wic: Remove unused plugin and error code

wic doesn't use it, so remove it.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove gpt_parser

wic doesn't currently use it, so remove.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove unused 3rdparty/commands

wic doesn't use anything but partition and bootloader, so remove the
rest.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove 3rdparty/urlgrabber

wic doesn't use it, so remove it.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove unused misc code

wic doesn't use it, so remove it.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove unused fs_related code

wic doesn't use it, so remove it.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove rpmmisc

wic doesn't use it, so remove it.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove proxy support

wic doesn't use proxy support, so remove it.

Also remove anything related to proxies in misc and conf, and while
we're at it, remove all the obviously unneeded code from those files -
it's easier to just remove it than to figure out the callchain to the
proxy code usages.

Basically the extra stuff relates to packaging, images, and config
files we don't use.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove grabber implementation

wic doesn't need to grab any urls, so remove it.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove fiemap support

wic doesn't use it, so remove it.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove bmap support

wic doesn't use it, so remove it.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove mic package managers

wic doesn't use the yum or zypp package managers, remove them.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove rt_util

More package-related stuff we don't need.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove mic chroot

mic chroot allows users to chroot into an existing mic image and isn't
used by wic, so remove it.

Removing chroot.py leads in turn to various plugin-loading failures
for a number of plugins that wic doesn't use either, so remove those
as well.

The existing source plugins refer to chroot but don't use it, so fix
those up.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove mic bootstrap

packaging bootstrap, not needed

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove packaging, config commands

Remove commands related to repos, packaging, configuration

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Remove unused custom commands

installer, repo, desktop-related stuff

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

wic: Make exec_cmd() error out instead of warn

The reason exec_cmd() warns but doesn't error out (broken parted)
doesn't really make sense, since the parted invocations don't even use
exec_cmd().  It really should just fail since by not doing so it's
actually enabling invalid images in some cases.

Also, since the return code is now always zero, there's no point in
having a return code, so remove it.  This represents a change in the
API, so we also need to update all callers.

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>

qemu: upgrade to 2.1

QEMU 2.1 comes with fixes and improvements.
See http://wiki.qemu.org/ChangeLog/2.1 for details.
- Added config for quorum support, depending on gnutls.
- pcie_better_hotplug_support.patch removed,
integrated upstream.
- Qemu-Arm-versatilepb-Add-memory-size-checking.patch updated
to 2.1 source code.
- no-strip.patch removed, no longer necessary due to code changes.

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

trace-cmd: Update to version 2.3.2

Current trace-cmd version 1.2 throws "recorder error in
splice input" error while recording a trace on kernel
version 3.10. Update the trace-cmd version to latest 2.3.2
to resolve this error and drop the redundant patches.

* patches dropped - not needed anymore
 - trace-cmd-Add-checks-for-invalid-pointers-to-fix-seg.patch - backport
 - trace-cmd-Do-not-call-stop_threads-if-doing-latency-.patch - backport
 - trace-cmd-Setting-plugin-to-nop-clears-data-before-i.patch - backport
 - trace-cmd-fix-syntax-error-of-shell.patch - backport

Signed-off-by: Yasir-Khan <yasir_khan@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

perl-rdepends: add CPAN, CPAN::Meta requirements

While testing libmodule-build-tiny-perl, it was discovered that
perl-module-cpan is missing some RDEPENDS.

* Run "perl -mCPAN::Meta" on a target, the following is missing:
  -- perl-module-parse-cpan-meta (Parse::CPAN::Meta)

* Run "perl -mCPAN" on a target, the following (and others) are
  missing:
  -- perl-module-file-glob (File::Glob)
  -- perl-module-config-git (Config_git)

* Also added missing modules from runtime-requires in
  https://metacpan.org/source/ANDK/CPAN-2.05/META.json

This patch adds them to perl-rdepends for  ${PN}-module-cpan

Signed-off-by: Tim Orling <TicoTimo@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

kexec-tools: reinstate 'kexec-tools' package

The vmcore commit emptied out ${PN} leading to things like:

Collected errors:
 * opkg_install_cmd: Cannot install package kexec-tools.

Let's do the right thing and make ${PN} an empty meta-package that drags in all tools like people expect it to do.

Signed-off-by: Koen Kooi <koen.kooi@linaro.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

ubootchart: delete ubootchart recipe

The Ubootchart recipe had known issues. Ubootchart itself is no longer updated
upstream. Ubootchart is also now redundant with Bootchart2.

