lttng-libust: enable Python agent when Python is available

Add the --enable-python-agent configure option to build the LTTng-UST
Python agent when Python 2 or Python 3 is also part of the BR image. The
LTTng-UST Python agent is a Python package which adds LTTng tracing to
Python's standard logging when imported.

Disable the Python agent otherwise.

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

lttng-libust: bump to version 2.8.1

The two current patches are removed because they were merged upstream.

The new patch is needed to build with uClibc. It was also submitted
upstream, but not merged as is yet. Defining NT_GNU_BUILD_ID to 3 when
it's not defined seems like it's something done in other projects. This
definition's value should never change (number of the build ID note
section in ELF objects).

LTTng-UST now has its man pages written in AsciiDoc and they get
converted to troff at build time. This is not needed in a Buildroot
image anyway.

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

configs/imx28evk: Generate the standard sdcard.img

Generate the standard sdcard.img format that can be directly copied
to the SD card.

Remove the custom create-boot-sd.sh script and update the readme.txt
file accordingly.

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
[Thomas: further tweaks to readme.txt.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

postgresql: bump version to 9.6.1

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

nano: bump version to 2.7.1

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

git: bump version to 2.10.2

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

linux: bump default to version 4.8.6

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

linux-headers: bump 4.{4, 8}.x series

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

package/opencv: convert diff patch to git patch

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

package/opencv: update sysctl patch

Fixes:
  http://autobuild.buildroot.net/results/845/845f5e906fda3e87a2fc439a62baede44842283b/

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

package/xen: fix build failure on x86 after bump to 4.7

Since the bump to 4.7, Xen fails to build because of a double definition
of __OBJECT_FILE__. This is due to (who would have guessed) the weirdness
of their buildsystem.

Fix that with a dirty hack: undefine the macro before defining it.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Alistair Francis <alistair.francis@xilinx.com>
[Alistair: Tested on ARMv7 (little endian) boot.]
Tested-by: Alistair Francis <alistair.francis@xilinx.com>
Reviewed-by: Alistair Francis <alistair.francis@xilinx.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/uboot: define help commands when using Kconfig

Expose the Kconfig-related help commands when U-Boot uses the Kconfig
configuration. Like others (like Linux, barebox, busybox), only expose
menuconfig, not all the configurators.

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

lighttpd: bump version to 1.4.43

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

package/mysql: fix SITE url, add sha256 hash

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

trinity: fix Config.in depends on logic

The comment from trinity Config.in currently does not show up in
menuconfig for glibc or uClibc toolchains with older headers.
This commit fixes the logic so that it shows up when either one of the
dependencies from trinity are not met.

Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

domoticz: bump to version 3.5834

Signed-off-by: Fabrice Fontaine <fabrice.fontaine@orange.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/leafnode2: fix SITE url

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

motion: fix static linking of mysql with enabled libz library

Fixes ([1]):

  checking for mysql_init in -lmysqlclient... no
  configure: error: MySQL support can't build without MySQL libraries

[1] http://autobuild.buildroot.net/results/6cf/6cf38698a2a9a041ff375c01ae8898eace17cfb2

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

qt: use $(TARGET_MAKE_ENV) when calling $(MAKE)

We drop the usage of the QT_CONFIGURE_ENV variable, because it isn't
defined anywhere, so it's always empty.

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

fmlib: use $(TARGET_MAKE_ENV) when calling $(MAKE)

And drop FMLIB_MAKE_ENV which is non-existant as pointed by Arnout.

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

dtc: use $(TARGET_CONFIGURE_OPTS) when calling $(MAKE)

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

perl: use $(HOST_MAKE_ENV) when calling $(MAKE)

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

perl: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

libselinux: use $(HOST_MAKE_ENV) when calling $(MAKE)

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

libselinux: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

lzip: use $(TARGET_MAKE_ENV) for configure as well

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

lzip: use $(HOST_MAKE_ENV) when calling $(MAKE)

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

cryptopp: use $(HOST_MAKE_ENV) when calling $(MAKE)

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

gupnp-av: bump to version 0.12.10

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

gssdp: bump to version 1.0.1

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

gupnp: bump to version 1.0.1

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

mobile-broadband-provider-info: bump to version 20151214

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

json-glib: bump to version 1.2.2

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

mfgtools: add gcc version dependency

MfgTools requires gcc >= 4.8 since it needs C++11 support.

