gdb: upgrade from 7.3.1 to 7.4

(From OE-Core rev: c24bd38fdde69e30981efc1c0e81f8aa65555130)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gmp: upgrade from 5.0.2 to 5.0.3

rebased configure.patch & amd64.patchto the newer code.
Removed sh4-asmfix.patch as it is not needed with the newer code.

(From OE-Core rev: 7c33a18ed5cdca6765d04ad9e4387001f72993a4)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

distro_track: Track Updates

openssl: Update to 1.0.0g
openssl: Update to 0.9.8t (gplv2)
gnutls: Update to 2.12.16
rxvt-unicode: Update to 9.15
gnupg: Add gplv2 version 1.4.7
file: Update to version 5.10
libidn: Update to 1.24
libdrm: Update to 2.4.30
xinit: Update to 1.3.2
xf86-input-keyboard: Update to 1.6.1
xf87-input-evdev: 2.6.0

Manual Checks:
util-linux - NO Update reason

(From OE-Core rev: cc3a6f625278d8ec9a61be1170823c3ec4302e61)

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

xf86-input-keyboard: Update to 1.6.1

(From OE-Core rev: 5960e647d18284fd42b4547dfeff3b1970b1d819)

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

xinit: Update to 1.3.2

(From OE-Core rev: 1b3db3597c2b644e61394014886082eb212eef1a)

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

libdrm: Update to 2.4.30

(From OE-Core rev: 4f9e5767b239335c47dbd57da6a21ce1f2bf686d)

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

libidn: Update to 1.24

(From OE-Core rev: ef2dcae84fe3a91f781ce906ea283d9f8d698af8)

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

file: Update to version 5.10

(From OE-Core rev: 260dd6a85d0f3fe64f7c4df5c4b99bb48b4e36ca)

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

gnupg: Add gplv2 version 1.4.7

(From OE-Core rev: daff80c595616de10f32e74e2881e851707c7ff9)

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

rxvt-unicode: Update to 9.15

(From OE-Core rev: 75789868df9cfc5cb74edfc3fe7d773cfa272ce6)

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

gnutls: Update to 2.12.16

(From OE-Core rev: 3439512952a64dc9537aa9842c64489219a76ee5)

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

openssl: Update to 0.9.8t (gplv2)

(From OE-Core rev: d58903b22aa879484a5eaa01efad5c031ea485fa)

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

openssl: Update to 1.0.0g

(From OE-Core rev: 263c39e624d6c8681e716d26718ce7502de8ae6d)

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

self-hosted-image: add package management for correct installation

(From OE-Core rev: 6ac8940f9d328afccd0176ae7c33053c3e565aca)

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

task-self-hosted: add python-pkgutil since it's required

(From OE-Core rev: fb5064de4d5d83b2f8b4d0aabea419f0681537f0)

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

task-self-hosted: add disk partitioning tools

(From OE-Core rev: 956d582f95f98690add05dc3f34fd520dedd8df0)

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

atk: add native / disable introspection / merge atk.inc->recipe / change license

Native support currently is appended in meta-oe. Native build failed with:

| /usr/share/gir-1.0/GLib-2.0.gir:9:1: error: Unsupported version '1.1'
| /usr/share/gir-1.0/GLib-2.0.gir:9:39: error: Unsupported version '1.1'
| error parsing file Atk-1.0.gir: Failed to parse included gir GLib-2.0
| make[3]: *** [Atk-1.0.typelib] Error 1

build tested in meta-oe / angstrom environment

(From OE-Core rev: a69108ea388fce9034ecc8670f1149fabae55e1b)

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

gconf.bbclass: Add proper RDEPENDS to packages with gconf postinststep.

Right now gconf bbclass adds both postinst and prerm steps, but it does
not ensure that packages involved have gconf in RDEPENDS. This can lead
to a situation where postinst/prerm steps fail because gconftool-2 is
not installed.

