iputils: Break libsysfs dependency

iputils drops a /bin/arping with a runtime linkage against libsysfs in
/usr.  Port Fedora 17 iputils-20071127-infiniband.patch, which inlines
access previously done by libsysfs.

(From OE-Core rev: 99c482124e9f476923f6d5cf0a9e5551507f9d14)

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

linux-yocto-custom: Clarify usage and clear COMPATIBLE_MACHINE

There has been some confusion over proper use of the linux-yocto-custom
recipe. It is not intended to build as is from meta-skeleton. It should
be modified via a bbappend file to provide a Linux kernel config at the
very least.

Update the commentary to make this requirement more explicit. Add some
additional detail about how to create a bbappend file and how and when
to modify the various variables.

Clear COMPATIBLE_MACHINE so bitbake will not attempt to build the recipe
unless the user explicitly adds there machine to the variable, which
should encourage them to read the recipe comments before attempting to
build it.

(From OE-Core rev: 8ac3d0566e793777ec4f6629ef752d6db6231ca9)

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

image-mklibs: pass correct libdir to mklibs

libdir should be specified, or else mklibs won't work for 64bit targets.
It wouldn't be able to find the libs.

    Traceback (most recent call last):
      File "<build>/bitbake_build/tmp/sysroots/i686-linux/usr/bin/x86_64-wrs-linux/mklibs", line 553, in <module>
        header = elf_header(find_lib(libraries.copy().pop()))
      File "<build>/bitbake_build/tmp/sysroots/i686-linux/usr/bin/x86_64-wrs-linux/mklibs", line 89, in elf_header
        raise Exception("Cannot find lib: " + obj)
    Exception: Cannot find lib:

(From OE-Core rev: d2cd2ccea8bc4d110647ba3bd202772e5407000a)

Signed-off-by: Jesse Zhang <sen.zhang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

python: Fix RPATH QA issue on non-gplv3 builds

On non-gplv3 builds, gdbm gets built different due to the different version
which triggers a different codepath in python's db support and then hence
triggers an invalid RPATH QA issue. This change extends the appropriate patch
to cover the code paths we need it to cover and avoid adding the problematic
RPATH.

(From OE-Core rev: 90540692c40719b757b80bdc6e78e679a3df83a7)

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

linux-yocto-3.4: Disable extra slang header search path

Add in a workaround to avoid host infection detection build failures
from the slang include directory in perf. I'll defer to Bruce to
fix this properly but we need a workaround now as this is breaking
builds.

(From OE-Core rev: 4fd4b2eafb5f4ff2ef85d7f5ff3238a41c34313b)

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

linux-firmware: Ensure PR goes forward, not backwards after the revert

(From OE-Core rev: 447819db343a93bcc1e079c2a33b592da20f6c1e)

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

Revert "linux-firware: use ${base_libdir} instead of /lib for packaging"

This change breaks firmware loading on 64 bit platforms. This therefore
reverts commit ce2665b6ae77770d69ffbc4d45c96ce829d16738.

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

kernel.bbclass: Revert ${base_libdir} change

The base_libdir change in 5b8a4798ea2ea7df66bb53c26448251ea7da3dd9
breaks the kernel build for 64 bit machines. Revert this part of the
change.

(From OE-Core rev: 351bdc59f8af915ae203af7799eedc0ce42785f2)

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

kernelshark: Inherit pythonnative

(From OE-Core rev: 2b23cd1c8f81de989bce1ecab2d5b4328328de86)

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

gdb-cross: Explicitly disable python

Python isn't in DEPENDS but can be autodetected from the sysroot. Explicitly
disable it to ensure deterministic builds.

(From OE-Core rev: b8dd9592f844d15cdc8631b56ec42888a07826c9)

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

subversion: Add missing build dependency on sqlite3

subversion needs an explicit dependency on sqlite3, otherwise it
does not build. Tested by building core-image-minimal.

(From OE-Core rev: 47aca0f0f79c30d1df1f89c710d3e354f436145d)

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

findutils-4.4.2: add bison-native to DEPENDS

[YOCTO #2858]

(From OE-Core rev: 09a023e777b1e05b3edbec2bc49232042f6bf088)

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

bison: updated to 2.5.1

The gets patch is not needed anymore in this version. The
manpage patch was updated for this version.

