wget: upgrade from 1.14 to 1.15

* Removed:
  - fix_doc.patch (backport)
* Add the HOMEPAGE.

(From OE-Core rev: c5ed2e443d8f3d6c31a3fa99cddc20b8f89f7d0e)

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

file: upgrade from 5.16 to 5.17

(From OE-Core rev: cf1a4a8b1a0ece2f75fab9d79c69ce2726ef78b9)

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

lttng-modules: update to 2.4.1 version

    *This updates lttng-modules for 2.4.1 and
     it also fixes the build with 3.13 Linux kernel.

(From OE-Core rev: 023c2401677f9af6895887d6e8bca2fd1d730985)

Signed-off-by: Chunrong Guo <B40290@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

grub-efi: Use a variable to specify built-in grub modules.

The previous behavior defines a static set of modules that are built
into the grub efi executable. This works fine for a limited set of boot
environments namely the standard linux/initrd. This patch conditionally
 assigns the same modules to a variable. This allows other meta layers
to add additional modules or completely override the defaults. The use
case driving this patch is the use of multiboot2 and related modules.

(From OE-Core rev: 597f8e0040ba3135220000b23767858c64b5c9b8)

Signed-off-by: Philip Tricca <flihp@twobit.us>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

openssh-sshd: host contamination fix

If you do a readelf -x .rodata /path/.../to/openssh/6.5p1-r0/packages-split/openssh-sshd/usr/sbin/sshd

You'll see two references to OE's sysroots/${BUILD_SYS} login and passwd binaries.
First one can be overridden with LOGIN_PROGRAM environment variable (see configure.ac),
second needs a cached variable definition.

(From OE-Core rev: 2127c80d0cc2a3a4d676bd3c0890454a175fba8e)

Signed-off-by: Matthieu Crapet <Matthieu.Crapet@ingenico.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

grub git: workaround debugedit problems

Debugedit errors out on bare metal binaries. The first version of this patch limited it to 64 bit targets, but the problem now shows up on 32 bit targets (minnowboard) as well.

(From OE-Core rev: f95c7611007ddda3fe36db90168c7d5de096f9cd)

Signed-off-by: Koen Kooi <koen.kooi@linaro.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

make menuconfig work for recent xfce environment

xfce terminal was renamed 'Terminal' -> 'xfce4-teminal' mainline end of 2012,
so the distros supporting 'Terminal' will dissapear. The distros not
mentionied in  __init__ do (e.g fedora 19 - tested) fail - or will fail
sooner or later.

(From OE-Core rev: d07f3812ec371da6f18fa1dd920cdde470bd89ad)

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

linux-yocto/3.14: aufs, edgerouter config and -rt

Updating the 3.14 SRCREVs to import the following changes:

  - enable AUFS: a missing Kbuild patch was preventing aufs from compiling
  - edgerouter: remove RTC configuration options
  - preempt-rt recipe. The patch for 3.14 is availble, so we populate the recipe.

(From OE-Core rev: 8e38a24a9d9cd44733a25d3de307907a788d643d)

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

recipe_sanity.bbclass: avoid error when running 'bitbake -e'

Running 'bitbake -e' without further arguments causes a stack trace on stderr:

| ERROR: Command execution failed: Traceback (most recent call last):
|   File "[...]/bitbake/lib/bb/command.py", line 99, in runAsyncCommand
|     commandmethod(self.cmds_async, self, options)
|   File "[...]/bitbake/lib/bb/command.py", line 405, in showEnvironment
|     command.cooker.showEnvironment(bfile)
|   File "[...]/bitbake/lib/bb/cooker.py", line 453, in showEnvironment
|     logger.plain("\npython %s () {\n%s}\n", e, data.getVar(e, envdata, 1))
|   File "[...]/bitbake/lib/bb/data.py", line 89, in getVar
|     return d.getVar(var, exp)
|   File "[...]/bitbake/lib/bb/data_smart.py", line 522, in getVar
|     return self.getVarFlag(var, "_content", expand, noweakdefault)
|   File "[...]/bitbake/lib/bb/data_smart.py", line 612, in getVarFlag
|     value = self.expand(value, cachename)
|   File "[...]/bitbake/lib/bb/data_smart.py", line 350, in expand
|     return self.expandWithRefs(s, varname).value
|   File "[...]/bitbake/lib/bb/data_smart.py", line 340, in expandWithRefs
|     raise ExpansionError(varname, s, exc)
| ExpansionError: Failure expanding variable can_delete_FILESPATH, expression was def can_delete_FILESPATH(cfgdata, d):
|     expected = cfgdata.get("FILESPATH")
|     #expected = "${@':'.join([os.path.normpath(os.path.join(fp, p, o)) for fp in d.getVar('FILESPATHBASE', True).split(':') for p in d.getVar('FILESPATHPKG', True).split(':') for o in (d.getVar('OVERRIDES', True) + ':').split(':') if os.path.exists(os.path.join(fp, p, o))])}:${FILESDIR}"
|     expectedpaths = d.expand(expected)
|     unexpanded = d.getVar("FILESPATH", 0)
|     filespath = d.getVar("FILESPATH", True).split(":")
|     filespath = [os.path.normpath(f) for f in filespath if os.path.exists(f)]
|     for fp in filespath:
|         if not fp in expectedpaths:
|             # __note("Path %s in FILESPATH not in the expected paths %s" %
|             # (fp, expectedpaths), d)
|             return False
|     return expected != unexpanded
|  which triggered exception AttributeError: 'NoneType' object has no attribute 'split'

Removing the commented second line in can_delete_FILESPATH() hides the error.

(From OE-Core rev: a84c36e0e5e8332ddc5a6c34e1f598d5cb87cee2)

Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

scripts/send-error-report: simple hack to use proxy from the enviroment

People behind a proxy couldn't send an error report to an upstream server,
this should fix the issue if they use a proxy that doesn't require authentication,
or one that uses basic http authentication and it's correctly exported in the enviroment.

