grub-efi: add oe's kernel name to the conf file

Our kernel's name is bzImage, we need add it to grub.d/10_linux.in so
that the grub-mkconfig and grub-install can work correctly on the
target.

(From OE-Core rev: 1dd9a16030bc2dc673c7b121dea6e78f2681f55e)

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

grub git: add oe's kernel name to the conf file

Our kernel's name is bzImage, we need add it to grub.d/10_linux.in so
that the grub-mkconfig and grub-install can work correctly on the
target.

(From OE-Core rev: 6677bb09ae64e878b12552caef999b60bc99d0e4)

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

grub 2.00: add oe's kernel name to the conf file

Our kernel's name is bzImage, we need add it to grub.d/10_linux.in so
that the grub-mkconfig and grub-install can work correctly on the
target.

(From OE-Core rev: c972e122066aa80550155feea619908f6d3c3176)

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

do_rootfs: Add PACKAGE_FEED_URIS as a vardep

A bit uncertain on where to put this, suggestions welcome.
Needed to automatically rebuild do_rootfs when PACKAGE_FEED_URIS
change.

(From OE-Core rev: c779bf78f2e1f66f3c8a6b02054e39bee4ea88a5)

Signed-off-by: David Nyström <david.c.nystrom@gmail.com>
Signed-off-by: David Nyström <david.nystrom@enea.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

do_rootfs: Added PACKAGE_FEED_URIS functionality

Adding a common interface to add predefined package manager
channels to prebuilt rootfs:es.

Adding PACKAGE_FEED_URIS = "http://myre.po/repo/, will
assume repo directories named (rpm,ipk,deb) as subdirectories
and statically add them to the rootfs, using the same PKG_ARCHs
as the build which produced the images.

Tested with RPM, IPK and DEB.

deb feed functionality seem broken, is anyone using this ?

(From OE-Core rev: 9b8811045546ad67b4695d980f09636d5506e50c)

Signed-off-by: David Nyström <david.c.nystrom@gmail.com>
Signed-off-by: David Nyström <david.nystrom@enea.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package.py: use subprocess.Popen for rpmdeps call

* I've noticed errors like this in log.do_package:

  DEBUG: Executing python function package_do_filedeps
  sh: 1: Syntax error: "(" unexpected
  sh: 1: Syntax error: "(" unexpected
  DEBUG: Python function package_do_filedeps finished

  which are actually caused by some filenames included in package
  containing '()' characters

  Maybe we should change meta/classes/package.bbclass to
  fail when some filedeprunner call fails like this and fix
  filedeprunner to escape '()' and other possibly dangerous chars
  it's called like this:
  processed = list(pool.imap(oe.package.filedeprunner, pkglist))

* don't use shell=True
* show the command when it fails and let do_package task to fail

(From OE-Core rev: 148c04c1bf39ca0d21288fdce61c51dc8e1c3226)

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>

insane: Special case kernel modules for x32 targets

The Kernel module packages for x32 target have 64 bit binaries, which
breaks the QA_check expecting all the packages to be 32bit.

Make a special case for kernel module packages for x32 targets, to avoid
this false error.

Fixes Bug:
[YOCTO #5903]

(From OE-Core rev: a4261ab92735abfba835e916931d2e920f335aaa)

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

package_manager.py: make list_installed() list pkg dependencies too

list_installed("deps") will now return the package dependencies.

(From OE-Core rev: b16cd9eaa8190dc678af9d98a9bc4fb690809c5f)

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

dhcp-client: fix invoke dhclient-script failed on Read-only file system

In read-only file system, '/etc' is on the readonly partition,
and '/etc/resolv.conf' is symlinked to a separate writable
partition.

In this situation, we should use shell variable to instead of
the temp file '/etc/resolv.conf.dhclient'.

[YOCTO #5624]
(From OE-Core rev: df793f4356b411cbb92445c4559c9b21eb6d99fc)

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

grub-efi: use autotools-brokensep

(From OE-Core rev: ee4a68dd11d6c18681a6d3fdbcd7428136eaaf62)

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

ncurses: use lnr instead of python code

Drop the use of custom Python code and instead use the new lnr script to
generate relative symlinks.

(From OE-Core rev: efe328d3713f60257358cab5d7c6a1d38d1a8d88)

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

scripts: add lnr (link relative)

lnr is a simple script to generate relative symlinks from absolute paths,
similar to "ln -r" but without requiring coreutils 8.16 (Ubuntu 12.04 and others
currently ship 8.13).

(From OE-Core rev: 6ae3b85eaffd1b0b6914422e8de7c1230723157d)

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

insane: check packages for absolute symlinks to the tmpdir

Add a sanity test that checks for symlinks in packages that point into the
TMPDIR on the host.

(From OE-Core rev: 9e28808a6d6f47dc10ad87b878c7e912c2bbe16f)

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

oe-selftest: force some values in local.conf for _sstatetests_noauto.py

1) Some values should have default values in order for tests like sstate relocation to be relevant
   - SSTATE_DIR needs to have default value
   - SSTATE_MIRRORS need to be unset (we need the sstate files to be created)
   - TMPDIR needs to be in default location

2) Added conf/auto.conf to list of ported files to secondary build directories

(From OE-Core rev: 3b98de79e14c44e70fd72ed94493a674444bdba2)

Signed-off-by: Corneliu Stoicescu <corneliux.stoicescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

oe-selftest: optimize code in _sstatetests_noauto.py module

1) Some of the checks made when creating new directories are unnecessary beacause
the test will fail anyway if os.mkdir() fails, with the appropriate error message.

