configs: atmel: add support for sama5d2 xplained boot from sd card

This defconfig allows to build all the components to boot from a sd card on a
SAMA5D2 Xplained board. For more details, please read the
board/atmel/readme.txt file.

Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

configs: atmel: add support for sama5d4 xplained boot from sd card

This variant of SAMA5D4 Xplained defconfig file allows to boot from a SD
card.

Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
[Thomas: remove host-samba from the package selection, not needed when
booting from SD card.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

configs: atmel: unify and update defconfig for Xplained boards

Unify and update the defconfig files for Atmel Xplained boards:
- use at91bootstrap as first stage bootloader instead of u-boot spl
- build all dtb variants
- bump to linux4sam_5.2

Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

gcc: fix libsanitizer build with _FILE_OFFSET_BITS=64

After the commit f67a4f50e2ace3a2 "gcc: preserve CXXFLAGS_FOR_TARGET"
target CFLAGS are passed to all libraries bundled with gcc. This breaks
libsanitizer on gcc-4.9.x when building with _FILE_OFFSET_BITS=64:

  error: size of array ‘assertion_failed__837’ is negative
  ...
  libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.cc:837:1:
  note: in expansion of macro ‘CHECK_SIZE_AND_OFFSET’
  CHECK_SIZE_AND_OFFSET(dirent, d_ino);

This issue is fixed in the libsanitizer mainline and in gcc-5.x.
There's no issue with gcc-4.8.x and earlier.

Reported-by: Julien CORJON <corjon.j@ecagroup.com>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Backported from: https://llvm.org/svn/llvm-project/compiler-rt/trunk@220328
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libgtk3: bump to version 3.20.0

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

pango: bump to version 1.40.0

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

adwaita-icon-theme: bump to version 3.20

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

atk: bump to version 2.20.0

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

libsoup: bump to version 2.54.0.1

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

icu: bump to version 57.1

0006-fix-Error-allocating-memory-for-pkgDataFlags.-compil.patch is
upstream so drop it.

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

gdk-pixbuf: create loader cache at build time

Create the gdk pixbuf loaders cache file at build time rather than run
time, since the root filesytem can be read-only.
We also didn't cater for systemd-based systems, so this solves two
issues at once.

The only side-effect is that we must now pull host-gdk-pixbuf into
target gdk-pixbuf dependencies, but generally users will have gdk-pixbuf
selected if they use libgtk2/3 which already pull in host-gdk-pixbuf.

And remove initscript correspondingly.

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

samba4: bump to version 4.4.0

libaio support is now automatic so drop the enable/disable (it will fall
back to pthread aio if libaio is not present).

0002-build-improve-stack-protector-check.patch is upstream so remove it.

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

python-pyopenssl: bump to version 16.0.0

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

python-cryptography: bump to version 1.3.1

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

python-idna: bump to version 2.1

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

nmap: bump to version 7.11

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

imagemagick: bump to version 6.9.3-7

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

tree: bump to version 1.7.0

Signed-off-by: Ricardo Martincoski <ricardo.martincoski@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

qt5base: disable c++ standard compiler flags for the host build

Apply upstream patch, that disables c++ standard compiler flags
unconditionally for host builds.

Fixes:
http://autobuild.buildroot.net/results/895/8958cfb71efdd3ce32b2145237e865ca8a6a6a7f/

Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Reviewed-by: Julien Corjon <corjon.j@ecagroup.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

gtkmm3: depends on libgtk3

packages shouldn't select libgtk3 directly, just depend on it, like for
libgtk2.
In the past libgtk3 didn't require any *GL backend and the dead-end
solution/last resort was the broadway (networked) gdk backend - though
not very useful it didn't require any funky dependencies.
But now we do. Fixes:
http://autobuild.buildroot.net/results/794/794c7ed221432e46a810fc281732ba417cd4cda3/

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

json-glib: bump to version 1.2.0

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

busybox: use md5 as default password algorithm

DES is terribly outdated and a security vulnerability.