Fixes the following autobuild issues:
http://autobuild.buildroot.net/results/c8cbc4e5296af91be5fa8b17871bedf08b6a2697
http://autobuild.buildroot.net/results/fadabb38d8dde6f2be85fbc286b060e2f3e82f65

Signed-off-by: Gary Bisson <gary.bisson@boundarydevices.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

ezxml: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

redis: bump to version 3.2.5

Adjust default config patch to match new major version.
localhost binding is now default so drop it, 300s keepalive is now
default as well (instead of 60s from the previous patch and 0 by
default) - which sounds reasonable so keep it.

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

linux: bump default to version 4.8.5

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

linux-headers: bump 4.{4, 8}.x series

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

moarvm: bump to version 2016.10

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

package/ffmpeg: bump version to 3.2

Support for libfaac was removed upstream:
https://github.com/xbmc/FFmpeg/commit/dc0f711459e0c682bf9f94ba38d26736e90cff45

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

package/x264: bump version

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

package/tvheadend: bump version

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

package/dtv-scan-tables: bump version

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

package/x11r7/xdriver_xf86-input-libinput: bump version to 0.22.0

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

package/x11r7/xlib_libXi: bump version to 1.7.8

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

network-manager: remove bogus --mandir option

The network-manager.mk passes --mandir=$(STAGING_DIR)/usr/man/, which
causes man pages to be installed in
$(STAGING_DIR)/$(STAGING_DIR)/usr/man and
$(TARGET_DIR)/$(STAGING_DIR)/usr/man.

This is definitely bogus and causes some build failures thanks to the
new check added by Yann E. Morin. This option was in
network-manager.mk since the addition of the package. Maybe it was
fixing a bug in the package logic back at the time, but nowadays, it
causes an installation at the wrong location. Without this commit, the
man pages are properly installed at the right location.

Fixes:

  http://autobuild.buildroot.net/results/92683749b0b8e726bc83781ca6d8e103428362ba/

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

package/x265: bump version to 2.1

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

bellagio: add patch to fix installation path of documentation

Fixes:

  http://autobuild.buildroot.net/results/c026551c6b3483d98dd2507c7529a0be8ff79007/

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

package/libglob: fix license typo

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

adwaita-icon-theme: install to staging

This is required for the upcoming evince package which checks for adwaita
presence via pkgconfig file, which normally resides in staging.

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

microblaze: fix musl support

For musl we need patches for bintils 2.25.1 and 2.26.1.
Binutils 2.27 and gcc 6.2.x does not work for microblaze,
even not for uClibc-ng or glibc.
For gcc 5.4.x the existing patch need reworking so that
musl and uClibc-ng is supported.

Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
[Thomas:
 - Add proper description for the binutils patches
 - Use BR2_microblaze instead of BR2_microblazeel and BR2_microblazebz]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/gpm: indentation cleanup

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

package/postgresql: indentation cleanup

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

ipset: bump to version 6.30

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linux-firmware: Include all bnx2x firmwares

Currently buildroot will create a broken system for any kernel version
not between 3.18 and 4.1, as the bnx2x firmware it includes is only
valid for that version of the upstream driver.

This change modifies the build so that all versions are included as there was a
strong preference for not adding options for each version, nor only supporting
the latest kernel:

On Fri, Feb 26, 2016 at 7:09 PM, Thomas Petazzoni wrote:
> I think adding new config to chose between different versions of a
> given firmware is going a bit too far. If we were to do that for all
> firmwares in linux-firmware, it would really increase the number of
> Config.in options too much.
>
> Shall I suggest to install both versions of the firmware? The firmware
> files are not that large, and if filesystem size is really a strong
> issue, it is always possible to clean up the non-required firmware
> files in a post-build script.

On Wed, Mar 2, 2016 at 7:47 AM, Arnout Vandecappelle wrote:
> No, this is certainly not good, it fixes 4.2+ and simultaneously breaks 4.1-.
>
> What is wrong with Thomas's proposal to remove the unneeded versions in a
> post-build script? If your size is so important that 640K matters, you would
> probably anyway want to keep only one of the six binaries instead of three
> different chip revisions.

It is up to the user to ensure they remove the versions they don't want.

Note that this increases the size of the root file system by about 8MB.

