bitbake: siggen.py: Insure .siginfo files writes into shared sstate cache are atomic

   Use tempfile.mkstemp to create a temporary file in the sstate dir and move it
   into place after closing.  The previous code would fail in the chmod() if two
   users were running jobs that touched the same signature file.

(Bitbake rev: ff11e9ac5eba2d957917664a7b91b1277d8ad548)

Signed-off-by: Jeffrey C Honig <jeffrey.honig@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: runqueue.py: Improve error output to be more useful when non-existent tasks are found

(Bitbake rev: 9be584272a63f48d8dc7c9f05b017d11250aa247)

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

gcc-cross-initial.inc: Use BFD ld as default always

When using gold as default linker in final toolchain
gcc-cross-initial still needs to use BFD ld since it
will link eglibc-initial

(From OE-Core rev: f643f886b16f586426c744afde080a99d71a9d58)

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

webkit-gtk: remove svn recipe

(From OE-Core rev: ecb18e2b6a8fc16b323f93c62ed5ff5b080c5dd0)

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

webkit-gtk: add 1.8.1 recipe from tarball

License checksums changed as new years were added.

(From OE-Core rev: e6c5f1edcccfc78259caac853361c14ebe6d2677)

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

gnutls: Fix build with automake 1.12.1

Added new patches:
   avoid_AM_PROG_MKDIR_P_warning_error_with_automake_1.12.patch
   gnutls_fix_for_automake_1.12.1.patch

more details in patch headers.

(From OE-Core rev: a7a07aff69d4bb44410721eecf33a60ac2166328)

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

lame: fix build with automake 1.12.x

automake 1.12.x has deprecated automatic de-ANSI-fication support.
Fix sources to work with automake 1.12.x

Added a new patch:
   lame-3.99.5_fix_for_automake-1.12.x.patch

(From OE-Core rev: 84a433e383e7b2e96932739c31b25956ad8dfc54)

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

sed-4.1.2: fix build with automake 1.12.x

automake has depricated use of @mkdir_p@, Fix the
issue by using @MKDIR_P@ instead.

added a new patch:
   sed-4.1.2_fix_for_automake-1.12.patch

(From OE-Core rev: b5e9b6db49770261e8394e676eec603125b51a44)

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

gawk-3.1.5: fix build with automake 1.12.x

Add the gawk-3.1.5_fix_for_automake-1.12.patch in SRC_URI.

(From OE-Core rev: 2d0c20a39be017e93303c7e7f9c11f32f737372d)

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

tar: fix build with automake 1.12.x

automake 1.12.x automatically deletes empty directories, so
the additional rmdir from the do_install_append fails.

cleanup the do_install_append for automake 1.12.x

Avoid this error:
| rmdir: failed to remove `/srv/home/nitin/builds/build-gcc47/tmp/work/i586-poky-linux/tar-1.26-r1/image/usr/sbin/': No such file or directory
NOTE: package tar-1.26-r1: task do_install: Failed

no PR bump as no change in the output.

(From OE-Core rev: c3c103abb533cc9aae178d41b1bd216165f3bc9a)

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

gpgme: fix build with automake 1.12.x

automake 1.12.x automatically deletes empty directories, so
the additional rmdir from the do_install_append fails.

cleanup the do_install_append for automake 1.12.x

Avoids this build issue:
| rmdir: failed to remove `/srv/home/nitin/builds2/build0/tmp/work/i586-poky-linux/gpgme-1.3.2-r0/image/usr/libexec': No such file or directory
| ERROR: Function failed: do_install

(From OE-Core rev: 4fc32e828113df50da154fcaf3eb3d56b9e72f09)

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

webgit-gtk: fix build with automake 1.12.1

Added a new patch:
  webgit-gtk_fix_build_with_automake_1.12.patch

More comments in the patch header.

(From OE-Core rev: f3ea68cdd2559d84d5b2cd95323d186fff14c7b7)

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

libidn: fix build with automake 1.12

Added a new patch
    avoid_AM_PROG_MKDIR_P_warning_error_with_automake_1.12.patch

More details in the patch headers.

(From OE-Core rev: c654e26007b4b014bc614fde15b7a8d9cf52d849)

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

telepathy-python: fix build with automake 1.12

Added a new patch:
telepathy-python_fix_for_automake_1.12.patch

More comments in the patch header.

(From OE-Core rev: aa3032aef2af5a91c1af77b9323fe762b055cd93)

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

libunistring: fix build with automake 1.12

Added a new patch:
   libunistring/libunistring_fix_for_automake_1.12.patch

More comments in the patch header.