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

altera: remove whitespace before colon

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

altera: improve readme to cover both supported boards

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

boot/uboot: compute CRC on SPLs for Altera SoC FPGA

Signed-off-by: Jan Viktorin <viktorin@rehivetech.com>
[Thomas: add Altera in the option name and description, drop reference
to Go being needed and to Maxime Hadjinlian's version of mkpimage
since a C version is now used.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

mkpimage: new host package

The tool helps to create a working SPL to boot Altera SoC FPGA boards.

The code of mkpimage is integrated directly from the Barebox repository
as stated in the 'mkpimage.mk' file.

This tool is *NOT* necessary for any other boards so far.

Signed-off-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Tested-by: Jan Viktorin <viktorin@rehivetech.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

granite: new package

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

libgee: new package

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

vala: add vala/valac wrapper

vala/valac can use gir and vapi data files installed by other packages,
but since these are normally installed to staging and host-vala looks
for them in the host directory (logically) this leads to failure.
So wrap them to call the real tool and add this information via
command-line parameters to them.

This is required for vala-in-vala bindings (vapi).

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

acpid: bump to version 2.0.27

Drop 0003-support-for-non-glibc-libcs.patch since it's upstream.

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

connman-gtk: bump to version 1.1.1

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

package/libldns: Fix build error with perl 5.22

Bug was already reported upstream, as solution upstream refered to this
commit: https://www.nlnetlabs.nl/bugs-script/show_bug.cgi?id=733#c2

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libglib2: bump to version 2.48.0

Switch to system/unbundled pcre since it's the default and recommended
by upstream now.
It's also good security practice since pcre patches won't get updated in
the bundled version inside glib so often.

As stated in glib's NEWS:

Overview of changes in GLib 2.47.5

* the system copy of PCRE is now used by default to implement GRegex.
  Configure with --with-pcre=internal if a system PCRE version
  is unavailable or undesired.

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

pcre: enable unicode properties for host variant

They're required for host-libglib2 and using system pcre is the
default/recommended with newer versions.

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

package/util-linux: fix lsipc segfault

lsipc segfault when no option is given.

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/quota: renumber patches

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

erlang: bump to version 18.3

Erlang/OTP 18.3 is a service release on the 18 track with mostly bug
fixes, but is does contain a number of new features and improvements as
well.

Some highlights of the release are:

  . New statistics info about runnable and active processes & ports.
    Call erlang:statistics with: total_run_queue_lengths |
    run_queue_lengths | total_active_tasks | active_tasks.

  . Time warp improvements: dbg:p/2 and erlang:trace/3 with
    monotonic_timestamp |strict_monotonic_timestamp.

  . Introduced a validation callback for heart.

  . The module overload in sasl has been deprecated.

  . ~90 contributions since 18.2

Signed-off-by: Frank Hunleth <fhunleth@troodon-software.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

gnuradio: fix compilation with cmake-3.5

Volk is a submodule of gnuradio, consequently PROJECT_{BINARY,SOURCE}_DIR must
be used instead of CMAKE_{BINARY,SOURCE}_DIR.

Patch backported from https://github.com/gnuradio/volk/pull/65

fix:

http://autobuild.buildroot.net/results/9a2c5ba503345d2d295b943776882c33f4f0f1d9/
http://autobuild.buildroot.net/results/17d1292adaaca5025bffb42191d9b2172246a514/
http://autobuild.buildroot.net/results/71f97ac20ea41a66daaadea5738b636bc45e805c/
http://autobuild.buildroot.net/results/0277af68c363af733c3e6f3197d3ef97597c5d7b/
http://autobuild.buildroot.net/results/ad5bccd2c4496859824f84327038eba5f6206d97/
http://autobuild.buildroot.net/results/f3dc0c224174cd8beddeaa8ea089320ddff621b5/
http://autobuild.buildroot.net/results/8b79a1e597d5f3f73d0534e90523692c6b8a18d2/
http://autobuild.buildroot.net/results/c0ee1bf791618da28e9faec9db30e7114475f211/
http://autobuild.buildroot.net/results/b44383ca61e1a235ba2be0f2d2e47ced7391798a/
http://autobuild.buildroot.net/results/ddee04bf8c477fb8ae6fe7f15b112485797dd96e/