Signed-off-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linux-firmware: Bump version

Signed-off-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

mmc-utils: bump version to 2cb6695

Change log since last version bump:

  2cb6695 Add missing -c for write_reliability
  e094e28 Fix return value check from set_partitioning_setting_completed()
  f90eac4 Add help texts for -c option
  752072d Add -c option for partitioning commands
  55380d9 Register being set is other than mentioned in printed message.
  3c02759 Fix reported offset for enhanced user area start address
  a3d3331 mmc-utils: Add Command Queue fields to Extended CSD
  0ca049f mmc-utils: Add ability to configure write protect on an eMMC device
  bb779ac mmc-utils: Add disable partition boot to 'mmc bootpart enable'
  89cd01e mmc_utils: add ffu support
  4af1749 mmc-utils: Merge the lsmmc tool into mmc-utils
  44f94b9 mmc_utils: doc: add man-page
  928ff07 mmc_utils: gp create: only allow partition [1-4] to be selected

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

toolchain-external.mk: fix ARCH_SUBDIR calculation

ARCH_SUBDIR is computed based on the value of ARCH_SYSROOT_DIR and
SYSROOT_DIR. For nested toolchains ARCH_SYSROOT_DIR is a subdir of
SYSROOT_DIR, so a sed command like this one...

  sed -r -e "s:^${SYSROOT_DIR}(.*)/$:\1:"

...basically removes the leading SYSROOT_DIR part from ARCH_SYSROOT_DIR.

But, for side-by-side sysroot toolchains ARCH_SYSROOT_DIR and
SYSROOT_DIR are at the same level, so the above sed command doesn't
make any effect.

This patch therefore improves the calculation of ARCH_SUBDIR to
clearly handle the three possible cases:

 - There is a single sysroot, or the selected architecture sysroot is
   the main one (i.e SYSROOT_DIR == ARCH_SYSROOT_DIR). In this case,
   ARCH_SUBDIR is empty.

 - There are side-by-side sysroots, such as
   SYSROOT_DIR=.../sysroot/mips-r2-hard/ and
   ARCH_SYSROOT_DIR=.../sysroot/mipsel-r2-hard/.

 - The arch-sysroot is nested, such as SYSROOT_DIR=.../sysroot and
   ARCH_SYSROOT_DIR=.../sysroot/armv4t/

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
[Thomas: improve the logic to handle the SYSROOT_DIR==ARCH_SYSROOT_DIR
case.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

bridge-utils: bump to version 1.6

Drop upstream patches.

Change download and homepage to kernel.org, and use .xz tarball.

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

rabbitmq-server: add missing host dependencies

Fixes:
http://autobuild.buildroot.net/results/0ec785b8fa876d879934af618851d77a3ce70f55
http://autobuild.buildroot.net/results/03eb39eb9498eb42202ac8b16c1ffc175c0b69d3
http://autobuild.buildroot.net/results/199b04408fa8d9c94b9bed42fdcaafb03278f03c

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

rabbitmq-server: DEPENDS->DEPENDENCIES

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

ser2net: bump version to 3.0

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

linux-headers: bump 3.{4, 18}.x and 4.1.x series

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

lft: bump to version 3.77

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

btrfs-progs: bump to version 4.8.2

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

libgtk3: bump to version 3.22.2

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

harfbuzz: bump to version 1.3.3

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

vala: bump to version 0.34.2

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

samba4: bump to version 4.4.7

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

ngrep: don't use pcap_restart

libpcap since version 1.8.1 does not export the pcap_restart symbol. Disable
use of this symbol.

Fixes:
http://autobuild.buildroot.net/results/647/647061a6fe8cffe3b083c911a986fa7683cdb8f6/
http://autobuild.buildroot.net/results/d90/d90e4d7c62e5b860f1f0ffb5e4dc545928f91734/
http://autobuild.buildroot.net/results/560/56076caede5c7bfc8a57da3a9dc47d0d22309e79/

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

toolchain-external: install libmvec.so from glibc toolchains

On some architectures (namely x86-64), glibc may provide a libmvec
library since glibc 2.22, which programs built with gcc OpenMP support
might get linked to.

In order for these programs to work on the target, we need to copy
this library to the target filesystem.

This commit takes care of this for the external toolchain
situation. Note that libraries listed in TOOLCHAIN_EXTERNAL_LIBS are
silently ignored if they don't exist. Therefore, we don't need to have
any condition on the architecture or glibc version.

For more details on libmvec, see
https://sourceware.org/glibc/wiki/libmvec.

Fixes bug #9111.

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

glibc: install libmvec.so when available

On some architectures (namely x86-64), glibc may provide a libmvec
library since glibc 2.22, which programs built with gcc OpenMP support
might get linked to.

In order for these programs to work on the target, we need to copy
this library to the target filesystem.