(From OE-Core rev: 015e3eca85c779cd78080f81c9d056be223465d1)

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

pango: fix dependencies to avoid build issue

this commit avoids this build issue:

| /srv/home/nitin/builds/build-gcc47/tmp/sysroots/x86_64-linux/usr/libexec/i586-poky-linux/gcc/i586-poky-linux/4.7.1/ld: cannot find -lelf
| collect2: error: ld returned 1 exit status
...
| make: *** [all] Error 2
| ERROR: oe_runmake failed
NOTE: package pango-1.28.4-r7: task do_compile: Failed

(From OE-Core rev: dafefa7bd4de518adb63ec19390b77a56cb65d8e)

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

libfm: fix build with automake 1.12.x

automake 1.12.x automatically deletes empty directories, so
the additional rmdir from the do_install_append fails.

cleanup the do_install_append for automake 1.12.x

Avoids this build issue:
| rmdir: failed to remove `/srv/home/nitin/builds/build-gcc47/tmp/work/i586-poky-linux/libfm-0.1.17-r4/image/usr/lib/gio/modules/': No such file or directory
NOTE: package libfm-0.1.17-r4: task do_install: Failed

(From OE-Core rev: 456b48aaf746f07492fda5a047453d35b3ae7fb7)

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

xf86-video-vmware: fix build with automake 1.12.x

automake 1.12.x automatically deletes empty directories, so
the additional rmdir from the do_install_append fails.

cleanup the do_install_append for automake 1.12.x

Fixes this build issue:

| rmdir: failed to remove
`/srv/home/nitin/builds/build-gcc47/tmp/work/i586-poky-linux/xf86-video-vmware-2_12.0.1-r17.1/image/usr/bin':
No such file or directory
NOTE: package xf86-video-vmware-2_12.0.1-r17.1: task do_install: Failed

No PR bump as no change in the output.

(From OE-Core rev: ae6cad80ea65b4e2ecbc15b94f66d874ee2400c7)

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

telepathy-glib: fix build with automake 1.12.x

automake 1.12.x automatically deletes empty directories, so
the additional rmdir from the do_install_append fails.

cleanup the do_install_append for automake 1.12.x

avoid this issue:
| rmdir: failed to remove `/srv/home/nitin/builds/build-gcc47/tmp/work/i586-poky-linux/telepathy-glib-0.18.1-r0/image/usr/bin': No such file or directory
NOTE: package telepathy-glib-0.18.1-r0: task do_install: Failed

no PR bump as no change in the output

(From OE-Core rev: bd8e8550c8d7ab7daea8d63d6850c22639e7fb85)

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

libgcrypt: fix build with automake 1.12.x

automake 1.12.x automatically deletes empty directories, so
the additional rmdir from the do_install_append fails.

cleanup the do_install_append for automake 1.12.x
avoid this issue:

| rmdir: failed to remove `/srv/home/nitin/builds/build-gcc47/tmp/work/i586-poky-linux/libgcrypt-1.5.0-r0/image/usr/sbin': No such file or directory
NOTE: package libgcrypt-1.5.0-r0: task do_install: Failed

no PR bump as no change in the output

(From OE-Core rev: c17511dee02428bb457d1ae08356ef2f8a7f9888)

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

evolution-data-server: fix build with automake 1.12.x

Added 2 patches:
   avoid_automake_errors.patch
   fix_for_automake-1.12.x.patch

More information in the patch headers.

(From OE-Core rev: 64c2aa566f022525c7ef0acd7a80359446675990)

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

librsvg: fix build with automake 1.12.x

automake 1.12.x automatically deletes empty directories, so
the additional rmdir from the do_install_append fails.

cleanup the do_install_append for automake 1.12.x

(From OE-Core rev: 2041d00e05ca9260e29b14fb3b640b7594de0102)

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

dates: fix build with automake 1.12.x

automake 1.12.x automatically deletes empty directories, so
the additional rmdir from the do_install_append fails.

cleanup the do_install_append for automake 1.12.x

(From OE-Core rev: 0d333c61c6ce74666ce87593dc2797ca45a97766)

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

avahi: fix build with automake 1.12.1

added a new patch : fix_for_automake_1.12.x.patch

More information in the patch header.

(From OE-Core rev: d2662016e5f767c1c2a2d264c10634db600ddc02)

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

pcmanfm: fix build with automake 1.12.x

Added a patch:
  pcmanfm_fix_for_automake_1.12.patch

More information in the patch header.

(From OE-Core rev: fee7861dcaef40ab86888a12fc2550d03463f98a)

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

gnutls: Fix build with automake 1.12.1

Added new patches:
   avoid_AM_PROG_MKDIR_P_warning_error_with_automake_1.12.patch
   gnutls_fix_for_automake_1.12.1.patch

more details in patch headers.