If people still want ubootchart around, it can be moved to meta-oe.

Ubootchart removed as part of the solution to [YOCTO #5893].

Signed-off-by: Max Eliaser <max.eliaser@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

core-image.bbclass: Correct comment about packagegroups.

Default core image actually includes packagegroup-base-extended, not
just packagegroup-base.

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

gstreamer1.0-plugins-good: fix the bug for gstrtpmp4gpay

set dafault value for MPEG4 without codec data in caps

https://bugzilla.gnome.org/show_bug.cgi?id=734263

Signed-off-by: Wang Zidan <b50113@freescale.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

bootchart2: create recipe for bootchart2

This recipe creates packages for the bootchart2 system-wide profiler daemon
and related utilities. It fetches the Git revision immediately past the one
corresponding to the 0.14.6 release of bootchart2. (0.14.6 had a systemd-
related bug that was corrected right after it was tagged.)

The recipe contains three packages:
   *   bootchart2 - The daemon itself.
   *   pybootchartgui - Python program to visualize and display the data
       collected by bootchart2 or compatible daemons such as the original
       bootchart.
   *   bootchartd-stop-initscript - A SysV init script to stop data collection
       when booting completes.

Depending on how you wish to use bootchart2, you may not end up having all
three of those packages installed on your image.

There is also a bootchart2-native variant, which is intended solely to provide
a native version of the pybootchartgui utility. The non-cross-compiled version
of the bootchart2 daemon has not been tested at all, don't use it.

The recipe is extensively documented. Read the comments at the beginning of
bootchart2_git.bb or else you'll have no idea how to use it.

This recipe is based on a recipe from meta-WebOS. The WebOS people had some
extra code (including patches against the bootchart2 code) to support the
Upstart init system. However, since upstream Poky does not support Upstart,
that stuff is being left behind. The WebOS people can write a bbappend to re-
add it.

Original recipe written by Wonhong Kwon of LG. Upstreamed as part of the
solution to [YOCTO #5893].

Signed-off-by: Max Eliaser <max.eliaser@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

v86d: fix for systemd to load uvesafb module correctly

The /etc/init.d/fbsetup script doesn't have any effect in a systemd
image. Its purpose is to load the uvesafb kernel module at boot.
This functionality could be achieved by adding a configuration file
under /etc/modules-load.d/ directory which would be parsed by the
systemd-modules-load.service.

[YOCTO #4420]

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

alsa-state: install init script only when 'sysvinit' is in DISTRO_FEATURES

The /etc/init.d/alsa-state is totally useless for a systemd image.
Its functionality has been replaced by alsa-state.service files.

So if 'sysvinit' is not in DISTRO_FEATURES, installing this script doesn't
make any sense.

[YOCTO #4420]

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

syslinux: fix reinstall error

Fixed:
make: *** No rule to make target `/path/to/sysroot/4.9.0/include/stdarg.h', needed by `cpio.o'.  Stop.
make: *** Waiting for unfinished jobs....
ERROR: oe_runmake failed

This happens when upgrade gcc from 4.9.0 to 4.9.1, and the .cpio.o.d isn't
regenerated when recompile (the compile happens when do_install), the content
of it are:

[snip]
cpio.o: /path/to/sysroot/4.9.0/include/stdarg.h
[snip]

And Makefile includes the .cpio.o.d file if it exists, so there would be
errors when /path/to/sysroot/4.9.0/include/stdarg.h doesn't exist.

Remove .*.d (a few .d files, its Makefile uses this wildcard) will fix
the problem.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

u-boot-mkimage: fix recompile error

Fixed:
make: *** No rule to make target `/path/to/sysroot/4.9.0/include/stddef.h', needed by `crc32.o'.  Stop.
make: *** Waiting for unfinished jobs....
ERROR: oe_runmake failed

This happens when upgrade gcc from 4.9.0 to 4.9.1, and the .depend isn't
regenerated when recompile, the content of it are:

[snip]
crc32.o: /path/to/sysroot/4.9.0/include/stddef.h
[snip]

And Makefile includes the .depend file if it exists, so there would be
errors when /path/to/sysroot/4.9.0/include/stddef.h doesn't exist.