Removing this code.

2) Moved the adding to tracked paths for deletion of temporary build directories
and sstate-cache directories to after they are created. This makes more sense
and prevents deletion of these directories if they allready exist and may contain
useful data.

(From OE-Core rev: d002bcf7f80fd9652d3f310957c99bd1e6c8b3a1)

Signed-off-by: Corneliu Stoicescu <corneliux.stoicescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

consolekit: fix console-kit-log-system-start.service startup

console-kit-log-system-start.service fails to to start if the
/var/log/ConsoleKit directory does not exist. Normally it is created
automatically but as we mount a tmpfs at /var/log, we need to add
a tmpfiles.d entry to create it.

(From OE-Core rev: 2a9a14bf400fe0c263c58aa85b02aba7311b1328)

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

nfs-utils: run rpc.statd as rpcuser:rpcuser instead of rpcuser:root

For security policy, change the group of running rpc.stdtd to rpcuser,
just like Radhat does.

(From OE-Core rev: 7f922a7b65690fcc110413b83953d466d46e977b)

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

rpcbind: add rpc user and run rpcbind daemon with rpc user

For security policy, change to run rpcbind daemon with rpc user
just like Redhat does, so set the --with-rpcuser to rpc and add rpc user.

(From OE-Core rev: 52f7a1b403eabc52ca104f752484ec1572f2d9dd)

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

perf: don't use oe.path.relative

Instead of using oe.path.relative, use the Python Standard Library function
os.path.relpath.

(From OE-Core rev: 01f640f2e878ef86db4138f422fdf74f5f41c8c5)

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

gcc-cross: don't use oe.path.relative

Instead of using oe.path.relative, use the Python Standard Library function
os.path.relpath.

(From OE-Core rev: 90c3a0401c566e26d89a5c0410b2a51fe27b95b2)

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

lib/oe: drop custom implementation of oe.path.relative

As we now require Python 2.7 and os.path.relpath() was added in 2.6 we can now
drop the reimplementation in oe.path.

oe.path.relative is simple now a wrapper that changes the order of the arguments
and it's use discouraged.

(From OE-Core rev: 1a03cd16401d2926bba902ffc5df30911b5c9394)

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

pango: Fix postinst by adding missing qemu-native dependency

* postinst_prologue calls qemu_run_binary but there is no
  dependency on qemu-native which causes postinst to sometimes
  fail which is fatal for building read-only rootfs

(From OE-Core rev: fc14983ae62dc4eb9f08e8f172ac51faaa6bcae2)

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>

libgcc: make sure symlinks are created in a valid directory

When adding extra symlinks, we have to make sure that the directory
that the links are created in is valid. Added a check for this.

This is an incremental addition to commit
97f2a81d6796ddaf7bbaab86c2ab9039673c732c

(From OE-Core rev: ea297e807f4f84c281e0f8c6c4470ee3f3fa949b)

Signed-off-by: Alexandru-Cezar Sardan <alexandru.sardan@freescale.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

python-numpy: fix build for qemumips and qemuppc

When building python-numpy for qemumips or qemuppc, the following error
appeared.

cp: cannot stat `xxx/python-numpy/1.7.0-r1/*config.h': No such file or directory

This is because for qemumips or qemuppc, there are no such files in SRC_URI.

This patch fixes this compiling error by adding necessary files to the SRC_URI.

(From OE-Core rev: 15582a25f964e53d28ca0b5c94df3a803c366fed)

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

gstreamer: Fix valgrind determism problem in the same way as gstreamer 0.10

(From OE-Core rev: 81006fe51340bba284131fa9ebc45da129b9a4ae)

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

Upgrade to systemd-stable v208

Additional changes in ptest code since now
we have directories and not only bunch of files
under test/ dir so a simple install does not
work anymore we have to cp the files

(From OE-Core rev: e201f291b269c70d732778b34de01529aca387b5)

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

classes/buildhistory: fix task signatures changing

Fix task signatures for recipes that add to SSTATEPOSTINSTFUNCS changing
when adding and removing INHERIT += "buildhistory" (really this time!)

This relies on the BitBake vardepvalueexclude feature, however it will
not fail without it - signatures will be changed in that case though.

Part of the fix for [YOCTO #5897].

(From OE-Core rev: 27c8a9a282358b9a8a330252ee2104b250777b38)

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

bitbake.conf: add new vardepvalueexclude varflag to BB_SIGNATURE_EXCLUDE_FLAGS

We don't want the value of this varflag itself entering any signatures,
ever.

Part of the fix for [YOCTO #5897].

(From OE-Core rev: 1497d3d4b10844aa19ce6dcceed25aa36454160f)

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

opkg-utils: Update to latest git master

The latest commit in opkg-utils allows packages created by opkg-build to be read
by dpkg-deb again.

(From OE-Core rev: 219944af2700ce9dbc425fac384cd32b0a802123)

Signed-off-by: Paul Barker <paul@paulbarker.me.uk>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gcc: Add upstream fix for gcc bug 58595

Fix for internal compiler error hit when building lttng-tools_4.2.0:

 kernel-consumer.c:324:1: internal compiler error: in gen_movsi, at
   config/arm/arm.md:5539

(From OE-Core rev: ec1d5bdf4cc0a7a3e4747b42b7b95805752bea07)

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

lttng-ust: Disable doc/examples in the build

Don't build the doc examples - we don't need them and in fact they
never successfully built in previous iterations of the lttng-ust
recipe anyway.

