdadm: Don't depend on ansidecl.h header

We don't DEPEND on binutils for ansidecl.h so ensure we should never use the header.
This makes builds determinstic and means something like:

bitbake binutils
bitbake mdadm -c compile
bitbake binutils -c clean
bitbake libxml2

doen't fail to build.

(From OE-Core rev: 5b51872ec4c2400ae60f434478f04e9aca3ccb7d)

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

directfb: fix Upstream-status -> Upstream-Status

(From OE-Core rev: 45b40d483c675369cdc164babd0db5c87adf0de4)

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

cleanup-workdir: update help text

Update the help text to tell user that the files and dirs under WORKDIR
which are not created by Yocto will be deleted.

(From OE-Core rev: dcd2ebc5d63965a7ad6e714406149f63ffb4a704)

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

cleanup-workdir: only deal dirs related to current arch

Some users may build for different archs under same workdir, so they
don't want to clean the dirs not related to current arch.

Run command 'bitbake -e' with selected packages to get the dirs related
to current arch then clean them.

Update the way to get the WORKDIR by parsing the IMAGE_ROOTFS by the
way.

(From OE-Core rev: a16727ebc341e0a0ce59a5200dc774cf672593ee)

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

cleanup-workdir: replace commands with subprocess

Use modules subprocess to run command instead of module commands.

(From OE-Core rev: 33f18965bbeeec47f694f2aa165e5e07eadb7ab7)

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

cleanup-workdir: update the way to check obsolete dirs

Update the way to check obsolete directories.

According to package and its version construct a list of all packages'
current build directory. If any directory under $WORKDIR/*/ is not in
the list will be removed.

At same time, all the files(vs. directory) under $WORKDIR and
$WORKDIR/*/ will be removed because they are not created by poky.

(From OE-Core rev: 4d2920dee32bbc5d12ed98234de096d28d29415b)

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

meta-skeleton: add linux-yocto-custom reference recipe

Adding a skeleton/refrence recipe for using a subset of the yocto
kernel tools against kernel git repositories.

Sample/reference configuration fragments, features and a patch are
provided and documented in the recipe.

From the recipe itself:

   Provides an example/minimal kernel recipe that uses the linux-yocto
   and oe-core kernel classes to apply a subset of yocto kernel
   management to git managed kernel repositories.

 Notes:

   kconfig(s): the kernel must be configured with a defconfig, or via
               configuration fragment(s). Either of these can be added
               via bbappend.
   patches: patches can be merged into to the source git tree itself, added
            using standard bbappend syntax or controlled via .scc feature
            descriptions (also via bbappends)

   example configuration addition:
            SRC_URI += "file://smp.cfg"
   example patch addition (for kernel v3.4 only):
            SRC_URI += "file://0001-linux-version-tweak.patch
   example feature addition (for kernel v3.4 only):
            SRC_URI += "file://feature.scc"

 Warning:

   Building the sample kernel tree (kernel.org) without providing any
   configuration will result in build or boot errors. This is not a bug
   it is a required element for creating a valid kernel.

[YOCTO #2397]

(From OE-Core rev: 4b6f29ddb00480896f47d96ea376f3a6f6d8451e)

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

libzypp: Fix build with uclibc

cstdio is included indrectly with eglibc based systems
but not with uclibc based systems and use of functions
like ::eof are then reported as warnings. Therefore
we include cstdio explicitly.

(From OE-Core rev: ded71e4904ed8df23fe4ca5c5e4516a8a6e9fc76)

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

busybox: Fix syslog configuration file

When update alternatives was modified, the syslog configuration file
alternative was incorrectly defined to be "busybox".  Fix this by
enabling the proper target file.

[YOCTO #2557]

(From OE-Core rev: 935f2c02809814de2b903a704707d7bc85bbefdf)

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

core-image-gtk-directfb, qt4e-demo-image: Remove module-init-tools from IMAGE_INSTALL list

As module-init-tools package does not exist anymore there's no reason to
have it in the IMAGE_INSTALL package list. If RPM package manager is
used, build will fail with:

Unable to find package module-init-tools (module-init-tools)!

[Yocto #2580]

(From OE-Core rev: 85d612737d050d4633013c706d9ad6b8169814af)

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

gcc-cross: argument list too long when do_install

There would be an "Argument list too long" error when the length of
tmpdir is longer than 190, this is caused by:

headers=`echo $(PLUGIN_HEADERS) | tr ' ' '\012' | sort -u`

The PLUGIN_HEADERS is too long before sort, so the "echo" can't handle
it, use the $(sort list) of GNU make which can handle the too long list
would fix the problem, the header would be short enough after sorted.
The "tr ' ' '\012'" was used for translating the space to "\n", the
$(sort list) doesn't need this.

This doesn't impact the output, so it doesn't need the PR bump.