(From OE-Core rev: 547ab3e32770e318b6d21334e4ecf17af2eef3ca)

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

jpeg: fix configure with automake 1.12.1

Bring back earlier jpeg fix. upstream has removed the fix from earlier release.

The support for automatic de-ANSI-fication has been deprecated in
automake 1.11.2, and will be removed altogether in automake 1.12

This avoids this error:
| configure.ac:24: automatic de-ANSI-fication support is deprecated
| autoreconf: automake failed with exit status: 1
| ERROR: autoreconf execution failed.
NOTE: package jpeg-native-8c-r2: task do_configure: Failed

(From OE-Core rev: 893ce9db79a5c4cca8a6b6ed6a9cec81777aba2f)

(From OE-Core rev: 626676ba039bc5d9427665a52cbe40a202a542c8)

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

automake: upgrade from 1.11.3 to 1.12.1

(From OE-Core rev: c6b70eff7614ea774c0d8e39520f76bdce0ed977)

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

eglibc: Add recipes for 2.16 release

Drop the patches that were either applied
or fixed differenly in 2.16

Add patches to fix ppc spe patches to match
eglibc 2.16

Fix CPPFLAGS to contain correct includepaths
so autoconf cache is generated correctly

(From OE-Core rev: 705fe4dab44e623350fc32efece87a977d53bfc4)

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>

mkelfimage: Fix cross build

This patch fixes building this tool in cross environment
It uses tools such as CC, HOST_CC, I386_CC etc.
and as long as you are building native it does not
matter since they all are same namely 'gcc' but
in cross compiling that wont work. So we make the
makefiles and configure scripts aware of that fact

Upstream hosts this program in a svn repo. So change
the SRC_URI to official svn repo. With git we were
checking out a lot more that just mkelfimage which
is not required.

There were unpackaged empty directories too so
we make sure they are removed as well.

Fixes Yocto bug# 2765

(From OE-Core rev: 0d828f19f7ef2d852bbe6da4510efea76a39b982)

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>

kernel.bbclass: Dont package kxgettext.o

kxgettext.o is generated when building ppc kernels
so we end up with packaging errors like

> ERROR: QA Issue: Architecture did not match (20 to 62) on
> /work/virtex5-poky-linux/linux-xilinx-2.6.38-r00/packages-split/kernel-dev/usr/src/kernel/scripts/kconfig/kxgettext.o

(From OE-Core rev: 77278bd23617834c7f90b3b24a9945ec081dbb38)

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>

guile: Fix missing depends and broken arm build

This patch adds a fix where arm endianness is detected correctly
secondly it adds a missing dependency on libatomics-ops
which was found when build in clean tmpdir and populating
it from a valid sstate.

(From OE-Core rev: b537b035c9b9c9a1174dcafc0252c2b779b17902)

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>

gcc: Switch to git SRC_URI

svn has been giving troubles with svn 1.7 upgrade
Remove duplicate definition of S and B variables
from gcc-4.7.inc

Use 4.7.1.0 for PV to maintain the upgrade path
Its an interim solution once we upgrade to 4.7.2
we will drop it.

(From OE-Core rev: 3f3b74b3fd6ba813dd69804706edcea145876e09)

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>

lib/oe/sstatesig.py: Exclude git-native from sstate checksum

gcc exposed this issue where cross gcc recipes were not having
same task checksums as libgcc or gcc-runtime the target recipes
which use same shared workdir and it was triggering the unpack
fetch and patch tasks to reexecute and hence the trouble

Now that we have more than 1 package to consider lets combine
the check

Thanks RP for help and this is on the line of patch
793ce6cd9aa632e0f13789c8293770a86085d28d

(From OE-Core rev: 72cc3df204c4b838cc0358a7ac989adfdd5c1760)

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>

qemu: Explicitly add --disable-gl-accel when gl is not enabled

Default seems to be to enable GL accelaration so when
gl is not specified in PACKAGECONFIG then add the
--disable-gl-accel to really disable it.

(From OE-Core rev: 039a2f63712cbe87eb68a8c8c0cc365b74651f5a)

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>

gcc: Refresh support for fsl E5500/E6500

These patches are backport from gcc mainline
they fix issues where eglibc was not compilable
for e5500+ due to missing vector intrinsics
It obsoletes the existing e5500 patch

(From OE-Core rev: b440aeeceb8c2e9f412d98793fb1dd55ff9d73f5)

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>

binutils: Backport the e5500/e6500 patches from mainline

This updates the E5500 and E6500 support patches
as they have been applied upstream binutils.