(From OE-Core rev: a8511ee80246b4e2caa353b87f4b586f1539e6d4)

Signed-off-by: Stefan Stanacar <stefanx.stanacar@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

icecc: don't create unnecessary 'ice' dirs in sysroots when disabled

* parsing ICE_PATH="${@icc_path(bb, d)}" causes "ice" directories
  to be created in every sysroot, that could be a bit confusing for
  people who inherit icecc.bbclass, but disabled it
* shorten ICECC_VERSION="${@icc_version(bb, d)}" path a bit
  by returning sooner when disabled
* remove ICECC_PATH and ICECC_ENV_EXEC from signatures, we assume that
  using icecc doesn't influence the output, so it shouldn't matter when
  user supplies own version of icecc or env script
* always compare ICECC_DISABLED with "1", boolean typed_value isn't used
  because documentation already mentions using empty value to keep icecc
  enabled and that's not valid boolean value when oe.data.typed_value is
  used:
  ERROR: ICECC_DISABLED: Invalid boolean value ''

(From OE-Core rev: c3e8bfe30685e2357a6eb3ba8f4a014c7dc9f58e)

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

shadow: fix building systemd with useradd-staticids.bbclass enabled

| groupadd: 'systemd-journal-gateway' is not a valid group name

Without useradd-staticids enabled, group 'systemd-journal-gateway' is created
by useradd and that seems not to care for GROUP_NAME_MAX_LENGTH which has 16 by
default.

(From OE-Core rev: 33c7892326de296cc6d143577be5b395ac887d91)

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

selftest: DiskMonTest: use a high value for free space

Using the free space of the host works when
oe-selftest is the only build running, but if something else
on the host remove things this will fail (as seen on AB).
Using an absurdly high value should fix this.

(From OE-Core rev: 413fc934fadbd3603b7f78dfd3cc0ac83bb0377f)

Signed-off-by: Stefan Stanacar <stefanx.stanacar@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

libav: Add libsdl to DEPENDS only when x11 is enabled

When we have opengl in distro features but not x11 and try to
build libav then it calls out to build libsdl which inturn has depependency on libglu
and libglu fails to build

ld: error: cannot find -lGL
| collect2: error: ld returned 1 exit status
| make: *** [libGLU.la] Error 1
| ERROR: oe_runmake failed

since libglu wants glx enabled in mesa and glx in mesa is (righly) enabled only when x11
is in distro features.

This breaks the dependency chain leading to this problem

(From OE-Core rev: fe9fe5ca5039743fc80d14f0518b95929c458b1a)

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

libnotify: add dbus-glib dependency

* | No package 'dbus-glib-1' found
  | configure:11865: error: Package requirements (gtk+-2.0 >= 2.18, glib-2.0 >= 2.6,
    dbus-1 >= 0.76, dbus-glib-1 >= 0.76 gmodule-2.0) were not met:

* introduced in:
  commit 5c1eeb5d16566521e45947b07fdcd9f552fec45b
  Author: Alexandru Palalau <alexandrux.palalau@intel.com>
  Date:   Thu Feb 27 12:05:41 2014 +0200

      libnotify: don't use gnome.bbclass

  because gnome.bbclass was adding gconf->dbus-glib dependency

* it was also causing other recipes to fail when they were depending
  on libnotify which has dbus-glib in pkg-config and dbus-glib was
  missing, e.g. firefox:
  | checking for libnotify >= 0.4... Package dbus-glib-1 was not found
  in the pkg-config search path. Perhaps you should add the directory
  containing `dbus-glib-1.pc' to the PKG_CONFIG_PATH environment
  variable Package 'dbus-glib-1', required by 'libnotify', not found
  | configure: error: Library requirements (libnotify >= 0.4) not met;
  consider adjusting the PKG_CONFIG_PATH environment variable if your
  libraries are in a nonstandard prefix so pkg-config can find them.

(From OE-Core rev: abb2571a6352fc33a8b78ab0886d443894cda3b6)

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

apr: remove the use of ${SHELL} to avoid bash/dash confliction

While multible hosts sharing a common sstate cache, the fist host using
bash as default shell and build apr, the second host using dash as
default shell and build apr-util, there was a failure in apr-util:
...
| /bin/sh: 0: Can't open i586-poky-linux-libtool
| make[1]: *** [dbm/apr_dbm_gdbm.lo] Error 127
| make[1]: *** Waiting for unfinished jobs....
| make[1]: Leaving directory
`tmp/work/i586-poky-linux/apr-util/1.5.2-r0/apr-util-1.5.2'
...

The quick way to reproduce the defect in Ubuntu 1204:
1. Create a new build
2. sudo dpkg-reconfigure dash <set bash as /bin/sh>
3. bitbake apr
4. sudo dpkg-reconfigure dash <set dash as /bin/sh>
5. bitbake apr-util

Remove the use of $(SHELL) in the apr could avoid this issue.

(From OE-Core rev: 53af71200b6f6d117e1e7dc2e7a7d21c6faab8f9)

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

build-appliance-image: Update to head revision

(From OE-Core rev: bec37f8b85b6dd1e4e1c1b9b215a7952ca2fc386)

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

bitbake: bitbake-user-manual-hello.xml: Edits to the "Hello World" Appendix.

Fixes [YOCTO #1973]

These are review changes applied from Paul Eggleton.

(Bitbake rev: 431181705f1bc3fc51439faa04a193debc1d03fd)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake-user-manual-hello.xml: First draft of "Hello World" appendix

Fixes [YOCTO #1973]

Added the complete "Hello World" appendix first draft beyond
what Bill Traynor had supplied.

(Bitbake rev: 176baaf3d0f033091f5ab9d9cfb2cec3d25a5442)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake-user-manual-hello.xml: Edits to environment for hello world.

Some general edits to the sections that get you set up to run
the "Hello World" example.

(Bitbake rev: 43286d31c22e6e2fa96ba1b82445e85db0a0f18c)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake-user-manual-metadata.xml: Added section on pathname syntax

Added a brief subsection to the "Basic Syntax" section that
describes the issue using the tilde character (~) when providing
a pathname as part of a statement.  BitBake does not expand
this character like the shell does.  We recommend to not use
it in pathnames.

