dbus: launch helper should be setuid root

Closes #6464

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

CHANGES: update with recent changes

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

directfb: Add patch to use 'gcc' for linking objects

Fixes
http://autobuild.buildroot.net/results/fab819eee4002a9c392c48c1ebaca5c5b6555567/

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

dvb-apps: Fix generate-keynames.sh script for cross-compilation

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

kexec: fix homepage URL

The old one gives 404 error.

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

mesa3d: fix download URL

The MesaLib 7.10.3 tarball has been moved on ftp.freedesktop.org from
pub/mesa/7.10.3/ to pub/mesa/old_versions/7.x/7.10.3/.
A bump to a newer release is only coming after Buildroot 2013.08, so for now
let us just change URL to let the download work properly.

Reported-by: Ryan Wilkins <ryan@deadfrog.net>
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

kmod: fix build failure on gcc < 4.6 caused by _Static_assert()

This commit introduces a patch to kmod that ensures _Static_assert()
is only used if available. The patch has been submitted upstream.

Fixes:

  http://autobuild.buildroot.net/results/9daf0f46642020591731e20d3bf9041ff6259846/

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

bash: fails in parallel build

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

config: reorder top-level menu

Reorder the entries in the top-menu, in a more significant order.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas De Schampheleire <patrickdepinguin+buildroot@gmail.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Acked-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

redis: fix build with ccache enabled

Some parts of redis fail to build when ccache is enabled, due to one of our
own patches. The construct
    make CC=$(CC) target1 target2
would result in
    make CC=ccache <tuple>-gcc target1 target2

and here <tuple>-gcc would be treated as a target and the compiler used
would be just 'ccache'.

Reported-by: Johan Sagaert <sagaert.johan@skynet.be>
Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

gcc-final: Drop obsolete code for handling lib64

Since 5628776c4a4d29d0715633ea463b64cc19e19c5a
"toolchain-external: fix lib64 symlinks"

lib64 is a symlink to lib so there is no reason to copy the libraries
to lib64 anymore. This fixes multiple building problems
for buildroot toolchains:

[...]
mv: 'output/host/usr/mips64-buildroot-linux-uclibc/sysroot/lib64/libuClibc-0.9.33.2.so'
and 'output/host/usr/mips64-buildroot-linux-uclibc/sysroot/lib/libuClibc-0.9.33.2.so'
are the same file
mv: 'output/host/usr/mips64-buildroot-linux-uclibc/sysroot/lib64/libutil-0.9.33.2.so'
and 'output/host/usr/mips64-buildroot-linux-uclibc/sysroot/lib/libutil-0.9.33.2.so'
are the same file
mv: 'output/host/usr/mips64-buildroot-linux-uclibc/sysroot/lib64/libutil.so.0'
and 'output/host/usr/mips64-buildroot-linux-uclibc/sysroot/lib/libutil.so.0'
are the same file
rmdir: failed to remove
'output/host/usr/mips64-buildroot-linux-uclibc/sysroot/lib64': Not a directory
[...]

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

qt5webkit: needs gui support in qt

Fixes http://autobuild.buildroot.net/results/439ce2c3c33a92966808071d4fc7231d50453c79/

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

sunxi-mali: bug fixes for pc and init script

fix bad escape sequences in init script and add correct lib for pc file.

Signed-off-by: Spenser Gilliland <spenser@gillilanding.com>
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

minidlna: Fix automagic dependency to libiconv

minidlna will detect and use the libiconv library if
the package is selected. In that case, we need to link
against -liconv.

Fixes the following linking problem:

tagutils/tagutils.o: In function `do_iconv.constprop.6':
tagutils.c:(.text+0x4d5c): undefined reference to `libiconv_open'
tagutils.c:(.text+0x4da0): undefined reference to `libiconv'
tagutils.c:(.text+0x4e08): undefined reference to `libiconv_close'

[Peter: wrap COMMON_LIBS line]
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

toolchain-external: fix lib64 symlinks

* Always link lib64 or lib32 to lib
* Only copy the architecture's lib directory to staging
* Also cleanup a couple of mkdirs (concerning some 'lib' directories).