[YOCTO #2591]

(From OE-Core rev: b9b75e6cf31991c046713194f75bc61027558464)

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

directfb: added missing Upstream-status

(From OE-Core rev: 63d1ae1c2828cda03828b4c3fdde037d7a23f1e4)

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

pango: added upstream-status status.

Added missing upstream status.

(From OE-Core rev: 5938eb43b3d7de65e58610f9d5044fadd42cc225)

Signed-off-by: Edwin Zhai <edwin.zhai@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

distrodata: cleanup after moving distro_tracking_fields

Remove some of the older fields that are not in the new list as we
are able to better automagically generate this directly from the
recipe files the extra files will go away.

To use this, one will have to include the appropirate files, such
as maintainers.inc, upstream_status.inc

(From OE-Core rev: f159ea18cf2a3f0c94d7324a2b63a6f0aabf73f2)

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

strace: update to 4.7

Removed the x32 specific patches, since they seem to be fully
integrated into 4.7. Also removed the sigmask patch, since the
new version doesn't seem to use sigmask anymore.

(From OE-Core rev: fec92fb65a0135b17fea49a239bd9a73c96bba51)

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

beecrypt: add patch for build issue on Fedora 17

../../x86_64-linux-libtool --tag=CXX   --mode=compile g++ -DHAVE_CONFIG_H -I. -I../.. -I../../include  -isystem/intel/poky/builds/f17/tmp/sysroots/x86_64-linux/usr/include  -isystem/intel/poky/builds/f17/tmp/sysroots/x86_64-linux/usr/include -O2 -pipe -Wa,--noexecstack -c -o Properties.lo Properties.cxx
x86_64-linux-libtool: compile:  g++ -DHAVE_CONFIG_H -I. -I../.. -I../../include -isystem/intel/poky/builds/f17/tmp/sysroots/x86_64-linux/usr/include -isystem/intel/poky/builds/f17/tmp/sysroots/x86_64-linux/usr/include -O2 -pipe -Wa,--noexecstack -c Properties.cxx  -fPIC -DPIC -o .libs/Properties.o
In file included from ../../include/beecrypt/c++/util/Hashtable.h:42:0,
                 from ../../include/beecrypt/c++/util/Properties.h:36,
                 from Properties.cxx:25:
../../include/beecrypt/c++/util/AbstractSet.h: In instantiation of 'bool beecrypt::util::AbstractSet<E>::equals(const beecrypt::lang::Object*) const [with E = beecrypt::util::Map<beecrypt::lang::Object, beecrypt::lang::Object>::Entry]':
Properties.cxx:228:1:   required from here
../../include/beecrypt/c++/util/AbstractSet.h:59:27: error: 'containsAll' was not declared in this scope, and no declarations were found by argument-dependent lookup at the point of instantiation [-fpermissive]
../../include/beecrypt/c++/util/AbstractSet.h:59:27: note: declarations in dependent base 'beecrypt::util::AbstractCollection<beecrypt::util::Map<beecrypt::lang::Object, beecrypt::lang::Object>::Entry>' are not found by unqualified lookup
../../include/beecrypt/c++/util/AbstractSet.h:59:27: note: use 'this->containsAll' instead
../../include/beecrypt/c++/util/AbstractSet.h: In instantiation of 'bool beecrypt::util::AbstractSet<E>::equals(const beecrypt::lang::Object*) const [with E = beecrypt::lang::Object]':
Properties.cxx:228:1:   required from here
../../include/beecrypt/c++/util/AbstractSet.h:59:27: error: 'containsAll' was not declared in this scope, and no declarations were found by argument-dependent lookup at the point of instantiation [-fpermissive]
../../include/beecrypt/c++/util/AbstractSet.h:59:27: note: declarations in dependent base 'beecrypt::util::AbstractCollection<beecrypt::lang::Object>' are not found by unqualified lookup
../../include/beecrypt/c++/util/AbstractSet.h:59:27: note: use 'this->containsAll' instead

[YOCTO #2593]

(From OE-Core rev: 93aa773c4c03c21a70d7d53cf93e43416e17007c)

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

gcc-target: ensure target packages have the correct virtual package names

Resolves this error during do_rootfs:
|  * satisfy_dependencies_for: Cannot satisfy the following dependencies for task-core-standalone-sdk-target:
|  * eglibc (>= 2.15) *
|  * opkg_install_cmd: Cannot install package task-core-standalone-sdk-target.
NOTE: package core-image-sato-sdk-1.0-r0: task do_rootfs: Failed

(From OE-Core rev: 86e3b90d1963fa6982869e005d7858222a8527d5)

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

eglibc: added ac_cv_path_ to CACHED_CONFIGUREVARS

On Fedora 17, bash has moved to /usr/bin/bash and the configure process finds it
on the host machine there, this ensures that it is set correctly for the target.

[YOCTO #2363]

(From OE-Core rev: 133c2a12d13028a3829c87a59a2ff4fe21f57260)

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

quilt: added ac_cv_path_BASH to CACHED_CONFIGUREVARS

On Fedora 17, bash has moved to /usr/bin/bash and the configure process finds it
on the host machine there, this ensures that it is set correctly for the target.