fix partially :

http://autobuild.buildroot.net/results/33748edd3555296c1d8a8b49c83bca6003781993/
http://autobuild.buildroot.net/results/eb099fcfc5f98daa9c06e24d45dbc97df19b0d4f/
http://autobuild.buildroot.net/results/4469c2ba7e4611c7841ebac980a2fb17720b0433/
http://autobuild.buildroot.net/results/71960b94efdfc3c00990315ceac9b6e871a1f076/
http://autobuild.buildroot.net/results/a40486033a535dd11a90eceec80b41fd350bb394/
http://autobuild.buildroot.net/results/415d8475aadfd2bdd8feeac973c11b51b0bd1f80/
http://autobuild.buildroot.net/results/11024182df6b604d558cba07f2f3898edcdf5a06/
http://autobuild.buildroot.net/results/c6c1bb98eb42ee1d23183e275ea02343b5753846/
http://autobuild.buildroot.net/results/046a0ecc4b6b0d6df60b4ce3e279123e1c15123f/
http://autobuild.buildroot.net/results/6cc844e7bd7761e2238f55989e71520b91dd2a9f/
http://autobuild.buildroot.net/results/c958d9a11a3ec1c6c3a5792193c3b0df0529c812/
http://autobuild.buildroot.net/results/c3f186ebfef53e60dc2c4c7c8a27e6b3771f001a/
http://autobuild.buildroot.net/results/1399e81c32ea60f9ba37e9b5c1e6bf07541fba81/
http://autobuild.buildroot.net/results/8e038ac8b6e43abc4c7d437f267e497d7f808926/
http://autobuild.buildroot.net/results/f7b4fb5d8a65ecf45361cded2572eaeba1ea4302/
http://autobuild.buildroot.net/results/aa08901dfd81f6845dc632462bc32d921130c6b1/
http://autobuild.buildroot.net/results/30d4322e0a9fc3d8b04c1ad779e63cf17196fd5c/
http://autobuild.buildroot.net/results/1402de7f2d7ab36bf92cf5abdcb7ae90f0d13604/
http://autobuild.buildroot.net/results/5b644a88f457a8c5dda7961c8ae8b7299a774e8f/
http://autobuild.buildroot.net/results/c264cc6a5266b2703e58e0222f5b79d67ca3d450/
http://autobuild.buildroot.net/results/6b5cbf50a9c7c1f275ad876c77892c72358f23b5/
http://autobuild.buildroot.net/results/8ee3626e2523032218e17d9e6644403e64768f2d/
http://autobuild.buildroot.net/results/ae330a15a579b0d26d801ebbeaa1abcd896df177/
http://autobuild.buildroot.net/results/c814797698adb4bba7044c2edbe0ca3655d94fd5/

Signed-off-by: Gwenhael Goavec-Merou <gwenhael.goavec-merou@trabucayre.com>
Tested-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

gvfs: bump to version 1.28.0

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

libgtk2: bump to version 2.24.30

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

libsoup: bump to version 2.54.0

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

glib-networking: bump to version 2.48.0

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

vala: bump to version 0.32.0

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

cairo: bump to version 1.14.6

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

gtkmm3: bump to version 3.18.0

Required to nicely match the previous libgtk3 major version bump.

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

libgtk3: bump to version 3.18.9

Mark the wayland backend as good again since the bump and consequent
protocol version match fixes it.

Drop upstream 0004-Fix-undefined-reference-to-get_xkb.patch

Drop unnecessary 0005-do-not-build-extract-strings.patch
(extract-strings doesn't exist any more).