Before this patch:

$ ls -ld host/usr/x86_64-buildroot-linux-gnu/sysroot/{,usr/}lib* target/{usr/,}lib*
drwxr-xr-x 2 samuel users 4.0K Aug 12 22:26 host/usr/x86_64-buildroot-linux-gnu/sysroot/lib/
drwxr-xr-x 2 samuel users 4.0K Aug 12 22:27 host/usr/x86_64-buildroot-linux-gnu/sysroot/lib64/
drwxr-xr-x 5 samuel users 4.0K Oct 30  2012 host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/
drwxr-xr-x 5 samuel users 4.0K Aug 12 22:27 host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib64/
drwxr-xr-x 3 samuel users 4.0K Oct 30  2012 host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/libexec/
drwxr-xr-x 2 samuel users 4.0K Aug 12 22:27 target/lib/
lrwxrwxrwx 1 samuel users    3 Aug 12 22:27 target/lib64 -> lib/
drwxr-xr-x 2 samuel users 4.0K Aug 12 22:27 target/usr/lib/
lrwxrwxrwx 1 samuel users    3 Aug 12 22:27 target/usr/lib64 -> lib/

$ find . -type l -xtype l # find broken symlinks
find: `./host/usr/x86_64-buildroot-linux-gnu/sysroot/lib64/lib': Too many levels of symbolic links
find: `./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib64/lib': Too many levels of symbolic links
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libnss_files.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libnss_nis.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libnss_compat.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libnss_nisplus.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libutil.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libthread_db.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libcidn.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libcrypt.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libm.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libnss_hesiod.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libnsl.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/librt.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libnss_db.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libanl.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libBrokenLocale.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libnss_dns.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libresolv.so
./host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/libdl.so
./target/etc/resolv.conf
./target/dev/log

After this patch:
$ ls -ld host/usr/x86_64-buildroot-linux-gnu/sysroot/{,usr/}lib* target/{usr/,}lib*
drwxr-xr-x 2 samuel users 4.0K Oct 30  2012 host/usr/x86_64-buildroot-linux-gnu/sysroot/lib/
lrwxrwxrwx 1 samuel users    5 Aug 12 22:36 host/usr/x86_64-buildroot-linux-gnu/sysroot/lib64 -> lib/
drwxr-xr-x 5 samuel users 4.0K Oct 30  2012 host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib/
lrwxrwxrwx 1 samuel users    5 Aug 12 22:36 host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/lib64 -> lib/
drwxr-xr-x 3 samuel users 4.0K Oct 30  2012 host/usr/x86_64-buildroot-linux-gnu/sysroot/usr/libexec/
drwxr-xr-x 2 samuel users 4.0K Aug 12 22:36 target/lib/
lrwxrwxrwx 1 samuel users    3 Aug 12 22:36 target/lib64 -> lib/
drwxr-xr-x 2 samuel users 4.0K Aug 12 22:36 target/usr/lib/
lrwxrwxrwx 1 samuel users    3 Aug 12 22:36 target/usr/lib64 -> lib/

$ find . -type l -xtype l # find broken symlinks
./target/etc/resolv.conf
./target/dev/log

Fixes http://autobuild.buildroot.net/results/23fb6b1479d2b5906b72c9437b06ab4700ff246d/

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

pulseaudio: fix NEON usage

The NEON intrinsics used by Pulseaudio are only available in either
-mfloat-abi=softfp or -mfloat-abi=hard, not in
-mfloat-abi=soft. Therefore having NEON support in the processor is
not sufficient, we also should *not* be using soft-float.

Moreover, looking at BR2_ARM_ENABLE_NEON to know if the processor has
NEON support is incorrect. This option is only here to allow the user
to tell whether the processor has NEON support or not, for the ARM
cores that only have optional NEON support. Instead, the
BR2_ARM_CPU_HAS_NEON option really indicates whether NEON is available
or not (it is either set automatically by the ARM cores that always
have NEON support, or when BR2_ARM_ENABLE_NEON is enabled by the user
to confirm that his ARM processor has NEON support).

This fixes build failures such as
http://autobuild.buildroot.org/results/675/675c96059be348b594cc92980bef391126931c83/build-end.log

[Peter: add comment about reason for soft abi check]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

package/Makefile.in: fix ABI name for EABIhf targets

The current code to set ABI on ARM does the following:

ifeq ($(BR2_arm)$(BR2_armeb),y)
... set ABI without 'hf' suffix ...
else ifeq ($(BR2_ARM_EABIHF),y)
... set ABI with 'hf' suffix ...
endif

But since $(BR2_arm)$(BR2_armeb) will always be 'y' in the cases where
BR2_ARM_EABIHF is 'y', it means that the 'else' part of the condition
will never be used.