[YOCTO #2363]

(From OE-Core rev: 958934049e4b3b84e5eaa9c6b0d672cf2c4a693e)

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

curl: enable ssl support

This patch enables ssl support for curl to allow git to clone from
https / ssl sites. We do not want to enable gnutls for native or
nativesdk, as it adds additional dependency and increase build time

[YOCTO #2532]

(From OE-Core rev: 653786096cb13560537905410e737e0dd9a9cf57)

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

busybox: Add check for wifi/bluetooth to enable rfkill

rfkill is a usefull tool when you have a device with bluetooth and wifi enabled

[YOCTO #2494]

(From OE-Core rev: f71258f0ef96802e714b6f12ecc8c17842d6cf87)

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

opkg 0.1.8: do_fetch failed since no checksum specified

* We use meta/recipes-devtools/opkg/opkg_svn.bb by default, the
  opkg_0.1.8.bb failed since no checksum specified.

* The add_vercmp.patch in both opkg/ and opkg-0.1.8/, remove the one
  in opkg-0.1.8/

This patch doesn't impact the output, so I think that we don't have to
increment the PR.

[YOCTO #2498]

(From OE-Core rev: 316fac638646a50149f36ec4687b5a2a6d2e83b1)

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

linux-yocto/3.4: create linux-yocto-rt recipe

Adding the 3.4 variant of the linux-yocto-rt recipe. This updates
to 3.4.1-rt9, and builds and boots on the supported targets.

(From OE-Core rev: 7f007994cbc5f188439ad2b26f60dbcc1c2763bd)

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

linux-yocto/3.4: update yaffs2 and drop rc6

Updating the 3.4 SRCREVs to pickup a yaffs2 update and the removal
of a feature that was required in the 3.2 kernel tree.

1/2 [
    meta: rc6: remove rc6 patches for snb

    The sandybridge rc6 patches are part of the released v3.4 kernel.
    Hence there is no need to keep these patches in the 3.4 linux
    yocto kernel repository.

    (From OE-Core rev: 7822565a3a3ae906112c88f7e59f396c7fdabd4c)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
]

2/2 [
    yaffs2: update core support

    Uprev yaffs2 to latest version as of 2012-05-29

    To include the following commits:

    662466a yaffs2 checkpointing: Add further checkpoint data sanity checking
    d9cae0a yaffs2: Make checkpoint block scanning skip bad blocks.
    eb12d56 Remove trace that should not be in mounting.
    9ee5d06 yaffs for u-boot: Cleanups to make patches acceptable for u-boot
    83d9993 Merge remote branch 'origin/master'
    07353dd Clean up checkin. Do not use
    00ebf30 Update timothy test Makefile
    1658295 yaffsfs: Allocate directory searches from a table rather than via malloc.
    8dae7f1 Fix compilation of python tests.
    14ff4e1 yaffs u-boot: Fix problem if an illegal nand chip number is entered.
    8b34846 Set up u-boot glue code and patching scripts.
    1a17e3a Add more descriptive comment for using the yaffs commands.
    80f8530 Add files to hook up yaffs direct to u-boot
    95ff7ef Merge branch 'master' of ssh://www.aleph1.co.uk/home/aleph1/git/yaffs2
    e49491a Put both the summary and checkpoint version stamps together, update checkpoint version
    468d72b Fix compilation of mkyaffsimage and mkyaffs2image
    87166df Fix resizing of large files.
    eeb0f17 Fix summary header validation
    d308180 Add a summary header
    7862c13 Remove extra line from banner.
    d14c6a6 Add script to handle common Linux and direct files
    6518138 Change Makefiles to support new way to structure yaffs_strxxx renaming for yaffs direct
    8429784 Update README about linux patching
    6ee2533 Update Makefiles for new sed-based way case-insensitive handling
    2369fad Set up new version of case insensitive code using sed
    4d8ae2f Add large file support changes to yaffs_vfs_single.c
    ce0a5fb Merge branch 'big-files':  Merge in large file support
    80d031a Add Waldemar's patches for Linux 3.2
    a7ae50c yaffs large file support: Fix max file size issue that prevented the scanning working.
    a7cb3a3 WIP Changes to tests
    2360446 yaffs large file support: Add more tests
    0ed6086 yaffs large file support: Saves and restores large files fine under yaffs direct.
    fc4028b large file storage WIP. Can save and verify a large file under yaffs direct.
    20de150 WIP large file support
    b4d93e2 First cut of changes to support large files.

