git.libre-soc.org Git - buildroot.git/log

package/libevas: switch dependency from mesa3d to _HAS_LIBGL

We now have a virtual package that represents availability of
full OpenGL.

This should be the end of this dependency hell epic, now. :-)

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Bernd Kuhls <berndkuhls@hotmail.com>
Cc: Paul Cercueil <paul@crapouillou.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/mesa3d: with a DRI driver, is a full OpenGL provider

Make mesa3d a provider for full OpenGL, as soon as a DRI driver is enabled.

Full OpenGL is only possible when:
- a DRI driver is enabled,
- and X.Org is enabled

Since DRI driver in fact depend on X.Org being enabled in the first place,
we can safely declare mesa3d as a full openGL provider as soon as at least
one DRI driver is enabled.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Bernd Kuhls <berndkuhls@hotmail.com>
Cc: Paul Cercueil <paul@crapouillou.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/libgl: introduce as a new virtual package for full-openGL providers

Currently, only mesa3d is such a full-openGL provider.

But it is easy to see a few more cropping up at the door:
- NVidia's binary blob
- ATI's fglrx binary blob

Having this virtual full-openGL package will help when those are getting in.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Bernd Kuhls <berndkuhls@hotmail.com>
Cc: Paul Cercueil <paul@crapouillou.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/xorg-server: do not force-select xf86driproto for mesa3d

This is already handled by mesa3d itself.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Paul Cercueil <paul@crapouillou.net>
Cc: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/mesa3d: only enable full OpenGL when a DRI driver is enabled

Full OpenGL is only provided when:
- a DRI driver is enabled,
- and X.Org is enabled

Since DRI driver depend on X.Org being selected, this means that we
can (have to) enable full OpenGL as soon as a DRI driver is enabled.

On the other hand, Gallium drivers, which currently do enable full
OpenGL, only really provide OpenGL EGL and GLES, not full OpenGL.

So we exclude Gallium drivers when checking whether to enable full
OpenGL in mesa3d.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Paul Cercueil <paul@crapouillou.net>
Cc: Bernd Kuhls <berndkuhls@hotmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/mesa3d: simplify the code enabling full-OpenGL

Currently, full-OpenGL is enabled by checking that at least one driver
is enabled. This is done by checking that the just-constructed lists of
drivers are not empty.

But we already have a variable that is set to 'y' as soon as a driver
is selected:

  - DRI drivers select BR2_PACKAGE_MESA3D_DRI_DRIVER

  - Gallium drivers select BR2_PACKAGE_MESA3D_GALLIUM_DRIVER

  - both BR2_PACKAGE_MESA3D_DRI_DRIVER and BR2_PACKAGE_MESA3D_GALLIUM_DRIVER
    select BR2_PACKAGE_MESA3D_DRIVER

So, BR2_PACKAGE_MESA3D_DRIVER is set to 'y' as soon as at least one
driver is selected. We can use that variable rather than comparing the
lists of drivers.

Also, rearrange the code in a more logical way, by moving the code that
enables OpenGL nearer to the code enabling EGL and GLES.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Paul Cercueil <paul@crapouillou.net>
Cc: Bernd Kuhls <berndkuhls@hotmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/mesa3d: add comment for when OpenGL EGL is not available

OpenGL EGL needs udev support, so add a comment stating so.

Also, reword the OpenGL EGL prompt to be in the same format as the
OpenGL ES prompt (and like the full OpenGL prompt that will shortly
be added, too.)

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Paul Cercueil <paul@crapouillou.net>
Cc: Bernd Kuhls <berndkuhls@hotmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/mesa3d: DRI drivers depend on X.org

It does not make sense to build DRI drivers without X.org.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Paul Cercueil <paul@crapouillou.net>
Cc: Bernd Kuhls <berndkuhls@hotmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/mesa3d: re-arrange menu items

Make mesa3d a menuconfig, ie. a sub-menu on the boolean option.

Move drivers and additional API selections out of their own submenus,
add a comment as separator.

This will make it easier to further re-arrange the config items in
the following patches, to fix-enhance the packaging to handle the
full-openGL provider stuff.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Paul Cercueil <paul@crapouillou.net>
Cc: Bernd Kuhls <berndkuhls@hotmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/x11-video-drivers: do not select, but depends on mesa3d

Currently, X.Org drivers that need mesa3d forcibly select it.