(From OE-Core rev: d798bd1d25ae0e57b65758c54f033afda96c14e1)

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

lttng-modules: Fix 3.14 bio tracepoints

The mainline 3.14 commit 'block: Astract out bvec iterator' broke the
lttng-modules tracepoints.  Fix them here.

(From OE-Core rev: c11b29ff4f24af0445c3c6a694b8dc2037dcd7e4)

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

lttng-modules: Exclude arm

lttng-modules and gcc-4.8 don't mix, according to the lttng ML
'current_thread_info() not respecting program order with gcc 4.8.x',
so remove it from arm builds.

(From OE-Core rev: ccf687de7b856dbe6f347956743f07ff05c2533a)

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

lttng-ust: Add version 2.4.0

This updates lttng-ust to 2.4.0, codenamed Époque Opaque, needed
for interoperability with lttng-modules and the 3.14 (dev) kernel.

(From OE-Core rev: b4ae73a86ccf62a508faa76dc349bbab02540c37)

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

lttng-tools: Add version 2.4.0

This updates lttng-tools to 2.4.0, codenamed Époque Opaque, needed
for interoperability with lttng-modules and the 3.14 (dev) kernel.

(From OE-Core rev: da5aaafa8c47d9daadb34adbddc93e0f18ec6259)

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

lttng-modules: Add version 2.4.0

This updates lttng-modules to 2.4.0, codenamed Époque Opaque, needed
for interoperability with the 3.14 (dev) kernel.

(From OE-Core rev: bc9a310f920ae17330c749f2263f884423e9b050)

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

js: remove

This is a very old version of Mozilla's Javascript engine, isn't
actually used by anything in OE-Core now that web is gone, and in public
layers outside of OE-Core is apparently only used by mediatomb within
meta-baryon.

(From OE-Core rev: 8b94925adf4428c4e155ef9fd79d33703968d9da)

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

zaurusd: remove

Despite the fact that it's currently specific to the Sharp Zaurus, we
kept this in OE-Core up to now as a potential basis for better handling
of device-specific customisations for a wider range of devices. Whilst
moving device-specific customisations to a more central point is still
a laudable goal, these days such functionality probably wouldn't be
implemented on top of zaurusd or even in the same way as zaurusd, and
thus it makes sense to leave its recipe to be maintained in the
meta-handheld layer (where a current recipe already exists.)

(From OE-Core rev: 3255b9de49c8b4d74cce3a2a0e750d3d3e698b59)

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

e2fsprogs: populate-extfs.sh: remove it

Remove it since we don't need it anymore, and there is one inside the
e2fsprogs.

[YOCTO #4083]

(From OE-Core rev: f3a95ca6886b55e5819b068bdbd2cceb882d91a6)

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

image_types.bbclass: use mke2fs -d to create the ext2/3/4 image

We used populate-extfs.sh which invoked the debugfs to create the image,
now the mke2fs' option "-d root-directory" can do the same thing, and
which is more faster, for example, the core-imag-sato:

* In the past:
$ time mke2fs -t ext4 rootfs.ext4
real    0m0.249s
user    0m0.036s
sys     0m0.132s
$ time populate-extfs.sh /path/to/rootfs/ rootfs.ext4
real    0m29.355s
user    0m10.637s
sys     0m5.544s

* Now:
$ time mkfs.ext4 -F -d /path/to/rootfs/ rootfs.ext4
real    0m6.338s
user    0m3.824s
sys     0m1.356s

More than 25 seconds are saved.

[YOCTO #4083]

(From OE-Core rev: 6338fe9ba338615317a4e7ccc99287b236ac77ae)

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

e2fsprogs: enable the "-d" related patches

These patches should be enabled at the same time, so enable them in a
single commit, and thus we can disable it easily.

[YOCTO #4083]

(From OE-Core rev: 671bf47101b8f7bd2e6e8b7920615f9e7837ea81)

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

e2fsprogs: mke2fs: update the manual for the -d option

Update the manual for the -d option

[YOCTO #4083]

(From OE-Core rev: 506774e828dff576ac91e8d328d8e84e58455147)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: debugfs: use the functions in misc/create_inode.c

* Use the functions in misc/create_inode.c, and remove the duplicated
  code.

* The CREATE_INODE_DEPS in the debugfs/Makefile.in is used for recording
  create_inode.o's depends which is from misc/Makefile.in, we have to
  recompile create_inode.o because we need it to print more messages when
  it is used by debugfs, just like we recompile e2freefrag.o, but it seems
  that the e2freefrag.o's depends in debugfs/Makefile.in is incorrect, it
  would not rebuild when its depends (e.g.: lib/config.h) is changed,
  which would cause unexpected errors. Make duplicated code in
  debugfs/Makefile.in and misc/Makefile.in is not a good idea, maybe we'd
  better define CREATE_INODE_DEPS in the top Makefile, I'd like to send
  another patch and fix the e2freefrag if you are fine with it.

[YOCTO #4083]

(From OE-Core rev: 460f577379ef2ecf42fb8b7f855c21400c276d4f)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: mke2fs: handle hardlinks

Create the inode and save the native inode number when we meet the hard
link (st_nlink > 1) at the first time, use ext2fs_link() to link the
name to the target inode number when we meet the same native inode
number again.

This algorithm is referred from the genext2fs.

[YOCTO #4083]

(From OE-Core rev: 174ec622bdc00fc26392704fdadeac2c067a33f6)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: mke2fs: add an option: -d root-directory

This option is used for adding the files from a given directory (the
root-directory) to the filesystem, it is similiar to genext2fs, but
genext2fs doesn't fully support ext4.

