core: introduce per br2-external NAME

This unique NAME is used to construct a per br2-external tree variable,
BR2_EXTERNAL_$(NAME)_PATH, which contains the path to the br2-external
tree.

This variable is available both from Kconfig (set in the Kconfig
snippet) and from the .mk files.

Also, display the NAME and its path as a comment in the menuconfig.

This will ultimately allow us to support multiple br2-external trees at
once, with that NAME (and thus BR2_EXTERNAL_$(NAME)) uniquely defining
which br2-external tree is being used.

The obvious outcome is that BR2_EXTERNAL should now no longer be used to
refer to the files in the br2-external tree; that location is now known
from the BR2_EXTERNAL_$(NAME)_PATH variable instead. This means we no
longer need to expose, and must stop from from exposing BR2_EXTERNAL as
a Kconfig variable.

Finally, this also fixes a latent bug in the pkg-generic infra, where we
would so far always refer to BR2_EXTERNAL (even if not set) to filter
the names of packages (to decide whether they are a bootloader, a
toolchain or a simple package).

Note: since the variables in the Makefile and in Kconfig are named the
same, the one we computed early on in the Makefile will be overridden by
the one in .config when we have it. Thus, even though they are set to
the same raw value, the one from .config is quoted and, being included
later in the Makefile, will take precedence, so we just re-include the
generated Makefile fragment a third time before includeing the
br2-external's Makefiles. That's unfortunate, but there is no easy way
around that as we do want the two variables to be named the same in
Makefile and Kconfig (and we can't ask the user to un-quote that variable
himself either), hence this little dirty triple-inclusion trick.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

core/br2-external: validate even more

A br2-external tree must provide external.mk and Config.in.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

core: offload handling of BR2_EXTERNAL into the script

Currently, we treat the case where we have no br2-external tree
(BR2_EXTERNAL is empty) differently from the case where we do have one
(BR2_EXTERNAL is not empty).

There is now no reason to treat those two cases differently:

  - the kconfig snippet is always generated appropriately (i.e. it would
    include the br2-external tree if set, or include nothing otherwise);

  - we no longer have a dummy br-external tree either.

Also, the Makefile code to handle BR2_EXTERNAL is currently quite
readable if at least a little bit tricky.

However, when we're going to add support for using multiple br2-external
trees simultaneously, this code would need to get much, much more complex.

To keep the Makefile (rather) simple, offload all of the handling of
BR2_EXTERNAL to the recently added br2-external helper script.

However, because of Makefiles idiosyncracies, we can't use a rule to
generate that Makefile fragment.

Instead, we use $(shell ...) to call the helper script, and include the
fragment twice: once before the $(shell ...) so we can grab a previously
defined BR2_EXTERNAL value, a second time to use the one passed on the
command line, if any.

Furthermore, we can't error out (e.g. on non-existent br2-external tree)
directly from the fragment or we'd get that error on subsequent calls,
with no chance to override it even from command line.

Instead, we use a variable in which we store the error, set it to empty
before the second inclusion, so that only the one newly generated, if
any, is taken into account.

Since we know the script will always be called from Makefile context
first, we know validation will occur in Makefile context first. So we
can assume that, if there is an error, it will be detected in Makefile
context. Consequently, if the script is called to generate the kconfig
fragment, validation has already occured, and there should be no error.
So we change the error function to generate Makefile code, so that
errors are caught as explained above.

Lastly, when the value of BR2_EXTERNAL changes, we want to 'forget'
about the previous value of the BR2_EXTERNAL_MK variable, especially in
the case where BR2_EXTERNAL is now set to empty, so that we do not try
to include it later. That's why we first generate empty version of
BR2_EXTERNAL_MK, and then assign it the new value, if any.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

core: get rid of our dummy br2-external tree

Now that we generate a kconfig snippet, we can conditionally include the
BR2_EXTERNAL's Config.in only when BR2_EXTERNAL is supplied by the user,
which means our empty/dummy Config.in is no needed.

As for external.mk, we can also include it only when BR2_EXTERNAL is
supplied by the user, which means our empty/dummy external.mk is no
longer needed.

Ditch both of those files, and:

  - only generate actual content in the Kconfig snippet when we actually
    do have a BR2_EXTERNAL provided by the user (i.e. BR2_EXTERNAL is not
    empty);

  - add a variable that contains the path to the external.mk provided by
    the user, or empty if none, and include the path set in that variable
    (make can 'include' nothing without any problem! ;-) )

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Romain Naour <romain.naour@openwide.fr>
Cc: Julien CORJON <corjon.j@ecagroup.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

core: do not hard-code inclusion of br2-external in Kconfig

Move the inclusion of br2-external's Config.in to the generated kconfig
snippet.