This commit takes care of this for the glibc package (used for the
internal toolchain backend). Note that libraries listed in
GLIBC_LIBS_LIB are silently ignored if they don't exist. Therefore, we
don't need to have any condition on the architecture or glibc version.

For more details on libmvec, see
https://sourceware.org/glibc/wiki/libmvec.

Fixes bug #9111.

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

package/opencv: Do not add libdl to LINKER_LIBS for static builds

Fixes
http://autobuild.buildroot.net/results/765/7657e01481995a4f0d725467e935a83928a59a04/

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

valgrind: bump version to 3.12.0

See [1] for release notes.

- removed 0003-mips-replace-addi-with-addiu.patch (applied upstream, see [2])

[1] http://valgrind.org/docs/manual/dist.news.html
[2] https://bugs.kde.org/show_bug.cgi?id=356112

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

core/pkg-generic: check proper package installation

Some packages misbehave, and install files in either $(STAGING_DIR)/$(O)
or in $(TARGET_DIR)/$(O) .

One common reason for that is that pkgconf now prepends the sysroot path
to all the paths it returns. Other reasons vary, but are mostly due to
poorly writen generic-packages.

And a new step hooks to check that no file gets installed in either
location, called after the install-target and install-staging steps.

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

package/xdriver_xf86-input-synaptics: fix install paths

Get rid of the borked sdkdir variable.

Fixes (partially):
    https://bugs.busybox.net/show_bug.cgi?id=8696

Reported-by: buildroot@netsolux.ch
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: buildroot@netsolux.ch
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/xdriver_xf86-input-mouse: fix install paths

Get rid of the borked sdkdir variable.

Fixes (partially):
    https://bugs.busybox.net/show_bug.cgi?id=8696

Reported-by: buildroot@netsolux.ch
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: buildroot@netsolux.ch
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/xdriver_xf86-input-libinput: fix install paths

Get rid of the borked sdkdir variable.

Fixes (partially):
    https://bugs.busybox.net/show_bug.cgi?id=8696

Reported-by: buildroot@netsolux.ch
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: buildroot@netsolux.ch
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/xdriver_xf86-input-joystick: fix install paths

Get rid of the borked sdkdir variable.

Fixes (partially):
    https://bugs.busybox.net/show_bug.cgi?id=8696

Reported-by: buildroot@netsolux.ch
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: buildroot@netsolux.ch
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/xdriver_xf86-input-evdev: fix install paths

Get rid of the borked sdkdir variable.

Fixes (partially):
    https://bugs.busybox.net/show_bug.cgi?id=8696

Reported-by: buildroot@netsolux.ch
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: buildroot@netsolux.ch
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

perl: update cnf/config.{sub, guess} using autotools hook

perl may fail to build on newer host architectures such as aarch64 due
to its inability to guess the host architecture to build 'miniperl',
which is built during the process. An error message looks like this:

  Configuring build-time miniperl for unknown
  ERROR: No $target defined (?!)
  ERROR: configure --mode=buildmini failed

This happens because the config.sub and config.guess files from perl are
not modified and may become outdated.
These files are normally updated automatically by a Buildroot hook for
autotools packages, to avoid problems like these.
Although perl uses the config.sub and config.guess files, it is not a
strict autotools package, so it is not defined as an "autotools-package"
in Buildroot and so it doesn't inherit the hook.
This commit makes perl borrow the hook from the autotools infrastructure
so that it can build on newer build architectures.
This has been tested by building it on an aarch64 host machine.

Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Acked-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

olsr: disable in musl configurations

olsr doesn't build on musl, so let's disable it. The issue has been
reported upstream at https://github.com/OLSR/olsrd/issues/6.

Fixes:

  http://autobuild.buildroot.net/results/74cff8fc87d95d70840394fc049fc038797b9888/

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

olsr: bump to 0.9.0.3

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

libpcap: bump to version 1.8.1

Drop upstream patch.

Also remove ac_cv_linux_vers caching since this check was remove in upstream
commit f3e011b18c55ae.

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libpcap: drop libnl headers patch

We can avoid host headers using a --with-libnl parameter to override the
default host headers path. The configure script adds /include suffix to the
include path. So we still need to manually add -I with the correct path to
CFLAGS because libnl headers are under the libnl3/ directory.

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

DEVELOPERS: add entry for mfgtools

Signed-off-by: Gary Bisson <gary.bisson@boundarydevices.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

mfgtools: new package

This package contains the Freescale manufacturing tool.
It is designed to program firmware to i.MX boards during production.
The communication is done over USB using the Freescale UTP protocol.