[YOCTO #4083]

(From OE-Core rev: 83357ef1f6797e48de4870ccd552951b0efb3b53)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: mke2fs: set owner/mode/time for the inode

Set the uid, gid, mode and time for inode.

[YOCTO #4083]

(From OE-Core rev: 4039dfc38241418d10e5255725ffa9dc5c948815)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: mke2fs: create directory

The do_mkdir_internal() is used for making dir on the target fs, most of
the code are from debugfs/debugfs.c, the debugfs/debugfs.c will be
modified to use this function.

[YOCTO #4083]

(From OE-Core rev: 30b6ad5067d7a1514c02994c31baf6ec4e5fbcb7)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: mke2fs: copy regular file

The do_write_internal() is used for copying file from native fs to
target, most of the code are from debugfs/debugfs.c, the
debugfs/debugfs.c will be modified to use this function.

[YOCTO #4083]

(From OE-Core rev: e9a82af1581bcf357dd771765b20c8a16a4ffeb1)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: mke2fs: create symlink

The do_symlink_internal() is used for creating symlinks, most of the
code are from debugfs/debugfs.c, the debugfs/debugfs.c will be modified
to use this function.

[YOCTO #4083]

(From OE-Core rev: 367948c87c682953e98015656814b89b64004c26)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: mke2fs: create special file

The do_mknod_internal() is used for creating special file which is
block, character and fifo, most of the code are from debugfs/debugfs.c,
the debugfs/debugfs.c will be modified to use this function.

[YOCTO #4083]

(From OE-Core rev: dcb86f12691b346262c0aadbe5b2f3671b4ecd22)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: mke2fs: copy files recursively

Use opendir() and readdir() to read the native directory, then use
lstat() to identify the file type and call the corresponding function to
add the file to the filesystem, call the populate_fs() recursively if it
is a directory.

NOTE: the libext2fs can't create the socket file.

[YOCTO #4083]

(From OE-Core rev: fd0d00dafcf497f23b1c3886d062c8a57e2aa515)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

e2fsprogs: mke2fs: add the ability to copy files from a given directory

We will add a -d option which will be used for adding the files from a
given directory to the filesystem, it is similiar to genext2fs, but
genext2fs doesn't fully support ext4.

* We already have the basic operations in debugfs:
  - Copy regular file
  - Create directory
  - Create symlink
  - Create special file

  We will move these operations into create_inode.h and create_inode.c,
  then let both mke2fs and debugfs use them.

* What we need to do are:
  - Copy the given directory recursively, this will be done by the
    populate_fs()
  - Set the owner, mode and other informations
  - Handle the hard links

TODO:
  - The libext2fs can't create the socket file (S_IFSOCK), do we have a
    plan to support it ?

[YOCTO #4083]

(From OE-Core rev: f050c510b070d919d50e491476e83f2b0ae2b7b2)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

kernel-yocto: fix diffconfig/kernel_configme breakage

The bbclass did the following:

    do_diffconfig[depends] += "virtual/kernel:do_kernel_configme"

This clearly introduces a cross-kernel task dependency if the recipe
inheriting this class isn't the preferred provider of virtual/kernel, which is
obviously wrong, but further, will break the build if a kernel-yocto based
kernel is parsed and not skipped, but virtual/kernel refers to
a non-kernel-yocto recipe, which would not have the do_kernel_configme task.

Work around this by adding the in-recipe task dep programmatically with
bb.build.addtask when do_diffconfig exists.

(From OE-Core rev: 0e6881146d87f0d214d80bc92e54c113906db63a)

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

kernel-yocto: Disable ANSI escape codes from git branch

'git branch' may use ANSI escape codes in its output (to provide colour)
which doesn't play well with commands expecting pure plain text, e.g.

    fatal: '^[[31mmaster^[[m' is not a valid branch name.

Use the --no-color option to ensure all branch names are plain text.

Cc: Bruce Ashfield <bruce.ashfield@gmail.com>
(From OE-Core rev: 87acfdb28380c26344a79a9dc0b4b403173bdc44)

Signed-off-by: Matt Fleming <matt.fleming@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

image.py, rootfs.py, package_manager.py: redirect stderr to stdout when calling check_output()

If a command executed with subprocess.check_output() fails, the
subprocess.CalledProcessError.output contains only STDOUT and the user
needs to check the log.do_rootfs to see any other details.

This commit forwards stderr to stdout so that, in case of failure, the
entire error output will be displayed in terminal.

[YOCTO #5902]

(From OE-Core rev: 4661f1441429952f30e19cebd93dc42ce31fd868)

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

wayland: Add target sysroot scanner m4 macro

With the new aclocals process, accesses to the macros is a lot stricter and
the native macros are not used. We need to ensure the modified macros are used
in target builds to get the correct scanner functionality. Inserting the
native macro into the target is the correct thing to do in this case.

This resolves build failures in libva.

(From OE-Core rev: ad9bd9ce164c9cb1212de27fec84ad0fe5214bb2)

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

layer.conf: update LAYERVERSION_core for core-image-full-cmdline

LAYERVERSION_core change to 4 means rename of core-image-basic to
core-image-full-cmdline.

(From OE-Core rev: d87585385bfe00dd5e7448ae7d5bc5dc7f9ac782)

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

Rename PACKAGE_GROUP variable to FEATURE_PACKAGES

Since tasks were renamed to packagegroups some time ago, this variable
name implies that its usage is necessarily related to them which is not
the case. Rename the variable to more closely represent what it does
(whilst still providing backwards-compatibility with a warning for
PACKAGE_GROUP).