(From OE-Core rev: 202420871785cbdbbf57adbe26eb0f649e57512b)

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>

gcc: Convert do_headerfix into a definitive patch

do_headerfix task is essentially editing configuration
headers in sources which can well be maintained as a
patch which is easier to spot errors (if any) than
dynamically edited source tree

(From OE-Core rev: ddd344e9acef88d614473581266569ce14716ccc)

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>

strace: Fix build on eglibc 2.16

This patch fixes two build issues seen with strace
one in general with eglibc 2.16 and another with x32 abi

(From OE-Core rev: 2ed0011adc1642e0fbb6bfe1d7ae6b70ce7dfedb)

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>

libpam: Fix build with eglibc 2.16

pam_unix_acct.c: In function '_unix_run_verify_binary':
pam_unix_acct.c:97:19: error: storage size of 'rlim' isn't known
pam_unix_acct.c:106:19: error: 'RLIMIT_NOFILE' undeclared (first use in
this function)
pam_unix_acct.c:106:19: note: each undeclared identifier is reported
only once for each function it appears in

(From OE-Core rev: e59a0bac95ce025a6b826be28ccc9e42ca4b5a29)

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>

valgrind: Fix build with eglibc 2.16

(From OE-Core rev: 5a59bf0778608b7c5e1073bf38d36b98d831db0d)

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>

boost: Upgrade recipe to 1.50.0

1.49.0 is not buildable with eglibc 2.16 since
it exposes a define TIME_UTC which is also used
in boost until 1.50.0 when it was renamed to
TIME_UTC_ so instead of porting needed patches
we upgrade boost to latest release

(From OE-Core rev: 20c00be5556db5e6680bcdb7d4ec6d4516f4c410)

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>

grub,guile,cpio,tar,wget: Fix gnulib for absense of gets in eglibc

eglibc 2.16 does not export gets anymore

(From OE-Core rev: 497d8d82b8e7e04a457a7fdd689c657903218c0d)

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>

bison: Fix for gets being removed from eglibc 2.16

(From OE-Core rev: ba9e4ee09f39c427c0d1ba2d0d4c4f59029e4bea)

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>

gettext,m4,augeas,gnutls: Account for removal of gets in eglibc 2.16

These recipes use gnulib which needs this change to use gets
when its defined and not otherwise. Until that change goes into
gnulib and then all these package upgrade gnulib in their sourcebase
we patch them

(From OE-Core rev: 3148460575b65cb681065e5850b8afb96056d712)

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>

udev: Fix build with eglibc 2.16

Include sys/resource.h to get definition of
PRIO_PROCESS

(From OE-Core rev: 6b74c8edac0a4a468741ab7cc4ecef74314d2043)

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>

diffutils: Fix build with eglibc 2.16

eglibc 2.16 has removed gets so we account for that

(From OE-Core rev: bd47833dea9e1b39ca19b02f06ba9194c6d52be8)

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>

coreutils: Fix build with eglibc 2.16

eglibc 2.16 has removed gets so we account for that

(From OE-Core rev: 84183aee7e6f81919ec44f3043c7d9feea01f202)

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>

glib: package bash-completion files in separate package and move ${PN}-codegen to .inc

* 2.32.4 has completion files in ${datadir}/bash-completion
  and dbus also packages bash-competion files in separate package

(From OE-Core rev: b070eb22372ad1ef5b21162d832772e560184cf8)

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

glib: upgrade to 2.32.4

* it's bugfix release, see http://ftp.gnome.org/pub/gnome/sources/glib/2.32/glib-2.32.4.news
  for list of fixed bugs

(From OE-Core rev: 85a930188300152bc9a1bbf640a2441f3177e26f)

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

libx11: fix makekeys build

* on hosts without libx11 installed it fails with
  | makekeys.c:33:27: fatal error: X11/Xresource.h: No such file or directory
  since oe-core rev 52d4ff7b35602faa52360a3e7a800891ef60c739
* upstream added -I$(top_srcdir)/include in 24283d40b1e4314c6647dda49d2a159833341a8b
  to ensure that right Xresource.h version is used, but we're building
  makekeys outside source tree (x11_disable_makekeys.patch), so we need
  to pass -I${S}/include in our makekeys build too
* building makekeys outside source tree isn't probably needed anymore
  and recipes should be updated to reflect that, upstream fixes:
  https://bugs.freedesktop.org/show_bug.cgi?id=22584
  http://cgit.freedesktop.org/xorg/lib/libX11/commit/src/util/Makefile.am?id=b2487d07f7b355f230a56e32c763acd08862735c
  http://cgit.freedesktop.org/xorg/lib/libX11/commit/src/util/Makefile.am?id=24283d40b1e4314c6647dda49d2a159833341a8b