(From OE-Core rev: 31997ae9188302f276241266590a1ae0f4316c3f)

Signed-off-by: Peter Tworek <tworaz666@gmail.com>
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

distro_tracking_fields: take sysklogd off Saul's plate

(From OE-Core rev: 969661790df0e6f18617b8e867011dcc9504eb36)

Signed-off-by: Joshua Lock <josh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sysklogd: various fixes so it starts/stops cleanly

This patch fixes several issues with the sysklogd recipe:
o Errors at start due to non-existent /var/log/news/ - every other log
  file is created in /var/log, not a sub-directory. Do the same for news
  logs.
o klogd would not be stopped due to pidfile recycling, give klogd its own
  pidfile
o preinstalls failed at rootfs creation time by trying to access the host
  root filesystem rather than a path relative to $D. Update the preinst to
  test for $D and do the right thing.

(From OE-Core rev: 111d1b8bb2b89e06091335fff6a917bbd9a1f66e)

Signed-off-by: Joshua Lock <josh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pciutils: fix configure script for dash

configure.patch was missing a closing quote in a sed expression which
causes script execution to error on less forgiving shells such as dash.

(From OE-Core rev: ad72484a2783afa948966263629006558fbde476)

Signed-off-by: Joshua Lock <josh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

udev: improve udev-cache robustness

 * allow udev-cache to be disabled at runtime (using
   /etc/default/udev-cache);

 * make cache invalidated if kernel, bootparams or device list
   changes;

(From OE-Core rev: 22b72b23653736436f10d394de36201c32630d5d)

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

external-csl-toolchain: avoid the hardcoded version/prefix

(From OE-Core rev: 4346516a6a719355d2793af289f93e14d3dedd0c)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

external-csl-toolchain: fixes for metadata changes

These allow the recipe to build again:

- add compilerlibs, g++, libgcc to the provides
- add linux-libc-headers-dev to the packages
- in libc-package, only sed the ldd.bash.in file if it exists, as the external
  toolchain is using that class as well
- shift the inherit location of the libc classes, as they were overriding the
  recipe's do_install
- use ?= for EXTERNAL_TOOLCHAIN, so the user can set it

(From OE-Core rev: 8fb1c84f4cadf8d7a061fd6d90d270c19b474bfe)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libc-package: fix typo in LOCALETREESRC default value

(From OE-Core rev: 73ff2f404392a6942ec89af15ec7f52fe2e629a6)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

external-csl-toolchain: fix critical PATH manipulation bug

(From OE-Core rev: 9ccf5400d4d22f6fb90f5d61ad89dd0ac6273e43)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: Add description into the cache and clean up cache duplication

With the addition of new cache domains, the summary, license and section
information is no longer requred in the core cache since its only used by
the hob UI. This patch removes the duplicated entries.

It also adds the DESCRIPTION field to the cache for the benefit of hob2.

(Bitbake rev: 33ffb2e99825cb643b148b3462c2d4cf33ff5f58)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package bbclass: allow per package PRIVATE_LIBS

If a recipe packages multiple versions of shlib (e.g. powervr drivers) we only want the shlib code to pickup $PN, not $PN-foo subpackages.
This keeps backward compatibility with the global PRIVATE_LIBS usage if no per package PRIVATE_LIBS are set for a given package. In other words: this doesn't break the firefox recipe.

(From OE-Core rev: ce46dbddef40ae3eef7238ac07438b15bd09e156)

Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

Fix the issue that adt-installer tar ball is not regenerated if sstate is on, and other minor bug fixes

(From OE-Core rev: 61da952fdc2996c27c56234c36116a69a23a378d)

Signed-off-by: Jessica Zhang <jessica.zhang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: remove unneeded patch and re-enable PARALLEL_MAKE

Koen Kooi suggested that the quotefix patch was not needed,
and the build of the recipe should now be multithread-safe.
My testing has confirmed this.

(From OE-Core rev: 731700ec390d680ee4d277312755bea0ab4aeeec)

Signed-off-by: Scott Garman <sgarman@zenlinux.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pkgconfig: Ensure pkgconfig RDEPENDS are tracked

Currently, pkgconfig dependencies get added to -dev packages which install
pc files however nothing in the system makes bitbake aware of these
dependencies so images can fail with pkgconfig being missing.