(From OE-Core rev: 2f11d409e20d90f2d619b4d92495c3fcd30584ab)

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

buildhistory: reduce the numeric delta to 10%

(From OE-Core rev: f6effbf62351761430438cfcce285c982271239c)

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

buildhistory: correctly handle reductions of numeric fields

(From OE-Core rev: 94fffb92e6a7a8021fb1acb06b9c4218703e23b2)

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

libpam: disable NIS to not link with libtirpc when it is available

I was checking ways to make incremental builds faster so I started using
sstate-cache and SSTATE_MIRRORS. But this gave me some nasty bug:

| Collected errors:
|  * satisfy_dependencies_for: Cannot satisfy the following dependencies
for php-cgi:
|  *    libtirpc1 (>= 0.2.2) *
|  * opkg_install_cmd: Cannot install package php-cgi.

I checked details:

In my previous build libtirpc got built before libpam so libpam found it
and linked. As a result packages depend on libtirpc1 but as there is no
such build dependency sstate handling code did not used libtirpc copy...

(From OE-Core rev: e629bdcd1bcb51f2d2101fb53daeac0bd29ab637)

Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gthumb: fix parallel build problems

It's possible with a parallel build for make to spin waiting for a file change
when it isn't going to.  Fix the build rules so this doesn't happen.

[YOCTO #2319]

(From OE-Core rev: dea6a0eb2bf06b3dc47e276a046d8dd828e47228)

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

babeltrace: Update to upstream stable versrion 0.12

Latest version is 1.0.0 but in rc4 for now.

(From OE-Core rev: 6172f1fed9850180a3a84ca62cc0741d132031e4)

Signed-off-by: Radu Moisan <radu.moisan@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

atk: package update 2.2 -> 2.4

(From OE-Core rev: 9202bc255eb6867b4b6dd1c0850c70a4d0f1fb76)

Signed-off-by: Radu Moisan <radu.moisan@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

classes/package: don't make locale pkgs depend on PN pkg if not in PACKAGES

If ${PN} is not in PACKAGES then don't add it as an RDEPENDS for each
locale package. This fixes the installation of gcc-runtime locale
packages, for example.

(From OE-Core rev: d36c3235b3022c07f064929f55114f808a7634f6)

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

classes/buildhistory: save metadata revisions

Always write the metadata revisions for each layer into a machine-
readable "metadata-revs" file so that you can potentially link changes
in the output back to changes in the metadata. (Unlike the existing
similar build-id file, this is not specific to image changes.)

(From OE-Core rev: 45f38d167b523d769c072474b36b3d9b31b4e9d9)

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

buildhistory_analysis: ignore removal of self-dependencies

The recent removal of self-dependencies for dbg/dev packages produced
a fair amount of noise, so filter it out since we know this is a
reasonable change.

(From OE-Core rev: 3ec608462e4cb1eb77b6addbae1bda84fd76e4d2)

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

buildhistory_analysis: tidy up duplicate split_version function

This function is now provided by bb.utils and since we have now bumped
the minimum bitbake version, we can switch to that one instead.

(From OE-Core rev: 67a4d15f0a8032190e92026677150162a492a053)

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

classes/buildhistory: record PKG/PKGE/PKGV/PKGR

Save PKG (the actual output package name, which is often different due
to debian renaming), and PKGE/PKGV/PKGR (which may be manipulated in
certain special cases e.g. gitpkgv.bbclass in meta-oe, the
external-sourcery-toolchain recipe, etc.) Note that these are only
written when they are different from the normal package name in the
case of PKG, or PE/PV/PR for the other variables.

Also, use PKGE/PKGV/PKGR instead of PE/PV/PR when comparing package
versions since these actually represent the version that the package
manager sees.

Implements [YOCTO #2787].

(From OE-Core rev: 65d7e9b2d4d8115ac9fd513c04f39a2df9556a5a)

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

classes/buildhistory: save preinst/postinst/prerm/postrm

Write the value of these package script variables into the packageinfo
so that any changes to them can be tracked (in separate files since they
are multi-line).

Inspired by an earlier patch from Andreas Müller <schnitzeltony@googlemail.com>

(From OE-Core rev: 988a417c857c01c87de6ba9602968059cf8d830f)

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

classes/buildhistory: remove unused functions

getlastrecipeversion and readRecipeInfo weren't called anywhere.
Remove them so we don't have to keep them up-to-date with future
additions to the recipe info file.