(From OE-Core rev: abdc37d181019b9dd21ad08ad998ea4ff2a3e264)

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

busybox: Include sys/resource.h for rusage definition

Seen with eglibc 2.16

(From OE-Core rev: 077bd603040ebd16f01b4fa20999fb71a9a43695)

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>

gdb: Replace struct siginfo with proper siginfo_t

Fixes errors like

/home/kraj/work/angstrom/sources/openembedded-core/build/tmp-eglibc/work/armv5te-oe-linux-gnueabi/gdb-7.4.1-r0.0/gdb-7.4.1/gdb/linux-nat.h:79:18:
error: field 'siginfo' has incomplete type

(From OE-Core rev: 6e5b37403dd4d0a9a249535d07b36bbae98cadd6)

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>

mirrors: Add debian archive snapshot mirrors to search list

I ran across this problem when apt 0.7.14 was not fetchable
from regular debian mirrors and existing snapshot mirror
did not have it either since we did not use proper syntax
so it was not hitting that.

(From OE-Core rev: 5a5f67e5486837b4d7e5a89cc27a5c8f8c8598ef)

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>

lsof: define linux C library type when using eglibc

lsof tries to compile a temp c source file and execute the binary to
determine linux C library type (file Configure, line 2689-2717).
It is inpracticable for cross-compilation and may have build issue on
some distros since it depends on host settings.

Fix below error when building for 64bit target on 64bit host:
[...]
| dsock.c:481:44: error: 'TCP_LISTEN' undeclared (first use in this function)
| dsock.c:482:45: error: 'TCP_CLOSING' undeclared (first use in this function)
[...]
| make: *** [dsock.o] Error 1

The actual issue exists in do_configure:
[...]
Testing C library type with cc ... done
Cannot determine C library type; assuming it is not glibc.

Which is in turn caused by missing 'gnu/stubs-32.h" when compiling
the temp c source file on host:
[...]
fatal error: gnu/stubs-32.h: No such file or directory compilation terminated.

file gnu/stubs-32.h is provided by 32bit glibc.

(From OE-Core rev: fbf22dfe1a0dca3fc7ac56e4fd274c0145efbc70)

Signed-off-by: Ting Liu <b28495@freescale.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

base.bbclass: add GNUmakefile to the list of possible makefile names

(From OE-Core rev: b753eab1028ba48dfdcdeefa07f3f30743f3ee45)

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>

crosssdk/nativesdk: Ensure EXTRA_OECONF_FPU is unset

If EXTRA_OECONF_FPU is left set, certain ARM variables related to hard-float
can get pulled in and trigger rebuilds of the crosssdk code. The best solution
is to simply force the variable to a known correct value for the SDK targets
currently supported in the same way as TARGET_FPU.

There is some slight rearrangement of the gcc code to ensure the variable is
always used to call the fpu function.

(From OE-Core rev: 410990445ada8cdcfaec4e6fa5791cee9a5b8983)

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

python: Fix python-nativesdk variable contamination issue

Without this, python-nativesdk-crypt can depend on openssl which is incorrect.

(From OE-Core rev: 5740696c81a9143504c1cb93e94c54add27896e4)

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

sstate.bbclass: Move crosssdk before cross to ensure correct variables are set

There was a bug showing up where the crosssdk recipes were being installed as machine
specific in the sstate-control directory. This turned out to be due to the architecture
fields used by sstate being set incorrectly. The problem is that the crosssdk inherits
the cross class. It therefore needs to be listed in this if statement block before
the cross check, not after.

This should resolve some package architecture issue of crosssdk sstate files.

(From OE-Core rev: 79fe28e6033273f9632ca7549d1599d3fd1463ef)

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

gmp: upgrade to 5.0.5

(From OE-Core rev: 9c8000fdf696e234301a67d51dfeccc9bd1fcace)

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

sysvinit-inittab_2.88dsf.bb: Allow multiple serial port consoles to be defined

Set SERIAL_CONSOLES if you want to define multiple serial consoles, also if
you need to check for the presence of the serial consoles you can also define
SERIAL_CONSOLES_CHECK to determine if these are present when you boot. This
will prevent error message that pop up when the serial port is not present.

SERIAL_CONSOLES = "115200;ttyS0 115200;ttyS1 115200;ttyEHV0"
SERIAL_CONSOLES_CHECK = "${SERIAL_CONSOLES}"

The above lines in machine.conf or elsewhere will have the effect of having
two serial consoles and removing any that are not present at boot

(From OE-Core rev: eecd65f188ba2c924626b7e1bf5c8e2eb51e9b59)

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

runqemu: Fix running qemu when build without gl

When gl is disabled in PACKAGECONFIG then we dont need
to check for supporting libs to be present before running
qemu.

