libselinux: use correct definition of ARCH

The Makefile of libselinux performs the following check:

ARCH := $(patsubst i%86,i386,$(shell uname -m))
ifneq (,$(filter i386,$(ARCH)))
TLSFLAGS += -mno-tls-direct-seg-refs
endif

Which means that if the host machine is an x86, then TLSFLAGS will
contain -mno-tls-direct-seg-refs. That command line option causes
libselinux to fail when building it for target architectures where the
compiler doesn't support that option, i.e. MIPS:

mips-img-linux-gnu-gcc: error: unrecognized command line option
‘-mno-tls-direct-seg-refs’

So to fix that problem we can set the ARCH variable to $(KERNEL_ARCH),
and then append it to the LIBSELINUX_MAKE_OPTS.

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

lttng-tools: Enable aarch64 build

Now that liburcu builds for aarch64, lttng-tools has all its dependencies met
for aarch64, so allow it to build for this architecture.

Signed-off-by: Ben Shelton <benjamin.h.shelton@intel.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

lttng-libust: Enable aarch64 build

With aarch64 support added to liburcu, lttng-libust has its dependencies met
for aarch64, so allow it to build for this architecture.

Signed-off-by: Ben Shelton <benjamin.h.shelton@intel.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

lttng-libust: Bump version to 2.6.2

Signed-off-by: Ben Shelton <benjamin.h.shelton@intel.com>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

liburcu: Add support for aarch64

Bring over a patch from OpenEmbedded to treat aarch64 like arm in the build
process, which allows the build to succeed.

Signed-off-by: Ben Shelton <benjamin.h.shelton@intel.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

liburcu: Bump version to 0.8.7

Signed-off-by: Ben Shelton <benjamin.h.shelton@intel.com>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

python-pyratemp: new package

add python simple template module

Signed-off-by: Steve Calfee <stevecalfee@gmail.com>
Reviewed-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

faifa: fix musl build issue

Fixes:
http://autobuild.buildroot.net/results/17b/17bde543db253c008079b04c5e341f804160f59c/
http://autobuild.buildroot.net/results/9ee/9eeb9d1497d94dfbda710e8bcdf781bc6d4b8724/

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

gst1-plugins-bad: depend on wayland instead of selecting it for wayland plugin

As discussed here:
http://lists.busybox.net/pipermail/buildroot/2015-September/139213.html

It makes more sense to use 'depends on' (and default to y) for a big
dependency like wayland.

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

gst1-plugins-bad: depend on directfb instead of selecting it for directfb plugin

As discussed here:
http://lists.busybox.net/pipermail/buildroot/2015-September/139213.html

It makes more sense to use 'depends on' (and default to y) for a big
dependency like directfb.

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

gst-plugins-bad: depend on directfb instead of selecting it for directfb plugin

As discussed here:
http://lists.busybox.net/pipermail/buildroot/2015-September/139213.html

It makes more sense to use 'depends on' (and default to y) for a big
dependency like directfb.

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libuv: bump to version 1.7.4

Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

barebox: bump to version 2015.09.0

Signed-off-by: Fabio Porcedda <fabio.porcedda@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

docs: use lists.buildroot.org to refer to the mailman archives

Now that mailman handles lists.buildroot.org lets use that everywhere in the
documentation as well.

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

docs/manual: use bugs.buildroot.org for bugtracker

Now that we have a bugs.buildroot.org CNAME for bugzilla lets use that in
the manual as well.

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

stress-ng: Depend on kernel headers >= 3.3

perf.c need PERF_COUNT_HW_REF_CPU_CYCLES generic PMU event which
is available only since 3.3 kernel version

Fixes:
  http://autobuild.buildroot.net/results/88909b118dc927f31ed5c3b5f3513734a88085a9
  http://autobuild.buildroot.net/results/dac6a85fdd721b959175caf5629b91dee717c3c6
  http://autobuild.buildroot.net/results/88909b118dc927f31ed5c3b5f3513734a88085a9

Signed-off-by: Julien Corjon <corjon.j@ecagroup.com>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

wpa_supplicant: fix static link with readline

When linking statically, the order of libraries on the
linker command matters, since readline depends on ncurses.