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

c-ares: bump to version 1.11.0

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

lvm2: bump version to 2.02.147

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

package/prboom: fix typo in help text

rendring -> rendering

Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/prboom: disable for Sourcery CodeBench SH 2012.09

Building prboom with the external Sourcery CodeBench SH 2012.09 toolchain
fails with an internal compiler error:

r_fps.c: In function 'R_StopAllInterpolations':
r_fps.c:296:1: internal compiler error: Segmentation fault

Make the toolchain unavailable for building prboom.

Fixes:
http://autobuild.buildroot.net/results/2f1/2f14075e22b420cc3fdaaa2231494f2829f040ec/
http://autobuild.buildroot.net/results/d30/d3030fea0b3b8693be640754774e50ac064f8c89/
http://autobuild.buildroot.net/results/b5b/b5b90a03fc47f3771803cb474e0ea9671b5c4667/

Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linux-headers: bump 3.4.x series

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

ntfs-3g: bump version to 2016.2.22

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

libqmi: bump to version 1.14.0

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

dfu-util: Bump version to 0.9

Signed-off-by: Julian Scheel <julian@jusst.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

dt: add patch to fix failures on newer toolchains

On newer toolchains (glibc >= 2.20) _BSD_SOURCE behaviour was deprecated
in favour if the _DEFAULT_SOURCE macro. See man 7 feature_test_macros.
Add patch from Fedora to also consider _DEFAULT_SOURCE. Fixes:
http://autobuild.buildroot.net/results/9e2/9e2126b0e68d0d59d37616a268adb810efd8281a/

Signed-off-by: Gustavo Zacarias <gustavo.zacarias@free-electrons.com>
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>

cgroupfs-mount: new package

This package consists of scripts that setup cgroups at boot without
doing any cgroup management or classification of tasks into cgroups

Signed-off-by: Niranjan Reddy <niranjan.reddy@rockwellcollins.com>
[Thomas:
 - rename to cgroupfs-mount to match upstream
 - add proper hash, since hashes should be added for github sourced
   packages
 - fix minor typos in the init script
 - fix the license file information.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

protobuf: fix detection of __atomic_*() built-ins

- Use the recently introduced BR2_TOOLCHAIN_HAS_ATOMIC boolean.

- Import an upstream patch to fix error handling when atomic operations
  are not detected. Without this patch the build fails due to a syntax
  error instead of showing the proper message.

- Add a patch to configure.ac to check if libatomic is needed and force
  linking to it (we will attempt to submit this upstream).

- Disable build for SPARC64 because it fails due to a missing definition
  of Atomic64.

On PowerPC, the __atomic_*() built-ins for 1-byte, 2-byte and 4-byte
types are available built-in. The corresponding built-ins for 8-byte
types, however, are implemented via libatomic, so requiring gcc >= 4.8.

In Buildroot, to simplify things, it was decided to require gcc 4.8 as
soon as the architectures has at least one __atomic_*() built-in variant
that requires libatomic.

Since protobuf most likely only uses the 1, 2 and 4-byte variants, it
*could* technically build with gcc 4.7. This is probably not a big deal,
and we can live with requiring gcc 4.8 on PowerPC to build protobuf. The
same restriction applies to SPARC.

The build for SPARC64 breaks even using the master branch of protobuf
due to undefined references to some NoBarrier_Atomic*() functions.

Signed-off-by: Henrique Marks <henrique.marks@datacom.ind.br>
Signed-off-by: Carlos Santos <casantos@datacom.ind.br>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

cairo, icu, webkitgtk24: use BR2_TOOLCHAIN_HAS_LIBATOMIC

This commit modifies the cairo, icu and webkitgtk24 packages to use
BR2_TOOLCHAIN_HAS_LIBATOMIC when appropriate.