(From OE-Core rev: d703e07ea5056624101e1bc48f10d25b602635e9)

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

classes/image: support package-management in IMAGE_FEATURES

Previously the "package-management" feature was only fully supported
when inheriting core-image.bbclass, which is not really ideal given that
this is the standard way of adding runtime packaging to an image in
OE-Core.

Part of the fix for [YOCTO #5424].

(From OE-Core rev: 18b8e526570b347d269fa47140949e4ea4800ac5)

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

core-image-basic / packagegroup-core-basic: rename to *-full-cmdline

This more accurately represents what this image and packagegroup are
intended to provide (a more complete command-line environment similar
to what you would find on a traditional Linux system), and avoids
confusion with the similarly named core-image-base and
packagegroup-base.

Fixes [YOCTO #2424].

(From OE-Core rev: 959396d5319df6d6599cc74077e85182768b92af)

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

PR bumps to remove PRINC use in meta-fsl-arm

Those recipes need PR bump so we drop PRINC from meta-fsl-arm layer.

(From OE-Core rev: b60c68a4ea937a647e13d69dfecefb7dcbda95a0)

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

bitbake: runqueue: Fix typo

slef.self is clearly meant to be self, fix typo.

Otavio spotted and reported, thanks.

(Bitbake rev: 316daad7928a58cdfc42e27b20e739f4dd74a02a)

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

bitbake: data: add vardepvalueexclude varflag

On rare occasions it's useful to be able to exclude a part of a
variable's value from the variable's signature; for example if you want
to add an item to a list sometimes and not have the signature of the
variable change depending on whether the item is in the list or not. The
initial intended use case for this in OpenEmbedded is to allow adding a
function to SSTATEPOSTINSTFUNCS in buildhistory.bbclass and not have
that change any task signatures (so adding and removing
INHERIT += "buildhistory" won't lead to any rebuilds).

Part of the fix for [YOCTO #5897].

(Bitbake rev: f803bf8cfefafcbe212442e66b301ccd9c5aa2a5)

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

recipes: bump PRs

* resolves following warnings:
  WARNING: Use of PRINC 17 was detected in the recipe meta-openembedded/meta-systemd/recipes-core/systemd/systemd-machine-units_1.0.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe meta-openembedded/meta-gpe/recipes-support/fbreader/fbreader_0.12.10.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe meta-openembedded/meta-gpe/recipes-support/fbreader/fbreader_git.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe meta-openembedded/meta-multimedia/recipes-multimedia/sox/sox_14.4.0.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe meta-openembedded/meta-oe/recipes-multimedia/mplayer/mplayer-common.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe meta-smartphone/meta-android/recipes-bsp/chroot-script/chroot-script_1.0.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-connectivity/avahi/avahi-ui_0.6.31.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-connectivity/bind/bind_9.8.1.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-core/systemd/systemd-serialgetty.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-core/sysvinit/sysvinit-inittab_2.88dsf.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-extended/screen/screen_4.0.3.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-extended/shadow/shadow-securetty_4.1.4.3.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-extended/shadow/shadow_4.1.4.3.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-graphics/libsdl/libsdl_1.2.15.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-graphics/packagegroups/packagegroup-core-x11-xserver.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-kernel/modutils-initscripts/modutils-initscripts.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-multimedia/libsndfile/libsndfile1_1.0.25.bb (or one of its .bbappends)
  WARNING: Use of PRINC 1 was detected in the recipe openembedded-core/meta/recipes-support/libcap/libcap_2.22.bb (or one of its .bbappends)
  WARNING: Use of PRINC 11 was detected in the recipe openembedded-core/meta/recipes-bsp/keymaps/keymaps_1.0.bb (or one of its .bbappends)
  WARNING: Use of PRINC 11 was detected in the recipe openembedded-core/meta/recipes-core/systemd/systemd-compat-units.bb (or one of its .bbappends)
  WARNING: Use of PRINC 12 was detected in the recipe openembedded-core/meta/recipes-core/initscripts/initscripts_1.0.bb (or one of its .bbappends)
  WARNING: Use of PRINC 13 was detected in the recipe openembedded-core/meta/recipes-core/base-files/base-files_3.0.14.bb (or one of its .bbappends)
  WARNING: Use of PRINC 2 was detected in the recipe meta-openembedded/meta-oe/recipes-navigation/navit/navit_svn.bb (or one of its .bbappends)
  WARNING: Use of PRINC 2 was detected in the recipe meta-openembedded/meta-oe/recipes-support/lvm2/lvm2_2.02.97.bb (or one of its .bbappends)
  WARNING: Use of PRINC 2 was detected in the recipe openembedded-core/meta/recipes-connectivity/portmap/portmap_6.0.bb (or one of its .bbappends)
  WARNING: Use of PRINC 2 was detected in the recipe openembedded-core/meta/recipes-extended/pam/libpam_1.1.6.bb (or one of its .bbappends)
  WARNING: Use of PRINC 2 was detected in the recipe openembedded-core/meta/recipes-graphics/packagegroups/packagegroup-core-x11.bb (or one of its .bbappends)
  WARNING: Use of PRINC 3 was detected in the recipe meta-openembedded/meta-efl/recipes-efl/efl/entrance_svn.bb (or one of its .bbappends)
  WARNING: Use of PRINC 3 was detected in the recipe meta-openembedded/meta-oe/recipes-multimedia/mplayer/mplayer2_git.bb (or one of its .bbappends)
  WARNING: Use of PRINC 3 was detected in the recipe openembedded-core/meta/recipes-bsp/formfactor/formfactor_0.0.bb (or one of its .bbappends)
  WARNING: Use of PRINC 3 was detected in the recipe openembedded-core/meta/recipes-connectivity/avahi/avahi_0.6.31.bb (or one of its .bbappends)
  WARNING: Use of PRINC 3 was detected in the recipe openembedded-core/meta/recipes-connectivity/dhcp/dhcp_4.2.5-P1.bb (or one of its .bbappends)
  WARNING: Use of PRINC 3 was detected in the recipe openembedded-core/meta/recipes-core/init-ifupdown/init-ifupdown_1.0.bb (or one of its .bbappends)
  WARNING: Use of PRINC 3 was detected in the recipe openembedded-core/meta/recipes-graphics/xinput-calibrator/pointercal-xinput_0.0.bb (or one of its .bbappends)
  WARNING: Use of PRINC 5 was detected in the recipe meta-openembedded/meta-oe/recipes-core/meta/distro-feed-configs.bb (or one of its .bbappends)
  WARNING: Use of PRINC 5 was detected in the recipe openembedded-core/meta/recipes-connectivity/bluez/bluez4_4.101.bb (or one of its .bbappends)
  WARNING: Use of PRINC 6 was detected in the recipe openembedded-core/meta/recipes-core/packagegroups/packagegroup-base.bb (or one of its .bbappends)
  WARNING: Use of PRINC 6 was detected in the recipe openembedded-core/meta/recipes-core/packagegroups/packagegroup-core-boot.bb (or one of its .bbappends)
  WARNING: Use of PRINC 6 was detected in the recipe openembedded-core/meta/recipes-graphics/xorg-xserver/xserver-xf86-config_0.1.bb (or one of its .bbappends)
  WARNING: Use of PRINC 7 was detected in the recipe meta-openembedded/meta-oe/recipes-navigation/gpsd/gpsd_3.7.bb (or one of its .bbappends)
  WARNING: Use of PRINC 7 was detected in the recipe openembedded-core/meta/recipes-core/udev/udev-extraconf_1.0.bb (or one of its .bbappends)