We'll soon switch mesa3d to being a provider of the full OpenGL virtual
package, to come in a later patch.

It is bad practice to select a provider, as it can generate configuration
inconsistencies.

So, switch all X.Org video driver that need mesa3d to actually depend on
it, rather than forcibly select it.

To be noted: xf86-video-mach64 already used a 'depends on' rather
than a 'select'.

At the same time, move the intel drivers dependencies to the top.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Bernd Kuhls <bernd.kuhls@t-online.de>
Cc: Paul Cercueil <paul@crapouillou.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/xdriver_xf86-video-ati: only select libdrm if the DRI driver is enabled

The xf86 radeon driver does not need libdrm, unless DRI is enabled too.

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

package/libdrm: radeon does not depend on x86

Thomas reported that the libdrm-radeon dependency on x86 (32- or 64-bit)
dates back to the time where it also depended on libpciaccess.

Now that the libpciacess dependency has been dropped, Thomas managed to
build a libdrm-radeon on something else than an x86 (an ARM for example).

Make libdrm-radeon no longer depend on x86

As a side effect, this fixes some potential "unmet direct dependencies"
from the xdriver_xf86-video-ati package (which was the initial problem
I was trying to solve.)

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

libglib2: fix compilation for aarch64 and arc

Copied from here - https://github.com/openembedded/oe-core/blob/master/meta/recipes-core/glib-2.0/glib-2.0/0001-gio-Fix-Werror-format-string-errors-from-mismatched-.patch

Initially posted here - http://lists.openembedded.org/pipermail/openembedded-core/2013-October/085174.html

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

kexec: bump to version 2.0.6

Remove two patches which have been included upstream.

Signed-off-by: Sven Neumann <neumann@teufel.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

xbmc: Fix compile error with missing host-gettext

autoreconf -vif lib/cpluff
autoreconf: Entering directory `lib/cpluff'
autoreconf: running: autopoint --force
Can't exec "autopoint": No such file or directory at /home/buildroot/output/host/usr/share/autoconf/Autom4te/FileUtils.pm line 345.
autoreconf: failed to run autopoint: No such file or directory
autoreconf: autopoint is needed because this package uses Gettext
make[1]: *** [lib/cpluff/configure] Error 1

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

qt5script: Check for QtGui before building the examples

Adding an upstream patch to fix an autobuild failure.

qsdbg and marshal examples need QtGui module. If QtGui is not installed
they fail to compile.

Upstream commit:
https://github.com/qtproject/qtscript/commit/e4453c92ee41943a3a1d67a725804a44e747f64e

Fixes:
http://autobuild.buildroot.net/results/0e6/0e63bc604f3ec422996eb0f5fd92f4437954ca04/

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

zeromq: bump to version 4.0.4

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

imagemagick: bump to version 6.8.9-0

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

sqlite: version bump 3.8.4.3

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

nftables: new package

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

libnftnl: 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>

libsigsegv: not available on Microblaze

Fixes:

http://autobuild.buildroot.org/results/dd5/dd583e1c11bba814cd0f124eeeffeca5657a0aa9/

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

package/mesa3d: fix improper dependency of blind option

Currently, the blind option BR2_PACKAGE_MESA3D_DRI_DRIVER depends
on !STATIC.

But this option is also selected by the various DRI drivers, and
none of them currently depend on !STATIC (although there is a comment
stating DRI drivers need !STATIC, there's nothing to enforce that).

So, we could well end-up with an inconsistent configuration, where some
DRI drivers are selected even though STATIC is set.

Enclose all DRI drivers in an 'if !STATIC' condition, remove the
dependency from the blind option, move the comment so it is nearer
the affected drivers, rephrase the comment to match the rules about
dependencies on toolchain features.

[Thomas: really use the right wording for the comment about the
dynamic library dependency.]

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Paul Cercueil <paul@crapouillou.net>
Cc: Bernd Kuhls <berndkuhls@hotmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

liblogging: enable systemd compatibility libs

If systemd is chosen as init system, liblogging will enable the
installation of compatibility libraries in systemd.