Fixes:

  http://autobuild.buildroot.net/results/ec4/ec4e48c0e4b8fa72d8bb7ef4ad67a166699c0b62/

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

toolchain: introduce BR2_TOOLCHAIN_HAS_LIBATOMIC

Until now, we were assuming that whenever you have gcc 4.8, libatomic
is available. It turns out that this is not correct, since libatomic
will not be available if thread support is disabled in the toolchain.

Therefore, __atomic_*() intrinsics may not be available even if the
toolchain uses gcc 4.8.

To solve this problem, we introduce a BR2_TOOLCHAIN_HAS_LIBATOMIC
boolean, which indicates whether the toolchain has libatomic. It is
the case when you are using gcc >= 4.8 *and* thread support is
enabled. We then use this new BR2_TOOLCHAIN_HAS_LIBATOMIC to define
BR2_TOOLCHAIN_HAS_ATOMIC.

As explained in the comment, on certain architectures, libatomic is
technically not needed to provide the __atomic_*() intrinsics since
they might be all built-in. However, since libatomic is only absent in
non-thread capable toolchains, it is not worth making things more
complex for such seldomly used configuration.

Note that we are introducing the intermediate
BR2_TOOLCHAIN_HAS_LIBATOMIC option because it will be useful on its
own for certain packages.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
[Thomas: improve Config.in comment using a suggestion from Yann.]

qt5: bump version to 5.6.0

Remove upstream committed patches (qt5base):
  - 0001-Force_egl_visual_ID_33.patch (see [1], [2])
  - 0003-xcb-egl-fixes.patch (see [3])
  - 0005-forkd-disable-eventfd-for-uclibc.patch (see [4], [5])
  - 0009-fix-eglfs-for-sunxi-mali.patch (see [6])
  - 0014-linux-oe-g-Invert-conditional-for-defining-QT_SOCKLE.patch (see [7])

Remove MX6_WORKAROUND from qt5base.mk since we do not need to tweak
EGL visual ID to 33.

Remove upstream committed patch (qt5webchannel):
  - 0001-Fix-builds-without-qml-module.patch (see [8], [9])

Fix Qt5Enginio to 1.6.0 version (see [10])

[1] https://bugreports.qt.io/browse/QTBUG-44290
[2] https://codereview.qt-project.org/#/c/108849/
[3] https://codereview.qt-project.org/#/c/115438/
[4] https://bugreports.qt.io/browse/QTBUG-47337
[5] https://codereview.qt-project.org/#/c/121988/
[6] https://codereview.qt-project.org/#/c/125837/
[7] https://codereview.qt-project.org/#/c/126725/
[8] https://bugreports.qt.io/browse/QTBUG-47360
[9] https://codereview.qt-project.org/#/c/122086/
[10] https://bugreports.qt.io/browse/QTBUG-50111

Signed-off-by: Julien Corjon <corjon.j@ecagroup.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/qt5webkit-examples: remove obsolete package

Signed-off-by: Julien Corjon <corjon.j@ecagroup.com>
[Thomas: add removed option to Config.in.legacy.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/qt5webkit: remove obsolete package

Signed-off-by: Julien Corjon <corjon.j@ecagroup.com>
[Thomas: add removed option to Config.in.legacy.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/qt5quick1: remove obsolete package

Signed-off-by: Julien Corjon <corjon.j@ecagroup.com>
[Thomas: add removed option to Config.in.legacy.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

zeromq: add explicit --without-pgm when pgm is disabled

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

zeromq: add norm support

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
[Thomas: add explicit --without-norm.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

norm: new package

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
[Thomas:
 - remove useless parenthesis enclosing waf calls
 - move the header copying outside of the 'cd $(@D)'.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/luajit: build host variant dynamically

While building libevas.so for the host-efl with luajit support, the link fail
when libluajit is build statically:

/usr/bin/ld: output/host/usr/lib/libluajit-5.1.a(ljamalg.o): relocation R_X86_64_32S against `.rodata' can not be used when making a shared object; recompile with -fPIC
output/host/usr/lib/libluajit-5.1.a: error adding symbols: Bad value
collect2: error: ld returned 1 exit status

Build luajit dynamically in order to build host-efl with luajit support.