(From OE-Core rev: c6825ec6e92e20bb64a051d1576f363c16e98d68)

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

base: Show PRINC value in the warning message

* it's useful to know by how much we need to bump PR in original recipe to preserve upgrade path

(From OE-Core rev: 6e1633eb246fee3dee154513efa410d0930ea386)

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

base.bbclass: Deprecate the PRINC logic

The PRINC logic is now deprecated, the PR server should be used to handle
the automatic incrementing of the PR (package release) field.  Add a warning
message to alert users to the change. The functionality will be removed
in the next release cycle.

(From OE-Core rev: 6f8d01e0097ded355a1eb38724f11ee65401392e)

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

build-appliance-image: upgraded to commit 54c2e993ec129563c7ae9f3fdee74ab60da28698

- Solves bugs related to bitbake;
- Improved image generation.

(From OE-Core rev: 6e4f68bdcd2c09766a50a98bca0b360314784adb)

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

mc: Add 'inherit pkgconfig'

mc's configure script uses pkg-config so we should 'inherit pgkconfig'
to ensure the sysroot is properly populated. Currently we are lucky as
mc DEPENDS on glib-2.0 which inherits pkgconfig but we should not rely
on this dependency chain.

(From OE-Core rev: 37c367df895497d3147d6b0956ff16db4fdc92ed)

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

oe-selftest: add test for image manifest file creation

(From OE-Core rev: 1542470fa66884d0f7bf67d92916675a8a9820cc)

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

gnutls: fix failure during do_compile

Add a Debian patch to fix a load of errors building the documentation
within do_compile e.g.:

| ./x509-api.texi:15: misplaced {
| ./x509-api.texi:15: misplaced }

(From OE-Core rev: b09a9a5f298596795f17243e5ffcf7dab295a8e6)

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

gnutls: Fixed bug that prevented the rejection of v1 intermediate CA certificates.

This patch is for the OE-Core master branch - it comes from upstream:

>From 467478d8ff08a3cb4be3034ff04c9d08a0ceba3e
From: Nikos Mavrogiannopoulos <nmav@redhat.com>
Date: Wed, 12 Feb 2014 16:41:33 +0100

For more info see:
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-1959
http://www.gnutls.org/security.html#GNUTLS-SA-2014-1
https://www.gitorious.org/gnutls/gnutls/commit/467478d8ff08a3cb4be3034ff04c9d08a0ceba3e

(From OE-Core rev: ea623966b9c48a640cac5c5a5415decca2329249)

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

gnutls: CVE-2014-0092 correct return codes

This patch is for the OE-Core master branch - it comes from upstream:

git://gitorious.org/gnutls/gnutls.git
branch: gnutls_2_12_x
commit: 6aa26f78150ccbdf0aec1878a41c17c41d358a3b
Author: Nikos Mavrogiannopoulos <nmav@gnutls.org>
Date:   Thu Feb 27 19:42:26 2014 +0100

For more info see:
http://www.gnutls.org/security.html#GNUTLS-SA-2014-2
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-0092
https://www.gitorious.org/gnutls/gnutls/commit/6aa26f78150ccbdf0aec1878a41c17c41d358a3b

(From OE-Core rev: 61a507c545aead7e49da1e8ad54dec07a648fdb5)

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

security_flags: Update to correctly link X modules

Remove the -z,now flag from linking

[YOCTO #5885]

(From OE-Core rev: 545986bfbfe20f2b6e8a46e88e2cc3007ca344e6)

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

sanity.bbclass: support wildcards in SANITY_TESTED_DISTROS

With this change, you can use shell like globbing expressions (as
supported by Python's fnmatch) for entries in SANITY_TESTED_DISTROS.
This makes it possible to say that, e.g. "all Debian 7 Wheezy releases
are supported" with the entry "Debian-7.*".