This change explictly adds in the pkgconfig RDEPENDS to the -dev packages
and hence makes bitbake aware of the dependency.

(From OE-Core rev: 23e773eb8b70a5e36270b2e6415512b7dc4f1e05)

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

opkg-utils: convert to git repo at git.yoctoproject.org

The svn.openmoko.org site seems to have become unstable, make a copy
and convert to git at git.yoctoproject.org to preserve history

[YOCTO #1867]

(From OE-Core rev: b44717c29c4f50917570039adf896680d24bb216)

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

bootimg: Do not force FAT32 on all images, it violates the FAT specification

Fixes [YOCTO #1940]

do_bootimg was performing the FAT overhead calculations assuming FAT32 and then
forcing the use of FAT32 with "-F 32" to mkdosfs. The FAT specification is clear
on cluster count being the determining factor for FAT size (even if the fs
string is set to FAT32, go figure). Syslinux follows this spec, and rightly so,
resulting in a failure on core-image-minimal:

syslinux: zero FAT sectors (FAT12/16)

Drop the "-F 32" from mkdosfs to allow it to select the appropriate FAT size
based on cluster count. Leave the FAT overhead calculation in FAT32. This will
result in a little extra padding for really small images, but not enough extra
to justify recalculating for FAT12 and FAT16.

Tested with a core-image-minimal build for atom-pc. do_bootimg completed
successfully, and the resulting image was FAT16.

(From OE-Core rev: 634137704dd1a205e377a1131ef708f1c981f6b2)

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

fetch2/git: Add workaround for clone using alternates problem

To quote my report of this to the git mailing list:
"""

I have a problem with git clone commands using alternates failing by
mixing up different repositories. I have a situation where I could end
up with both:

/srv/mirrors/repo
/srv/mirrors/repo.git

as bare clones.

I then try cloning "repo" with alternates with the command:

$ git clone -s -n /srv/mirrors/repo /tmp/foo
Cloning into /tmp/foo...
done.

$ cat /tmp/foo/.git/objects/info/alternates
/srv/mirrors/repo.git/objects

Note how I'm now referencing repo.git, not repo. This doesn't work as
expected giving some very bizarre results when actually using the
repository.

I appreciate this is a rather bizarre corner case but its one that is
breaking the build system I work with. Ideally people would use a
consistent URL for the same repository but we have an example where they
haven't and this really shouldn't break like this.

Looking at the code, the cause seems to be

clone.c:get_repo_path():
        static char *suffix[] = { "/.git", ".git", "" };

since its looking in order for:
 repo/.git (fails)
 repo.git (suceeds, incorrect)
 repo (never looked at)

I'm not sure what would break if that order were to change, swapping the
last two options.

I can "force" the issue by running:

git clone -s -n /srv/mirrors/repo/ /tmp/foo

but this results in the slightly odd looking:

$ cat /tmp/foo/.git/objects/info/alternates

/srv/mirrors/repo//objects

which does at least work.
"""

This patch adds the trailing slash to ensure the correct repository is
referenced at the expense of some ugliness in the alternates file.

(Bitbake rev: d978e7b35550e3785c7c567ffe4c40a3c3947450)

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

cooker: add missing shutdowns to the exception handling

(Bitbake rev: c9f58ef6b897d3fa5b0d23734b5f2cb3dabb057a)

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

prserv: Do not ping PRService if not required

[YOCTO #1942]
Ping the PRService only if PRSERV_HOST and PRSERV_PORT are set.

(Bitbake rev: 20f24de0bdafac21f5d8a58701f977efa7041288)

Signed-off-by: Lianhao Lu <lianhao.lu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake-layers: list overlayed classes in show-overlayed

Classes (.bbclass files) can be overlayed in a layer although they are
currently located by BitBake in a different way (via BBPATH instead of
using layer priority) and thus it is useful to be able to see when this
is in effect and which layer's class is actually being used.