Signed-off-by: Michel Thebeau <michel.thebeau@windriver.com>
]

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

pybootchartgui: Fix the filename and add a default format

* Fix teh output filename to make it easy to use
* Add a default output format (svg)
* Fix the usage message
* Fix the version to v1.0.0

Currently, the help messages are:

$ ./pybootchartgui.py --help
Usage: pybootchartgui.py [options] /path/to/tmp/buildstats/<recipe-machine>/<BUILDNAME>/

Options:
  --version             show program's version number and exit
  -h, --help            show this help message and exit
  -i, --interactive     start in active mode
  -f FORMAT, --format=FORMAT
                        image format: svg, pdf, png, [default: svg]
  -o PATH, --output=PATH
                        output path (file or directory) where charts are
                        stored
  -s NUM, --split=NUM   split the output chart into <NUM> charts, only works
                        with "-o PATH"
  -n, --no-prune        do not prune the process tree
  -q, --quiet           suppress informational messages
  --very-quiet          suppress all messages except errors
  --verbose             print all messages

[YOCTO #2403]

(From OE-Core rev: 138c2c31e41e3f1803b7efbedf78326d71821468)

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

pybootchartgui: split the output chart into multiple ones

Split the output chart into multiple ones to make it more readable, it
only works with "-o path", which means that it doesn't work if the user
doesn't want to save the chart to the disk. For example:

$ ./pybootchartgui.py /path/to/tmp/buildstats/core-image-sato-qemux86/201205301810/ -f svg -s 5 -o /tmp/
bootchart written to /tmp/bootchart_1.svg
bootchart written to /tmp/bootchart_2.svg
bootchart written to /tmp/bootchart_3.svg
bootchart written to /tmp/bootchart_4.svg
bootchart written to /tmp/bootchart_5.svg

[YOCTO #2403]

(From OE-Core rev: 04a34899e1c15a70babd97a3a59ccb9f8af05bad)

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

pybootchartgui: make the build profiling in pictures

The original patch is from Richard, I rebased it to the up-to-date
upstream code, here are the original messages from him:

We have just merged Beth's initial buildstats logging work. I was
sitting wondering how to actually evaluate the numbers as I wanted to
know "where are we spending the time?".

It occurred to me that I wanted a graph very similar to that generated
by bootchart. I looked around and found pyboootchartgui and then hacked
it around a bit and coerced it to start producing charts like:

http://tim.rpsys.net/bootchart.png

which is the initial "pseudo-native" part of the build. This was simple
enough to test with.

I then tried graphing a poky-image-sato. To get a graph I could actually
read, I stripped out any task taking less than 8 seconds and scaled the
x axis from 25 units per second to one unit per second. The result was:

http://tim.rpsys.net/bootchart2.png
(warning this is a 2.7MB png)

I also added in a little bit of colour coding for the second chart.
Interestingly it looks like there is more yellow than green meaning
configure is a bigger drain on the build time not that its
unexpected :/.

I quite enjoyed playing with this and on a serious note, the gradient of
the task graph makes me a little suspicious of whether the overhead of
launching tasks in bitbake itself is having some effect on build time.
Certainly on the first graph there are some interesting latencies
showing up.

Anyhow, I think this is the first time bitbake's task execution has been
visualised and there are some interesting things we can learn from it.
I'm hoping this is a start of a much more detailed understanding of the
build process with respect to performance.

[YOCTO #2403]

(From OE-Core rev: 6ea0c02d0db08f6b4570769c6811ecdb051646ad)

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

pybootchartgui: add the original code

This is from:
http://pybootchartgui.googlecode.com/files/pybootchartgui-r124.tar.gz

Will modify it to make the build profiling in pictures.

Remove the examples since they would not work any more, and they cost
much disk space.

[YOCTO #2403]

(From OE-Core rev: 1f0791109e1aed715f02945834d6d7fdb9a411b4)

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

coreutils: Fix a typo in v6.9 (bracket not braket)

(From OE-Core rev: 66cc81d2a5f5c74e4a44e79521a791f402deeb26)

Signed-off-by: Andrei Gherzan <andrei@gherzan.ro>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libpcre: updated to version 8.30

Tested by compiling lighttpd. The license file changed because the
copyright notice was updated to cover 2012.

(From OE-Core rev: 1ae74c5b1485ac438b8f83b2a7d162ef63b089d2)

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

distro_tracking_fields: move to meta-yocto as seperate files

We are going to take a phased approach of breaking up the
distro_tracking_fields, first is to remove it from oe-core,
then create new files in meta-yocto.  We are doing this because
the distro_tracking_fields is getting unweildly and some of
the data can be part of a burn down list instead continually
stored, thus it will get split up.

(From OE-Core rev: e8808c85a6bba8422cd82631b2bbc367543e4cdd)

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

gcc-configure-target: Fix sysroot option breakage I introduced (sorry)

(From OE-Core rev: ba380813c0198f95b1e786aa7f49b80c68cfe9b9)

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

methodpool: Improve method already seen error message

The current error message is confusing, this improves it to explain the problem
and the possible ways to resolve it.