(From OE-Core rev: 22bf6284576ccee607d7bd197420a29242df11c0)

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

classes/buildhistory: ensure old package info is removed

If a package is removed from PACKAGES, ensure that the package info file
and directory are removed from buildhistory so that we don't have stale
data lying around.

(From OE-Core rev: 223b183197b363edef0c11a25bad33555fca5a15)

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

classes/buildhistory: remove obsolete flat package history feature

Remove the ability to set BUILDHISTORY_KEEP_VERSIONS = "1" to save the
package history as flat versioned files rather than relying on git to
keep previous versions of the package information. git has proven to
work quite well in this capacity.

(From OE-Core rev: ccedfd250620fc562988ba730ad5717b107a9d3e)

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

curl: disable ldap/ldaps explicitly

* openldap from meta-oe is autodetected and then libldap-2.4-2 runtime dependency added to curl and almost all meta-efl recipes

(From OE-Core rev: 666b67179d4492e4d950b94457ebf8cac6454f7d)

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

mesa: Use 'require' instead of 'include'

These include files are really required, not optional.

(From OE-Core rev: 1aa6c4a083e406de0119f92e9654068d6cc2e4de)

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

mesa: Update to 8.0.4 (latest stable version)

- The license file has changed, but the project still has the same
  license
- Patches have been rebased on top of the 8.0 branch. crossfix.patch
  and crossfix-mklib.patch have been merged as they address the same
  problem

(From OE-Core rev: e4039eb74b20e96d4b8837cd58cf2d13d091e1ad)

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

mesa: enable the Graphic Buffer Manager library

(From OE-Core rev: be882417cf432fe79a843cea4439503f286206cf)

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

mesa: add --enable-shared-glapi, and package it in libglapi

(From OE-Core rev: d856f4c18425b9e3e5a0898d7f29cffc1998ca12)

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

mesa: move glu.pc to libglu-dev

(From OE-Core rev: 98975edaff2cc4cc417dee2c7c530e602ff2e4cc)

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

mesa: format the packages list nicely

(From OE-Core rev: ca9c3f57c72f443b9efe5f0bd5cb9a7bb169ed30)

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

mesa: no need to depend on python-native, the class does that

(From OE-Core rev: dd2a11fa262ced8e9f06413e69bbe1cb1b3e538b)

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

linux-firware: use ${base_libdir} instead of /lib for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: ce2665b6ae77770d69ffbc4d45c96ce829d16738)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

kernel.bbclass: use ${base_libdir} and ${sysconfdir} instead of /lib and /etc

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: c859717f522098e7c68df578d59011f68122dc2c)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

lsb: use ${base_bindir} and ${sysconfdir} instead of /bin and /etc

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 8b94cd639c02bcc653312575592673c36df0ec92)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

usbinit: use ${sysconfdir} instead of /etc for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 3068d5c44667bc238f2331d36cb116403560c83f)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

lsbinitscripts: use ${sysconfdir} instead of /etc for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 3fce22d81cc27629c7c3fd4c174b68c8ced468cf)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

rsync: use ${sysconfdir} instead of /etc for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: bc45d990938c1b1d761cee6e90464d22f854a2ab)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

man: use ${sysconfdir} instead of /etc for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: e044be34601dcc8eeca7f96d628964bebf562e2e)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

chkconfig: use ${sysconfdir} instead of /etc for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 7b77d4372ace6aa2400e3328875b39b4f67476d0)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

qemu-config: use ${sysconfdir} instead of /etc for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 998b94adbbc0f0f036c5816fbf6ee594dba90c1c)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xserver-nodm-init: use ${sysconfdir} instead of /etc for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: aada9260af9dcac42c21e189fdf6698e509c26a4)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

builder: use ${sysconfdir} instead of /etc for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 43a7286ed5f628f04597ce7e4bea0d91bc95b09b)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

x11-common: use ${sysconfdir} instead of /etc for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 5082056a4671570ca4578da2f622b3cc6f9f2dda)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libpam: use ${localstatedir} and ${sysconfdir} instead of /var and /etc

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 7556e60bf23c07646594a1704b3db7dfc29e631d)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

openssh: use ${localstatedir} instead of /var for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: b7134e8f76026d15a5e35c6a2e5d77d5e48c2787)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

at: Use variables from the environment to override the makefile defaults

This allows us to remove the docdir worksarounds and fixes other
potential relocation issues at the same time as cleaning up the
recipe.