(Bitbake rev: f6493e4bad005a82580380d800ebf4c438292f5b)

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

bitbake-layers: add show-recipes subcommand

Add a show-recipes subcommand which lists all available recipes, with
the layer they are provided by. You can optionally filter the output by
recipe name (PN).

(This is a generalised version of the show-overlayed subcommand.)

(Bitbake rev: 05e86ba966f5a26721891c82b21afa48768a67cc)

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

bitbake-layers: improve show-overlayed output

Make the following improvements to the show-overlayed subcommand:

* Show recipes that are overlayed when the version is higher or lower,
  not just when it is the same. This gives a much better picture of the
  influence each layer is having over the metadata used for building.
  This can be disabled with the -s option if you just want to see
  recipes with the same version as before.
* Default to showing name (PN), layer and version rather than the full
  path and filename. The old style formatting can be used by specifying
  the -f option.
* Mark skipped recipes as such in the output, and print them in the
  correct sorted place in the list rather than at the end
* Prefix/suffix title line with === so it can be filtered out easily in
  shell scripts if desired

(Bitbake rev: 43b473275d3cb2e60a14e4a52cdc4654b3f4e5e7)

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

bitbake-layers: use directory name as layer name

It turns out that using the collection name as specified within
layer.conf (i.e. what gets added to BBFILE_COLLECTIONS) as a name to
refer to the layer is not particularly useful, since layer creators
aren't necessarily setting these to a meaningful value - e.g. OE-Core
uses "normal", meta-oe uses "openembedded-layer", etc. In any case,
BitBake uses the directory name in its list of configured layers in the
system information presented upon starting a build, so let's just do the
same here and avoid confusion.

Also rename the get_append_layer function to get_file_layer since it is
in no way specific to bbappends.

(Bitbake rev: 35d2c1c618826e961dbf4b9889b829f469346d74)

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

bitbake-layers: use dashes in subcommands

Rather than the slightly awkward underscores, use dashes in subcommands
e.g. show-layers instead of show_layers. (The old underscored forms
continue to be accepted however.)

(Bitbake rev: 6d311ddc1be04ae5bd0a1ebee94b44968e8a3f27)

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

rootfs_ipk: remove runtime_script_required usage

* it was introduced in 87780fc09b066525e47d0f50ee5497db54d304cd
* then partially removed in 2feba313c991170747381c7cf821a45c2cd04632
* so remove this use too as runtime_script_required is not initialized anymore and results in
  run.do_rootfs.6328: line 235: [: -eq: unary operator expected

(From OE-Core rev: 6d3eac57bdba8e2582c210a2f82a3a4546f68581)

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

gcc-runtime: fix override from poky-lsb to linuxstdbase

(From OE-Core rev: 1f5de4b36ad5853e489e47078c830f12a1c71ce9)

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

default-providers: add kbd as PREFERRED_PROVIDER for console-tools

kbd is used a primary provider for console-tools

(From OE-Core rev: 9a88125c546e6bcbec683eb736e232236c38a4f9)

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

gdb-common: Drop elfutils from DEPENDS

(From OE-Core rev: 8330205eeb605354c139605618255940e3b312d4)

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

linux-yocto: update to v3.0.18 + features

The -stable team released v3.0.18, so it becomes the new base for the
linux-yocto_3.0 recipe.

Along with the 3.0.18 update this kernel refresh brings in the following
changes:

  59314a3 meta/beagleboard: Using CONFIG_PANEL_GENERIC_DPI=y
  b168325 crownbay: use emgd-1.10
  281b80f kver: bumping to v3.0.18
  0d5d0dd common-pc*: add SMP and virtio

(From OE-Core rev: b58d330a88a64ac8ff82362cc90cc817f38b82cd)

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

linux-yocto: add kernel 3.2.2 support

Adding initial support for the linux 3.2 kernel.

Although the branching and naming has changed slightly, the same BSPs and
features exist in this tree as have existed in previous linux-yocto versions.

Notable features of this update are:

  - streamlined branch naming
  - configuration and functionality split in the standard kernel
  - creation of the 'tiny' kernel type branch
  - audited kernel configurations with optional features documented
    in the meta branch
  - feature and BSP refreshes
  - 3.2-rt support refresh
  - qemu ppc graphics support

Build and boot testing has been performed on all emulated targets
(sato and core).