(Bitbake rev: 8e4c5cee932c3f38d4147c59612d76ca0b002727)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake-user-manual-ref-variables.xml: Edits to the BBPATH variable.

I took the stuff about running bitbake outside the build
directory and removed the note syntax.

(Bitbake rev: 783f29a593ce16a37b275b02f71b3ec918ef8c91)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake-user-manual-ref-variables.xml: Fixed BBPATH example

The example used in the note in the BBPATH variable description
had incorrect syntax.  Spaces existed in the BBPATH = statement.
I have removed them.

(Bitbake rev: 896a1c102a4afda0e5d07d009bd7918044f88269)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: user-manual-hello.xml: General edits

I worked through the chapter and made some minor edits based
on my user experience.  Part of what I did was to update
the console output for the examples.

(Bitbake rev: 13bf6cf71ff49572c3413435a33d87e5b6a8df0d)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake-user-manual-hello.xml: Fixed two broken links.

(Bitbake rev: 7a0948beba2c882de31ec9b98c7740bba230aed9)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake-user-manual-metadata.xml: New section on anonymous Python functions

Per Paul Eggleton's suggestion, I added a new section on
anonymous Python functions into the "Functions" section.
I also updated the intro text to account for the added
type of functions.

(Bitbake rev: 983d03c1a082e2b83187f0788e61a7941670b242)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake-user-manual-metadata.xml: Edits to flexible inheritance section.

Fixes [YOCTO #5472]

Applied review edits from Paul Eggleton to this section.
Minor edits and some re-writing.

(Bitbake rev: 7259d9d40aad8254751f7674653cd362a9023054)

Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: doc: Rename user-manual -> bitbake-user-manual

This manual gets combined with other manuals and in that context, it helps
a lot if its seen as the Bitbake User Manual. Renames are a pain but
this is worthwhile so that other docs can correctly be combined with this
one. This also clarifies things like google search results which is helpful.

(Bitbake rev: 452a62ae0c2793e281d6769fd3e45500a74898d6)

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

scripts/send-error-report: use a real server as the default

Now that there is a public instance [1] of the error report web interface [2],
we should set it as the default.

[1] http://errors.yoctoproject.org/
[2] http://git.yoctoproject.org/cgit/cgit.cgi/error-report-web/

(From OE-Core rev: 90a955761db4724f0222527a129a3904d6d8aa6e)

Signed-off-by: Stefan Stanacar <stefanx.stanacar@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

toaster.bbclass: do not fail on non-existent files

Toaster may look up inexistent file paths in the build history
for packages that have been referenced but not built.

This triggers a failure, and this patch recovers by deleting
the reference to the non-built packages.

[YOCTO #6063]

(From OE-Core rev: 6bc4a0311cc3cd3f696d9dbca8fc0ef4e3bc340b)

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

u-boot: fix beaglebone boot issue with large kernel images

Fix beaglebone boot issue with large kernel images overwriting Device Tree.
See very detailed comments inside the patch.

The original patch is being reviewed upstream and is targeting mainline U-boot
version 2014.07. This is the adaptation of the patch for 2013.07 version we use

(From OE-Core rev: cd495307d233b81ebeb43198d13bbd4b3ad7407f)

Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake: Update to version 1.23.0 for master

(Bitbake rev: bb4980c63db386ce7d30d9a6b86e9f3861b3bc3a)

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

bitbake: bitbake: Update to version 1.22.0

(Bitbake rev: a8a32de0fb74ff63e5490e3b5e4419146a8f7367)

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

bitbake: cache: don't trigger reparse on recipes with wildcards in SRC_URI

Since we now get wildcards in the file checksum list in the cache, we
need to ignore them when checking to see if they still exist. This
fixes connman-gnome reparsing on every bitbake execution in OE-Core.

(Bitbake rev: 340c250fc664414ab2715a454bedbd19e8efe103)

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

linux-yocto/3.10: intel BSP configuration updates

Updating the SRCREV for the following configuration changes:

  7df9ef8ee47d minnow: Add minnow-drivers-extra fragment
  4f12b62002ca meta: smp.scc: increase default NR_CPUS to 64
  de70ccd19f86 intel-common: change intel-corei7064-preempt-rt-scc filename
  e3780cd93809 meta: mohonpeak: remove branch 'mohonpeak'

(From OE-Core rev: 20625df0f88b1948bfc05be0d2cbdaa201c3b2f0)

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

linux-yocto/3.14: beagleboard black: USB, DRM config tweaks

Importing the meta SRCREV to pick up the following two fixes:

  928d7b2ddad0 beaglebone: enhance USB support and enable MUSB modules
  fbe38387ac61 beaglebone: enable DRM for HDMI output

(From OE-Core rev: c95ae50dc0817fee291ca981c41cb556bf7db582)

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

linux-yocto/3.14: intel configuration fixes

Importing the intel configuration fixes from Darren Hart:

   a82ccc635e7a meta: Purge retired BSPs chiefriver, sys940x, and atom-pc
   b50ba1e37c26 x86: Drop X86_32 configs
   8aa0a946f528 x86: Move MTRR config into x86 common fragments
   ddae217aefee common-pc: Remove SMP from common-pc*-cpu fragments
   f3ad83aa4c39 x86: Consolidate common x86* CPU features

(From OE-Core rev: 399d3d51f35fd8782367e82c7b6e641f32396f80)

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

linux-yocto/3.10: intel-common: Add preempt-rt ktype targets

Updating the meta SRCREV to incorporate the intel common preempt-rt kernel
support.

(From OE-Core rev: 63de385a2dce1727d6d4b42dfdf8cb38d8ae5792)

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

linux-yocto/3.10: v3.10.35 and emenlow boot fixes

Updating the 3.10 SRCREVs to pick up 3.10.35 and a LTSI cherry pick that
fixes a boot issue with the emenlow.

(From OE-Core rev: f0990363aaeb3f8f7a8e02fc2acb0af2171e4eba)

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

linux-yocto/3.4: update to v3.4.85

Updating the SRCREVs to integrate the latest korg -stable update.