Also, passing HOST_LDFLAGS is needed so that a proper rpath is added
to the resulting binaries. Note that it is expected to have
HOST_LDFLAGS assigned to TARGET_LDFLAGS, it's just that luajit has a
somewhat non-conventional naming of variables. This is needed to avoid
the following error from the rpath sanity check:

***
*** ERROR: package host-luajit installs executables without proper RPATH:
***   output/host/usr/bin/luajit

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

icu: remove dev files

Remove files that aren't used for runtime purposes.

Size savings delta +450 KB.

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

connman-gtk: new package

It was added to the networking category close to connman for the ability
to tweak connman features without going around (i.e. add support for
wifi and so on).

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

libglib2: improve dev file handling

Remove source schema files from target
(/usr/share/glib-2.0/schemas/*.dtd and *.xml) at target finalization
time.
This is because other packages install schemas as well to avoid
duplicating file removal efforts all around.

Add a schema compiler run on target finalization as well - we previously
didn't do this since not many packages use it, however the upcoming
connman-gtk package does require a valid gschemas.compiled to work
properly.
We also do this at target finalization to avoid each package handling
this on their own.

Purge some extra binaries and files from the target, like the schema
compiler to save additional space since they're unused at run time.

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

Revert "package/gcc: fix libgcc build for xtensa"

This reverts commit c44cf2cc97df1729577263d2b0cb5263e8cdc4c1.
Now that xtensa gas don't move literals into .init and .fini this fix is
no longer needed.

See upstream commit
https://sourceware.org/git/?p=binutils-gdb.git;a=commitdiff;h=4111950f363221c4641dc2f33bea61cc94f34906,
which was backported to all supported binutils version, under the name
*-xtensa-fix-.init-.fini-literals-moving.patch.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
[Thomas: add more details in the commit log, as suggested by Yann
E. Morin, and using information provided by Max.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/pax-utils: switch to autotools

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[yann.morin.1998@free.fr: independently done a patch similar to the one
 by Thomas]
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linux: build and install kernel selftests

This patch adds the ability to compile and install the kernel
selftests into the target at /usr/lib/kselftests. The rationale behind
/usr/lib is that the selftests have subdirectories where they are
installed which makes them unsuitable to be placed in /usr/sbin as
this would result in /usr/sbin/kselftests/x/y/z. While the selftests
aren't libraries either, they don't achieve much as a standalone
binary so they can be considered to be a 'library of tests' making
/usr/lib sensible.

The selftests require that the kernel headers be installed into the
kernel build tree as some of the selftests have a hardcoded CFLAGS to
include kernel headers (CFLAGS += -I../../../../usr/include/). This is
most easily achieved by using the make ... headers_install inside the
kernel build dir.

This is likely to be a rarely used debugging/performance feature for
development and unlikely to be used in a production configuration.

Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Tested-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Thomas:
 - remove bash as a build dependency, it is only a runtime dependency.
 - fix typo in the Config.in help text, and rewrap
 - add missing 'depends on BR2_USE_MMU' dependency for the comment.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/jack2: add comment about celt and eigen

The optinal dependency on celt is not added, because we only have
celt-0.5.1 and jack2 requires celt >= 0.5.2 (even though it does look
for celt >= 0.5.0, it does not build with celt- < 0.5.2). Since we
cannot upgrade celt (for now), we just never make jack2 depend on
celt051, and it won't find it either, as it just searches for celt.pc
and we have celt051.pc.

As well, the dependency on eigen is only useful in conjunction with
gtkiostream, for which we do not have a package. So, we don't need to
depend on eigen.

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

jack2: allow to choose the mode to operate jackd

jack2 allows to use three modes:
- A legacy jackd service alone.
- A DBUS jackd service alone.
- A mixture between the standalone jackd and the DBUS jackd.