Fix this by appending 'hf' to the ABI variable when BR2_ARM_EABIHF is
selected.

[Peter: put EABIhf handling under arm/armeb conditional for consistency]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

configs/mpc8315erdb: bump to the latest and greatest versions

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

configs/p1010rdb: bump to the latest and greatest versions

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

openssh: fix tab/spacing in init script

Several of the lines in S50sshd script have a strange mix of spaces
and tabs, that at least do not look consistent with neighboring lines.
This patch makes the spacing consistent, and also strips the trailing
spaces.

Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

linux-fusion: bump version to 9.0.2

Fixes build with recent kernels.

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Tested-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Acked-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

python-setuptools: Add missing $(TARGET_DIR) to PYTHONPATH

The $(TARGET_DIR) variable is required when building
python-setuptools for the target otherwise the build system detects
the host installation which leads to permission error problems
like these:

Setuptools installation detected at /usr/lib64/python2.7/site-packages
Renaming /usr/lib64/python2.7/site-packages/setuptools-0.9.8-py2.7.egg-info to
/usr/lib64/python2.7/site-packages/setuptools-0.9.8-py2.7.egg-info.OLD.1377005697.88
OSError: [Errno 13] Permission denied

Moreover, remove the PYTHONPATH variable for host variant since it's
not needed.

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Tested-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Acked-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

apply-patches: run patch in batch mode

If the file to be patched is missing, then `patch' will interactively
ask for a file to be patched. This is annoying in e.g. the autobuilders
because they have to wait for a timeout instead of failing.

Giving the '-t' (batch mode) option to patch fixes this: it will skip the
missing file, and return a non-zero exit code. So the build cleanly
fails.

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Acked-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

Makefile: fix kernel modules test

Commit fde2605765f (Makefile: test before search for kernel modules)
changed the way we strip kernel modules, but it fails when modules aren't
available (as test -d returns with a non zero exit code).

Fix it by including the test -d call in a proper shell conditional.

Reported-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

Makefile: test before search for kernel modules

The 'find $(TARGET_DIR)/lib/modules' used to find and strip kernel
modules fails when no kernel modules have been installed. While the
'|| true' prevents the entire build from failing, there are still some
error messages displayed, which is not nice.

Instead, test if the directory exists before doing the find. We also
remove the '|| true' in order to really abort the build if a
problematic error occurs.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Luca Ceresoli <luca@lucaceresoli.net>

openssl: don't build tests

These require shared libraries and we save a tiny amount of build time.
Fixes:
http://autobuild.buildroot.net/results/88e/88e5e07e94e49879907186ff35ce66944f867ec2/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

acl: add patch to fix static installation

Just like 'attr', 'acl' doesn't use automake to control the
build/installation of its components, and the static-only installation
process was not installing libacl.a. We add a patch that fixes this.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Gustavo Zacarias <gustavo@zacarias.com.ar>

attr: add patch to fix static installation

When attr is built static-only, it forgets to install its libattr.a
file, which leads to the build failure of packages such as 'acl' that
rely on attr.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Gustavo Zacarias <gustavo@zacarias.com.ar>

gcc: fix libstdc++_pic for AVR32

It is not clear why 300-libstdc++-pic.patch pre-applied to
gcc-4.2.2-avr32-2.1.5 seems to have come from some older version
of gcc (as early as 4.0.3). This older patch incorrectly refers to
object files to be included in the library libstdc++_pic.a as *.o,
while their location seems to be in .libs/*.o (see the contents at
e.g. https://github.com/ZigFisher/Midge/blob/master/toolchain/gcc/4.0.3/300-libstdc%2B%2B-pic.patch ).
The same patch for gcc 4.1.0 refers to the files as .libs/*.o (see
https://github.com/ZigFisher/Midge/blob/master/toolchain/gcc/4.1.0/300-libstdc%2B%2B-pic.patch ).
This patch corrects rules in Makefile.in.