(From OE-Core rev: c6fe4b287c204d31a9ab86e3f38336d8828903ee)

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

sstate: Silence warnings when switching machines

We recently added multiple provider warnings. These could get issued
from -cross recipes due to the way these currently overlap in the
native sysroot. Filter out these warnings for now, until such times
as we improve the cross recipes so they don't have an overlapping
namespace.

(From OE-Core rev: d62f8d428831cc62b395f4c764c1e876353ddede)

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

selftest: DiskMonTest: use POSIX output for df

Without "-P" df  splits the output on multiple lines and breaks the test

(From OE-Core rev: 81a81807d34e4b2ea28d8ba9c29e5b48a14bbc1f)

Signed-off-by: Alexandru Palalau <alexandrux.palalau@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gcc: Fix a race over unwind.h

There are two places unwind.h is installed, even by the Makefile's admission.
Disable one of them to prevent build failure races.

(From OE-Core rev: d3b02218dcfedda8e4efb43b8fa6d13af8d91f78)

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

vala: inherit pkgconfig

Although somebody we depend on might cause pkgconfig-native
to be built, we should not rely on that.

(From OE-Core rev: ed4b26cf312360e0eb1b0fbec5b039b939d9823f)

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

perf: Fix sysroot option to CC handling

If you build perf in tree /xxx/treea, then cleansstate perf and build it
in /xxx/treeb having deleted treea, the build will fail, unable to find libc.

The problem is that the --sysroot option passed in through CC is missing.
This works fine if the default sysroot is ok, if it isn't, things will fail.
In 1.7 we'll start poisoning the default sysroot in gcc to catch this kind of
issue however that doesn't fix the problem with perf.

The problem is that various Makefiles set CC = $(CROSS_COMPILE)gcc. The
easist fix for now is to sed out the problematic Makefile lines. Its worth
noting the tools/lib/traceevent Makefile has a much more funky way of setting
CC which works for us and may be the way we need to fix the other Makefiles
upstream.

This fixes build failures we're occasionally seen on the autobuilders.

(From OE-Core rev: d9bd1ac6d91de932dd3e2fcac9da77e0a7c09f55)

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

kernelshark/trace-cmd: fix syntax error of shell

Delete "<<<" syntax of bash in Makefile, else we would get following error:

        Syntax error: redirection unexpected

[YOCTO #6112]

(From OE-Core rev: c2892fa27d9918aa4ce00a106a254ddfd44aa0f1)

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

openssh: fix sshd_config_readonly creation

The readonly sshd config sshd_config_readonly needs to be created from
the installed sshd_config as make install will adjust the paths in
the config file. This fixes the path for sftp-server being correct
in sshd_config but incorrect in sshd_config_readonly.

(From OE-Core rev: 400b4bce34ffb76e500e2195104cc200218aa4c3)

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

image.bbclass: improve sed expressions for ssh_allow_empty_password()

The sed expression was also replacing documentation text containing
PermitRootLogin in the line so "PermitRootLogin yes" was specified
twice.

(From OE-Core rev: 0572196158d9505a624bdee71760978f284728b0)

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

README.hardware: update with Texas Instruments Beaglebone instructions

Replaces outdated Beagleboard instructions with Beaglebone Black (and White).

(From meta-yocto rev: 283b51b0149f7e5838bd1c8465451897baf0bf44)

Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

edgerouter: clarify diskboot instructions

There was some missing details in how to boot from the edgerouter USB
storage. With this update, we have the information required to boot from
either the vfat, or ext3 partition.

[YOCTO #6113]

(From meta-yocto rev: ed2eba333d13cc544648169d06bc47c7e2bbb3f2)

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

bitbake: toaster: Allow toaster to start without pytz

This patch allows toaster to start without pytz.

Django can work with or without pytz, but in the
time zone fix I mistakenly added a hard dependency
on this module.

This patch eliminates the hard dependency.

(Bitbake rev: 40027a6e093c3b7480bfaccbd57e0e613d9a7b71)

Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: add URI search path list

Display the URI paths from SSTATE_DIR and SSTATE_MIRRORS
in the Task page.

[YOCTO #5996]

(Bitbake rev: 6d07b531736c2b304da2dfe661239fd3612b0541)

Signed-off-by: David Reyna <David.Reyna@windriver.com>
Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: sort on size in detail pages

[YOCTO 5778]

Implements the features described in the attachment to bugzilla 5778
   - new global changes to the format of size data, and
   - adding sorts by selected columns to specific detail pages.

Although new pagination and row search capabilities are shown on the
screen shots for the 5778 attachment, those features are specified in
a different bugzilla entry 5777 and are not implemented in this commit.

Also, the 5778 spec includes table sorting for the recipe package
detail page, but sorting for that page was not implemented in this
commit due to complications with sorting then returning to a page that
is only one URL fragment in a template.

The scope of file changes are described below.
Changes to support new 'size' field column formats...
    default.css - added sizecol class style (right justified)
    projecttags.py - changed filtered_filesizeformat to allow
        ".0" suffixes

Changes that add class 'sizecol, span2(as spec'd) ' to <th> and/or
<td> size columns were made to...
    dirinfo.py,
    package_built_dependencies.html,
    package_included_dependencies.html,
    recipe.html,
    bpackage.html, and
    target.html

More significant changes to support detail page table sorting
are:
    - tablesort.html: New created to implement the sort icons,
        directions, and table headings, and
        suppress sort handling if 'disable_sort' in context,
        without search or pagination elements ingrained
        in basetable_top. Confining the changes to this small file
        reduces the impact (testing and risk) on the larger set of
        files that arleady include basetable_top/bottom files.
    - view.py: Modified the following view functions with
        - trivial changes for size formatting to the views: target,
        - changes to package_built_detail, package_included_detail,
            package_included_reverse_dependencies to handle the sorting
            implementation as well as moving headings and size
            formatting for size columns from templates to the views.
    - Implementation of the detail sorting using above in:
            package_built_detail.html,
            package_included_detail.html, and
            package_included_reverse_dependencies.html
        to include the tablesort heading setup, format the size column,
        and iterate over the new sorted objects, suppressing sorts if
        table row count less than 2.