This patch adds the possibility to choose between these three modes,
while the legacy jackd was the only option before.

In addition, the jack_control tool is removed when the DBUS jackd is not
available, as this tool is used to control it.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
[yann.morin.1998@free.fr:
  - rebase
  - only use two booleans, not a choice
  - python is a runtime-only dependency
  - use python3 if enabled, fallback to python
  - simplify post-install condition
]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Thomas: fix minor typo.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/multicat: fix HAVE_CLOCK_NANOSLEEP with uClibc

multicat uses clock_nanosleep() conditionally if HAVE_CLOCK_NANOSLEEP is
defined. Unfortunatly, multicat does not use autoconf to check if
clock_nanosleep() is really available, but defines it if __APPLE__ is not
defined.

With uClibc clock_nanosleep() is not available if the C library is configured
without NPTL support (sets __UCLIBC_HAS_THREADS_NATIVE__) or
__UCLIBC_HAS_ADVANCED_REALTIME__ is not set.

Fixes:
http://autobuild.buildroot.net/results/84f/84f8302ee94e6e6849f6b26cb697c2f9986bafe5/
http://autobuild.buildroot.net/results/9e0/9e07a98e87755c992254295c46edb0dace967a21/
http://autobuild.buildroot.net/results/4b0/4b07d276d7c09f2fc2d63c8c37c92519cfe36dbc/
http://autobuild.buildroot.net/results/84f/84f8302ee94e6e6849f6b26cb697c2f9986bafe5/
and many more.

Upstream status: Pending
https://mailman.videolan.org/pipermail/multicat-devel/2016-March/000129.html

Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

weston: bump to version 1.10.0

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

wayland-protocols: new package

For wayland 1.10+ they're split from wayland core and required by
weston.
Hide it behind the wayland package since it's useless independently even
though it can be "built" and installed.
And it makes no sense to install to target since they are DTDs used at
build-time.

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

wayland: bump to version 1.10.0

wayland-scanner host/cross support is now native so drop the funky
trick.

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

libepoxy: move outside x11r7 scope

Move it outside x11r7 scope since it can be built without it with a few
patches (sent upstream via github pull request).

Motivation is simple, it's a dependency for newer gtk3 versions which
can work with a wayland or broadway backend, and having a full x11 stack
is pointless for that scenario.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
[Thomas: split the first patch into four separate patches, since
that's how they were submitted upstream.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/mesa3d: depends on BR2_TOOLCHAIN_HAS_SYNC_1

Fixes
http://autobuild.buildroot.net/results/399/399d76ac70c3933604067e2c6dcb4086b50acfff/
http://autobuild.buildroot.net/results/70c/70cd299600985d7cf00c5b449f7e75284160551a/
http://autobuild.buildroot.net/results/b4a/b4ab89c3628dd648683e25f9e3c6238f13871aba/
http://autobuild.buildroot.net/results/33d/33d6dc8e1f57b5a02b07c50e5cefa272c6d992fc/
http://autobuild.buildroot.net/results/7f3/7f32a9577ee2e50af38d4f60de50d938792b66a3/
http://autobuild.buildroot.net/results/c98/c98b74d1d1342ca18fb161d010e94d45462ac6f1/
and others

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

xdriver_xf86-video-ati: needs dri3proto for glamor

We need to match xserver here, since we need glamor support there in
order to enable it here, hence need to match the dependencies.
For xserver it needs dri3proto+libepoxy, however if only libepoxy is
present then glamor would be disabled for the server but enabled for the
driver, leading to build failure.