When liblogging will be updated to depend on libsystemd.so, this
selection will have to be removed.

Signed-off-by: Eric Le Bihan <eric.le.bihan.dev@free.fr>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

systemd: add option to enable compatibility libs

Since systemd 209, some libraries have been merged into libsystemd.so:

- libsystemd-daemon.so
- libsystemd-id128.so
- libsystemd-journal.so
- libsystemd-login.so

A new configuration menu entry has been added to enable the installation
of compatibility pkg-config files, so that programs that depend on them
can still be built.

[Thomas: as suggested by Yann during the review, add an else clause
with --disable-compat-libs.]

Signed-off-by: Eric Le Bihan <eric.le.bihan.dev@free.fr>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

luajit: needs for 32bit archictectures a 32bit host gcc

For 32 bit archictectures the luajit package needs a host compiler
able to generate 32 bit code using the "-m32" option so check if that
option is supported.

Signed-off-by: Fabio Porcedda <fabio.porcedda@gmail.com>
Cc: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: François Perrad <francois.perrad@gadz.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

toolchain: add a hidden config option to enable the toolchain package

The usual way to enable a package using the package infrastructure is to
use a config option so instead to add the toolchain package to the
TARGETS variable in the Makefile add a config option like all the other
toolchain packages.

[Thomas: remove comment that no longer made sense in the main
Makefile, and add a comment above the new hidden Config.in option to
explain what it is useful for.]

Signed-off-by: Fabio Porcedda <fabio.porcedda@gmail.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

host-libxml2: Prefer python2 when python3 is also installed

Fixes mesa3d build errors:
http://autobuild.buildroot.net/results/d10/d105a0b3ca11fad34f9a2dae0dae9bd041d918a6/
http://autobuild.buildroot.net/results/d71/d7121443715024d15a66ff1abf1261803c10cd35/
http://autobuild.buildroot.net/results/ce6/ce64164d76972f82acab277afc9c95a876c6433e/

checking for python2... python2
checking python2 module: libxml2... no
configure: error: failed to find required module libxml2
make: *** [/home/test/test/1/output/build/mesa3d-10.0.4/.stamp_configured] Error 1

mesa3d needs python2 bindings installed by libxml2. To enforce their build
option BR2_PACKAGE_HOST_LIBXML2_PYTHON is used exclusively by mesa3d so this
patch does not affect other packages.

Without this patch host-libxml2 installs the python bindings in
$(HOST_DIR)/usr/lib/$(PYTHON3_VERSION_MAJOR) using this defconfig:

BR2_PACKAGE_MESA3D=y
BR2_PACKAGE_PYTHON3=y

because $(HOST_DIR)/usr/bin/python points to $(HOST_DIR)/usr/bin/python3:

Quote from host-libxml2 configure log:
Found python in /home/fli4l/br2/buildroot/output/host/usr/bin/python
Found Python version 3.4

HOST_$(PACKAGE)_NEEDS_HOST_PYTHON does not work here because libxml2 does
not use the python-package infrastructure.

libxml2-python2.patch extends the python detection code in host-libxml2 to
first look for python2, to keep the patch small I did not update the indentions:

Found python2 in /home/fli4l/br2/buildroot/output/host/usr/bin/python2
Found Python version 2.7

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Tested-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

Makefile: add PATH to the USER_HOOKS_EXTRA_ENV

This can be useful for post-{build,image} scripts, in case some host-tools
were specifically built to be used by these scripts.

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Cc: "Yann E. MORIN" <yann.morin.1998@free.fr>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

pkg-cmake.mk: add PATH in the configure command environment

Because BR_PATH is not exported in the environment beforehand running
cmake, it is necessary to add it on the cmake configure command.

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

infra: remove unused {TARGET, HOST}_PATH definition

Since the variables TARGET_PATH and HOST_PATH are not used anymore,
let's remove them.

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

*.mk: replace (TARGET|HOST)_PATH by BR_PATH

Thanks to the 2 previous patches of the series, BR_PATH contains
all locations in which host-packages may install programs.

This patch replaces the occurrences TARGET_PATH and HOST_PATH with
BR_PATH, everywhere these variables are used in the *.mk files.