(Bitbake rev: d16126e9abfffde66ab70865a81997322847d44e)

Signed-off-by: Dave Lerner <dave.lerner@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: hide tasks without order or outcome

Hide the incomplete tasks, those without order numbers and/or outcomes.

[YOCTO #6060]

(Bitbake rev: bdbba81715765681a0404fa82f28e471b241051a)

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

bitbake: toaster: regex alternation filter caused django error

The combination of a regex filter specification that uses alternate,
plus a search string, plus multiple search_allowed_fields, leads to
a Django fatal error. Replace this regex filter for variables in
local files with a simpler 'contains' against the project's directory
plus a '/conf/' string.

Alex rebased this on top of fix for #6048.

[YOCTO #5962]

(Bitbake rev: fd57128dc3a35ca87031f3df1a531a085e89baf0)

Signed-off-by: David Reyna <David.Reyna@windriver.com>
Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: fix filtering query for multiple filters

This is a fix for using multiple expressions in filters.
Three different issues are touched:
* added an explicit error message for incorrect filter usage
* changed the value separator to something that will pass
through from the browser to the actual code
* changed the "and" operator for combining Q from a dubios
lambda function to the standard operator.and_

(Bitbake rev: 845b081fc108c656f04d4a70afa4695defc13c9f)

Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: keep number of rows across searches

This patch will make Toaster remember the selected
number of rows to be displayed in tables across different
searches.

The number of rows setting will not be remembered if the
the user leaves the page and subsequently returns to it.

(Bitbake rev: a84f296591be26972b808e98816e9d92cadf3eb5)

Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: show unique set-in files in configvar

When listing the set-in files in the configure varaible table, only show
each file once in that summary view.

[YOCTO #6048]

(Bitbake rev: e69a57fc3302ed8fac21631ba95ea66e93b243e5)

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

bitbake: toaster: disable 'size over total' sort

[YOCTO 6061]

Disable size-over-total table column sorts in the image installed
packages view because that field is a computed field and only model
fields are sortable in the current table toaster implementation.

(Bitbake rev: ba6937c9bffcf81f71ef6fa9f0d29fbbd6e17b2d)

Signed-off-by: Dave Lerner <dave.lerner@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: do not load all available timezones

This patch makes sure that we only load pytz-recognized
timezones. Pytz is used to transform the timezone information
for the database queries, and needs to be able to deal with
the TIME_ZONE value that we set up.

[YOCTO #6093]

(Bitbake rev: bfe67472e3ee778b78ef004b2153fa88b3807b92)

Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake-selftest: add tests for local fetching

Add some explicit tests for unpacking local files to the appropriate
location. Some of these tests are actually testing for broken behaviour;
these have been called out in the comments, and associated bugs have
been filed.

(Bitbake rev: ca921c773c52392a5a338b2f493ad38c8132f708)

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

bitbake: bitbake-selftest: fix help message to include command line

(Bitbake rev: 3016af554d9cc67aec2429b96d8d085b88315e77)

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

libpam: fix 64-bit pam plugins not installed when add to 32-bit image

While the BSP is configured as a 64-bit kernel and 32-bit userspace,
add a 64-bit version of libpam to the filesystem, there was a failure:
...
| Computing transaction...error: Can't install
pam-plugin-unix-1.1.6-r2@lib64_x86_64: no package provides libpam-lib64
|
| Saving cache...
...

While using 'lib64' as the multilib suffix of libpam RPROVIDES , the
RPROVIDES was overridden by map_depends_variable in classextend.py.
...
 $RPROVIDES_lib64-libpam [2 operations]
   set data_smart.py:429 [finalize]
     " libpam-${baselib}"
   set classextend.py:71 [map_depends_variable]4532
     "lib64-libpam"
 computed:
   "lib64-libpam"
RPROVIDES_lib64-libpam="lib64-libpam"
...

Rename the suffix could fix this issue.
[YOCTO #4532]

(From OE-Core rev: 77e3d60fa00a41424fe65977b2bf307727a5a26c)

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

yocto-bsp: Update templates to 3.14 kernel

Add 3.14 kernel support, and remove support for 3.4.

(From meta-yocto rev: 342dce5873a9f1844fd7ce52a1c2656dcfb43053)

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

eglibc: __slow_ieee754_sqrt{, f} functions for ppc e6500-32b

    *fix the following error:
    |/sysdeps/powerpc/powerpc32/e6500/fpu/e_sqrt.c:138:1:
    | error: redefinition of '__ieee754_sqrt'
    | __ieee754_sqrt (double x)

(From OE-Core rev: 83a49c0fc5f6835c1385acb501380a0e351e4ac2)

Signed-off-by: Chunrong Guo <B40290@freescale.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

busybox: fix a sh link wrong

When both bash and busybox be installed, without ash support
in busybox,if bash is installed before busybox in the final stage,
even if ALTERNATIVE_PRIORITY of bash > ALTERNATIVE_PRIORITY of busybox,
the symlink from /bin/sh to bash can be yet overwritten by busybox.

(From OE-Core rev: 6ef650359cc2a49376eb5ca92bc97b34cdd82862)

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

python: Fix CVE-2014-1912

A remote user can send specially crafted data to trigger a buffer overflow
in socket.recvfrom_into() and execute arbitrary code on the target system.
The code will run with the privileges of the target service.

This back-ported patch fixes CVE-2014-1912

(From OE-Core rev: 344049ccfa59ae489c35fe0fb7592f7d34720b51)

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

libx11: fix invalid preprocessing directive errors

Backport the patch to fix invalid preprocessing directive errors.

nls: always use XCOMM instead of # for comments in Compose.pre files

[YOCTO #6116]

(From OE-Core rev: 9d142a7f523f89cd65bef2cd6ce75e4f4500711b)

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

freetype: fix multilib header conflict - ftconfig.h

ftconfig.h conflicts between 32-bit and 64-bit versions.

(From OE-Core rev: 590446f6b191b32efc462c3cb7ac7cce4c897b05)

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

e2fsprogs: fix multilib header conflict - ext2_types.h

ext2_types.h conflicts between 32-bit and 64-bit versions.