[YOCTO #2530]

(Bitbake rev: d316f28ed725ff40daa8771c1aa224ac46d5b224)

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

fetch2/local: Ensure quoting is handled correctly from the url

The commit 420eb112a4f (fetch2: quote/unquote url paths) caused a regression
since local file urls containing quoted characters are no longer being handled
correctly. This fixes things so the url is dequoted correctly, fixing various
fetcher failures with recipes like "gtk+".

[YOCTO #2558]

(Bitbake rev: 95c0595d66b3f8a5f0716662ba2a878600f312ea)

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

fetch2: Fix missing output from stderr in fetcher logs

There are actually two problems to fix

1) The exception for bb.process.ExecutionError must be processed first
   because it is a derived from the bb.process.CmdError class and
   we never reach the ExecutionError otherwise.

2) The stderr needs to be printed as well as stdout to determine
   the root cause of a fetch failure.

The example I have is that I got a log that looked like:

--
ERROR: Function failed: Network access disabled through
  BB_NO_NETWORK but access requested with command
   /usr/bin/env wget -t 5 -nv --passive-ftp --no-check-certificate -P
   /localds 'http://downloads.yoctoproject.org/[...CLIPPED...] url None)
--

That really didn't tell me much, but with this patch I get error above
plus the following:

--
STDERR: /net/[...CLIPPED...]kernel-tools.git: Read-only file system
--

(Bitbake rev: af2133a04e1f4b22b181adf9c71f89c439bb88cf)

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

Hob: fixed issue about 'select a base image' combo not sensitive

The combo view is not sensitive, after select recipes come back.
so fixed this issue.

[YOCTO #2500]

(Bitbake rev: debcea9c063f7bddcba1d4ed47edbe173ef1d9cf)

Signed-off-by: Liming An <limingx.l.an@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

Hob: fixed an issue about no 'set_page' function in hob failed page show

Due in mistake deletion in previous patches, if building failed, will report
the error of no 'set_page', so add that function.

(Bitbake rev: b046f4316ee7f2afaa0bb723e4216714d9cb87a0)

Signed-off-by: Liming An <limingx.l.an@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

image-writer: tweak some strings

Tweak some strings in accord with the script name 'image writer', and
remove extra noun 'button'.

(Bitbake rev: d2d65335e8fc2417ebd3910f645cc00a4abfbeaa)

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

ui/crumbs/hig.py: simplify subprocess call

According to Darren's suggestion, subprocess.call() will wait until the
subprocess complete. So simplify the code, use subprocess.call() to wait
command complete.

(Bitbake rev: 3dea9d5a5af82d99d337238981807a4cb1bd5b93)

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

ui/crumbs/hig.py: tweak strings

Add comments for DeployImageDialog 'standalone' part of codes to make it
easy to read.
And tweak some strings.

(Bitbake rev: b0e604d3d6dbfcaeac0a2e0e4399aeb4327f8cbc)

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

connman: bump PR and add some comments

The comments are for informing anyone who wants to change USERADD_PARAMs
to make sure it's in sync with the one in xserver-nodm-init, in order
for rootless X to work.

(From OE-Core rev: c4d0b1c5438039dab07103c151c0c5a76a4c9499)

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

sanity.bbclass: warn the user if BBPATH contains wrong entries

If BBPATH references the working directory, the user is warned and asked
to fix the problem.

[Yocto #1465]

(From OE-Core rev: 73f10ae955ba217078fa2c5288b736ae8a30b184)

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

webkit-gtk: Apply work around for all PowerPC targets

The current patch for bug #1570 only applies to qemuppc but should be
applicable for all PowerPC targets.  Also update the patch so that
only one language backend, either ICU or PANGO, is built.

Also remove some old customizations (dependencies on darwin) as these
should now be handled in a layer specific .bbappend file.

(From OE-Core rev: 2782caf0f07f61cab92d2981c5e82914f22bd700)

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

connman: make useradd parameters consistent with rootless X

Both connman and rootless X share the same xuser. However, useradd
parameters are not consistent and if connman runs useradd first,
rootless X will be broken.

[Yocto #2467]

(From OE-Core rev: 28087e7a9ddfd2d118aa907a29ed522bb8cbfc16)

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

systemtap: update Upstream-Status and use correct syntax

patch was applied [1]

http://sourceware.org/git/gitweb.cgi?p=systemtap.git;a=commit;h=41df51cb6a6d3f09f8b57cb216dffa5ba335e326

(From OE-Core rev: 7fbe687da742be638d8c19f522d22fe9d6e8147a)

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

gcc-configure-common: Use build time sysroot for target gcc

Otherwise it assumes / as sysroot on build host and starts
searching for directories in there.