The project is maintained on NXPMicro Github repository:
https://github.com/NXPmicro/mfgtools

Signed-off-by: Gary Bisson <gary.bisson@boundarydevices.com>
[Thomas:
 - remove dependency on host-cmake, this is handled by the
   host-cmake-package infrastructure
 - add quotes around $(HOSTCXX)
 - add CPOL.htm to <pkg>_LICENSE_FILES.
 - tweaks to the readme.txt file.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

kvm-unit-tests: add gcc version dependencies

On x86 (both i386 and x86_64), the kvm-unit-tests code uses
__builtin_reachable, which appeared in gcc 4.5. This commit adds the
relevant dependencies to kvm-unit-tests:

 - on i386, we need gcc 4.5 as the target compiler

 - on x86-64, we need gcc 4.5 as the host compiler, because we use the
   host compiler to build kvm-unit-tests for this architecture

Fixes:

  http://autobuild.buildroot.net/results/21340a7e6494e23adb2fe8dc6253ab220c380fec/

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

Config.in: add BR2_HOST_GCC_AT_LEAST_4_5

In preparation to the addition of new dependencies in the
kvm-unit-tests package, add the BR2_HOST_GCC_AT_LEAST_4_5 symbol.

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

python-pyqt5: fix the build with Qt 5.6.2

This commit adds a patch to the python-pyqt5 package that fixes the
build with Qt 5.6.2.

It has been verified to fix:

  http://autobuild.buildroot.net/results/8ca61ec09cf0e6577d72fe5d809867aa284c8f1e/

And it should also fix:

  http://autobuild.buildroot.net/results/4e0c5776264a7382c842f04e333fe4ded7d14f8a/
  http://autobuild.buildroot.net/results/0306710e0f39591dba74f24aff40f7f094bcfa6e/

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

kvm-unit-tests: Bump version

This version bump includes several fixes for x86 and a significant fix
to powerpc transactional memory tests.

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

kvm-unit-tests: Fix x86_64 to use host compiler

kvm-unit-tests requires a multilib compiler for x86_64 as it compiles
32bit boot code.

This patch uses the BR2_HOSTARCH_NEEDS_IA32_COMPILER option to enforce
this and avoids using TARGET_CROSS for x86_64 targets and uses the host
32bit (and 64bit) capable compiler.

Signed-off-by: Cyril Bur <cyrilbur@gmail.com>
[Thomas: reword comment in .mk file.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

board: fix indentation in zybo readme.txt

Signed-off-by: Matt Kraai <kraai@ftbfs.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

tcpdump: bump version to 4.8.1

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

file: bump version to 5.29

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

package/ti-sgx-km: indentation cleanup

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

package/nvidia-driver: indentation cleanup

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

util-linux: fix host build when ncurses is not installed

lib/colors.c uses HAVE_LIBTINFO to decide whether term.h and ncurses.h
should be included. However, the configure.ac check is bogus: it
checks for tinfo using pkg-config, and then if it fails, checks with
AC_CHECK_LIB(). So, if you have tinfo installed, but not the
corresponding development package, the pkg-config test will fail, but
the AC_CHECK_LIB test will succeed, even though the headers are not
available.

To address this, we explicitly tell host-util-linux that tinfo is not
available, like we're already doing for ncurses.

Fixes:

  http://autobuild.buildroot.net/results/3c447e601a7b728860e78cabf2191ab206e6480d/

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

boot-wrapper-aarch64: Allow users to select the PSCI SMP boot method

Signed-off-by: Hollis Blanchard <hollis_blanchard@mentor.com>
[Thomas:
 - remove "default n"
 - pass explicit --disable-psci]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

support/download: make the git wrapper more robust

Currently, there are two failure paths in the wrapper:

  - if the tar fails, then the error is ignored because it is on the
    left-hand-side of a pipe;

  - if the find fails, then the error is ignored because it is a
    process substitution (and there is a pipe, too).

While the former could be fixed with "set -o pipefail", the latter can
not be fixed thusly and we must use an intemediate file for it.

So, fix both issues by using intermediate files, both to generate the
list of files to include in the archive, and generate the archive in a
temporary tarball.

Fixes the following build issue, where the find is failing for whatever
unknown reason:
    http://autobuild.buildroot.net/results/20f/20fd76d2256eee81837f7e9bbaefbe79d7645ae9/

And this one, where the process substitution failed, also for an unknown
reason:
    http://autobuild.buildroot.org/results/018/018971ea9227b386fe25d3c264c7e80b843a9f68/

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