(From OE-Core rev: e5bae3426856f9d2fdb604278154b6242011d103)

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

bitbake: bitbake: knotty: Clear footer before outputting to stderr

With the recent change to split the log output to stdout and stderr,
error messages that appeared while the footer was printed got all
messed up. This was because the messages to stderr was output _after_
the footer, then clearFooter() tried to remove the footer but removed
the error message and parts of the footer.

(Bitbake rev: 4fafea4fa69542b491e84463f6eae0d5bf645673)

Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: fetch2: fix traceback when a wildcard matches a directory

If there is a directory matching a wildcard in SRC_URI when getting file
checksums, we should recurse into that instead of producing an error.

(Bitbake rev: ae87b7eb414e3d5eefd2effec7b30c22d2186b02)

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

bitbake: fetch2: handle wildcards correctly when recording file checksums

The Local fetcher's localpath is returning the parent directory for a
wildcard match; we need to handle this and add the wildcard
specification so that we checksum the correct files.

Fixes [YOCTO #6127].

(Bitbake rev: 4a90edd62c16cbf41b5b93280e155077564c774a)

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

openssl: Upgrade to v1.0.1g

The trigger for the upgrade was the serious "heartbleed" vulnerability
(CVE-2014-0160). More information:
http://www.itnews.com.au/News/382068,serious-openssl-bug-renders-websites-wide-open.aspx

Dropped obsolete patches, because the new version contains them:
        0001-Fix-for-TLS-record-tampering-bug-CVE-2013-4353.patch
        0001-Fix-DTLS-retransmission-from-previous-session.patch
        0001-Use-version-in-SSL_METHOD-not-SSL-structure.patch

Modified 2 patches (small changes), in order to apply properly:
        initial-aarch64-bits.patch
        openssl-fix-doc.patch

Addresses CVEs:
    http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-0160
    http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-0076

(From OE-Core rev: ff52836e1838590eeec7d7658e15b21d83cf8455)

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

cmake: Improve method for not building ccmake

In commit 4bee0a93ed985b38c6b4eb605d8e16f5d7c82d51 I introduced an
unnecessary patch to do something that can easily be done without
patching.

The argument to disable building ccmake can be passed to configure provided
it is preceded by "--".

(From OE-Core rev: eaf176eaabb4c558ad76512b30b28ec97fd90bc6)

Signed-off-by: Mike Crowe <mac@mcrowe.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package.bbclass: Add CONFFILES to list of package specific variables

Changes to CONFFILES should change the sstate checksum. To make that happen,
it needs to be listed in the list of package specific variables, therefore
add it.

(From OE-Core rev: 9db71fa03b9d5f5307b2d09e7aa89f46f622aa09)

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

toaster.bbclass: image file is missing a "/"

Relaced the key with the join between path and file name.

[YOCTO #6090]
(From OE-Core rev: 85ffc93becb31772107a5a63b09fd3c16160f3ca)

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

documentation.conf (LICENSE): tiny addition

Examples:

LICENSE = "MIT"
LICENSE = "GPLv2 & LGPLv2.1"
LICENSE = "AFL-2 | GPLv2+"
LICENSE = "MIT & BSD-3-Clause & BSD-2-Clause & PD"
LICENSE = "GPLv2 & (LGPLv2.1 | MPL-1.1 | BSD)"
LICENSE = "GPLv2+ & LGPLv2+ & ( (GPLv2+ & LGPLv2.1+) | (GPLv3+ & LGPLv3+) )"

(From OE-Core rev: f53b6770554d868d287bed0d15194fdb9ed6a779)

Signed-off-by: Matthieu Crapet <Matthieu.Crapet@ingenico.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: hob: add "recipes/images/" to BBFILES when Hob is launched

The path for "recipes/images/" was not added in BBFILES when Hob had
to search for an image recipe. Therefore, it could not find it and an
error occurred.
This path needs to be added when Hob is launched.

[HOB #6086]

(Bitbake rev: 35c67281775b08925957c32663d587d486944e0e)

Signed-off-by: Irina Patru <irina.patru@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: fetch2: avoid cache ignoring missing files

Previously, if a file listed in SRC_URI was missing at the time of
parsing, and then was added and bitbake run again, it would not be
picked up because the recipe was cached without it. If we allow the file
to be added to the list of files to checksum, then it will be checked
for and found on the second run.

Fixes [YOCTO #4790].

(Bitbake rev: 71da822762cb298261cccdfa54b9c0fea02c3c5d)

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

bitbake: fetch2: Fix bug in file checksum generation

For a while its been puzzling me why connman-gnome rebuilds as often as it
does. It turns out you can trigger this with a new checkout of the metadata.

The SRC_URI that is causing the problems is:

SRC_URI = "file://images/*"

and rather oddly the results in checksums for a file "." being added to
the tree, e.g.:

('.', 'ab48a68186f0e0f277c21ef4cb390b4b')

The problem is that when iterating files lists, the checksum variable can
become set yet we don't break the out from the for loop, which leads to
odd (and non-deterministic) entries being added into the file checksum list.
The exact item added probably depends on the order of items on the disk.

Before this change, bitbake-diffsigs on connman-gnome:do_fetch would report:

This task depends on the checksums of files: [
('connman-signal-03.png', 'f6c16aee57b37b73793a2f1dea433ffa'),
('connman-signal-02.png', 'ad0cd22710c097d8174121fc1023c3be'),
('connman-signal-01.png', '8842bd83d2fa9ba56480df34c727c629'),
('null_check_for_ipv4_config.patch', 'a23271e41c9fe81551244d875106af10'),
('connman-signal-05.png', '808589e7e8d502b44c7b007e9e68d48c'),
('connman-signal-04.png', 'ab48a68186f0e0f277c21ef4cb390b4b'),
('null_check_for_ipv4_config.patch', 'a23271e41c9fe81551244d875106af10'),
('0001-Removed-icon-from-connman-gnome-about-applet.patch', 'e2d8269357c1e8c84291606da24eea85'),
('0001-Removed-icon-from-connman-gnome-about-applet.patch', 'e2d8269357c1e8c84291606da24eea85'),
('.', 'ab48a68186f0e0f277c21ef4cb390b4b')]