(From OE-Core rev: 72264df07d7e57ba8232aae90b76ab220fdb2d22)

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

kern-tools: import config audit and branch manipulation updates

Updating the kern-tools SRCREV to import the following changes:

 a7a9930 kgit-meta/scc: allow multi-line defines
 e3442c6 kern-tools: add required and optional kernel configuration processing
 b56b67b configme: extract redefinition and mismatch information
 6118eaf kern-tools: remove branchname assumptions

These are required to support the kernel configuration policy changes
implemented in the yocto 3.2 kernel repository.

(From OE-Core rev: 5a3759f4b23361dd2c1968ce51e5c3695a661972)

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

tzdata_2011n: provide pkg_postinst and timezone defaults

* Introduce DEFAULT_TIMEZONE variable ('Universal', present in base package)
* and add /etc/timezone during do_install.
* While there, fix hardcoded references to $exec_prefix (/usr).
* Add the pkg_postinst script (adapted from Gentoo).
* Create /etc/localtime during postinst.
* Bump PR

(From OE-Core rev: 5304ce429f2c05857f04fbe3b47a067983ca96be)

Signed-off-by: Andrea Adami <andrea.adami@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

classes: replace 'Poky' with 'OE-core'

(From OE-Core rev: 1fdcd32520a05465b0d54c062f28bac9cdf74a20)

Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

alsa-utils: make alsa-utils-alsactl recommend alsa-states

This will drag in the asound.state from the BSP that alsactl uses.

(From OE-Core rev: 6a1c75971d35b6a8272741a4f3e25ef46f7ce305)

Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

alsa-state: fix RDEPENDS assignment

(From OE-Core rev: 358ad279d525ca3c806cbb6ceadb2d6170468d80)

Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

qt4-tools-nativesdk-4.8.0: fix build

the build is actually failing because host g++ and ld are used to
compile and link

(From OE-Core rev: 4ccae37db1aa77a1d15098c3720ea6e2d383fbdc)

Signed-off-by: Eric Bénard <eric@eukrea.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

genext2fs: fix inode computation

This patch fixes a problem with the computation of inodes
based on the bytes_per_inode options, for a larger FS (> 2G)
the inode count would go negative and a smaller default
count would be used, this would cause the FS to run out of
inodes.

(From OE-Core rev: df5e886be059da35fb69710c79227cc768f1c58e)

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

image_types.bbclass: refactor cpio images generation.

* Add common check for the presence of an /init
* symlink, being 'touch' would fail with it
* and 'touch -h' is not yet generally supported
* (depends on distro on build host).

(From OE-Core rev: ab282056148f3152f96b6990e09270ae5405da9d)

Signed-off-by: Andrea Adami <andrea.adami@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

useradd.bbclass: Fix missing quote

(From OE-Core rev: d7b13cd42ab8d5f44f97e119b73ec2e363677d26)

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

useradd: Ensure dependencies are only added for target recipes, not native or nativesdk

(From OE-Core rev: 63d006b2d3fc2223c74f81b91f70f5c841108c80)

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

distro_tracking_fields: fix the typos in the field of maintainer.

(From OE-Core rev: fba0459da7f274ae284d36b7fdbf2f9d10a52610)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libtool 2.4.2: fix missing quote in FILES_libtldtl

The missing quote prohibits OE from packaging it correctly, leading to:

|  * satisfy_dependencies_for: Cannot satisfy the following dependencies for task-gnome:
|  *    libltdl7 (>= 2.4.2) *   libltdl7 (>= 2.4.2) *   libltdl7 (>= 2.4.2) *   libltdl7 (>= 2.4.2) *   libltdl7 (>= 2.4.2) *   libltdl7 (>= 2.4.2) *   libltdl7 (>= 2.4.2) *

for incremental builds and binary feeds.