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

arch/arm: add Cortex-M4 entry

This commit adds the option to select the Cortex-M4 ARM core, in the
same family as Cortex-M3. This will be useful to enable the internal
toolchain backend for this ARM core, and provide some defconfigs for
Cortex-M4 platforms.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

arch/arm: Cortex-M3 provides only Thumb-2

The Cortex-M cores only support Thumb-2, not Thumb. In fact, Thumb-2
is a superset of Thumb, and we could have a single option for both in
Buildroot, since -mthumb on ARMv4/v5 means original Thumb, while
-mthumb on ARMv7 means Thumb 2. However, for clarity, it makes sense
to have two separate options. But in this case, Cortex-M3 should not
advertise that it supports Thumb, as in fact selecting Thumb would
generate Thumb-2 code.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

arch/arm: introduce and use BR2_ARM_CPU_ARMV7M

All ARM cores should select a BR2_ARM_CPU_* option. Currently, the
cortex-m3 does not, which this commit fixes by introducing a
BR2_ARM_CPU_ARMV7M option.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/minidlna: unavailable for static builds

minidlna has long standing issues with static-only builds:
http://patchwork.ozlabs.org/patch/501844/
http://patchwork.ozlabs.org/patch/506310/

This patch fixes
http://autobuild.buildroot.net/results/d73/d7370d0f8d1c7c028996feab0ea4f36edbf8a698/
http://autobuild.buildroot.net/results/575/5752be7c1820579425ba24178482ab0d6eea7b62/
"pack_utils.c:(.text+0x9a8): multiple definition of `free_metadata'"

The multiple definition can be fixed of course but instead of digging
deeper and deeper into this mess let's disable static builds.

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

lirc-tools: needs host-python3

As stated in NEWS "Build scripts now requires python3". Fixes:
http://autobuild.buildroot.net/results/f41/f417a17ad24eec8f72402a634588c6b3af8293be/

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

qemu-m68k: add new board

Ethernet is not working, but at least you get a shell and
can test applications for m68k.

Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

arch/m68k: re-enable the architecture

This allows to build a m68k toolchain with uClibc.

Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/multicat: add patch to fix musl build

Add a patch to fix two musl build issues:
  1) `struct udphdr` requires definition of `_GNU_SOURCE`
  2) `mode_t` requires including <sys/types.h>

Fixes:
http://autobuild.buildroot.net/results/9f1/9f1a3f3aab854f1bd20f1064391e613457312022/

Upstream status: Pending [1]

[1] https://mailman.videolan.org/pipermail/multicat-devel/2016-March/000126.html

Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/multicat: add patch to fix static build

Building multicat in a static context fails with:
  aggregartp.c:381: undefined reference to `pthread_self'

Fix this by adding -lpthread to LDLIBS.

Fixes:
http://autobuild.buildroot.net/results/e69/e699cf6b97bc97f2e8e51aa8b86f0e16af70b5b0/
http://autobuild.buildroot.net/results/201/2019f11dcc43cb0db13ceb4e3a575c1a91573d4c/
http://autobuild.buildroot.net/results/fa9/fa98afec377a1e8fce64af5897d1f7af45804069/
and maybe more.

Upstream status: Pending [1]

[1] https://mailman.videolan.org/pipermail/multicat-devel/2016-March/000125.html

Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

binutils: disable plugins for static links

Plugins want to use dlopen even when configured with --disable-shared.
Add --disable-plugins when STATIC_LIBS is selected.
Fixes:
  http://autobuild.buildroot.net/results/a59e5858d777dcee3d8f7fb6e82ee4735b520155/

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

package/dovecot-pigeonhole: bump version to 0.4.13

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/dovecot: bump version to 2.2.22

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/libva-intel-driver: bump version to 1.7.0

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/libva: bump version to 1.7.0

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/kodi: needs host gcc >= 4.6

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

Config.in: add symbols for BR2_HOST_GCC_AT_LEAST_4_6

This option is needed to fix the Kodi build with older host gcc
versions.

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/kodi: fix segfault during playback of mpeg2 content using vaapi

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

dropbear: bump to version 2016.73

some new runtime options, minor fixes, and fixes for issues found by
various code analyze and lintian tools.

Signed-off-by: Alexander Dahl <post@lespocky.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>