Signed-off-by: David du Colombier <0intro@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

linux-headers: bump 3.18.x series

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libyuv: Fix aarch64 support

aarch64 is a NEON-enabled arch according to the source code[1] but the
files containing the NEON-specific functions[2] were not added to the
build using CMake.

Fixes
http://autobuild.buildroot.net/results/c8d/c8d9bff33484b598ff85a26771ce51295bec550d/
http://autobuild.buildroot.net/results/7f3/7f3a09157c4c385434458b6aafec202376e75706/
http://autobuild.buildroot.net/results/991/9917d7bb18305c6a6bcbd9fe2aa986d452f7d4a4/
http://autobuild.buildroot.net/results/6bf/6bf9e48b53f77b8a7bb8a667ea3f63298a4685ce/
http://autobuild.buildroot.net/results/ccf/ccff5d1cb3de8642e3fa24a6b01b31a503df67f9/
http://autobuild.buildroot.net/results/0c5/0c559f1b1b9bc9f2e2a131e34f034ce4e34fa75b/
http://autobuild.buildroot.net/results/bf3/bf324985e9bbcd61196304de1d0313a0e216a0a8/
http://autobuild.buildroot.net/results/d55/d552e5e46c263ff9860192a993ee9ed11d7783a6/
http://autobuild.buildroot.net/results/2d7/2d73ccdfd5d06bd6dfdc4e303fe48a554460fa1c/
http://autobuild.buildroot.net/results/972/97207ed46a519368e242f0621e4363d522e21ee8/
http://autobuild.buildroot.net/results/d84/d84d54d85183296c57c9131d946e4517ddcaa84d/
http://autobuild.buildroot.net/results/c18/c189149b8d2a0242da06ec68863db23cd0fa65b8/
http://autobuild.buildroot.net/results/73a/73a026644179708692c43166a8ffba3ce0e19d95/

[1] https://freeswitch.org/stash/projects/SD/repos/libyuv/browse/include/libyuv/row.h#263
[2] https://freeswitch.org/stash/projects/SD/repos/libyuv/browse/BUILD.gn#101

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

linux-headers: bump 3.10.x / 3.14.x / 4.1.x series

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

blackbox: remove deprecated package

Also update legacy as well.
And remove the BR2_DEPRECATED_SINCE_2014_08 symbol since it's unused
now.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

linux-headers: remove deprecated versions 3.{0, 11, 13, 15}

Remove for the upcoming 2015.11 release, they've been deprecated for a
year now.
Also remove some previously forgotten selection removals in legacy.

[Peter: drop 3.0.x kernel patch]
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

dhrystone: pass HZ=100 to fix build issue on musl

Fixes:
http://autobuild.buildroot.net/results/1b6cb1b40431bf484db2d99bfdc8237cbae6a9d7/
http://autobuild.buildroot.net/results/20e669d6e4bc3d872a355ef658ddad8f34676624/

musl doesn't provide sys/param.h to figure out the value of HZ, but as we
know we are building for Linux, and on Linux the clock_t values returned
from times(2) are defined in terms of USER_HZ which is always 100, we can
just as well tell dhrystone to use that directly.

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/valgrind: disable for static-only builds

Fixes:
http://autobuild.buildroot.net/results/3b6/3b6f253f30afaee428ea0fc559be210dcec61dae
http://autobuild.buildroot.net/results/60e/60e5a49360e2ee9482b4d548789f874a6502c429
http://autobuild.buildroot.net/results/5cc/5ccc2e48fa4deb90ab214c82e8921ec494555457
http://autobuild.buildroot.net/results/452/4524184b407bb6f9b55f8f85c8e0b475ff105420

Valgrind ordinarily works by preloading a shared object to replace malloc,
free and friends. Some combinations of toolchain/libc allow this to build
with -static, but others complain.

Since it makes little sense to try to build this on a static-only system,
this patch disables it on static platforms entirely.

[Peter: move comment to top of file to fix menuconfig indentation]
Signed-off-by: Brendan Heading <brendanheading@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

mtd: fix static build with uuid

The uuid library might depend on gettext. Use pkg-config to discover uuid
dependencies at build time, and edit Makefile accordingly to avoid missing
symbols when linking statically.

Based on a patch from Vicente Olivert Riera.