Afterwards:

This task depends on the checksums of files: [
('connman-signal-03.png', 'f6c16aee57b37b73793a2f1dea433ffa'),
('connman-signal-02.png', 'ad0cd22710c097d8174121fc1023c3be'),
('connman-signal-01.png', '8842bd83d2fa9ba56480df34c727c629'),
('null_check_for_ipv4_config.patch', 'a23271e41c9fe81551244d875106af10'),
('connman-signal-05.png', '808589e7e8d502b44c7b007e9e68d48c'),
('connman-signal-04.png', 'ab48a68186f0e0f277c21ef4cb390b4b'),
('null_check_for_ipv4_config.patch', 'a23271e41c9fe81551244d875106af10'),
('0001-Removed-icon-from-connman-gnome-about-applet.patch', 'e2d8269357c1e8c84291606da24eea85'),
('0001-Removed-icon-from-connman-gnome-about-applet.patch', 'e2d8269357c1e8c84291606da24eea85')]

which is correct and deterministic without the "." entry.

(Bitbake rev: f9416e76e272ba3249abff099f6f3a47fe82e03e)

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

test-dependencies.sh: Redirect stderr

* newer bitbake is printing some messages to stderr, we want to log
  them as well

(From OE-Core rev: f442c15aaeb8c0641093e92f2b832dfaa2d9a486)

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

systemtap: Move to current HEAD

This moves the systemtap recipe to the current HEAD commit, in order
to get users past some bugs affecting ppc and arm.

Fixes [YOCTO #6016]

(From OE-Core rev: efac179f243459c53d0431fbcce92a3561b40956)

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

sysvinit: fix the bootlogd init script header

fix the bootlogd init script header, to make chkconfig be able to work
on bootlogd

(From OE-Core rev: d1a7e5dbaa1217b692b46e2756c318e5aaa34f9d)

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

busybox: Specify '-nostdlib' when linking to a .o file.

Ensure that linking to the built-in.a file uses the 'nostlib'
option.  Submitted to busybox mailing list here:

    http://lists.busybox.net/pipermail/busybox/2014-March/080730.html

This has been accepted into upstream busybox so it will not be needed
for future versions.

(From OE-Core rev: fdfff57304e6641ef5a8db0fb456bfd46bf1e2dd)

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

busybox: Use CCLD

Use the CCLD variable to ensure proper tuning parameters.
Notably when building on a x86-64 host with an i686 toolchain
there is an error building built-in.o because it is trying to
link 32-bit and 64-bit object files:

     i686-pc-linux-gnu-ld --sysroot=/work/dmoseley/Mentor/amd-2014.05/build.genericx86-64-external/tmp/sysroots/genericx86-64     -r -o applets/built-in.o applets/ap
     i686-pc-linux-gnu-ld: Relocatable linking with relocations from format elf64-x86-64 (applets/applets.o) to format elf32-i386 (applets/built-in.o) is not support
     make[1]: *** [applets/built-in.o] Error 1

(From OE-Core rev: 088d2f0185da6648305b4719ee9cb0d2f2b37c6f)

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

useradd_base.bbclass: avoid the traffic crowding

Currently, it would sleep 1 second when fail to add the user, this maybe
not enough when we use the sstate cache, as my test shows below, nearly
all the useradd actions are doing in the same minute when mirror from
ssate cache, and it would fail when the load is high, I got these time
by adding strace before the useradd for debugging:

2014-03-31 14:48:22.978079781 +0800 /tmp/log/pulseaudio.4.c
2014-03-31 14:48:22.028079813 +0800 /tmp/log/pulseaudio.1.c
2014-03-31 14:48:21.949079816 +0800 /tmp/log/pulseaudio.3.c
2014-03-31 14:48:20.903079852 +0800 /tmp/log/pulseaudio.2.c
2014-03-31 14:48:20.006079883 +0800 /tmp/log/nfs-utils.9.c
2014-03-31 14:48:18.876079923 +0800 /tmp/log/xuser-account.9.c
2014-03-31 14:48:18.824079924 +0800 /tmp/log/pulseaudio.0.c
2014-03-31 14:48:17.826079959 +0800 /tmp/log/xuser-account.8.c
2014-03-31 14:48:17.766079961 +0800 /tmp/log/nfs-utils.8.c
2014-03-31 14:48:16.794079995 +0800 /tmp/log/xuser-account.7.c
2014-03-31 14:48:16.735079997 +0800 /tmp/log/nfs-utils.7.c
2014-03-31 14:48:14.719080066 +0800 /tmp/log/xuser-account.5.c
2014-03-31 14:48:14.677080068 +0800 /tmp/log/nfs-utils.5.c
2014-03-31 14:48:12.621080139 +0800 /tmp/log/nfs-utils.3.c
2014-03-31 14:48:11.589080175 +0800 /tmp/log/nfs-utils.2.c
2014-03-31 14:48:10.242080221 +0800 /tmp/log/builder.0.c
2014-03-31 14:48:09.523080246 +0800 /tmp/log/nfs-utils.0.c
2014-03-31 14:48:09.488080248 +0800 /tmp/log/openssh.0.c
2014-03-31 14:48:09.485080248 +0800 /tmp/log/rpcbind.1.c
2014-03-31 14:48:07.590080313 +0800 /tmp/log/rpcbind.0.c
2014-03-31 14:28:15.437121590 +0800 /tmp/log/avahi.0.c
2014-03-31 14:18:19.067142238 +0800 /tmp/log/dbus.0.c

The nfs-utils and xuser-account are failed to add the user.

The useradd command needs two locks, passwd.lock and group.lock, it may
get one, but can't get another one if we look into these .c files, sleep
1 second is not enough, it needs more seconds, the reason is that, if
succeed, it doesn't have any side effects, if failed, we need wait for
more seconds rather than make it more crowding.

I've tried to use "sleep 5", but it didn't make much better since they
would sleep and wake up nearly at the same time, I also tried to use
"sleep <RANDOM seconds between 1 and 10>", that didn't make much better
,either.