Signed-off-by: Alexander Lukichev <alexander.lukichev@gmail.com>
Acked-by: Simon Dawson <spdawson@gmail.com>
Tested-by: Simon Dawson <spdawson@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

generatelocales: fix 'cut' invocation

'cut' needs to be invoked with the -s option to make sure it doesn't
print anything when the delimiter isn't found. This is particularly
important for the charmap detection, because UTF-8 is appended if
the charmap is empty. But without -s, it will never be empty.

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

strongswan: fix typo in tnccs_20 dependencies

It looks like there is a typing error in dependencies of tnccs_20 in strongswan
configure script. Add a patch to solve it.

Resolve issues detected here:
    http://autobuild.buildroot.net/results/b069e304c42660e7f7eba0b358f95dbaf879df64
    http://autobuild.buildroot.net/results/7355ddbaa213f4fcfd1b6d37b25bed1aab525723

Signed-off-by: Jérôme Pouiller <jezz@sysmic.org>
Tested-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/libbsd: further reduce the set of supported archs

Support for a,out on ARM will be dropped in Linux 3.11 (and is already
missing in the -rc).

m68k is marked BROKEN in Buildroot, so it is never ever tested.

The x86 familly is the only one left with support for a.out, now.

Reported-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

qt5: change upstream URL

The Qt5 upstream URL has changed, which leads to build failures in the
autobuilders.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

Prepare for 2013.08-rc2

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

CHANGES: update for 2013.08-rc2

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

configs/qemu: bump relevant config versions

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linux: bump 3.10.x stable version

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linux-headers: bump 3.{0, 4, 10}.x stable versions

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

webkit: Fix comment dependency for BR2_PACKAGE_LIBGTK2

The comment should only be displayed if BR2_PACKAGE_LIBGTK2
is not selected.

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

webkit: Restrict supported architectures

See Source/WTF/wtf/Platform.h. Webkit is only supported for
arm, armeb, i386, mips, mipsel, powerpc, sh, sparc, x86_64.

[Thomas P: propagate dependency to the midori package.]

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

netatalk: requires threads

Fixes:
http://autobuild.buildroot.net/results/e19/e191a9e0768b632b84c3a8a8201538bd5fbe3708/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linux: don't take HEAD as default for git repositories

Specifying a floating tag like HEAD for a repository version is bad practice,
as it results in non-reproducible builds. This patch removes the default
assignment of HEAD as version when a custom git repository is used for the
Linux kernel.

Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linux: bump 3.10.x stable version

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linux-headers: bump 3.{0, 2, 4, 10}.x stable versions

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

toolchain-external: Restrict Sourcery CodeBench toolchains for MIPS/n32

n32 is not supported in Sourcery CodeBench toolchains.

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

ffmpeg: fix typo in BR2_arm1176jzf_s

Signed-off-by: Ryan Coe <bluemrp9@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

ltrace: disable libunwind for uClibc toolchains

ltrace requires that libunwind is built with backtrace() support.
For the internal uClibc toolchain we don't enable it, and for external
uClibc toolchains we can't know.
It's also unavailable for static uClibc toolchains.
So just disable libunwind support for uClibc toolchains in general.
Fixes:
http://autobuild.buildroot.net/results/ee0/ee037a19590fb85c64f97f78f74bcfd4d7766706/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

wayland: needs threads

Fixes:
http://autobuild.buildroot.net/results/383/383c6ac297250f87db9d03afa38d45b2c1adcb47/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libcec: needs threads

Fixes:
http://autobuild.buildroot.net/results/16a/16ad8b64e68477dcd663d07c600fd1be3164df81/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

aircrack-ng: fix static build

When static build and sqlite3 support are enabled together,
link flags needs to be defined in the right order.

Fixes http://autobuild.buildroot.org/results/a74/a74ced69052c1d0a91dbe32483fd0612d1bf24a

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

readline: use normal/default install target

Since the custom INSTALL_TARGET doesn't consider static builds it breaks
horribly and there's no need with the current version since it works
just fine with a little varnish to remove samples. Fixes:
http://autobuild.buildroot.net/results/5a6/5a62c38ec621d49230d76981db6024035e88804a/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

bash: fix static build

The static build needs some trickery, so make it happen. Fixes:
http://autobuild.buildroot.net/results/2b12868f8297a96d7abf7fcf526ca5eace73d66d/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

opencv: gtk support doesn't work with directfb

This was dropped on 4b3d949b8af240bda451b0a4e57d60153188c2e4 when
libgkt2 was updated. Fixes:
http://autobuild.buildroot.net/results/817/8173b7fd3d6af79a7f2c04bf36059a4429127516/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