Fixes:
http://autobuild.buildroot.net/results/2dd/2ddd714446d9eb75701bd48c117dc5bbbd291a76/
http://autobuild.buildroot.net/results/3af/3af4ef252c8cb2b22134d91ce48e5930c60ec437/
http://autobuild.buildroot.net/results/648/648341fce974e8f016eee57c95d7511a886eef49/

[Peter: drop MTD_MKFSUBIFS_LIBS variable]
Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Tested-by: Gergely Imreh <imrehg@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

toolchain-external/CodeSourcery MIPS 2015.05: fix lib-names headers

The CodeSourcery MIPS 2015.05 toolchain has some missing headers we need
to create manually in order to avoid compilation errors. A bug has been
already reported and fixed upstream, and the fix will be included in the
next release.

Fixes:
  http://autobuild.buildroot.net/results/bea/bea76392dec5c8e1bcea8be990ad109c6d27e947/
  http://autobuild.buildroot.net/results/64f/64f2b6b6e60d5c2d9537ad6891211cda6baaf6d5/
  ...and many more.

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

perl: fix static build

see http://autobuild.buildroot.net/results/502/502e6a6280451acc48e5f500e774b2bd04c7229e/

Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

gnutls: bump to version 3.3.18

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

toolchain-external: clarify the comment about *-gcc-ar... LTO wrappers

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

toolchain-external: trivial clean up of messages

Before this commit, the output of the toolchain-external build steps
looked like this (abbreviated for clarity):

>>> toolchain-external undefined Building
>>> toolchain-external undefined Installing to staging directory
>>> toolchain-external undefined Copying external toolchain sysroot to staging...
>>> toolchain-external undefined Building ext-toolchain wrapper
mkdir -p output/host/usr/bin; cd output/host/usr/bin; for i in ...
/usr/bin/gcc -O2 -Ioutput/host/usr/include -DBR_SYSROOT='...
if test -f output/host/usr/bin/i686-pc-linux-gnu-gdb ; then mkdir -p ...
>>> toolchain-external undefined Fixing libtool files
>>> toolchain-external undefined Installing to target
>>> toolchain-external undefined Copying external toolchain libraries to target...
if test -e output/target/lib/ld-uClibc.so.1; then ln -sf ld-uClibc.so.1 output/target/lib/ld-uClibc.so.0 ; fi
if test -e output/target/lib/ld64-uClibc.so.1; then ln -sf ld64-uClibc.so.1 output/target/lib/ld64-uClibc.so.0 ; fi

All the long lines with conditions and loops in them are not usefull,
so put $(Q) in front of them. The line with mkdir can better be split
on a separate line so the cd stands out more. There are two redundant
semicolons that can be removed. The installation of gdbinit could
use an extra message so the user can see what is going on.

After this commit, the toolchain-external build steps look like this:
>>> toolchain-external undefined Building
>>> toolchain-external undefined Installing to staging directory
>>> toolchain-external undefined Copying external toolchain sysroot to staging...
>>> toolchain-external undefined Building ext-toolchain wrapper
/usr/bin/gcc -O2 -Ioutput/host/usr/include -DBR_SYSROOT='...
>>> toolchain-external undefined Installing gdbinit
>>> toolchain-external undefined Fixing libtool files
>>> toolchain-external undefined Installing to target
>>> toolchain-external undefined Copying external toolchain libraries to target...

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

gettext: bump version to 0.19.6

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

samba: bump to version 4.3.0

New patch status: sent upstream.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

mdadm: bump to version 3.3.4

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

squid: bump to version 3.5.8

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

connman: bump to version 1.30

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

conntrack-tools: bump to version 1.4.3

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libnetfilter_conntrack: bump to version 1.0.5

Patches upstream so drop them along with autoreconf.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libtasn1: bump to version 4.6

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

gnupg2: bump version to 2.0.29

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libgcrypt: bump version to 1.6.4

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

git: bump version to 2.5.2

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

util-linux: show utilities alphabetically ordered

The list of util-linux utilities we can find on its Config.in file it
seems to be alphabetically ordered in purpose. However, there are some
of them which are in the wrong place. This patch fixes that.