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

Makefile: add $(HOST_DIR)/sbin to BR_PATH

Extend BR_PATH because a few host-packages install programs in this
location.

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

Makefile: introduce BR_PATH

Since the HOST_PATH and TARGET_PATH variables almost contain the same
things, let's factorize this in a single BR_PATH.

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

mtd: fix build against recent uClibc 0.9.33

Commit fae7cc33a1 added a local rpmatch() implementation. However, the
rpmatch() implementation was backported to the 0.9.33 branch, thus breaking
the Buildroot default toolchain. Rename the local rpmatch() to avoid
collision.

Fixes:
http://autobuild.buildroot.net/results/51f/51fe39bd942e4d8d0045ac810708719b9459e21f/

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

host-nodejs: use g++ as linker

Fixes:
http://autobuild.buildroot.net/results/f6b/f6b80ae32acf7f89bb1c12999679ff1be9733d44/

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

agent++: need dynamic library support

Required by SNMP++.

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

snmp++: need dynamic library support

Fixes:
http://autobuild.buildroot.net/results/7b9/7b9ec11514219e5f315df0b30204f431b4c4f8a1/

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

package: add Boehm-Demers-Weiser gc library

This is needed for applications like Zile

[Thomas:
- use $(...) instead of ${...} to reference LIBGC_VERSION
- add missing dependency on libatomic_ops dependencies
- fix DEPENDANCIES -> DEPENDENCIES
- replace += by = in dependencies definition
- add host-pkgconf to the list of dependencies
- fix license, it's a X11-style permissive license
- fix license file, COPYING didn't exist, use README.md instead]

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libatomic_ops: introduce a BR2_PACKAGE_LIBATOMIC_ARCH_SUPPORTS hidden option

In preparation to the introduction of packages that select
libatomic_ops, this commit adds an hidden
BR2_PACKAGE_LIBATOMIC_ARCH_SUPPORTS option which allows to easily
depend on the architectures that libatomic_ops is available for.

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

host-python-pyrex: use the HOST_*_NEEDS_HOST_PYTHON infrastructure

python-pyrex does not support python3.

So, using the *_NEEDS_HOST_PYTHON infrastructure will make sure we won't try
building it for python3. Besides, it will automatically take care of adding the
right host-python package to the dependency list.

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

nodejs: force python interpreter

nodejs build system is based on python, and it does not support python3.

This patch makes sure python2 is used for the build.

Note that, setting PYTHON=... at configure time or in the make
environment is not enough to override all the hard-coded python call, so
we have to sed some python scripts to avoid being screwed.

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

nodejs: use the standard configure/make options

Also cleanup leading whitespace.

Note that overloading LD in required to avoid linker failures.

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

xserver_xorg-server: Compile fix for mips64

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

xserver_xorg-server: Bump version to 1.15.1

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

xproto_xproto: Bump version to 7.0.26

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

flann: new package

[Thomas: changed license from BSD to BSD-3c, to be more specific.]

Signed-off-by: Davide Viti <zinosat@tiscali.it>
CC: minimod@morethan.org
CC: patrickdepinguin@gmail.com
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

jquery-mobile : new package

Since version 1.4 jquery-mobile includes the core-jquery.
No need for extra dependencies.

[Thomas: remove trailing spaces in Config.in file, and rewrap help
text.]

Signed-off-by: Sagaert Johan <sagaert.johan@skynet.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

jquery-ui: switch to sources.buildroot.net

Switch to sources.buildroot.net cached zipfile.
Upstream changed the zipfile and the internal directory structure
without notifying anyone or changing the filename thus causing breakage.
And we can't use github release tarballs since they are not minified.
And we shouldn't reflect this in the install hook since an old cached
tarball wouldn't work, and we'd break anyone using the old package file
with the new one too. Thanks upstream! Fixes:
http://autobuild.buildroot.net/results/43a/43a16a44a26b38f4c6a2f352d8bfe5e07af6b2e3/

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

libcurl: drop polarssl support

As of curl 7.36.0 it doesn't support polarssl < 1.3 any longer. Fixes:
http://autobuild.buildroot.net/results/d82/d82c3618e9dde3da7e36ba2b58545a9a8de5e442/

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

openvpn: bump to version 2.3.3

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

flac: needs wchar

Fixes:
http://autobuild.buildroot.net/results/873/873a1b3551c5fe04fa477fd682ff91fecdd41606/

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

fdk-aac: Disable for MIPS I, II, III and IV

This package uses the 'msub' instruction which is available only in
mips32 and mips64 ISAs.