imagemagick: fix magick++-config fixup error

Magic++ bindings are built only with C++ and WCHAR toolchains.
Add a WCHAR toolchain check for the magick++ config fixup.
Looking into the future the fixup shouldn't bail on a missing file so we
can avoid awkward kludges for packages that have many options and config
files.
Fixes:
http://autobuild.buildroot.net/results/33a/33ac4b17866a64379b7bab3c0549f6e075c98dde/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

uclibc: arc: Add susv3 legacy alias bcmp

This fixes the building of screen package for ARC

Reported-by: Mischa Jonker <mjonker@synopsys.com>
Cc: Mischa Jonker <mjonker@synopsys.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

toolchain-external: don't create gdb symlink when building host-gdb

The external-toolchain infrastructure creates symbolic links for all
tools in the host directory. However, when buildroot builds its own
version of a cross debugger (BR2_PACKAGE_HOST_GDB), and the toolchain
also provides a cross debugger, there would be two symbolic links for
gdb in the host directory, which is confusing.
An example use case is where the external toolchain only provides a
64-bit gdbserver (e.g. Cavium Networks SDK) but the target is completely
32-bit (e.g. n32 ABI). In this case, using gdbserver on target requires
copying a bunch of 64-bit libraries to the target as well, just for gdb.
In this case, one can let buildroot build both gdbserver as cross-gdb
(both in 32-bit).

This patch modifies the symlink creation so that no gdb (or gdbtui)
symlink is created if buildroot is going to build a cross-gdb.

Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

legacy: add a legacy option for BR2_ARM_OABI

As Arnout noted, we removed the BR2_ARM_OABI option without adding the
corresponding option in Config.in.legacy, so users upgrading to
Buildroot 2013.08 with an OABI configuration would get migrated
automatically to EABI without any warning.

This commit introduces such a legacy option, so that such users would
be explicitly warned about the removal of OABI support.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

udev: s/BR2_TOOLCHAIN_USE_THREADS/BR2_TOOLCHAIN_HAS_THREADS/

Bug fix.

Signed-off-by: Spenser Gilliland <spenser@gillilanding.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

manual: minor typos and formating fixes

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

manual: tips: document how to build the manual

Signed-off-by: A.R.D. <contact@team-ard.com>
Cc: Willy Lambert <lambert.willy@gmail.com>
Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

manual: contribute.txt: add patch revision section

Improve the contribute manual section by adding an explanation about patch
review and version.

The section now provides advices in how to respond maintainers requests and how
to proceed on replying them.

[Thomas: further small modifications.]
Signed-off-by: Vinicius Tinti <viniciustinti@gmail.com>
Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

strongswan: fix build error 'no rule to make target libtls.la'

In some strongswan configurations, an implicit dependency could be missing.
This commit patches strongswan to force this dependency when needed.

Fixes several autobuild failures, like:
http://autobuild.buildroot.net/results/643f5b33973d5b12ffe6d84254bf3be9ed81d0ad/
http://autobuild.buildroot.net/results/24bcdcb83119e416c92458fa9ff2f5269957a743/

Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

strongswan: needs host-pkgconf

Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Tested-by: Jérôme Pouiller <jezz@sysmic.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libffi: Add -mno-compact-eh when compiling for Mips

Fix compilation on Mips. Resolve issues detected here:
    http://autobuild.buildroot.net/results/f0c9db496233f53c26e92294d7305aa4511ec7b3/
    http://autobuild.buildroot.net/results/8968690c248df86b040218867f92b573721e872c/
    http://autobuild.buildroot.net/results/a35db8e6a49d022133b486b12f6a8e40b3c95b6e/

Also remove previous special case handling defined in libffi.mk.