(From OE-Core rev: 6612873dc59054e6d37fa7488226218bfb759127)

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>

image_types.bbclass: Default XZ_COMPRESSION_LEVEL to -e -6

Having XZ_COMPRESSION_LEVEL on -e -9 and -T 0 will make xz eat more
than 6Gb memory. Reduce this to -6 to make xz to use about 471Mb
on the tested machine.

(From OE-Core rev: cfd201ecda95d00bc5d70aa5045c649b98ffe8b8)

Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Andrei Gherzan <andrei@gherzan.ro>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

image_types.bbclass: Add XZ variable to set number of threads to be used while compressing

Default this variable to 0. This will make xz use as many threads as there are CPU
cores on the system.

(From OE-Core rev: 81c8e8a81e90ce95042e9fff8bac6fa011c416d1)

Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Andrei Gherzan <andrei@gherzan.ro>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

image_types.bbclass: Fix COMPRESS_CMD for xz to redirect compressed data to file

Having -c modifier makes xz to output the compressed data to stdout. In this
way the needed data will be in the do_rootfs log.
Redirect data to ${IMAGE_NAME}.rootfs.${type}.xz .

(From OE-Core rev: 7ceea61636d6a6f886002e3c29941da3356157ac)

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

cpan.bbclass: Fix CCFLAGS.

We should not just replace CCFLAGS with CFLAGS while compiling, because
this may cause run-time errors with perl's DynaLoader.pm.

Tested on qemux86 with new libnet-libidn-perl bb recipe:

root@qemux86:~# perl -e "use Net::LibIDN"
Not a CODE reference at /usr/lib/perl/5.14.2//DynaLoader.pm line 213.
END failed--call queue aborted at /usr/lib/perl/vendor_perl/5.14.2//Net/LibIDN.pm line 213.
Compilation failed in require at -e line 1.
BEGIN failed--compilation aborted at -e line 1.

(From OE-Core rev: 855211ae7a224f96663862d4a0e58a90d945dd48)

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

puzzles: convert to use SVN

Recently the tarball has been changing due the fact it gets rebuilt
nightly and timestamps are getting updated and changed internal to
the tarball.  Moving back to SVN to try and address this.

(From OE-Core rev: 8f30d41ede28a86564c402ce5b467e51aec6b15d)

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

Update proto -> protocol for recipes

(From OE-Core rev: 247fbed5a93800c44dc36e403f13bdc8dd704553)

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

xz: update to 5.1.2alpha

This contains bugfixes from the 5.0.3 and 5.0.4 releases

(From OE-Core rev: bbdbb70192e16fe351f9cf8f781110faf839be51)

Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

task-core-sdk.bb: add libgomp and libgomp-dev by RECOMMENDS

(From OE-Core rev: b2b682331cc1ec54448b5002ff94672d09fe9e7c)

Signed-off-by: Zhenhua Luo <b19537@freescale.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

python-dbus: upgrade to 1.1.0

Tested by compiling core-image-basic

(From OE-Core rev: 39b2a455ab6b7c38b782e1b3e2811c3d9e4de7af)

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

classes/rootfs_deb: use more reliable check for package existence

It turns out "apt-cache showpkg" does return some information when a
package does not exist but another package recommends it, which can
occur for empty *-dev packages; so use "apt-cache policy" with a
different line count instead.

(From OE-Core rev: ac62761a9cacdfd0225d2db5a75584e6c85469a8)

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>

linux-libc-headers: updated to 3.4.3

Tested by building core-image-sato-sdk.

(From OE-Core rev: d67e47bd457de90b19428245644dc64ecf4ba11a)

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

package.bbclass: Allow overriding of debugedit starting path

In the usual case, the computed path used for debugedit to fix
up path references for the target filesystem is correct. However,
prebuilt binary components, such as prebuilt toolchains, can
have debug paths that do not reflect paths within the local
build directory. Providing an override lets us continue to use
the standard debugedit logic in package.bbclass.

(From OE-Core rev: 32b1fde106bf423873a4a56e1111f230494e2d4a)

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

runqemu/mips: adjust runqemu script to support mipsel machine

(From OE-Core rev: a12e2409af246d4c1d0e0c9a2457fb9c850bd8ca)

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

slang: change char type to signed char in macros

Fix [YOCTO 2726]

C language has 3 distinct char types:
         char
         unsigned char
         signed char
A char has the same range of values as signed char on X86,
but same as unsigned char on ARM which made Slang's typecast()
and array_sort() unable to work for char value on ARM, since
Slang is assuming "char" as "signed char".