[Thomas: replicate the new dependency to the comment.]

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

flite: new package

[Thomas: rewrap Config.in help text, and use proper format for wchar
comment according to the Buildroot manual.]

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Cc: Eric Jarrige <eric.jarrige@armadeus.org>
Cc: Julien Boibessot <julien.boibessot@armadeus.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

mutt: Add options for imap/pop3 support, enable optional iconv/OpenSSL support

[Thomas: keep a normal 'config' instead of 'menuconfig' for mutt.]

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

mutt: Bump version to 1.5.23

removed upstream applied patch mutt-02-automake.patch
http://dev.mutt.org/trac/changeset/0488deb39a35

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

openpgm: make its build system python2/python3 friendly

This patch fixes build issue when python3 is enabled, and host python2
interpreter is not available.

Fixes:
http://autobuild.buildroot.net/results/01f/01f886a073439c0639ed93e596b68bcadf1a5824/

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Cc: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

omniorb: force python interperter

Omniorb build-system requires python2 interpreter.

By default, it looks for python program, which may fallback on the system
python interpreter in case python is disabled and python3 is enabled.

So, this patch enforces the python interpreter to python2 built by
Buildroot.

Fixes:
http://autobuild.buildroot.org/results/b4f/b4f3a2602ba0224ac3253c4bf6ed87ec045df772/

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

toolchain-external: add Linaro ARM big endian toolchain

Linaro has started to release ARM big endian toolchains, so we
integrate this toolchain in the external toolchain logic of
Buildroot. Since ARM big endian is probably going to be a lot more
uncommon than ARM little endian, we will only support one version at a
time of this toolchain.

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

toolchain-external: bump Linaro AArch64 toolchains

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

toolchain-external: remove useless info in help text of Linaro ARM toolchains

The information "To use this toolchain, you must disable soft float
usage." which was visible in the help text of Linaro toolchains is no
longer useful, since those toolchains are only visible when the ARM
EABIhf ABI is selected, which by design is not compatible with
soft-float.

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

toolchain-external: bump Linaro ARM toolchains

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

support/graphs: fix comparisons agains None

As Samuel said:
    In Python, None is a singleton, and it is recommended to use "is" or
    "is not" for testing them [1].

    [1] http://legacy.python.org/dev/peps/pep-0008/#programming-recommendations

Reported-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

gd: bump to version 2.1.0

Drop obsolete/applied patches. Refresh the rest, and add sequence numbers.

Add a patch fixing build against uClibc when UCLIBC_HAS_LONG_DOUBLE_MATH is
missing.

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

fswebcam: bump version to 20140113

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

fswebcam: disable static build

fswebcam doesn't play nicely with static build as it doesn't track its
indirect dependencies in link time. Fixing this requires tracking all indirect
optional dependencies, which is tedious and error prone. Just disable static
build.

This build failure didn't show on the autobuilder for some reason.

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

toolchain-external: Introduce kernel headers sanitization

The Nios-II Sourcery external toolchain (the only Nios-II we currently
support) exports broken kernel headers. In particular, these kernels should
be exported using the "headers_install" rule which applies a set of fixes
on the kernel headers so they are suitable for userspace usage.

In order to fix this, add a post-install hook to perform the header fixes
ourselves. The result is equivalent to apply the "headers_install" rule.

Fixes:
http://autobuild.buildroot.net/results/c32/c32ad4bac5f651502e551f7733f702afaa0e742a/

Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

lttng-libust: Disable liblttng-ust-dl with uClibc.

According to uClibc commit [1], dlinfo is not available in recent uClibc (>0.9.33),
but is available in older version of uClibc (<=0.9.33) whith a different prototype
than dlinfo() in glibc.

To be able to use LTTng UST with uClibc, we need to disable the Dynamic Linker
Tracing functionality [2].