(From OE-Core rev: 3be575babcdb12f95f4882d3c42bdf3ba1ec1900)

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

mingetty: use ${base_sbindir} instead of /sbin for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: f8f33550f5cc263673cbd481a567116514603b8d)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

systemtap: use ${datadir} instead of /usr/share for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 095f9c056373a252c0ba2679d8e6f4ad932a958d)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xkeyboard-config: use ${datadir} instead of /usr/share for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: c0ebcac02023705aead0d2588224a62d64fe1752)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

ubootchart: use variables instead of hardcoded paths

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 73ee5587a46751ed39a181b3ff2a12d4a3086c21)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

useradd-example: use ${datadir} instead of /usr/share for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: b93a9dadc9a2754d4e33929bda8dba46dd20b694)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

cronie: use variables instead of hardcoded paths

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 7b13be661c869eb4bf88d16a928a97edd5bc203b)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

lsbtest: use ${bindir} instead of /usr/bin for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 0a8003a3542d17332488ca715619990b222ceb71)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

sudo: use ${bindir} and ${sysconfdir} instead of /usr/bin and /etc

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: a9c73d9d3dd5f7fbee82d7ed3cb88a80aa3cbfbd)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

lsbsetup: use ${bindir} instead of /usr/bin for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 84fd332ab2f2c512109fc0a080d03533883fa235)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

alsa-state: use ${sbindir} instead of /usr/sbin for packaging

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 72fd0993d8643d7ca52d954d65395585fdf34b03)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

xinetd: use ${sbindir} and ${sysconfdir} instead of /usr/sbin and /etc

It is considered good practice to use the build system provided
variables instead of directly specify hardcoded paths.

(From OE-Core rev: 6f0eebbd12d79a86dd4a79f87b6be4758439a3f5)

Signed-off-by: Javier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

python: Fix to support the python module _bsddb built with db 5.3

_bsddb module in python 2.7 could be built only with db version
between 4.1 and 4.7. A patch was added to avoid build warning
about this for [YOCTO #1937] but not actually fixed it.

This patch enable _bsddb module be built with db 5.3, and remove
--disable-statistics from the DB5_CONFIG to fix segmentation fault
when using _bsddb module in python.

[YOCTO #2749]

(From OE-Core rev: 11267f8a1ccf65142988b095351a84b0fa0fcbcf)

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

psplash: Update to latest psplash version

* Update to the latest version of psplash which fixes a segmentation fault
  bug

(From OE-Core rev: 662637decc23888c9b0da93807257943bab14583)

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

glib-2.0-nativesdk: Add dependency on glib-2.0-native

when we build for nativesdk it needs tools from installed
glib-2.0-dev package. We zeroes out compile_resourses but
then genmarshal came up so instead of going about disabling
them one by one lets depend on glib-2.0-native since
thats the right thing to do.

(From OE-Core rev: 5416bf849ceefb49f06bedfa5e67595e9c68582a)

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

populate_sdk_rpm.bbclass: Depend do_populate_sdk task on createrepo-native

Now that we have moved the rpm-createsolvedb.py
script to createrepo-native for good reason of
python-native dependency, we have to make sure
that it is staged before its used, unlike before
where it was under scripts/ dir and always existed
outside the realm of package management.

Fixes build error when doing meta-toolchain builds from
scratch

|
/home/kraj/yocto/poky/build/tmp/work/ppc64e5500-poky-linux/meta-toolchain-1.0-r7/temp/run.populate_sdk_image.26712:
line 610: rpm-createsolvedb.py: command not found
| DEBUG: Python function do_populate_sdk finished
| ERROR: Function failed: populate_sdk_image (see
/home/kraj/yocto/poky/build/tmp/work/ppc64e5500-poky-linux/meta-toolchain-1.0-r7/temp/log.do_populate_sdk.26712
for further information)

(From OE-Core rev: 90cfe33d30ef4f9491bd52e7965c5279a2e2650c)

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

eglibc-locale: bump PR after fc4b138948656d482603b80f8c8e86fdbb185df4

(From OE-Core rev: 17685eb72f335f53caf3cadaadcf479d394f8792)