[YOCTO #5265]

(From OE-Core rev: 1e527136e2ac274735a25b957e0391f48b18beba)

Signed-off-by: Olof Johansson <olof.johansson@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

parted: add patch to handle deprecated readline function

(From OE-Core rev: ebaa5485abda86691b0eeadaf689d75072357178)

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

readline: drop readline-only-enable-meta-key-for-a-single-call-read.patch

The 6.3 has fixed this issue.

(From OE-Core rev: 86f3674b402c7f3f3e0e82043936fb24d5fb5b7e)

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

readline: update configure-fix.patch

The configure-fix.patch is used to patch configure.in in 6.2 (or 5.2),
but configure.ac in 6.3

(From OE-Core rev: 11798d94419392dc5639a770792aaee0b7920035)

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

readline: upgrade to 6.3

The readline-6.2-patches is obsolete for 6.3

(From OE-Core rev: 66bc6f4127e88db18e260c71d181aecfd58c7999)

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

ppp: drop if_pppol2tp.h patches

The old version of Linux if_pppol2tp.h was removed, so droped the
related patches.

Here is the commit from git://github.com/paulusmack/ppp.git
...
commit c41092dd4c49267f232f6cba3d31c6c68bfdf68d
Author: Paul Mackerras <paulus@samba.org>
Date:   Sun May 20 16:48:17 2012 +1000

Remove old version of Linux if_pppol2tp.h

This has been in the Linux kernel source now for long enough that we
can rely on getting a usable version from /usr/include on all
distributions that we care about.  The version we have here had started
to lag behind what is in the Linux kernel, causing compilation errors
due to struct pppol2tpv3_addr being undefined.  Removing our local
version means we will use what is in /usr/include instead.
...

(From OE-Core rev: 5ea7b47c10bf9e4a1741c165de39456f1cb4e6ff)

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

ppp: drop enable-ipv6.patch

IPV6 has been enabled by default in 2.4.6

(From OE-Core rev: 14896ef9e3eb091591d25776f8f6a590294a874e)

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

ppp: upgrade to 2.4.6

(From OE-Core rev: d05b1d06b5b8b84c2a40639ad28e31cd683f89a4)

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

bash: update build-tests.patch

(From OE-Core rev: 3c234df240a11903ef3588a2c078dcbce4ca1719)

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

bash: upgrade to 4.3

The bash-4.2-patches is obsolete.

(From OE-Core rev: 31eb09a888729fcfd17d02f2a47375e10e87f79a)

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

mtdev: upgrade to 1.1.5

Removed the following patch(es):
 * fixsepbuild.patch (changes included in release)

(From OE-Core rev: a881acdcc4df8e29bc2107ef5c04ad96024ce2fb)

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

package_manager.py: RpmPM: don't add smart channel if already added

Make sure the channel was not already added, before attempting to add.

[YOCTO #5890]

(From OE-Core rev: 0574f5390fb8958db5864f41a7bd8037ac046ac5)

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

Revert "lib/oe/package_manager.py: RpmPM: fix issue with multilib builds"

Converting a list to a set does not keep the order of the items, as they
were added, which might result in the wrong packages being installed in
the final image...

This reverts commit 12f47c23df8c109676f66d580d666a3147f3b046.

[YOCTO #5890]

(From OE-Core rev: 495cce9c9e63484c740aa669facb995b0ef74f85)

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

gstreamer: Fix valgrind header detection

--disable-valgrind isn't enough, it also looks for the header since the binary
may be present without the -dev headers. Add in a configure value to
ensure the header is disabled as well.

(From OE-Core rev: 4f42aefed6f6edb8d057b204b82e4a05fff2dff0)

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

sstate: Ensure SSTATEPOSTINSTFUNCS is accounted for in signatures

Bitbake can't interpret the python execution of these functions so
include the dependency manually.

[YOCTO #4507]

(From OE-Core rev: 39c53d26a1ea0859ca30456b6af0a28b2ea16769)

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

image.bbclass: add BAD_RECOMMENDATIONS/NO_RECOMMENDATIONS rootfs vardeps

Changes to BAD_RECOMMENDATIONS or NO_RECOMMENDATIONS require the rootfs
for the image to be rebuilt.

[YOCTO #5898]

(From OE-Core rev: f271ec4748d604d956f2d011af1deb7f7e3a68f9)

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

testimage: add task level lock

For machines other than qemu it will not be okay to run
multiple testimage tasks in parallel.

(From OE-Core rev: 9560305a9e28316438cb57421afc7877890c4b76)

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

sstate-cache-management.sh: Don't use echo -n

* when debug is enabled it causes more confusion than what it's gaining
  without debug enabled

(From OE-Core rev: 53899da20cd439ae53974e507f3dd86cd6879455)

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

sstate-cache-management.sh: Allow to remove old .siginfo and .done files for all tasks