A specific test on dlinfo() prototype is performed to enable or disable this
functionality.

This patch supersede the one added in abf58f46cedacace1db53de5cc12a8fe9f7fbd82
that wrongly disable liblttng-ust-dl even if dlinfo() comme from glibc.

Fixes:
http://autobuild.buildroot.net/results/a6c/a6c33dd7ec2a36a50c5ea74b989a371d6c85e899/build-end.log

[1] http://git.uclibc.org/uClibc/commit/?id=f3c9dc499c5c787ddd8c4320f2d44d2ae6e40c22
[2] http://lists.lttng.org/pipermail/lttng-dev/2014-February/022423.html

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

lttng-tools: mark as affected by gcc bug 58595

Fixes:

http://autobuild.buildroot.org/results/6a6/6a66a389035939af542e172dbe2027ffa671e28a/

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

toolchain: add option to declare toolchains affected by gcc PR 58595

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

gcc: add fix for PR 58595 to gcc 4.7.3

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

gcc: add fix for PR 58595 to gcc 4.8.2

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

linux: bump to version 3.14.1

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

linux-headers: bump 3.{4, 10, 13, 14}.x series

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

graphs: make graphs with lots of packages nicer to look at

Some magic numbers obtained with trial-and-error and successive
iterations, to eventually get a nice graph.

[Thomas: remove excessive spaces in expressions.]

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

support/graph-depends: don't show toolchain dependency for all packages

Al packages depend on 'toolchain'. Currently, 'graph-depends' graphs this
dependency. The resulting graph is thus cluttered with less-than-useful
information.

Instead, do not graph the 'toolchain' dependency for any package, save
for the fake 'all' package. The graph is now a bit more readable.

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

manual: document BR2_GRAPH_DEPTH

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

support/graph-depends: add option to limit the depth of the graph

Currently, the complete dependency chain of a package is used to
generate the dependency graph. When this dependency chain is long,
the generated graph becomes almost unreadable.

However, it is often sufficient to get the first few levels of
dependency of a package.

Add a new variable BR2_GRAPH_DEPTH, that the user can set to limit
the depth of the dependency list.

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

support/graph-depends: use argparse to parse argv[]

Currently, we are using a crude, ad-hoc parsing of argv[].
This is a limiting factor to adding new options.

Use argparse instead, and introduce a single argument for now:
--package, -p PACKAGE

In the (near) future, we'll be able to add more option arguments,
such as depth-limiting for big graphs.

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

Makefile: rename non-user-facing variable

Variables should be prefixed with BR_ when they are not user-facing.

As a side effect, the new variable is prettier than the previous one. :-)

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

configs: bump raspberrypi_defconfig to use the latest stable kernel

Bump raspberrypi_defconfig to use the latest revision from the stable
rpi-3.10.y branch, which is based on v3.10.36.

Bump the kernel headers used for the toolchain at the same time.

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

package/rpi-firmware: bump revision

Of utmost importance, this bump allows to overclock the RPi to even
higher frequencies: ARM > 1200MHz, core > 600MHz, RAM > 600MHz. :-)
To be used with caution, of course! :-)

Fixes and enhancements for:
  - clocks : allow higher overclocks
  - audio  : PCM channel expansion/conversion
  - video  : stutter on streams without b-frames, choice of scaling /kernel/
  - camera : padding-removal, new YUV and BGR modes
  - decoder: timestamps fixes, frames interval tracking
  - encoder: misc jpeg encoding fixes, save inline motion vectors
  - display: don't require EDID on HDMI

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

package/rpi-userland: bump revision

Fixes and enhancements for:
  - camera : preview, white balance, padding-removal
  - encoder: save inline motion vectors
  - misc   : locking in GPU<->CPU dialog (?), tvservice eye-candy

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

vlc: select libvorbis when opus is also selected

If BR2_PACKAGE_OPUS is selected then vlc will be built with support for
opus passing "--enable-opus" option to the configure script. But when
the configure script checks for the opus library it also checks for the
ogg library at the same time, so if only opus has been installed and not
ogg, then it will fail with an error message like this one:

checking for OPUS... no
configure: error: Library ogg opus needed for opus was not found

Fixes:
http://autobuild.buildroot.net/results/4da/4da8e7cfb35c351e5d623835118512248d331d16/

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

iproute2: bump to version 3.14.0

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

ruby: bump to 1.9.3-p545

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

mtd: fix build against uClibc

Add a rpmatch() implementation when uClibc is missing one.