(From OE-Core rev: ed6968b731e8ebc161fa59e18a9284bddf86c61b)

Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

useradd.bbclass: Add explict setscene dependencies to ensure correct ordering of setscene tasks

(From OE-Core rev: ffc7bbcf0011de3f1f6e8d95f1de0b8f7164fa51)

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

useradd.bbclass: Ensure pseudo can load in the pseudo unloaded case

In the do_populate_sysroot_setscene case, pseudo has been unloaded and we need
to reload it. This code change ensures all the pseudo options are specified
so pseudo loads correctly.

It also improves some of the comments so all the different contexts are listed.

(From OE-Core rev: 76345cd61c9523ce6755ef8e923dec37800b7a98)

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

useradd.bbclass: Execute user addition code before do_package_setscene, not after do_populate_sysroot_setscene

The user addition needs to happen before the do_package files are extracted
by do_package_setscene since those are the ones we need to preserve the file
ownership information for. This patch ensures this happens.

(From OE-Core rev: 34282c1b996ef008384af456735692d66ddabc13)

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

sstate.bbclass: Add support for sstate preinst functions

(From OE-Core rev: f2b0a71b3100a0d2ceb80300d7f3823a31eb907a)

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

runqueue.py: Add inter setscene task dependency handling

This is being added to resolve setscene race issues where we do have
particular dependencies required between setscene tasks. This allows
specific dependencies to be specified. This allows us to fix the races
in sstate with the useradd class in OE-Core.

Any tasks being depended upon have their reverse dependencies cleared to
ensure we don't have circular references.

(Bitbake rev: e1b157d26374a70e6274edcb4c0b9f3bc48f765c)

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

runqueue.py: Fix missing setscene dependencies

When constructing the setscene inter-dependencies, we need to account for all task,
not just the last one found. This patch corrects this oversight and ensures all
dependencies are added, not just the first one found.

(Bitbake rev: b9b5b5129d066e1ff7d3effda116afc3c6657beb)

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

data/runqueue.py: Fix commented out test code after variable name changes

(Bitbake rev: 175c5c8a5e51365aa127562e8598e84f5cd36495)

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

shadow-sysroot: Simplfy login.defs installation

Currently if shadow-sysroot is installed from a sstate package,
shadow_sysroot_sstate_postinst is looking in ${D} for login.defs which
isn't a valid for an install from sstate.

The easiest and most correct way to fix this is to override the standard
sysroot_stage_all function to install the files correctly and then
sstate installs work as expected. This simplifies the code as an added
bonus.

(From OE-Core rev: 6437e7315a0f2e077f9d69e4b65b48280ea8edbc)

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

libgcc: Add INSANE_SKIP of staticdev for libgcc-dev and libgcov-dev

Skip the staticdev QA test since the static archives belong
with the -dev.

(From OE-Core rev: 3de810a7f745b253d4b63eea353a31514d64a2e3)

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

insane.bbclass: Add test for static .a libraries in non-staticdev packages

This check verifies that non staticdev and non pic packages do not
have static libraries in them. There are a few exceptions for the
libgcc-dev and libgcov-dev packages which are whitelisted, also the
*_nonshared.a need to be part of their respective -dev packages.

This is currently a warning to work out any other issues in the upper
layers but will be moved to ERROR at somepoint in the future.

[YOCTO #663]

(From OE-Core rev: 4f31eb53fb0b1bfdf8be05356e57df607a6e82a9)

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

pm-utils: Add grep to RDEPENDS to avoid command failures.

Second version of the patch that adds grep to RDEPENDS.

Fixes [YOCTO #1887]

(From OE-Core rev: 6d4ff87b935d61aae63260262c0f48fe7e9d2a48)

Signed-off-by: Marc Ferland <ferlandm@sonatest.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

linux-tools: don't build perf when GPLv3 in INCOMPATIBLE_LICENSE

As binutils is required by perf to build and is GPLv3 licensed adding
GPLv3 to INCOMPATIBLE_LICENSE will cause linux-yocto to be skipped.

Long term we should look at moving perf to a separate recipe but as a
short term solution this patch will ensure that when GPLv3 is in
INCOMPATIBLE_LICENSE perf is not built and it's dependencies are not
added to build.