* this allows to remove them even for tasks without sstate archives (e.g. rm_work, multilib_install, ..)
  whole list of tasks found in one of my sstate-cache directories:
  sstate:xxx_boostconfig.tgz ... 6
  sstate:xxx_bundle_initramfs.tgz ... 31
  sstate:xxx_compile.tgz ... 47894
  sstate:xxx_compile_kernelmodules.tgz ... 31
  sstate:xxx_compile_ptest_base.tgz ... 68
  sstate:xxx_configure.tgz ... 48247
  sstate:xxx_configure_ptest_base.tgz ... 68
  sstate:xxx_deploy.tgz ... 8
  sstate:xxx_evacuate_scripts.tgz ... 58
  sstate:xxx_fetch.tgz ... 7045
  sstate:xxx_generate.tgz ... 812
  sstate:xxx_generate_qt_config_file.tgz ... 576
  sstate:xxx_generate_toolchain_file.tgz ... 219
  sstate:xxx_install.tgz ... 48273
  sstate:xxx_install_locale.tgz ... 58
  sstate:xxx_install_ptest_base.tgz ... 68
  sstate:xxx_kernel_checkout.tgz ... 14
  sstate:xxx_kernel_configcheck.tgz ... 34
  sstate:xxx_kernel_configme.tgz ... 17
  sstate:xxx_kernel_link_vmlinux.tgz ... 30
  sstate:xxx_make_scripts.tgz ... 42
  sstate:xxx_multilib_install.tgz ... 29
  sstate:xxx_package.tgz ... 13848
  sstate:xxx_package_write_ipk.tgz ... 4615
  sstate:xxx_packagedata.tgz ... 4613
  sstate:xxx_pam_sanity.tgz ... 3
  sstate:xxx_patch.tgz ... 7883
  sstate:xxx_populate_adt.tgz ... 28
  sstate:xxx_populate_ide_support.tgz ... 33
  sstate:xxx_populate_lic.tgz ... 1718
  sstate:xxx_populate_sysroot.tgz ... 5094
  sstate:xxx_prepare_qmake_symlink.tgz ... 6
  sstate:xxx_prepsources.tgz ... 168
  sstate:xxx_rm_work.tgz ... 107630
  sstate:xxx_rm_work_all.tgz ... 52563
  sstate:xxx_rootfs.tgz ... 13
  sstate:xxx_sanitize_sources.tgz ... 8
  sstate:xxx_sizecheck.tgz ... 31
  sstate:xxx_strip.tgz ... 31
  sstate:xxx_uboot_mkimage.tgz ... 31
  sstate:xxx_unpack.tgz ... 7857
  sstate:xxx_unpack_extra.tgz ... 6
  sstate:xxx_validate_branches.tgz ... 14

* with this patches most of these files are removed as well
  Figuring out the sstate:xxx_compile ... (0 from 0 .tgz files for
  compile suffix will be removed or 42310 from 47795 when counting also
  .siginfo and .done files)

(From OE-Core rev: 66c6505e7ee0ea2efa13d1257982f721b26d83ae)

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

sstate-cache-management: improve shown numbers of removed files

* deleted files were counting .siginfo and .done files, but total_files were counting only .tgz archives
  so it could show confusing output like:
  Figuring out the sstate:xxx_packagedata.tgz ... (21098 from 15162 files for packagedata suffix will be removed)
* split the counts and show both
* fix .sigdata sometimes used instead of .siginfo
* this will also show new "suffixes" which have only .siginfo files but
  no .tgz, e.g.:
  Figuring out the sstate:xxx_rm_work.tgz ... (0 from 0 .tgz files for
  rm_work suffix will be removed or 0 from 107630 when counting also
  .siginfo and .done files)
  because now we're storing siginfo files from all intermediate tasks,
  but they are not being removed by this script

(From OE-Core rev: ec881997c748ed8bfb3fc75797367ce3599bd5b4)

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

bitbake.conf: Drop -fpermissive

Drop the -fpermissive C++ compiler flag. We've had this around for years, most
code should have been fixed long ago. Its possible some recipes may fail
however we can (and should) just use the flag where needed.

An OE-Core world build seems to work just fine with this change.

(From OE-Core rev: 24dd8e129447013ee98609f3892ec414b1b21340)

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

oprofile: Partially fix separate builddir support

It doesn't make separate build completely work but its worth
merging as one step in the right direction.

(From OE-Core rev: 2c4942ab36d66de9c189e8c116195f67f2894a08)

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

alsa-tools/autotools: Ensure that aclocal files can be present with AUTOTOOLS_COPYACLOCAL variable

Introduce a AUTOTOOLS_COPYACLOCAL variable which forces the copy of the aclocal
files even when a configure.ac/.in file isn't present.

Use this new feature in alsa-tools.

(From OE-Core rev: 07db5222a970e40bff51c5df793f0021b0e2aba8)

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

autotools: Limit aclocal files to those in dependencies

We still occasionally see race issues with cp-noerror, and it copies
too many files, we should limit the the m4 files to those explicitly in
the DEPENDS for the recipe.

This change takes advantage of the BB_TASKDEPDATA data from bitbake to
only copy in those files listed in the manifest of the recipes in DEPENDS.

I've had this testing locally for some time, its ready for wider
review/testing.

(From OE-Core rev: 79ea036de331bde65a88fb777647dc099ef05acf)

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

openjade: Fix configure assumptions about file moves

This means that do_configure prefunc code can see the files in question
and ensures aclocal files get copied in.

(From OE-Core rev: eab0adf47ea0d78d34f5ce608f4039a158877184)

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

bzip: Add extra files as a specific task

This means that do_configure prefunc code can see the files in question
and ensures aclocal files get copied in.

(From OE-Core rev: 42f983152659c9ba60b0642e1acd65eb42d4b090)

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

Revert "ncurses: use ln -r to generate relative symlink"

ln -r isn't available in older versions of coreutils such as those
present in Ubuntu 12.04. We'll find an alternative solution.

This reverts commit 65c2439ab6dcbd4c52439cdc96acb8ba222c196b.

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