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

util-linux: bump version to 2.27

Removed 0002-program-invocation-short-name.patch since the changes it
does are now upstream:

  https://kernel.googlesource.com/pub/scm/utils/util-linux/util-linux/+/37edac9a0ca83f311279d2e49bacd6ac80310dc8

This release includes a security fix:

CVE-2015-5224 - chfn, chsh file name collision due to incorrect mkstemp
use if compiled without libuser.
[thanks to Qualys Security Advisory team; qualys.com]

Release notes here:

  https://www.kernel.org/pub/linux/utils/util-linux/v2.27/v2.27-ReleaseNotes

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Acked-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Tested-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

mesa3d: bump version to 10.6.7

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Acked-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

gdb: prevent building the documentation

Force gdb to not build the documentation. This way we avoid depending on
host-texinfo. This is a temporary fix until upstream accepts a proposed
--disable-docs configure option.

Since the documentation will not be build at all, we can remove the
parts related to host-texinfo and MAKEINFO in the gdb.mk file.

Fixes:

  http://autobuild.buildroot.net/results/dd5/dd50ed99abb2c8495def826866b184030953f90e/

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Brendan Heading <brendanheading@gmail.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

setools: fix race condition

As reported upstream:

  https://github.com/TresysTechnology/setools3/issues/5

setools tries to compile a .c file before being generated and then it
produces an error like this one:

error: policy_scan.c: No such file or directory

Until upstream fixes the problem we can workaround this by using MAKE1
as the SETOOLS_MAKE.

Fixes:

  http://autobuild.buildroot.net/results/119/1196ac448c77320b59b63b6133d2d1af1245bf69/

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

system: make systemd's kernel headers comment fit

The comment line that explains that systemd needs a glibc toolchain and
kernel headers >= 3.10 doesn't fit in the dialog box, which makes it
look as if headers >= 3.1 are needed.

Abbreviate the comment a little by removing an 'an'. It still doesn't
fit completely (only one trailing * is shown), but at least it's
readable.

While we're at it, change the quoting style to " like we always use.

Reported-by: eschu on IRC
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/usbmount: run-parts directory fix

usbmount.mk created/installed run-parts scripts in
/etc/usbmount/usbmount.d.  However, usbmount executes run-parts on
/etc/usbmount/mount.d for UDEV add action and on /etc/usbmount/umount.d
for UDEV remove action. Change usbmount.mk to create the directories
referenced by usbmount and install the run-parts scripts appropriately.

Signed-off-by: Doug Kehn <rdkehn@yahoo.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

directfb-examples: bump to 1.7.0

This commit bumps the directfb-examples package to use the upstream
1.7.0 version.

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

directfb-examples: do not install to staging

DirectFB examples are just examples, there is no reason to install
them to the staging directory.

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

directfb-examples: remove options for each example

For some historical reasons, the directfb-examples package has one
sub-option for each example, and the .mk files go through some great
lengths to install only the needed data files for each example.

This is a bit silly for several reasons:

 * In no other Buildroot package we do offer such a fine-grained
   selection of what the package installs;

 * directfb-examples are examples, they are therefore typically not
   meant to be used on a final production system, but more during
   development or for demonstration. Those are situations where size
   generally isn't that critical (all installed, the package takes 2.4
   MB on the rootfs);

 * Those options and the .mk file logic has to be updated whenever the
   package is bumped;

 * Users can always use a post-build script to keep only the examples
   they are interested in.

All the options are moved to Config.in.legacy. However, we could
debate on whether this is actually useful, since the main option
BR2_PACKAGE_DIRECTFB_EXAMPLES will now install a superset of whatever
was selected before through sub-options.

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

linux-fusion: use a more traditional style to define LINUX_FUSION_MAKE_OPTS

Use a single assignment to define the value of LINUX_FUSION_MAKE_OPTS
instead of a sequence of += assignments.

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

linux-fusion: bump to 9.0.3, use .tar.xz

This commit bumps linux-fusion to the latest upstream version, and
switches to use the more space efficient xz-compressed tarball.

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

sawman: remove package

This commit removes the sawman package, since the corresponding
functionality has been merged in DirectFB itself since 1.7.x.

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

divine: remove package

This commit removes the divine package, since the corresponding
functionality has been merged in DirectFB itself since 1.7.x.

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

directfb: add new TIFF option

[Thomas: don't enable option by default.]

Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

directfb: add options for divine and sawman

Since DirectFB 1.7.x, the divine (virtual input) and sawman (window
manager) features have been merged in DirectFB itself rather than
being shipped as separate tarballs. This commit creates additional
options to enable those features.

Upcoming commits will get rid of the existing divine and sawman
Buildroot packages.

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

directfb: remove BR2_PACKAGE_DIRECTFB_UNIQUE option

This option depends on BR2_UNIMPLEMENTED (which does not exist) since
2009. Since BR2_UNIMPLEMENTED does not exist, it was not possible to
enable BR2_PACKAGE_DIRECTFB_UNIQUE, which means that nobody cared
about making it available, and that we don't need to add any
Config.in.legacy handling to remove this option.

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

directfb: remove broken X.org support

The BR2_PACKAGE_XSERVER option never existed, so it was never possible
for DirectFB to pass --enable-x11. Moreover, doing this without having
at least some dependency on certain X libraries will clearly not work.

Consequently, let's remove this bogus X.org support, and disable X11
support unconditionally (until someone is interested enough to bring
this back to life).

The option was actually named BR2_PACKAGE_DIRECTFB_XSERVER but was not
used by the .mk file.

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

directfb: do not use inexistant configure options

The --disable-sysfs and --disable-explicit-deps options are not
supported by DirectFB anymore, so don't use them.

[Thomas: rework commit log.]

Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

qt: fix directfb-1.7.7 compile errors

Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

directfb: mark as available only for gcc >= 4.5

The new DirectFB version does not build with gcc 4.3 from the Blackfin
toolchain. One of the reason is that va_copy has some issues, which
were fixed in gcc 4.4.0
(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=36799). There are also
some other issues, which were fixed by a patch proposed by Peter
Seiderer at
http://lists.busybox.net/pipermail/buildroot/2015-February/120281.html.

However, it probably doesn't make a lot of sense to carry patches that
are not upstream for such old compilers. Instead, this commit takes
the action of making DirectFB available only on toolchains using gcc
>= 4.5, which was tested with the Arago toolchain. gcc 4.4 could
potentially work, but wasn't tested (it is no longer supported by the
internal toolchain backend, and we don't have any toolchain based on
gcc 4.4), so we take the safe decision of requiring at least gcc 4.5.

[Peter: add comment explaining toolchain dependenc as suggested by Vincente]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

directfb: bump to 1.7.7

In addition to doing the bump, this commit changes the --enable-fusion
option to --enable-multi-kernel, as was renamed upstream.

The patches were just refreshed, except
0003-remove-redefinition-of-__typeof__.patch, which was removed since
it was a backport from upstream, and is therefore now part of DirectFB
1.7.7.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libsoxr: bump to version 0.1.2

Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

xtables-addons: bump to version 2.8

With support for linux kernel 4.2

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/stress-ng: new package

[Peter: license is GPLv2+]
Signed-off-by: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

stunnel: bump to version 5.23

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

ffmpeg: bump to version 2.8

0016-ranlib.patch is not needed anymore since the problem that it fixes
is already handled upstream:

  https://github.com/FFmpeg/FFmpeg/commit/2c77ca459e9c9aceee5ed93f8dd8de09e9fbf8df

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

sunxi-tools: bump version

Among other changes brings A23/A33/A83T/H3 support. The makefile uses
pkg-config to find libusb, so ensure that is available.

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

toolchain-external: finish removal of SH2A toolchains

Commit c68c365d2937920df1f2062e4821ebb6d21d195e ("toolchain-external:
remove CS sh2 toolchains") removed the definitions of the
BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_SH2A_201103 and
BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_SH2A_201009, but did not actually
remove the code that was using those options.