I think that a better ways is sleep more and more seconds (up to 10
seconds) when failed, this can't fix the problem that they may do the
actions at the same time, but the logic is: if it is not crowding, sleep
less time should be OK, otherwise sleep more and more time.

Here is the testing result which seems much better:
2014-04-03 14:09:56.605185284 +0800 dbus.0.c
2014-04-03 14:09:39.899185862 +0800 rpcbind.5.c
2014-04-03 14:09:38.400185914 +0800 distcc.4.c
2014-04-03 14:09:35.206186025 +0800 pulseaudio.1.c
2014-04-03 14:09:33.979186067 +0800 rpcbind.4.c
2014-04-03 14:09:33.364186089 +0800 pulseaudio.0.c
2014-04-03 14:09:33.360186089 +0800 distcc.3.c
2014-04-03 14:09:30.996186171 +0800 avahi-ui.0.c
2014-04-03 14:09:30.298186195 +0800 distcc.2.c
2014-04-03 14:09:29.905186208 +0800 rpcbind.3.c
2014-04-03 14:09:29.410186226 +0800 avahi-ui.2.c
2014-04-03 14:09:28.239186266 +0800 distcc.1.c
2014-04-03 14:09:27.298186299 +0800 xuser-account.0.c
2014-04-03 14:09:27.032186308 +0800 distcc.0.c
2014-04-03 14:09:26.836186315 +0800 rpcbind.2.c
2014-04-03 14:09:25.846186349 +0800 nfs-utils.1.c
2014-04-03 14:09:25.752186352 +0800 avahi-ui.1.c
2014-04-03 14:09:24.779186386 +0800 builder.0.c
2014-04-03 14:09:24.746186387 +0800 rpcbind.1.c
2014-04-03 14:09:23.916186416 +0800 openssh.1.c
2014-04-03 14:09:23.848186418 +0800 nfs-utils.0.c
2014-04-03 14:09:23.594186427 +0800 rpcbind.0.c
2014-04-03 14:09:22.609186461 +0800 ppp-dialin.0.c
2014-04-03 14:09:21.817186488 +0800 openssh.0.c

[YOCTO #6085]

(From OE-Core rev: 9cedc786ba132935748bdca8bc33b56c366b531e)

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

bitbake: toaster: added covered task list

if a task has a 'covered' indication, the list of tasks that
covered the task are computed and displayed. amended to add tooltip.

[YOCTO #5925]

(Bitbake rev: bb05ee13f53f10988579b6238802327732041d0c)

Signed-off-by: Farrell Wymore <farrell.wymore@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: correct package count

The package count was incorrect because it was counting anonymous
packages. the full path of the image files was shortened to just
the filename.

[YOCTO 6087]
[YOCTO 6091]

(Bitbake rev: 06b190b2c23799bd2c9749be28e11bf5d59ed4fc)

Signed-off-by: Farrell Wymore <farrell.wymore@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Fixes to the build dashboard

* Add "on" before the date in the build status

* Get rid of the extra space before the "and" in the
build status

* Make the errors and warnings links in the build
status show the message content

* Make sure that the "Images" section in the left navigation
appears only for successful builds

* Link the number of packages installed to the "Packages
included" tab in the image information page

* Remove unnecessary inline styles in the dashboard
template

* Make sure target names in the h1 are listed in
alphabetical order

(Bitbake rev: 644a38e9a90728af52ebda9846d2037dd6831b41)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Make tables happy in Chrome

There is a known layout issue with Bootstrap tables in
Chrome. This change applies a css fix for it, and changes
the span classes in the variables table to improve
formatting.

(Bitbake rev: a7fee015b1cc34351d5ad8b854fc6bf368416400)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Fix the empty packages alert

* Remove a tag in all capitals
* Remove an unnecessary div
* Fix a typo

(Bitbake rev: 1ca139ca157d62cee9c48b22063e77614fa34325)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Fix "Recent builds" links

* Link the number of errors to the #errors section in the
build dashboard

* Link the number of warnigns to the #warnings section in
the build dashboard

* Link the build time to the build time page

* Move some javascript that applies only to builddashboard.html
to that template from main.js

(Bitbake rev: 94c61d0d9ff24e0c781ea0a8e4bfea4c4ac94d4e)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: catch IndexError error in dictionary lookups

In the get_dict_value filter, there is the edge case where halted
builds can lead to IndexError errors in dictionary lookups, so we
need to catch those.

[YOCTO #6067]

(Bitbake rev: f6fcce974ce3b145bc472cd4e9721d56191828a4)

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

bitbake: toaster: Changes to help text

All help text strings have been reviewed by the
technical writer, and some of them by Paul Eggleton.
This patch implements their suggested changes.

(Bitbake rev: ce89530b178be2f3202d45523ef1340e00df05be)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Remove html validation errors in tasks.html

Remove the duplicated class attribute and the name
attribute.

Also, make sure that the row blue highlight works
without the name attribute when you land on the
tasks table after clicking the order link in a
task details page.

The commit also fixes a typo in default.css.

[YOCTO #6033]

(Bitbake rev: 471234f1ab1dbcd736a892720e99a305363db5ff)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toasterui: save warnings from non-build context

Some warnings and errors may come from a non-build context,
and they were not saved, even if they were counted for the current
build.

This patch saves these messages in memory until we have the
entire build context available.

[YOCTO #5642]

(Bitbake rev: fc7a74e7961775b5d7ff25298abed10138d24dc9)

Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toasterui: avoid variable name conflict

We rename a local variable in as to prevent a conflict with
a similary named function parameter.

(Bitbake rev: f4e57f794651c4894600445e843ca9d5e104cd84)

Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: Update API

Make sure the REST API includes the latest changes
to the database schema.

(Bitbake rev: fb3d1d189f010488a8726872b01313857697751b)

Signed-off-by: Belen Barros Pena <belen.barros.pena@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: toaster: link task order to right tasks page

When linking from a task order number to the All Tasks
page, automatically display the correct page for that link
anchor.

[YOCTO #5933]

(Bitbake rev: e5b6681432b627eb73bfa766d3154162206f374e)

Signed-off-by: David Reyna <David.Reyna@windriver.com>
Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>