Remove .*.d (a few .d files, its Makefile uses this wildcard) will fix
the problem.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

kernelshark: fix recompile error

Fixed:
make: *** No rule to make target `/path/to/sysroot/4.9.0/include/stddef.h', needed by `parse-events.o'.  Stop.
make: *** Waiting for unfinished jobs....
ERROR: oe_runmake failed

This happens when upgrade gcc from 4.9.0 to 4.9.1, and the
.parse-events.d isn't regenerated when recompile, the content of it are:

[snip]
parse-events.o: /path/to/sysroot/4.9.0/include/stddef.h
[snip]

And Makefile includes the .parse-events.d file if it exists, so there
would be errors when /path/to/sysroot/4.9.0/include/stddef.h doesn't
exist.

Remove .*.d (a few .d files, its Makefile uses this wildcard) will fix
the problem.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

kexec-tools: fix recompile error

Fixed:
NOTE: make -j 32
make: *** No rule to make target `/path/to/sysroot/4.9.0/include/stddef.h', needed by `kexec/kexec.o'.  Stop.

This happens when upgrade gcc from 4.9.0 to 4.9.1, and the kexec/kexec.d
isn't regenerated when recompile, the content of it are:

[snip]
kexec/kexec.o: /path/to/sysroot/4.9.0/include/stddef.h
[snip]

And Makefile includes the kexec/kexec.d file if it exists, so there
would be errors when /path/to/sysroot/4.9.0/include/stddef.h doesn't
exist.

Remove kexec/kexec.d and other similar files will fix the problem.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

logrotate: fix recompile error

Fixed:
NOTE: make -j 32
make: *** No rule to make target `/path/to/sysroot/4.9.0/include/stddef.h', needed by `logrotate.o'.  Stop.

This happens when upgrade gcc from 4.9.0 to 4.9.1, and the .depend isn't
regenerated when recompile, the content of the .depend are:

[snip]
logrotate.o: /path/to/sysroot/4.9.0/include/stddef.h
[snip]

And Makefile includes the .depend file if it exists, so there would be
errors when /path/to/sysroot/4.9.0/include/stddef.h doesn't exist.

Remove .depend will fix the problem.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

libnewt: fix recompile error

Fixed:
NOTE: make -j 32
make: *** No rule to make target `/path/to/sysroot/4.9.0/include/stddef.h', needed by `test.o'.  Stop.

This happens when upgrade gcc from 4.9.0 to 4.9.1, and the .depend isn't
regenerated when recompile, the content of the .depend are:

[snip]
test.o: /path/to/sysroot/4.9.0/include/stddef.h
[snip]

And Makefile includes the .depend file if it exists, so there would be
errors when /path/to/sysroot/4.9.0/include/stddef.h doesn't exist.

Remove .depend will fix the problem.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

busybox: Add Upstream-Status to patch

Signed-off-by: Saul Wold <sgw@linux.intel.com>

sysvinit: allow stack size configuration from rcS

For certain swap/overcommit settings (e.g. when overcommit is disabled
on a real-time system), we need to limit the stack size used by
initscripts.  When the STACK_SIZE environment variable is set (usually
in /etc/default/rcS), ulimit the stack size to the value specified.

Make the stack size ulimit a soft limit, which allows the user to
increase the stack size where required without having to run the
respective application as root.

Signed-off-by: Scot Salmon <scot.salmon@ni.com>
Signed-off-by: Gratian Crisan <gratian.crisan@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Acked-by: Rich Tollerton <rich.tollerton@ni.com>
Acked-by: Brad Mouring <brad.mouring@ni.com>
Acked-by: Bill Pittman <bill.pittman@ni.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

wpa-supplicant: use PACKAGECONFIG for ssl selection

Select between openssl or gnutls as ssl implementation via
PACKAGECONFIG instead of explicitly adding both via DEPENDS.

Signed-off-by: Yasir-Khan <yasir_khan@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

lz4: update to 120

Update package to version 120
Add Makefile to LIC_FILES_CHKSUM check as it changes with every rev.

Signed-off-by: Armin Kuster <akuster@mvista.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

boot-directdisk.bbclass: use local HDDDIR

Use $HDDDIR rather than ${HDDDIR} in build_boot_dd(), otherwise may
errors:
- Set these in local.conf:
  IMAGE_FSTYPES += "live"
  IMAGE_FSTYPES += "vmdk"
  AUTO_SYSLINUXMENU = "1"
  NOHDD = "1"

$ bitbake core-image-sato
  DEBUG: Executing shell function build_boot_dd
  install: cannot create regular file
  `/path/to/core-image-sato-1.0/hddimg//vesamenu.c32': No such file or directory
  WARNING: exit code 1 from a shell command.