[RP: Tweak to add missing = for consistent syntax]
(From OE-Core rev: d132bcc41d9956f8cc171e34232d991e2b444689)

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

lsof: Fix cross 32 -> 64bit ranlib host contamination problem

When building on a 32bit host OS and building 64 bit binaries,
the cross ar and ranlib must be called, or you get an error
as follows:

   x86_64-wrs-linux-gcc    -m64 --sysroot=/opt/build/tmp/sysroots/qemux86-64 -o lsof dfile.o dmnt.o dnode.o dproc.o dsock.o dstore.o arg.o main.o misc.o node.o print.o proc.o store.o usage.o util.o -Wl,-O1 -Wl,--hash-style=gnu -Wl,--as-needed -L./lib -llsof
   ./lib/liblsof.a: could not read symbols: Archive has no index; run ranlib to add one
   collect2: error: ld returned 1 exit status

(From OE-Core rev: 5c257ee55f9e0dd66fc1bbd6e9c08918e8820178)

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

openjade-native: Ensure we reautoconf the package

Currently since configure.in in is in a subdirectory, we don't reautoconf the
recipe. We really need to do this, to update things like the libtool script used
and fix various issues such as those that could creep in if a reautoconf is
triggered for some reason. Since this source only calls AM_INIT_AUTOMAKE to gain the
PACKAGE and VERSION definitions and that macro now errors if Makefile.am doesn't
exist, we need to add these definitions manually.

These changes avoid failures like:

----
| ...
| DssslApp.cxx:117:36: error: 'PACKAGE' was not declared in this scope
| DssslApp.cxx:118:36: error: 'VERSION' was not declared in this scope
| make[2]: *** [DssslApp.lo] Error 1
----

(From OE-Core rev: 6f0d830ed047f700702645d454054dfd6fea6f25)

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

cmd1.bbclass: Ensure ncurses is built and used for menuconfig tasks

Currently, the task just exits if something goes wrong. This adds the
ncurses-native dependency. It also adds a small delay before closing the
window so any messages displayed there can be seen.

Trying to get the kernel build system to correctly find and link with
our copy of ncurses is some kind of nightmare. I ended up having to add
it to HOST_LOADLIBES globally for this task which is rather nasty but I
couldn't find any other way.

[YOCTO #2513]

(From OE-Core rev: fe417e8a4d625c6933de72163d2fee52ac47f571)

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

qt4: Drop 4.7.4

We have 4.8.1 now and 4.7.4 is obsolete.

(From OE-Core rev: cb945328fe019c246c520904c02ca29d3798df07)

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

rpm: Ensure pkgconfig deps don't stray from the package being built

Force pkg-config to only look in the install directory, and not be
able to fall back to system pkg-config directories for resolution.

(From OE-Core rev: 7afcfeb1db00c800485b2f3a89dad1ed84083979)

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

package_rpm: Verify that the package manifests are not empty

An empty manifest will cause a failure condition.  We ensure that any install
actions occur only with a manifest larger then size 0.  Also ensure that padding
is added to the end of the manifest, instead of the beginning to enable this
size check.  (Padding is required for very small manifest files..)

(From OE-Core rev: 9b115181a8b10ec1eb8eac355405695f63e2e011)

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

createrepo: For compatibility w/ RPM 5.4.9 remove RPMVSF_NOSIGNATURES flag

The RPMVSF_NOSIGNATURES flag was removed from RPM5 - 5.4.9 as a result of
additional -requires- package validation to help protect from malicious
packages.  The createrepo "dump" script has been updated to no longer use
this flag.  This remains compatible with prior versions of RPM.

(From OE-Core rev: 24786ccbbdd1188aad4d8508e1757d38798b6f81)

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

rootfs_rpm.bbclass: Update database configuration file

Move the creation of the database configuration files to just before we
perform the install actions.  This ensures they will exist even in SDK or
other non-target filesystem images.

(From OE-Core rev: 113e015afd46e46c9ba0413ca15f0316fa2eec24)

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

rpm: Bump PR numbers on componets that require RPM

With the recent RPM uprev, libzypp, sat-solver and zypper should be
rebuilt to ensure they get the right BerkleyDB and rpmdb interfaces.

(From OE-Core rev: a57002c854fed27d9c06a360da7133637afec2d6)

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

package_rpm: Add capability to directly query packageorigin

Add functionality to RPM to directly query the packageorigin (path) from
the resolver database, instead of having to do this via an indirect method.

This results in a minor performance improvement.

(From OE-Core rev: a8ff3141fd78442bf328c9d3a489db88ad27486b)

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

ossp-uuid: Add oss-uuid package and RPM 5.4.9 integration

RPM 5.4.9 now strongly encourages you to have the ossp-uuid library available.

Add this recipe, and change RPM to use the uuid functionality.

(From OE-Core rev: 8a2b55c5b7c544ec15113bc21e5c62276449c9c3)

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

rpm: Upgrade RPM to 5.4.9

Beside upreving RPM, add necessary integration patches to libzypp.

Also change the configuration of RPM to support PACKAGECONFIG flags.
RPM is highly configurable, the default configuration is good for
minimal OE-Core use.