Fixes [YOCTO #1879]

(From OE-Core rev: ce61f9031b54067bffa304dab90c31278631dcdf)

Signed-off-by: Joshua Lock <josh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

uclibc-git: Update to latest RC

Document the patches

(From OE-Core rev: 746e8ffa66850bf9050cd6baf94eb76c492eb535)

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

eglibc-2.15: Add upstream status for x86_fenv.patch

(From OE-Core rev: 64a8f93cb6a7cd86f966b8f53227246deebce3a0)

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

uclibc: Dont package all .a in uclibc-dev

only *_nonshared.a are to be packaged in uclibc-dev
rest can go into uclibc-staticdev

(From OE-Core rev: aa799ba49833bdf6acbcd4b34f0605050c938175)

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

util-linux: Change SRC_URI to a fetchable location

(From OE-Core rev: 140ca6f3e690f34940fd60809f2d03fabb27105d)

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

ltp: Update recipe and fix building on uclibc

(From OE-Core rev: 4e8644584aa8826e9cbf9ff0dc23b8c395978013)

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

distro-tracking-fields: update entries for clutter-gst-1.8

(From OE-Core rev: 5801f480b74465696389783dc651ec48ff42e715)

Signed-off-by: Joshua Lock <josh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

clutter-gst: update to 1.4.6

(From OE-Core rev: a440654076008a107fff2245eff7791a4abf5abf)

Signed-off-by: Joshua Lock <josh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

cogl: update upstream status for build_for_armv4t.patch

This has patch has been merged into upstreams git repository and will be
available in the next stable release.

(From OE-Core rev: 77f5e99a821ad30b859a402bdc55c495741b24cc)

Signed-off-by: Joshua Lock <josh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gcc-cross: Don't reference machine specific variables

gcc-cross is installed into a package architecture specific directory
and is not meant to be machine specific. This patch replaces MACHINE_ARCH
with PACKAGE_ARCH to ensure this is really the case.

This was found by examining sstate checksums.

(From OE-Core rev: b3ca57418f47b16ed0a63d3291bf7d31e3ca5a45)

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

bootimg: Fix a math thinko in the block count calculation

Fixes [YOCTO #1852] ... again.

The conversion from sectors to blocks was multiplying by 2 instead
of dividing by 2. Blocks are 1024 bytes, sectors are 512 bytes. The
result was images being much larger than intended.

Reported-by: Tom Zanussi <tom.zanussi@intel.com>
(From OE-Core rev: b35384fa3ca96b31c63d764322215abced2066e4)

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

lib/oe/sstatesig.py: Exclude common machine specific dependencies from task signatures

Where we have machine specific recipes with well defined behaviour, it makes
no sense to rebuild recipes with these as dependencies whenever the machine
changes. This patch lists those well behaved recipes and excludes them from
the task signatures so we can change MACHINE without invalidating existing
PACKAGE_ARCH binaries.

(From OE-Core rev: 07e34778fc74126af1380bf249fd34a5e3df12c2)

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

shadow: Split securetty into a separate recipe

The securetty file is machine specific whilst the rest of the shadow recipe
is not. Unfortunately making the recipce machine specific is both inefficient
and also causes dependency problems since parts of the system such as the useradd
code depend upon it and this introduces a machine specific element to sstate
checksums which should not be machine specific.

To resolve this, this patch separates out the file into a separate recipe
meaning the machine specific components are isolated.

(From OE-Core rev: 318133b5202632c6957c2aade22b1ef7af929f23)

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

gcc-common.inc: Fix comment

(From OE-Core rev: a4a98b6d6273d4e716afbff0880d0ac2ff7be0a9)

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

gcc-common.inc: Ensure WORKDIR doesn't adversly affect sstate checksums

Changes in WORKDIR were changing the do_patch stamps between gcc-crosssdk
and gcc-cross. This excludes the variable since in this case, we don't
need want those changes to affect the signatures.