Now we change "char" as "signed char" explicitly in
util/mkslarith2.sl, and use it to regenerate src/slarith2.inc

(From OE-Core rev: 0a1d842e2db63265a6a6708b1101e1f52f223fc9)

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

gconf: Disable gtk support

There are only a couple of helper utilities within gconf that need gtk+ as a
dependency and those are unused and pretty useless. We might as well drop
the dependency on gtk and allow more parallel builds by reducing dependency
bottlenecks.

(From OE-Core rev: 6facee8443966b646cd1e72f14ae13e58a13f621)

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

rpm_5.4.9.bb: fix builds on systems with older libc

| dbconvert.o: In function `rpmdb_convert':
| /local/jenkins/jobs/yocto-upstream/workspace/label/master/machine/p4080ds/poky/master/tmp/work/x86_64-linux/rpm-native-5.4.9-r45/rpm-5.4.9/tools/dbconvert.c:126: warning: the use of `tempnam' is dangerous, better use `mkstemp'
| dbconvert.o: In function `main':
| dbconvert.c:(.text+0x923): undefined reference to `htobe32'
| dbconvert.c:(.text+0xaa4): undefined reference to `htole32'
| dbconvert.c:(.text+0xac9): undefined reference to `htole32'

(From OE-Core rev: add2c772cd404e8bea4828959fcb5ee33c35c048)

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

mkelfimage: Upstream fixed packaging issue so upgrade to e1e6a91

Address QA issue: Files/directories were installed but not shipped.
New revision contains patch to mkelfimage removing
/usr/share/mkelfImage/elf32-i386 directory from Makefile install target.

http://review.coreboot.org/#/c/1195/

(From OE-Core rev: 7dd161816ccbb03db4f86902f96d3e86e09ccb64)

Signed-off-by: Raymond Danks <ray.danks@se-eng.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

package install failure on systems without /sbin/ldconfig

Package install failures due to issues in post install rules.

package_do_shlibs only looks for libraries in a directory "lib", this should be
modified to the variable baselib.

ldconfig_postinst_fragment failure observed on systems without /sbin/ldconfig
        [ -x /sbin/ldconfig ] && /sbin/ldconfig
results in the post install rule returning a failure. Modify to
        if [ -x /sbin/ldconfig ]; then /sbin/ldconfig; fi

(From OE-Core rev: 3f85ab6d0d05a4e34fe73fa343d2b3b942fc69b7)

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

conf/machine: replace TUNE_CONFLICTS with TUNECONFLICTS

* it wasn't consistent with other machine configs
* reported 2 months ago..
  http://lists.linuxtogo.org/pipermail/openembedded-core/2012-May/022154.html

(From OE-Core rev: 3fec966531059b4b21f40be3b22a60edf88c5190)

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

perf: add perl, python to DEPENDS

perf has perl and python compile-time dependendencies, add them.

(From OE-Core rev: 1b40dac2c501ae8c8d812cf2866fe7917f144a19)

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

linux-yocto/3.0: perf scripting support

Updating the 3.0 SRCREVs to pickup the recent perf scripting fixes.

  cef17a1 perf: add sgidefs.h to for mips builds
  3258da3 perf: change --root to --prefix for python install
  1f9c25b perf: add 'libperl not found' warning
  f2b1bf1 perf: use pkg-config instead of python-config
  568e05f perf tools: Add support to install perf python extension
  7e47f81 Revert "perf: hard-code NO_LIBPERL/NO_LIBPYTHON"

(From OE-Core rev: 20405bdab18b7a2b101f818e72863d289d5945e4)

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

bitbake: fetch2/__init__.py: Warn user if SRC_URI is using "proto" and not "protocol"

As well, if "proto" is used, get the associated value as "protocol"

(Bitbake rev: 53e6b630f0463d2d07cdaa9c9eb36794dc9b6b69)

Signed-off-by: Andrei Gherzan <andrei@gherzan.ro>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: fetch2/bzr.py: Use "protocol" parameter to get the protocol

(Bitbake rev: 630876b40ed181312e84f902c4cfb97361afbe81)

Signed-off-by: Andrei Gherzan <andrei@gherzan.ro>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: fetch2/hg.py: Use "protocol" parameter to get the protocol

(Bitbake rev: f4a780a7f50fdd8f2bd75888dad790bcfe95b873)

Signed-off-by: Andrei Gherzan <andrei@gherzan.ro>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: fetch2/osc.py: Use "protocol" parameter to get the protocol

(Bitbake rev: 01040efce82cfbaa76f46d9f95c984be8f9ea810)

Signed-off-by: Andrei Gherzan <andrei@gherzan.ro>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

bitbake: bitbake: Abort build if runtime dependency conflict

Currently if there are multiple preferred providers available for
a runtime dependency, bitbake will print an Error message and let
the build go on. Anyways the build should abort while any Errors
occured.