This will ultimately allow us to use more than one br2-external tree.

Offload the "User-provided options" menu to the generated Kconfig
snippet. We can also move the definition of the Kconfig-version of
BR2_EXTERNAL into this snippet.

We introduce an extra check that was not present in the previous code,
to check that we do have permission on that directory. Prevciously, it
was handled as a side effect of not being able to cd into there, but it
is cleaner to check it expressly.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/go-bootstrap: bump version to 1.4.3

This bump includes
https://go.googlesource.com/go/+/a527bdbda39c48fa772d8d54eb6b849f240442c1

which fixes a gcc6-related compile error

cmd/6c
/home/buildroot/br2/output/build/host-go-bootstrap-1.4.2/src/cmd/6c/txt.c: In function 'gmove':
/home/buildroot/br2/output/build/host-go-bootstrap-1.4.2/src/cmd/6c/txt.c:995:28: error: left shift of negative value [-Werror=shift-negative-value]
     f->vconst |= (vlong)~0 << 32;
                            ^~
/home/buildroot/br2/output/build/host-go-bootstrap-1.4.2/src/cmd/6c/txt.c:1045:28: error: left shift of negative value [-Werror=shift-negative-value]
     f->vconst |= (vlong)~0 << 32;
                            ^~
cc1: all warnings being treated as errors

not yet caught by autobuilders using this defconfig:

http://autobuild.buildroot.net/results/394/394e22be0ef986463e97b3040dad8f978262732c/

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

package/linux-pam: bump to version 1.3.0

removed 0003-Conditionally-compile-per-innetgr-availability.patch,
applied upstream:
https://git.fedorahosted.org/cgit/linux-pam.git/commit/?id=835d64947996b7cc96fe187f9b3103db36dddf77

This bump includes
https://git.fedorahosted.org/cgit/linux-pam.git/commit/?id=01e0038fa55581c4afc9d63b6180d2ea77ba2940
which fixes
http://autobuild.buildroot.net/results/394/394e22be0ef986463e97b3040dad8f978262732c/

Removed comment in Config.in about broken --enable-static-modules
configure option because it was removed upstream:

https://git.fedorahosted.org/cgit/linux-pam.git/commit/?id=a684595c0bbd88df71285f43fb27630e3829121e

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

gcc: remove BR2_GCC_SUPPORTS_GRAPHITE

The blind option BR2_GCC_SUPPORTS_GRAPHITE was used to distinguish gcc
versions that support the graphite loop optimizer. But since a while
already, all the versions we support do support graphite. So this symbol
isn't needed anymore.

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

gcc: gcc arc supports graphite

The ARC version of gcc does support graphite. It was probably just
forgotten when the BR2_GCC_VERSION_ARC symbol was introduced.

While we're at it, also remove a redundant newline.

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

gcc: remove BR2_GCC_NEEDS_MPC

The blind option BR2_GCC_NEEDS_MPC was used to distinguish gcc versions
that rely on the mpc library and the ones that don't. But since a while
already, all the versions we support do need the mpc library. So this
symbol isn't needed anymore.

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

dependencies: check for Math BigInt/BigRat Perl module

As for MakeMaker Perl module, add a check for Math BigInt/BigRat Perl
Core module, which are needed by mpv.

Fixes:
Waf: Entering directory `output/build/mpv-0.17.0/build'
[  1/198] Compiling version.sh
[ 12/198] Processing ebml_types.h: demux/ebml.c demux/demux_mkv.c -> build/ebml_types.h
[ 13/198] Compiling demux/ebml.c
Can't locate Math/BigRat.pm in @INC (you may need to install the Math::BigRat module)

Waf: Entering directory `output/build/mpv-0.17.0/build'
[  1/198] Compiling version.sh
[  2/198] Compiling TOOLS/osxbundle/mpv.app/Contents/Resources/icon.icns
[  3/198] Compiling video/out/x11_icon.bin
[  4/198] Compiling etc/input.conf
[  5/198] Compiling sub/osd_font.otf
[  6/198] Compiling video/out/opengl/nnedi3_weights.bin
[  7/198] Compiling player/lua/defaults.lua
[  8/198] Compiling player/lua/assdraw.lua
[  9/198] Compiling player/lua/options.lua
[ 10/198] Compiling player/lua/osc.lua
[ 11/198] Compiling player/lua/ytdl_hook.lua
[ 12/198] Processing ebml_types.h: demux/ebml.c demux/demux_mkv.c -> build/ebml_types.h
[ 13/198] Compiling demux/ebml.c
[ 14/198] Compiling DOCS/man/mpv.rst
Can't locate Math/BigInt.pm in @INC (you may need to install the Math::BigInt module)

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: François Perrad <francois.perrad@gadz.org>
[Thomas et al.: make those perl modules only required when mpv is enabled.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

dependencies: check for MakeMaker Perl module

ExtUtils::MakeMaker is one of the Perl Core modules usually packaged in
Perl package for a Debian/Ubuntu based system.
For a Fedora based system, each Perl Core modules have their own RPM
package. So install only Perl package is not enough.