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

bitbake.conf: exclude whole MACHINEOVERRIDES from OVERRIDES vardeps

* whole MACHINEOVERRIDES can change e.g. between MACHINES with different arm architecture, causing allarch packages to reexecute do_package
  bitbake-diffsigs ../shr-core/tmp-eglibc/stamps/all-oe-linux/xserver-nodm-init-2.0-r16.do_package.sigdata.90e760a8f6cecbd87cb2e95f1237e3cc ../shr-core/tmp-eglibc/stamps/all-oe-linux/xserver-nodm-init-2.0-r16.do_package.sigdata.9eeccfd15f25032b3b6b132534660fff
  basehash changed from 7618e17d3fda05d1f15246e6800ca0f0 to 97bc4dc8c1521c535bd96b2aa62d8a03
  Variable MACHINEOVERRIDES value changed from ${MACHINE}${@bb.utils.contains("TUNE_FEATURES", "armv5", ":armv5", "" ,d)}${@bb.utils.contains("TUNE_FEATURES", "armv4", ":armv4", "" ,d)}:${MACHINE_CLASS} to ${MACHINE}${@bb.utils.contains("TUNE_FEATURES", "armv7a", ":armv7a", "" ,d)}${@bb.utils.contains("TUNE_FEATURES", "armv6", ":armv6", "" ,d)}${@bb.utils.contains("TUNE_FEATURES", "armv5", ":armv5", "" ,d)}${@bb.utils.contains("TUNE_FEATURES", "armv4", ":armv4", "" ,d)}:${MACHINE_CLASS}

(From OE-Core rev: d205aa2b42a4676b1ba587c0375cd6ac7e18e9ce)

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

qt4-native: fix creating /include directory in build host root

* same fix as
  http://git.openembedded.org/openembedded-core/commit/?id=7d3fd68480d54a8ef0d91a772df3506cd45ef2c2
  is needed for native too

(From OE-Core rev: 7f7f42769cfae509ca5366715d2d4fd9ef5c4784)

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

tcl: Add ${bindir_crossscripts}/tclConfig.sh to sysroot stage

tclConfig.sh is changed in do_install for cross compile and
is installed to STAGING_BINDIR_CROSS, but if SSTATE_DIR is set
and tcl is from sstage, tclConfig.sh can't be found in
STAGING_BINDIR_CROSS, add ${bindir_crossscripts}/tclConfig.sh
to sysroot stage can fix it.

[YOCTO #2891]

(From OE-Core rev: 7a97abb3f1c2ff8fd2dc7c59f866430cf99ab82d)

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

gdk-pixbuf: fix parallel install issue

This patch fixes parallel install issue that lib libpixbufloader-png.la
depends on libgdk_pixbuf-2.0.la which will be regenerated during insta-
llation, if libgdk_pixbuf-2.0.la is regenerating and at the same time
libpixbufloader-png.la links it, the error will happen.

Error message is:
* usr/bin/ld: cannot find -lgdk_pixbuf-2.0
* collect2: ld returned 1 exit status

Make an explicit dependency to the libs install targets would fix this
issue.

[YOCTO #2883]

(From OE-Core rev: e6fb5a25a098bd639be4b176f58e552404232c93)

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

busybox: fix du-k-works test case

[YOCTO #2896]

The du-k-works test seems to assume that du -k on a directory with two files
totaling 80 kilobytes should report either 80k or 88k. In ext4 filesystm however,
both busybox' du -k and coreutils' du -k reports a size of 84k which causes the
test to fail.

Intergrate the patch from:
http://git.busybox.net/busybox/commit/?id=a5ee090e8651692545514a81a16c6cde3a2dc577

(From OE-Core rev: a022fb6b69388f28f3130e71125b64e4c154d504)

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

gcc-cross-initial: Ensure it uses an isolated sysroot

If we don't do this, a stale limits.h may be detected in STAGING_DIR_TARGET
which would result in a different limits.h getting generated by gcc-cross-initial
that references it. The referenced limits.h will then not get found by eglibc-initial
causing rather strange build failures.

The simplest solution is to create a temporary sysroot containing only the things
gcc-cross-initial should care about and this results in a correct limits.h file
regardless of what else may have been built.

(From OE-Core rev: 9c304eae0724474902fe2f3150adc6af115af9ba)

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

gcc: Bump PR since there have been several gcc changes and various problems reported and this should flush anything stale out

(From OE-Core rev: a9d0cbe1d84bb26fc1a1f48764fe514cf9f9c548)

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

bitbake: data_smart: Fix unanchored regexp causing strange parsing issue

If this regular expression is unanchored, it would accept strings like:

do_install_append1
do_install_appendsomelongstring

and treat them like they were do_install_append. Clearly this isn't desirable.
Only one instance of this type of issue was found in OE-Core and has been fixed
so correcting the regexp should be safe to do.