[YOCTO #2734]

(Bitbake rev: 5f81a714f4fca785780bef555b419f0250e5ec1c)

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

connman-gnome: remove unused patches, they were all merged upstream

(From OE-Core rev: bf134eaf752cd5db6cce79ad4c1ae0f25d46b08e)

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

libgomp: add libgomp (openmp) library, and build for powerpc targets by default

(From OE-Core rev: b76a7bc8dcb7aed7d6f026e77a226837004c50af)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gcc: gcc-cross-canadian: use correct location for libraries for powerpc64

This fixes the issue where gcc invokes the linker with an incorrect -L
library location and gives up because it can't find libraries. It was
looking in a /lib folder instead of /lib64

(From OE-Core rev: 1b428f0151a5b86332b501871bf4a925b43580a5)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

gcc-configure-common.inc: use --with-long-double-128 on powerpc to comply with ABI

(From OE-Core rev: 243539efec94e79dade2525d0ab3586eea48c529)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

openjade-native_1.3.2.bb: fix typo and change the deps exclusion to correct var

(From OE-Core rev: 4be17f018f8dd0ed576498badc03ee13095b2263)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

dtc.inc: fix for libdir == /usr/lib64

On 64bit systems dtc will still install libaries in /usr/lib
unless we havet this override

(From OE-Core rev: b60579e66601ba60669a9e1194409fcd8b80530e)

Signed-off-by: Matthew McClintock <msm@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

valgrind: fix default.supp missing issue

    When run valgrind, following error appears:
    ==2254== FATAL: can't open suppressions file "/usr/lib/valgrind/default.supp"

(From OE-Core rev: 0b3261d513cdad80174a9b9e804981c50bcb7ca2)

Signed-off-by: Zhenhua Luo <b19537@freescale.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

directfb: disable imlib2 explicitly

When HOST machine installed imlib2, compiling directfb in OE will
direct this library, and try to use it. This is wrong, since there is
no imlib in OE, so we should disable it explicitly here.

Error as following:
| idirectfbimageprovider_imlib2.c:64:20: fatal error: Imlib2.h: No such
file or directory

Config.log as following (which is wrong):
configure:19602: checking for imlib2-config
configure:19620: found /usr/bin/imlib2-config
configure:19633: result: /usr/bin/imlib2-config

(From OE-Core rev: fbdee213465f38791f3dd6c2663f2fe295be3596)

Signed-off-by: Dennis Lan <dennis.yxun@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

subversion: do_install failed (parallel issue)

install-neon-lib needs libsvn_delta-1.la which will be regenerated
during libsvn_delta-1.la's installation, if libsvn_delta-1.la is
in regenerating and at the same time install-neon-lib links it, the
error willl happen.

The error message is:
/bin/ld: cannot find -lsvn_delta-1
collect2: error: ld returned 1 exit status

This is a parallel issue, so it doesn't happen often.

Note:
The autoreconf doesn't generate build-outputs.mk, it would be generated
by autogen.sh (use build.conf as the input), but autogen.sh isn't
suitable for cross compiling, so both modified build-outputs.mk and
build.conf.

[YOCTO #2727]

(From OE-Core rev: ce37c45abb4cf43e5009867f695982de2eb33450)

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

gnupg: ensure deterministic build by disabling libcurl

gnupg has it's own fake curl, since we use gnupg with zypper, there does not
seem to be a strong reason to add curl to the depends list.

(From OE-Core rev: 44ea2cf548ca20488449d347f8f3b7d49f7e7493)

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

perf: add perf-tui feature

Add a new feature named 'perf-tui'.  Adding this into the
PERF_FEATURES variable in perf.inc will enable the perf TUI (Text-base
UI) user interface on a target, which adds libnewt and turns on the
perf text UI options in perf, if perf is included in an image.

If 'perf-tui' isn't named as a feature (the default), the perf TUI
will be disabled and unavailable.

(From OE-Core rev: 4fd8a550886f02189e4ed127d0a2f16e92f8474c)

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

perf: add perf-scripting feature

Add a new feature named 'perf-scripting'.  Adding this into the
PERF_FEATURES variable in perf.inc will enable perf scripting on a
target, which will turn on all the language bindings currently
available in perf (Perl and Python), if perf is included in an image.

If 'perf-scripting' isn't named as a feature (the default), all perf
language bindings will be disabled and unavailable.

(From OE-Core rev: fc0661041436013b9099dbd659994a2b8b292c19)

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