So this commit removes the parts of the code that are currently dead
due to this: the definition of the prefix of those toolchains, the
hashes, and the URLs.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/upmpdcli: bump to version 0.12.0

Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libupnpp: bump to version 0.12.1

Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

docs/website: add a space between attributes

Error highlighted by the W3 HTML validator:

Error: No space between attributes.
At line 71, column 64
e="width:100%;"href="/download

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

docs/manual: add the "mark as superseded" hint

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

strongswan: bump to version 5.3.3

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

harfbuzz: bump to version 1.0.3

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

lvm2: bump to version 2.02.130

Release announcement:

https://www.redhat.com/archives/lvm-devel/2015-September/msg00048.html

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

udisks: fix build with newer glibc versions

Backport an upstream patch to fix the build with newer glibc versions.
This is how the failure looks like:

job-drive-detach.c: In function 'sysfs_exists':
job-drive-detach.c:44:15: error: storage size of 'statbuf' isn't known
   struct stat statbuf;

Fixes:

  http://autobuild.buildroot.net/results/ab3/ab36c2a1d2ca41a4d9ec6bab7d8d45678dbc5dcd/

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

udisks: bump to version 1.0.5

This version depends on libgudev when using systemd, otherwise it fails
with an error like this one:

checking for GUDEV... no
configure: error: Package requirements (gudev-1.0 >= 147) were not met:

[Peter: add to _DEPENDENCIES, not _CONF_OPTS]
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

docs: website: update for 2015.08.1

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

scancpan: remove hack for Module-Build

with Perl 5.22, Module-Build is no longer a core module

Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

perl: bump to version 5.22.0

Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libsoundtouch: needs c++

Fixes
http://autobuild.buildroot.net/results/a92/a92fcb94492edfdf15c1e18c13b03d75c0210382/
http://autobuild.buildroot.net/results/d1d/d1d9a17a93f5ab61b4fbab67e0b61e79709d411f/
http://autobuild.buildroot.net/results/d55/d5548c5dbbb94bef8eccf614962d0a4fe5c91133/
http://autobuild.buildroot.net/results/798/798f103e066f94314e5664b8bcad7337e40a3fd7/
http://autobuild.buildroot.net/results/484/48407da02451382a1cf7e0c16efa7db8a338ded0/
http://autobuild.buildroot.net/results/4a0/4a0ac131884c0bc231c9f1dfd356cb533fbe5f0b/
http://autobuild.buildroot.net/results/162/1629703bc3160143875a1771f1ac25a079ec94ff/
http://autobuild.buildroot.net/results/be1/be1b69d0394de37615d3a3d2d4c4e25125b71ec2/
http://autobuild.buildroot.net/results/32b/32b9512de01af6417d2db668dc2d75953119f8ab/
http://autobuild.buildroot.net/results/1c0/1c0d433d435639f4355098fc6915c3588f922cad/

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

uclibc: fix coreutils cut bug, enable needed config symbols

Symbols needed for all versions/flavours of uClibc.

Fixes https://bugs.buildroot.org/show_bug.cgi?id=4099

Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

dos2unix: update homepage link

Use the maintainer page that is much more informative than freecode.com.

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

uclibc: symlinks are now automatically created by uClibc-ng makefile

So the workaround can be dropped now that we've moved to uClibc-ng 1.0.6.

[Peter: Extend commit message]
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

uclibc: fix coreutils cut bug, enable needed config symbols in uclibc-ng config

Fixes https://bugs.buildroot.org/show_bug.cgi?id=4099

Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

uclibc: update to 1.0.6

Fixes boost compile problem with ARC.
New function mkostemps required for efl are available.
Static sudo compiles are possible.
Full changes:
http://mailman.uclibc-ng.org/pipermail/devel/2015-August/000463.html

Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libldns: Fix static build

http://autobuild.buildroot.net/results/a0e/a0e686cbe5104f48c9ab05aef7b2989abdb23821/
http://autobuild.buildroot.net/results/527/527b6167fcade1333d8bedffca21e32e410ed8c9/
http://autobuild.buildroot.net/results/0f2/0f229b6806b5cf18d5b56f80088bba7dc3c6eb28/
http://autobuild.buildroot.net/results/bad/bad1c0590868c1e97f647172f85b112ff2604994/
http://autobuild.buildroot.net/results/7db/7dba36ae44560e65a00daf912abe8b43fa8a060d/
http://autobuild.buildroot.net/results/3bf/3bf24e3e845ad8ebcfd0469787e7b7342ecfe3c5/
http://autobuild.buildroot.net/results/7c6/7c6985518bb068ef1764dfb44c59c9f583f07c71/
http://autobuild.buildroot.net/results/211/2117de3dc3d1168bdba0c771b54b3d998e3d62ce/
http://autobuild.buildroot.net/results/908/90893b2cdbbc02d75be5cbc09b8a244d5b05d386/