This because it uses the ${HDDDIR} which is set in bootimg.bbclass, use
local HDDDIR which is set in build_boot_dd() will fix the problem.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

remake: fix build error

Fix build error:
...
unknown command `colophon'
unknown command `cygnus'

Signed-off-by: Maxin B. John <maxin.john@enea.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

tzdata: update to 2014f

md5sum on asia file changed do to severel tzone updates

Signed-off-by: Armin Kuster <akuster@mvista.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

tzcode: update to 2014f

Signed-off-by: Armin Kuster <akuster@mvista.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

ghostscript: Remove bogus gsfonts reference from DESCRIPTION

[YOCTO #5586]

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

documentation.conf: Small tweak to INITRD variable

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

code cleanup for INITRD variable handing

* Remove an unnecessary check
* Instead of ignoring, report the errors

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

base-files: set dynamic COLUMNS via resize command

By default, COLUMNS is set to 80. If possible, run 'resize' to
determine what the current dimensions are. This avoids the final
part of long lines overlap the start of the same line.

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

busybox: Make busybox connman aware

This change makes busybox aware of both connman and the standard dhcp
client, while there is a caveat about using busybox's ifup/ifdown, this
change will make the 3 possibly dhcp clients behave better when busybox
is used.

Enable the CONFIG_FEATURE_IFUPDOWN_EXTERNAL_DHCP feature to busybox will
search for various external clients and then add connman as an external
client.

[YOCTO #6521]

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

oelint.bbclass: Check for ${PN} or ${P} usage

Check for ${PN} or ${P} usage in SRC_URI or S.
We should use ${BPN} or ${BP} instead to avoid breaking multilib.

[YOCTO #5427]

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

oelint.bbclass: add patch checking

Check that all patches have Signed-off-by and Upstream-Status.

[YOCTO #5427]

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

oelint.bbclass: make oelint bbclass work

Remove LICENSE, MAINTAINER, PRIORITY and valid SECTION checking.
Convert tab indentation into four-space.

[YOCTO #5427]

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

curl: remove unused patch

Remove unused "remove_inappropriate_file_from_rel.patch"

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

perf: fix indentation

* multiline variables should use spaces not tabs for indentation
* do_configure is also using wrong indentation, but I'm not fixing
  this one (still hoping that we'll eventually fix styleguide to
  use 4 spaces everywhere)

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

perf: explicitly disable libunwind

* it's not included in DEPENDS and it's detected from sysroot

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

Add doc entry for "ptest-pkgs" in IMAGE_FEATURES list

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

scripts/send-error-report: fetch /Errors/ instead of /.

If HTTP_PROXY or http_proxy is set when the send-error-report script
is run, it will check to see if fetching / on the specified server
returns 200 without the proxy set. If it does it will assume that the
proxy is not needed. However this check can never work because
fetching / always redirects to /Errors/ in the current code and
thus returns code 301. This is fixed by fetching /Errors/ instead of /.

[YOCTO #YB6576]

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

gstreamer1.0-omx: Remove unnecessary dependency on plugins-bad.

This also corrects an erroneous overwriting of the dependencies
from the inc file.

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

kexec-tools: Create separate package for vmcore-dmesg

The kexec-tools recipe already specifies separate packages for kexec and
kdump. Thus, it follows that a separate package can also be used to install
vmcore-dmesg granularly.

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

slang 2.2.4: fix the iconv existence checking

When checking whether there is iconv, the configure file always check
the host env.

Now we make it working properly by adding correct prefix for cross-
compiling environment.

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

cups: regen configure so cups-no-gcrypt.patch has an effect

The existing patch backports a change to the m4 source file, but without
a forced autoconf before configure it has no effect.

Signed-off-by: Peter A. Bigot <pab@pabigot.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

cairo: explicitly disable LTO support by backporting patch which removes it

* cairo-native was failing to build in gentoo with gcc-4.9 and LTO
  enabled, more details in upstream bug
  https://bugs.freedesktop.org/show_bug.cgi?id=77060

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

tcp-wrappers: install /etc/hosts.allow and /etc/hosts.deny

Signed-off-by: Yao Xinpan <yaoxp@cn.fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

tcp-wrappers: fix do_install when $prefix == $base_prefix

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

formfactor: Grammar fix in recipe comment.

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

test-dependencies: don't override failed-recipes.log

* when dependency changes are reported in OUTPUTC/failed-recipes.log
  it can be the same as final output file which is later overwritten
  because OUTPUTC == OUTPUT_BASE by default
* use similar format for messages as insane_qa check

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

nss*: Replace hardcoded "/etc" with "${sysconfdir}"

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

alsa-lib: pcm_local.h: include <time.h> to enable CLOCK_MONOTONIC

CLOCK_MONITONIC is defined in <bits/time.h>, add <time.h> before <sys/time.h>.