(Bitbake rev: 23bd5300b4a99218a15f4f6b0ab4091d63a602a5)

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

bitbake: Fix script location after mishandled merge

(Bitbake rev: 0f3293c2196a30bc52bf1eebfae87d8477880572)

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

gcc-cross-initial: Fix typo causing limits.h not to be present

(From OE-Core rev: f26014b1b2e7ae0a23829487ca0c0bc05043b5d7)

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

tzdata: Upgrade tzdata to upstream 2012d

(From OE-Core rev: 89012452873181c0ad801022ec3b4851715a2e0a)

Signed-off-by: Radu Moisan <radu.moisan@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: cooker: fix OVERRIDES in BB_SIGNATURE_HANDLER (to override value from DISTRO.conf from local.conf)

We really need to pass the finalised data store into the parser init function
(and hence the siggen init function). This ensures any value changes get
passed into the correct code.

(Bitbake rev: 19efc6081c15a59bb8f5aaf8478650a2732cafe8)

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

bitbake: fetch2/local: Add search paths to the debug log to improve log usefullness

(Bitbake rev: 2054c7d99933c1523d4b5c7f65d37c69b8472e47)

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

bitbake: fetch2/local.py: Provide better debug output when fetch of a local file fails

When a fetch failure occurs for a local file, this patch ensures we print the
locations searched making it easier for the user to debug the problem.

(Bitbake rev: a461adbc5f09b41c771a7603370f6f2d1299ae8e)

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

bitbake: runqueue.py: Clean up runqueue exception catching in the normal task failure case

(Bitbake rev: a0bc58031d4eb31f8587171e870ecad059af5098)

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

texi2html: Fix perl location on recent distros

This fixes errors like:
| error: Failed dependencies:
|       /bin/perl is needed by texi2html-5.0-r1.i586

(From OE-Core rev: d4c27021ffc813732526ab9ae6969e5ae0bdf7e8)

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

gnome-desktop: Fix python location on recent distros

This fixes errors on distros with python as /bin/python of the form:

| error: Failed dependencies:
|  /bin/python is needed by gnome-desktop-2.32.1-r5.i586

(From OE-Core rev: 02c5299cf6f1c347e2653409b32be8fd39876ac9)

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

runqemu-gen-tapdevs: add UID to CL options

runqemu-if* now requires UID as a command line option which means
runqemu-gen-tapdevs needs to take it as an option.

(From OE-Core rev: a624ec02c4acec83efbe66406f22abafbdce5d63)

Signed-off-by: Elizabeth Flanagan <elizabeth.flanagan@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

Complete recipe enablement

RP: The list of recipes in multilib.conf needs to go away and we need
to just be able to extend all recipes with the multilib class.
Tested by building and running lib32-core-image-sato-sdk.

[YOCTO #1563]

(From OE-Core rev: 5d691f1bc95dfb9109b07827b33957b1fa5b2fa4)

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

bitbake: contrib/dump_cache.py: dump recipe -> packages mapping from bb_cache.dat

Add the dump_cache.py to dump the "recipe -> packages" mapping for
target recipes form bb_cache.dat:

* Usage:
  dump_cache.py bb_cache.dat

* The format is:
    recipe_path pn pv packages
  For example:
    /path/to/gzip_1.5.bb gzip 1.5 gzip-dbg gzip-staticdev gzip-dev gzip-doc gzip-locale gzip

* Only save the mapping for the target recipe

* We can extend this to dump other informations when needed.

* Put this script to bitbake/contrib/ (not to oe-core) is because it
  needs the bb.cache.

[YOCTO #2741]

(Bitbake rev: 75a7caf6f2d9f4399c95b9249db1b3bc5a48dc61)

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

bitbake: bitbake-layers: fix help for show-overlayed to match behaviour

show-overlayed actually lists the same recipe in multiple layers
regardless of priority, so change the help description to match.

(Bitbake rev: 17217ff707e41780cb473b9cdd2621dd18b982c2)

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

bitbake: Hob: change the Build failure scenario as ui design

As ui design 'build-fail-spec.pdf' to change the GUI

[YOCTO #2183]

(Bitbake rev: f9b81e44b2e71b4de6729bd2c69b25fd619b5fed)

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

bitbake: Hob: log for Hob and allow users to show logs after successful build

If users build images in Hob, record logs and allow users to retrieve the lo
after successful build.