(From OE-Core rev: 0de85d2ca3bc9674323302a35995dfddb6044a71)

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

guile: fix a quote in description

(From OE-Core rev: 29118b57c7804a237e0c8fef41fcf032476bfb10)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

ltp: disable elf binary stripping

Otherwise the ltp tests ldd01 & nm01 fail due to missing stripped
information.

This Fixes these bugs:
[YOCTO #1922]
[YOCTO #1923]

(From OE-Core rev: 075266fa8bf12719c58a21c75a35fede5608dcdf)

Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

distro_tracking.inc: update the info for recipes upgraded recently

This patch is to update the information for the recipes upgraded recently, including strace, pciutils, ed, ofono, iproute2, bluez4, bluez-hcidump, kexec-tools, tremor, iptables, polkit, libxcb, xcb-proto, pixman and util-macros.

And add the new recipe for mobile-broadband-provider-info, which is depended on by ofono v1.3.

(From OE-Core rev: 10b255bd881e6f63c7eb7140aa798fcb2093db89)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libsoup: update its info in distro_tracking.inc

(From OE-Core rev: 01ed085b6b2be063ee9249f3ad3f3c03bd19490b)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

strace: upgrade to 4.6

(From OE-Core rev: 2b81cf3d76241572c299a0a58a3d0ca022bef9a0)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

pciutils: upgrade to 3.1.9

(From OE-Core rev: 745bdcd3421747b050493079101f6f5ba7830daf)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

ed: upgrade to 1.6

This patch is to upgrade ed to 1.6, and the change of license checksum is because

the diff between two ed.h files:

3c3,4
<     Copyright (C) 1993, 1994, 2006, 2007, 2008, 2009, 2010

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

ofono: upgrade to 1.3

This patch is to upgrade ofono to 1.3 and the v1.3 ofono depends on mobile-broadband-provider-info.

(From OE-Core rev: d691b150d1222a23422658d6dc4b72a3307bc77f)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

mobile-broadband-provider-info: bring a new recipe into poky

This patch is to bring a new recipe mobile-broadband-provider-info into poky because the latest version ofono v1.3 depends on it.

And the patch is imported from the OpenEmbedded server (http://git.openembedded.org/openembedded/tree/recipes/mobile-broadband-provider-info) as of commit ids

commit 9d612c4bb9bef813bdef9e198efc5949d51905eb
commit 84bf98dc6876ef6be8a4ee3744924e922253b53e
commit 30a913149de95ecafe4ef87bd50cfdbd6737fe1e
commit 7d2be53866d317bb032c63548f504f63107b4c2c

And upgrade to the latest.

(From OE-Core rev: 851299c8dd2bf8012078a2c78f79dc1b9fdc3323)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

iproute2: upgrade to 3.2.0

This patch is to upgrade iproute2 to tag 3.2.0.
And the tag is not fully testd by the upstream and has build error.
We use the next commit of tag 3.2.0 which fixs the error.

(From OE-Core rev: f2a7b225b5a243e0a364be678a71bdc3b0fa99e0)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bluez4: upgrade to 4.98

(From OE-Core rev: 8d3188a4a69214f23d8b7f8d972bdd35f574196b)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bluez-hcidump: upgrade to 2.2

(From OE-Core rev: cfa2e9ef3f52f8956d05c2b12ca9984fc08ef238)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

kexec-tools: upgrade to 2.0.3

This patch is to upgrade kexec-tools to 2.0.3.
And since fix_for_compiling_with_gcc-4.6.0.patch is there in 2.0.3, it is removed.

(From OE-Core rev: aea75907ff0ad11b7a17910f74491278e55bbc10)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

tremor: upgrade to 18163 (20120122)

(From OE-Core rev: 67ae9bbbab31b69f37e3ee0c1c969e43da034699)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

iptables: upgrade to 1.4.12.2

This patch is to upgrade iptables to 1.4.12.2, and introduce a patch not to check unknown symbols.
Otherwise, when it is compiled, it will report "libxtables.so.7" from LD_PRELOAD cannot be preloaded.

(From OE-Core rev: 27ed7024cf2ee9c9f84246fd931bc390cb638851)

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>