[Thomas: add Signed-off-by from Jérôme inside the patch itself.]
Signed-off-by: Jérôme Pouiller <jezz@sysmic.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

sunxi-cedarx: fix armel and internal toolchain build issues

Fixes autobuilder issues
http://autobuild.buildroot.org/results/a06/a06fe1f2fa2cb57d57524eff5f7b88fe36fdbf68
http://autobuild.buildroot.org/results/328/328b36c3cbf952cccabf16a25fc13f0c3515e8bd

Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Spenser Gilliland <spenser@gillilanding.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

gcc: fix !BR2_LARGEFILE builds

see http://lists.busybox.net/pipermail/buildroot/2013-August/076278.html

Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Tested-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

configs: lock kernel version for atngw100_defconfig

The 3.10.x kernels fail to build with the following message:

vsprintf.c:(.text+0x17e0): relocation truncated to fit: R_AVR32_16N_PCREL against symbol `_ctype' defined in .text section in lib/lib.a(ctype.o)

While this is most likely a toolchain issue, but since AVR32 isn't a
well-maintained architecture, there's not much we can do.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

Fix ARM uclibc build with !BR2_UCLIBC_ARM_BX.

When BR2_UCLIBC_ARM_BX is not set, UCLIBC_OPT_UNSET gets called with an extra
argument.  Remove it.

Signed-off-by: Justin Maggard <jmaggard10@gmail.com>
Acked-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

cppcms, libcec: fix license typo

Signed-off-by: Jerzy Grzegorek <jerzy.grzegorek@trzebnica.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

cairo: add patch to fix Blackfin build failure

Fixes
http://autobuild.buildroot.org/results/1b1/1b1e4ad2f00de1e0d1cc9045822751c3882c7493/build-end.log.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

boost: fix 64-bit build due to ambiguous format call

This commit adds a patch imported from the upstream bug tracker, fixing a
problem with boost builds failing for 64-bit, like in autobuild [1].
Refer to [2] for the actual autobuild error.

Upstream did not yet accept this patch, because it is not portable enough
(not OK for Windows builds) [3], but this should not withhold us from fixing
the error in buildroot.

[1] http://autobuild.buildroot.net/results/0f6c2f99b05f57adb557c2ddd686b5072f5e9ecf/
[2] http://lists.busybox.net/pipermail/buildroot/2013-August/076222.html
[3] https://svn.boost.org/trac/boost/ticket/6851

Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

manual: mention LIBFOO_EXTRACT_CMDS

[Samuel: add precision about non-tar archive]
[Thomas: reword with more details. Fix commit title]
Signed-off-by: Stephan Hoffmann <sho@relinux.de>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

git: needs MMU support

Fixes
http://autobuild.buildroot.org/results/b5454e62c6ea18a9d30148c704927e820f29a7d4/build-end.log.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

eglibc: add comment explaining why we pass -O2

While reviewing the eglibc package, Peter mentionned that it would be
good to add a comment that explains why we explicitly build eglibc
with -O2 instead of using the user selected optimization level. This
commit add such a comment.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

opengl: libgles and libegl also provided by gpu-viv-bin-mx6q

The gpu-viv-bin-mx6q package selects BR2_PACKAGE_HAS_OPENGL_EGL and
BR2_PACKAGE_HAS_OPENGL_ES, so when it is enabled, Buildroot believes
that OpenGL and EGL support is available.

However, both libgles.mk and libegl.mk do not add the dependency on
gpu-viv-bin-mx6q, so when pulling the libgles or libegl dependencies,
the build fails due to the absence of an OpenGL implementation. This
commit fixes that.

Fixes the build failure at
http://autobuild.buildroot.org/results/dbd/dbd938914883a9e205f967f7b4b4a8a7dc7be117/build-end.log.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

gnupg: security bump to version 1.4.14

Fixes CVE-2013-4242.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libgcrypt: security bump to version 1.5.3

Fixes CVE-2013-4242.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

samba: security bump to version 3.6.17

Fixes CVE-2013-4124.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

zeromq: fix build on some compilers due to missing newline

Fixes http://autobuild.buildroot.net/results/645b357d470b75baa9a93eb5be4f1dc5c8c337fa/

Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

Prepare for 2013.08-rc1

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

CHANGES: update for 2013.08-rc1

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linux: update help text of BR2_LINUX_KERNEL_INSTALL_TARGET

The BR2_LINUX_KERNEL_INSTALL_TARGET option not only copies the kernel
image to /boot, but also the DTBs, if such DTBs are generated by the
build process. It makes sense to mention that in the help text of this
option.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

Install DTB as part of images install command

When installing DTB to the images directory, do this as part of the image
install commands rather than the target install commands.

Signed-off-by: Reuben Dowle <reuben.dowle@4rf.com.az>
Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

Fix building initramfs kernel with appended dtb

The target for building the kernel with a cpio initramfs is not
calling the append dtb commands, creating a final kernel image
without an appended dtb. Instead it needs to call kernel make on
the intermediate target, then call the append dtb commands
to produce the final image as the primary kernel build target
does.