Signed-off-by: Wang Zidan <b50113@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

alsa-lib: fix hw_ptr exceed the boundary

For long time test case, the hw_ptr will exceed the boundary, then cause
the avail size wrong.

Signed-off-by: Wang Zidan <b50113@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

alsa-lib: use get32_labels for multi-source

The PCM route plugin can assign the destination value from average of
multiple sources with attenuation. This requires the read of each
channel value, sums and writes the resultant value in the requested
format.

get_labels gives the value as is only with endianness and signedness
conversions, but put32_labels assumes that the value is normalized to
32bit int and it shifts down to the dest format. In addition, the
current code lacks get_labels entries for the 24bit formats.

For fixing these bugs, this patch replaces the read with get32_labels
and use always 64bit int for sum.

Signed-off-by: Wang Zidan <b50113@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

psmisc: Typo in fuser makes -M on all the time

psmisc: Typo in fuser makes -M on all the time

  It adds a patch to fix a major issue on fuser that behaves as if -M option is always used.

  More info on psmisc website here :
  http://sourceforge.net/p/psmisc/code/ci/3638cc55b4d08851faba46635d737b24d016665b/

Signed-off-by: Patrice Bouchand <pbfwdlist@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

tcmode-default.inc: allow override of binutils-crosssdk

Sometimes, binutils-crosssdk needs to be override,
like is the case for producing Darwin cross SDK.

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

local.conf.sample.extended: fix example for EXTRA_USERS_PARAMS

The `groupdel' command doesn't support '-g' option, so remove it.

[YOCTO #6575]

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

autotools: Exclude SDK_OS from autotools task signatures

The change to handle SDKs of different types resulted in the signatures
becoming SDK specific which wasn't intended. Exclude the variable from the
hash values to avoid this.

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

package: Fix darwin shlibs code

We need to scan binaries as well as libraries for dependencies.

Also ensure if its not an object file (as found by otool), we handle
this case.

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

package: Convert dylib handling from .la to otool

Currently, the darwin shlibs detection is done by parsing the .la file
dependency fields. This is very old code and is incomplete in some
cases so convert to using otool -l and otool -L to correctly load
the rpath and dependency information.

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

package: Fix pkgdest for darwin shlibs code

When the code was rewritten we forgot to strip the pkg that is present
in the pkgdest path. This was fixed in the linux version of the code
but not the darwin one, this matches the fix.

Without this, the provider paths are broken.

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

autotools: Fix warning for odctools-crosssdk

odcctools-crosssdk doesn't use the suffixed naming the rest of crosssdk does
and this results in a annoying build warning. Avoid this.

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

sstate: Allow switching between linux and non-linux SDK builds within the same tmpdir

Currently if you try and switch between linux/darwin/mingw SDK builds in the
same TMPDIR, things break. This is due to sstate not reflecting the SDK_OS in
the manifest names.

Since they are different, reflect this in the manifest naming and allow
this to work.

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

bdwgc: Upgrade 7.2d -> 7.4.2

Update the SRC_URI to the new site, same as HOMEPAGE. The libatomic
pieces were removed and some of the READMEs, allowing simplification
of the project license which is clear now.

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

local.conf.sample.extended: update for preferred ABI

[YOCTO #4073]

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

opkg-utils: Update SRCREV

opkg-build now checks whether tar supports the '--format' option before using
'--format=gnu' so that packages can be build with both Busybox tar (no
'--format' option) and GNU tar (defaults to posix format unless told otherwise
on some distros).

Signed-off-by: Paul Barker <paul@paulbarker.me.uk>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

mesa-demos: upgrade to 8.2.0 and allow building without GLEW/GLU

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

tune-mips*: Ensure tunes are inherited in order

Without this, you are not able to use mips32r2 on a mips64 based tune.

We want to be able to do a tri-lib system of mips64, mips64-n32 and mips32r2.

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

gcc: Fix gcc-multilib-config comparison

Fix an issue on a multilib configuration that contains more then 1 multilib.

I.e. on MIPS64:

DEFAULTTUNE = "mips64"
MULTILIBS = "lib32n:mips64_n32 lib32:mips32"

While normally you'd use 'libn32', the above is legal.

With the startswith code, the system will look to expand the 'lib32' element
and find the 'lib32n' instead, and will result in a warning:

lib32 doesn't have a corresponding tune. Skipping...