(From OE-Core rev: bf94103d4c532ffdfdcdcc6d27c9f65f7824f8f8)

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

rpm: Upgrade RPM to 5.4.8 (db to 5.3.15)

RPM 5.4.8 requires db 5.3.x, so both are upgraded together.

(From OE-Core rev: c5898ef3fc3820ff9c44bc5b1b16e5def64aa877)

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

gnomebase: add GNOME_COMPRESS_TYPE variable

Upstream Gnome projects are starting to migrate to the .xz compress format,
so we need to add this to allow recipes to override the default of .bz2 as
the upstreams make the transition.

[YOCTO #2241]

(From OE-Core rev: 0eddc176e73e74d4b1a7d3163d65f619fcd4baf0)

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

nasm: updated to 2.10.01

Tested by compiling syslinux

(From OE-Core rev: ba22e4fcf07eb8007f6d500908d503f1d3dbdd16)

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

menu-cache: upgrade to 0.3.3

Other changes:
- licence MD5 for menu-cache.h has changed because some whitespaces
  were removed;

(From OE-Core rev: fe1e9b40d5d0dd8d4daa2751f4c86db26936fc30)

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

ethtool: upgrade to 3.2

Other changes:
- the HOMEPAGE URL since project hosting chenged to kernel.org;
- licence MD5 for ethtool.c has changed because a new copyright
  line from Sun has been added: "Portions Copyright (C) Sun
  Microsystems 2008"

(From OE-Core rev: ca1400b51bab2451bd6025a7b79ddabb1ac3d295)

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

tzdata: upgrade to 2012c

Other changes:
- the licence header shifted one line down, hence the start/end lines
  used to compute the MD5 sum were incremented by 1;

(From OE-Core rev: 2f1daa8e25bfc72d1fd75788a8e70b193accc0cb)

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

freetype: upgrade to 2.4.9

(From OE-Core rev: 6dc49f00248ffb937de45c4d075907774b4cd8fc)

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

linux-libc-headers: set default LINUXLIBCVERSION to 3.4

The 3.4 kernel is released, and is the default for qemu* builds, so
we can safely update the default libc-headers version to 3.4.

Built and booted for qemu*

(From OE-Core rev: 3e57510bb11b350fbe15cae2fb5bf851956061ac)

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

v86d: Fix cross link using host-cross ar

If building on 32bit host and creating 64bit libraries, the target
package builds should not invoke the 32bit hosts's ar.  Specifically
you will get an error message like:

x86_64-linux-gcc  -m64 --sysroot=/opt/qemux86-64/tmp/sysroots/qemux86-64 -Wl,-O1 -Wl,--hash-style=gnu -Wl,--as-needed -Llibs/x86emu v86_x86emu.o v86_mem.o v86_common.o  v86.o -lx86emu -o v86d
libs/x86emu/libx86emu.a: could not read symbols: Archive has no index; run ranlib to add one
collect2: error: ld returned 1 exit status

(From OE-Core rev: 4bb926513084dfa6ebf5598a177239315530276e)

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

libnewt: Fix cross link using autoconf detected AR

If building on 32bit host and creating 64bit libraries, the target
package builds should not invoke the 32bit hosts's ar.  Specifically
you will get an error message like:

x86_64-linux-gcc    -m64 --sysroot=/opt/qemux86-64/tmp/sysroots/qemux86-64 -g -o test test.o libnewt.a -Wl,-O1 -Wl,--hash-style=gnu -Wl,--as-needed -lslang
libnewt.a: could not read symbols: Archive has no index; run ranlib to add one
collect2: error: ld returned 1 exit status

(From OE-Core rev: a02878d05e6b57f2455228785ea5f213a62ed976)

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

gitignore: add wildcard to match toplevel patch files

To support the basic workflow of trivial patches:

 git format-patch HEAD~.. ; git send-email --to foo@bar.com 0001-foo.patch

We don't want git status reporting on patches lying in the top
level dir in this case.

Cc: Richard Purdie <richard.purdie@linuxfoundation.org>
(From OE-Core rev: f9c2d97d549e848127b741b967d90288c60f4d43)

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

augeas: Fix compilation failure

on uclibc we see this failure
too few arguments to function 'escape'

(From OE-Core rev: 4ea9c79cdcf0ffdc833b7a63e7a2b8388d6bf6e6)

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

task-self-hosted.bb,task-core-lsb.bb: Conditionalize pth inclusion

pth is not portable to uclibc therefore we need to exclude it for
uclibc based systems.