[Peter:
 - Move workaround under openssl conditional
 - Only add host-pkgconf for static builds
 - Always use lib target]
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libssh: needs MMU support

Fixes
http://autobuild.buildroot.net/results/405/40562b1b68bb7c57a37907eb16fcae88448f5bc6/
http://autobuild.buildroot.net/results/194/194ee885dd195706bded3ed524ff1685f4f6d4b5/
http://autobuild.buildroot.net/results/55c/55c00309bc7497529d28e48dce14299adeb11ede/
http://autobuild.buildroot.net/results/5d4/5d4e40fc4bb9c5a7ede26434c3d6ab325c631ecc/
http://autobuild.buildroot.net/results/8be/8bead9cfd826843482ccc37e9c1290b8774c8de0/
http://autobuild.buildroot.net/results/401/40100e6b07c3057ba427975ffdf3c3834c6fb385/
http://autobuild.buildroot.net/results/d70/d7083a8b197e1833ddaa30620ec1037ed15ec0ce/
http://autobuild.buildroot.net/results/6a2/6a20f29e65f446165b21cba3d55294e6ea446ff2/
http://autobuild.buildroot.net/results/718/718fdc0e5c67fd5cb1bceaec97f39285c159fbda/

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/php: security bump to version 5.6.13

Link to release announcement:
http://php.net/archive/2015.php#id2015-09-04-2

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/{mesa3d, mesa3d-headers}: bump version to 10.6.6

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

dos2unix: remove host-gettext dependency from host package

This shortens build time.

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

dos2unix: remove WCHAR dependency

Add a patch fixing wchar build failure.

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

dos2unix: bump to version 7.3

Cc: James Knight <james.knight@rockwellcollins.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

dos2unix: add target support

Allow the `dos2unix` utility to be built and installed on the target
system.

[baruch: properly handle target gettext]
Signed-off-by: James Knight <james.knight@rockwellcollins.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/bash: indicate getcwd is malloc-supported

When Bash attempts to find the current working directory, it uses a C
library call `getcwd` to resolve it. When cross-compiling, the
configuration process cannot determine if the target system's C library
can support an "unfixed" path length. Therefore, Bash will fallback to a
size of `PATH_MAX` for determining the current working directory. When
using OverlayFS (and possible other file systems), this becomes an issue
since file paths can commonly exceed standard `PATH_MAX` length. This
typically results in the following error appearing:

   error retrieving current directory: [...]

Common C library `getcwd` calls can default to a higher limit (usually
the system's page size). The current configurable C libraries (as of at
least 2015.08) support a zero (0) size buffer length. Most use the
system's page size; musl, being an exception, which defaults to
`PATH_MAX` (as Bash was doing). Since these C libraries support
allocating buffer space with a zero (0) provided size, the following
configuration change allows Bash to support getting a larger-length'ed
working directory on target's that support it.

Signed-off-by: James Knight <james.knight@rockwellcollins.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

bind: security bump to version 9.9.7-P3

Fixes:
CVE-2015-5722 - denial-of-service vector which can be exploited remotely
against a BIND server that is performing validation on DNSSEC-signed
records.
CVE-2015-5986 - denial-of-service vector which can be used against a
BIND server that is performing recursion and (under limited conditions)
an authoritative-only nameserver.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/linux-fusion: slight simplification

Using an intermediate variable to "store" LINUX_VERSION_PROBED is
unnecessary, because they are both recursively-expanded variables, and
the `make kernel-version` code will anyway be used in both places it is
needed; storing in an intermediate variable will not make that a single
expansion of the sub-shell.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

packages: ensure linux supports modules even when not using kernel-module

Some packages build kernel modules without using the kernel-module infra
(because they use custom build systems); they do not automatically get
the kernel to support modules which is ensured when using the infra.

It must be done manually for all those packages, whenever they are
enabled.

Note: the nvidia-driver case does not need the ifeq-block other packages
use, because it is already enclosed in a more stringent ifeq-block.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Peter Korsgaard <jacmet@uclibc.org>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Noé Rubinstein <nrubinstein@aldebaran.com>
Cc: Jan Viktorin <viktorin@rehivetech.com>
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Tested-by: Jan Viktorin <viktorin@rehivetech.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>