Fixes:
http://autobuild.buildroot.net/results/171/17187c3c9ba2b9df2bb6ac103e9a54706f07c28b/

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

cppcms: needs threads

Fixes:
http://autobuild.buildroot.net/results/d61/d618de68c8a50a932e85bdefb23c40e86790dc6c/

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

libatomic_ops: mark as supporting aarch64

[Thomas: fix typo in the patch: BR2_aarch64 -> BR2_aarch64.]

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

rtai: install to staging

Building RTAI userspace applications requires having the RTAI headers
and libraries installed to the staging directory, so this patch adds
RTAI_INSTALL_STAGING = YES.

In addition, this patch fixes up the rtai-config script, but does that
manually, because the rtai-config script doesn't match the convention
that the common <pkg>_CONFIG_SCRIPTS logic is designed for. It also
specifies a special directory for the RTAI headers, so that they don't
get mixed up with kernel headers.

Finally, it updates the RTAI download address to use the new correct
location.

[Thomas: fix commit log.]

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

python3: correctly fix PYTHON3_PATH

Commit 6251ccf03254f832c6e6ba0203ad46c18f217ca9 ('python3: fix the
value of PYTHON3_PATH') tried to fix the value of PYTHON3_PATH, but
did it incorrectly: it changed PYTHON_VERSION_MAJOR to
PYTHON_VERSION3_MAJOR, while it should have been
PYTHON3_VERSION_MAJOR.

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

pciutils: allow installation of static library when BR2_PREFER_STATIC_LIB

[Thomas: rework according to Arnout comments.]

Signed-off-by: Andy Kennedy <andy.kennedy@adtran.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

znc: new package

ported patch from Cédric Chépied:
http://patchwork.ozlabs.org/patch/306896/

[Thomas: added comment explaining why we don't use <pkg>_AUTORECONF,
and use a more conventional syntax for <pkg>_DEPENDENCIES and
<pkg>_CONF_OPT.]

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

pacage/xbmc-addon-xvdr: new package

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

package/xbmc-pvr-addons: new package

[Thomas: fix minor typo in comments "his own files" -> "its own files"]

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

exim: use a more standard build-time configuration

Buildroot currently ships a very minimal build configuration file for exim,
which disables most optional features. This is not coherent with the runtime
configuration file, taken verbatim from the exim distribution, which enables
some of these features.

The visible symptom is an error during boot that prevents exim from starting:

Exim configuration error in line 541 of /etc/exim/configure:
router dnslookup: cannot find router driver "dnslookup"

In order to fix this problem, we change the way exim is configured at build
time. Instead of blindly copying a minimal Buildroot-provided configuration
file, we now copy the exim-provided one and then tweak it to change the needed
options. This actually makes the configuration closer to standard exim.
As the amount of tweaking is remarkable, we also define a few macros to make
it easier and more readable.

This new approach was suggested by Bernd Kuhls.

Reported-By: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

linphone: arc: Undefine __arc__ when building for ARC processors

This patch is similar to 821a6d63. Libexosip2 defines symbol __arc__, which
gets in contradiction with same symbol defined by the GCC for the Synopsys
DesignWare ARC cores. Previous patch undefined __arc__ for libosip2 and
libexosip2, however linphone includes headers from libexosip2, which causes
build failure unless same patch is applied to it.

This patch fixes
http://autobuild.buildroot.net/results/0d2e19d40d6d0df4048e6ebae5f77022f89ca7cb/

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

rsh-redone: make rsh-redone.mk more Buildroot conventional

This commit slightly adjusts rsh-redone.mk to better match the usual
Buildroot conventions:

* Variables are named in upper-case, prefixed by the name of the
   package.

* No re-use of one _CMDS definition in another, and instead usage of
   a RSH_REDONE_MAKE_FLAGS variable to pass identical flags betwen
   build and install steps.

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