(From OE-Core rev: 6482f276533f3a177e07def9d616958d9bbc10c6)

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

cmake.bbclass: Add OECMAKE_C_LINK_FLAGS and OECMAKE_CXX_LINK_FLAGS variables

In some cases we need to specify linker flags and right
now we do not have a way to communicate that to cmake
based systems. cmake defines CMAKE_C_LINK_FLAGS and CMAKE_CXX_LINK_FLAGS
for these needs. This patch therefore defines two local variables
namely OECMAKE_C_LINK_FLAGS and OECMAKE_CXX_LINK_FLAGS which
can be altered by recipes to tweak linker flags

(From OE-Core rev: 9e00a74749ba8e1cf0d15efe8e16af60b189e080)

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

gcc-common: Replace use of TARGET_ARCH with TRANSLATED_TARGET_ARCH

TARGET_ARCH is poured into TRANSLATED_TARGET_ARCH
for gcc-cross family or gcc-crosssdk family
of recipes we have to check for TRANSLATED_TARGET_ARCH

(From OE-Core rev: 03f7322b1f4e1b1cf465a46cba3b82573384ab9e)

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

gcc-common: Enable hardfloat ABI when target is arm

We need to check target to be arm before enabling hard-float
ABI. There are crossdk targets or candian-cross targets built
for arm and we should not enable it for those class of recipes.

(From OE-Core rev: f48bb903218e14ba15dab452e8bb43ba25b9ea44)

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

uclibc: Move to latest tip of master

This gets the important fix for eventfd syscall
handling which was exposed by glib 2.32 upgrade
without this patch core-image-sato would get stuck
and wont boot all the way through

Same patch is needed for 0.9.33 as well

(From OE-Core rev: 80fa01c6ea618ea32ab40d151d8be2347ede30b3)

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

gnupg-2.0.19: Depend on npth for uclibc

(From OE-Core rev: 5bb9af52f59b91ff979d06178eeadbf91907bbe1)

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

gpgme: Depend on npth for uclibc

(From OE-Core rev: 854b4de95fe3a25a877506d762092962652b6770)

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

npth: Add recipe

npth is modern version of pth
gnugp 2.1 will drop pth in favor of npth

npth works with uclibc too whereas pth does not

(From OE-Core rev: 1b6bebf02be5bb5879c06d6f79b434ee054717a7)

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

libtool: Let -fstack-protector passed to link step

linker should add -lssp to linker cmdline when
-fstack-protector therefore add this knowledge to libtool
otherwise packages will fail to link

(From OE-Core rev: d1756ff379ab310bfa6323d8643b7e9d764f94bf)

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

pth: Mark incompatible with uclibc

pth does not work with uclibc

(From OE-Core rev: b7c12dd4424f0702c21b1b856af4578d9a4abc90)

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

binutils_2.22: Backport to fix bogus textrels

This is reported by prelinker which does not
go well with textrels.

(From OE-Core rev: 7b1caeca8e5c0dd066bb940fe6d2dff4d22bf816)

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

linux-yocto/3.2: fix meta SRCREV typo

Remove a stray character at the end of the meta SRCREV.

(From OE-Core rev: 009f7734e880084a5b01d8377cb006dc82f1d278)

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

hob2: create a standalone deploy image tool

[Yocto 2388]

Create a deploy image tool using the existing dialog including
DeployImageDialog and ImageSelectionDialog.

This tool writes bootable images to USB devices, and it can be run
directly without hob.

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

hob2: update DeployImageDialog for seperated tool

Part of [Yocto 2388]

Update class DeployImageDialog to get ready for a standalone deploy
image tool. The standalone tool can be run directly without hob, and
add a button to select image file. So adjust the layout of
DeployImageDialog.

(Bitbake rev: 399cfbaf36ccd4b934e25f915e64b87f32a3eb82)

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

ui/crumbs/hig.py: check deploy process return value

Update function response_cb of DeployImageDialog to get deploy process
return value. According the return value tell user that deploy image
successfully or not.

(Bitbake rev: f78f6d43a68e0f1dc4d3e4164eed453fcb9c22a8)

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

ui/crumbs/hig.py: remove extra spaces

Remove extra spaces at the end of line or blank line.

(Bitbake rev: 70b75a05a806fecf8be15b9ba66fec2eb87dcce4)

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

ui/crumbs/hig.py: fix run time error

Commit 094742bed2fc01d55f572da946fcfa7a48521401 re-implement the
function popen_read(). If there is no USB device, it crashes with
"ExecutionError: Execution of 'ls /dev/disk/by-id/usb*' failed with exit
code 2:"

Replace popen_read() way with glob module to get the USB devices.

(Bitbake rev: 0c43fe72e3c6a12ac19173d8cbbad81af21c2d85)

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

ui/crumbs/utils.py: import module bb

When I try to call function which_terminal() of utils.py directly,
it complains NameError: global name 'bb' is not defined

So import module bb

(Bitbake rev: 9e370d9b1022112be2ac8f73feb182b1bba179b0)

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

Hob: fixed the index issue when set actived item for base image combo

The combo item index dosn't be increased as the insert a dummy item, when
execute the upadate base image combo operation.

[YOCTO #2500]

(Bitbake rev: 59e19634a84fcb4c34b92cdcf7a9ea807c9abb63)

Signed-off-by: Liming An <limingx.l.an@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>