Signed-off-by: Nathan Ford <nford@westpond.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linux: Fix uImage with appended DTs generation

Fixes bug #5516 - appended device tree blobs on uImage fails

Before version 3.7 of the kernel, building the zImage and then the
uImage will rewrite the zImage in the process, removing the device tree
we just appended.

Use mkimage to append the device tree to the uImage and rebuild the
headers directly.

Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

dropwatch: link with libintl if present

binutils libbfd links to libintl if present, so dropwatch needs it too.
Ugly hardcoded LDFLAGS, but then that's what the dropwatch Makefile
does. Fixes:
http://autobuild.buildroot.net/results/549/54917e7943143e47263b60b4eee3dfc6f0801407/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

dropwatch: fix build failure with binutils 2.23.1+

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libgpgme: handle argp-standalone

libgpgme has an internal implementation of argp parsing for when the
libc lacks such an implementation (i.e. uClibc).
However if it detects argp.h the internal implementation is disabled and
for a uClibc-based toolchain that breaks the build.
This happens when argp-standalone is installed since libgpgme never
tries to link to libargp.

Make it link against libargp, which also saves some target size. Fixes:
http://autobuild.buildroot.net/results/33c/33c5776d2b3f7fa276a27f7602eb71c4d77c7b62/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

group file: define groups expected by udev

udev-182 requires several groups to be resolvable at udev startup,
including disk, cdrom, floppy, tape, audio, video, lp, tty, dialout,
and kmem.  Only some of these are in the default skeleton's group
file, So let's add the missing groups, and plugdev too.

This avoids getting these logs in /var/logs/messages:

Jan  1 00:00:08 buildroot daemon.err udevd[37]: specified group 'dialout' unknown
Jan  1 00:00:08 buildroot daemon.err udevd[37]: specified group 'kmem' unknown
Jan  1 00:00:08 buildroot daemon.err udevd[37]: specified group 'video' unknown
Jan  1 00:00:08 buildroot daemon.err udevd[37]: specified group 'lp' unknown
Jan  1 00:00:08 buildroot daemon.err udevd[37]: specified group 'floppy' unknown
Jan  1 00:00:08 buildroot daemon.err udevd[37]: specified group 'cdrom' unknown
Jan  1 00:00:08 buildroot daemon.err udevd[37]: specified group 'tape' unknown

Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

Makefile: add variable print capabilities

Add a new goal that dumps all of our variables.

Taken from http://www.cmcrossroads.com/ask-mr-make/6521-dumping-every-makefile-variable

Signed-off-by: Emeric Vigier <emeric.vigier@savoirfairelinux.com>
[yann.morin.1998@free.fr: drop per-variable print; refresh patch]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libedit: really needs wchar support

wctype.h is always included, even in configure tests.
Fixes:
http://autobuild.buildroot.net/results/ed8/ed85b9d688ad8389895e3ecd7a6a18151efc328d/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

readline: make description more enlightening and fix indentation

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

qt5jsbackend: needs ARM v5+

Just like nodejs and other v8-based javascript engines for ARM interwork
it needs the BLX instruction which is only available on V5+ cores.
Fixes:
http://autobuild.buildroot.net/results/853/853cd29bf427902b2cb82a101964a4e45714cb86/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

cppcms: new package

[Thomas: reformat header and Config.in help text, add dependency on
glibc since cppcms uses <monetary.h> functions that aren't available
in uClibc.]

Signed-off-by: Nicolas Ménégale <nicolas.menegale@openwide.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libnspr: Add dependency on !BR2_arc

ARC is not supported yet in libnspr.

v2: Propagate to ecryptfs-utils as well

Signed-off-by: Mischa Jonker <mjonker@synopsys.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

strongswan: scripts need charon or tools

Fixes:
http://autobuild.buildroot.net/results/007/007e75362cbe4693e36d5a7ab96c539d1f3b00fc/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

strongswan: security bump to version 5.0.4

Fixes CVE-2013-2944.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/uclibc: add mkostemp() backport

Needed for modern versions of systemd and udev.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/uclibc: add execvpe() backport

Needed for modern versions of systemd and udev.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/uclibc: number patches for 0.9.32.1

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>