The logs are generated if and only if:
      - users do "just bake"
      - users do "build image" after "build packages"
      - users do "build packages" only

[YOCTO #1991]

(Bitbake rev: 291289787bb042b99f0599babc2d67c220aadb87)

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

bitbake: fetch2: print checksums when they are different then expected

* in form which can be copied into a recipe without modification
* like oe-classic did since:
  http://git.openembedded.org/openembedded/commit/?id=68abc465559a68e9201c9525be3c4acc6297eaed
* it shows them in right form when they are missing completely, but in
  more verbose form when different
* it needs to print that only when checksums were requested, e.g.
  fetching from sstate mirror sets both md5mismatch and sha256mismatch,
  but your checksums shouldn't be shown

(Bitbake rev: 9eb34fc866775fd8310759a0111f232a9dc98981)

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

bitbake: fetch2: handle broken symlinks in local mirror handling

If a file:// mirror is being used, the fetcher will create a symlink to the
local file. However, if the local file gets removed, that link will be dead,
and os.path.exists() returns False in that case, so it tries and fails to
recreate the link. Now we unlink such a dead link if it exists.

(Bitbake rev: 229ed3857e826e3e215e843cb51f729c1e13ed37)

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

bitbake: knotty: ensure that directory for BB_CONSOLELOG exists before using it

(Bitbake rev: 1e06d8012868ba5a31503dc99cbf18570be629d9)

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

bitbake: fetch2/__init__.py: Add NoChecksumError exception

Without the new exception, when the system is configured to use premirrors,
but not allow network access (via BB_NO_NETWORK), when a recipe was lacking a
checksum the wrong error message(s) were being generated.

Instead of complaining about trying to perform network access, if the system
was able to find the item in the premirror, it should inform the user of the
lack of checksums, and the two SRC_URI fields they should use to update their
recipe.

(Bitbake rev: cb10e9c03a3f96d94e27e18330009616dde5e2b3)

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

guile: updated to 2.0.6

Some patches (disable_goops_optimizations, fix_cross_compilation) don't
seem to be needed anymore. Others (change-install-data-hook, mark-unused-modules)
were updated to work on 2.0.6. Tested by building and running guile under
QEMU with core-image-sato-sdk and also builing guile on MIPS.

(From OE-Core rev: 310f169d3d89f3a4fc6a540974a30c7eb565db3a)

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

busybox: udhcpc shouldn't be a service

udhcpc is invoked from the networking service, so it shouldn't exist
standalone. If /etc/init.d/busybox-udhcpc is present, it will cause
udhcpc to be started two times.

And since the script doesn't support the stop action, it will cause
udhcpc to be started even when system is being shutdown.

Remove /etc/init.d/busybox-udhcpc to resolve the issues.

[YOCTO #2840]

(From OE-Core rev: 8329c4679ec0ce319d2a81d755a0da5b05474688)

Signed-off-by: Jesse Zhang <sen.zhang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: hob: deal event DiskFull

Part of [Yocto #2168]

When bitbake runqueue is teminated by disk monitor, it will send event
DiskFull. Update to handle it.

(Bitbake rev: a691de10d8b8de552628b277e78d47c607ee3a46)

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

bitbake: monitordisk: fire event DISKFULL when terminate build

Part of [Yocto #2168]

Add a event DiskFull to descript the termination by disk monitor.

Update check() to fire the event DiskFull when terminates the build.
This could help UIs to deal this scenario and show more information to
end user.

(Bitbake rev: 0a0fca3d94a1db6458ae21501a66461a334410ed)

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

dbus: Ensure dbus-nativesdk doesn't RPROVIDE dbus-x11

dbus-x11 should not RPROVIDE dbus-x11 as this is incorrect and confuses
builds. This fixes the nativesdk case.

(From OE-Core rev: 09e3ebdb77c090ddc45fed7c6acc1591cd4b3595)

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

python-setuptools: Fix setuptools install libdir != /usr/lib

--install-lib= is required to be passed via distutils install
otherwise the install location gets set to whatever the python-native
location is.

(From OE-Core rev: d8ca8b21c6959e772fa1b437ce498b7759fae0ab)

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