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

sanity.bbclass: Add ability to verify toolchain flags

When attempting to use a binary toolchain, such as meta-mentor,
we want the ability to verify that the CCARGS, ASARGS and LDARGS
contain the necessary and appropriate flags.

This change specifically verifies, if set:
   TUNEABI_REQUIRED_CCARGS_tune-<tune>
   TUNEABI_REQUIRED_ASARGS_tune-<tune>
   TUNEABI_REQUIRED_LDARGS_tune-<tune>

Each of these, will be processed by the class and verified that the
selected tune's CCARGS, ASARGS, and LDARGS contains the listed item.  This
can be used to validate that the user has not accidently or otherwise
missed an argument.  Note, conflicting arguments are not verified.

Without verification it's possible for a misconfiguration to go
undetected, presenting runtime and debugging errors.

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

init-install: Skip CDROM devices during probe

Signed-off-by: Drew Moseley <drew_moseley@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

init-install: Strip partition number from live_dev_name

This is needed in case the boot disk was created with mkdiskimage.
In that case the parameter passed is a variant of /dev/sda4 which
includes the partition number.  Without this change this install script
will offer to install onto the live media.

Signed-off-by: Drew Moseley <drew_moseley@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

init-install-efi.sh: Verify /sys based files exist before displaying them

Some mmc cards do not have all the data files in /sys/block
populated.  Check for existence before displaying the files
to avoid erroring out of the install process.

Signed-off-by: Drew Moseley <drew_moseley@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

gcc: Upgrade 4.9.0 -> 4.9.1

Drop patches which are already available in 4.9.1

Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

package_manager.py: set preferred ABI for rpm

When using the RPM packaging backend to generate a rootfs there needs to
be a way to configure the preferred ABI to resolve ELF file conflicts.

Currently RPM resolves ELF file conflicts with the last-installed wins.
Using SMART it's difficult to know what the last installed will be.

There are three specific policies that can be selected:
1: ELF32 wins
2: ELF64 wins
3: ELF64 N32 wins (mips64 or mips64el only)

Another option "0" is uncontrollable, which means that if two are being
installed at once Elf64 is preferred, but if they're being installed in
two different transactions, last in wins, so we don't document it.

Add RPM_PREFER_COLOR to let the user config the preferred ABI.

[YOCTO #4073]

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

libnl: Upgrade to 3.2.25

Remove one patch and rebase another

Signed-off-by: Saul Wold <sgw@linux.intel.com>

libcheck: Upgrade to 0.9.14

Signed-off-by: Saul Wold <sgw@linux.intel.com>

lsbinitscripts: Upgrade to 9.55

Signed-off-by: Saul Wold <sgw@linux.intel.com>

help2man-native: Upgrade to 1.46.1

Signed-off-by: Saul Wold <sgw@linux.intel.com>

oeqa/utils/decorators.py: add import os

An 'import os' was omitted here while testing the previous decorators using runtime tests that import the os module before this one. Unfortunately oe-selftest fails because of this missing import.

Signed-off-by: Corneliu Stoicescu <corneliux.stoicescu@intel.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

qemu.inc: Allow optional use of pkg-config from the HOST

Currently, if pkg-config isn't installed on the build system, this code
can cause an error. We don't need to require this, only use it if its
present so allow the test to fail gracefully.

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

gstreamer1.0-omx: upgrade to 1.2.0

Signed-off-by: Carlos Rafael Giani <dv@pseudoterminal.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

gstreamer1.0-libav: upgrade to 1.4.0

* Backported patch removed, since it is already included in this release

Signed-off-by: Carlos Rafael Giani <dv@pseudoterminal.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

gstreamer1.0-plugins-ugly: upgrade to 1.4.0

Signed-off-by: Carlos Rafael Giani <dv@pseudoterminal.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

gstreamer1.0-plugins-bad: upgrade to 1.4.0

* eglglessink was replaced by glimagesink
* sndfile plugin has been ported to 1.0
* webp support added since libwep recipe has been added to meta-multimedia
* mfc was replaced by v4l2videodec, which does not need special flags
* cdaudio was removed
* directshow plugin was replaced by winks

Signed-off-by: Carlos Rafael Giani <dv@pseudoterminal.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>

gstreamer1.0-plugins-good: upgrade to 1.4.0

* xvideo options removed, since the V4L2 XV support never became upstream
  and is pretty much dead now

Signed-off-by: Carlos Rafael Giani <dv@pseudoterminal.org>
Signed-off-by: Saul Wold <sgw@linux.intel.com>