Fixes:
>>> host-libxml-parser-perl 2.41 Configuring
[...]
perl `which perl` Makefile.PL
Can't locate ExtUtils/MakeMaker.pm in @INC (you may need to install the ExtUtils::MakeMaker module)

Add a new Perl module check in dependency.sh.

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: François Perrad <francois.perrad@gadz.org>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/qt: Avoid X11 header collision with OpenGL/EGL

Including X11 headers via EGL/egl.h leads to a collision of defines
between Qt and X11. To fix this qt5 added
  DEFINES += MESA_EGL_NO_X11_HEADERS
in various files:

https://github.com/qtproject/qtbase/blob/dev/src/platformsupport/eglconvenience/eglconvenience.pri#L19
https://github.com/qtproject/qtbase/blob/dev/src/plugins/platforms/minimalegl/minimalegl.pro#L10
https://github.com/qtproject/qtbase/blob/dev/src/plugins/platforms/eglfs/deviceintegration/eglfs_x11/eglfs_x11.pro#L6
[...]

This patch fixes a similar compile error in qt4:
http://autobuild.buildroot.net/results/9b0/9b0ed92984eead89d148eaa71aff6e2f1e117837/
http://autobuild.buildroot.org/results/4fd/4fd57553336d36439fab32a7221fc8639758e887/
http://autobuild.buildroot.org/results/5cb/5cb8072f8220f7203fa79936eedcc43eab53ec66/

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

Remove BR2_DEPRECATED

The BR2_DEPRECATED logic is a lot less useful than the legacy handling,
because the symbols just disappears without warning to the user. For
example, we had a few defconfigs that were using deprecated symbols
(which were not actually used because BR2_DEPRECATED wasn't set) so
these didn't build the expected code anymore.

Also, the idea behind BR2_DEPRECATED is that you can easily revive it
again if there is interest. However, it is relatively easy to revert
the removal of a package as well.

The deprecation is also more effort because it has to be removed twice:
once when deprecating, and once when really removing.

It doesn't make sense to add a legacy entry for BR2_DEPRECATED. Users
who actually used it will get legacy warnings instead.

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

wvstreams: remove deprecated package

We will remove BR2_DEPRECATED, so remove this deprecated package.

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

wvdial: remove deprecated package

We will remove BR2_DEPRECATED, so remove this deprecated package.

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

webkitgtk24: remove deprecated package

We will remove BR2_DEPRECATED, so remove this deprecated package.

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

torsmo: remove deprecated package

We will remove BR2_DEPRECATED, so remove this deprecated package.

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

sstrip: remove deprecated package

We will remove BR2_DEPRECATED, so remove this deprecated package.

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

linux-headers: remove deprecated version 4.2 and 4.3

We will remove BR2_DEPRECATED, so remove this deprecated package.

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

kodi-addon-xvdr: remove deprecated package

We will remove BR2_DEPRECATED, so remove this deprecated package.

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

ipkg: remove deprecated package

We will remove BR2_DEPRECATED, so remove this deprecated package.

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

gcc: remove deprecated 4.7.x

We will remove BR2_DEPRECATED, so remove this deprecated option.

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

binutils: remove deprecated 2.24.X

We will remove BR2_DEPRECATED, so remove this deprecated option.

[Peter: drop binutils patches]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

configs/freescale_mpc8315erdb: remove U-Boot build

Binutils was locked down to 2.24 because U-Boot 2014.01 didn't build
with more recent binutils. However, this doesn't work anymore because
binutils 2.24 is deprecated.

Current U-Boot 2016.09.01 doesn't work out of the box because the
u-boot-nand.img support has been removed.

Since updating U-Boot is generally not needed anyway, we just remove
the U-Boot support.

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

configs/freescale_p1010rdb_pa: remove U-Boot build

Binutils was locked down to 2.24 because U-Boot 2014.01 didn't build
with more recent binutils. However, this doesn't work anymore because
binutils 2.24 is deprecated.

Current U-Boot 2016.09.01 does build successfully, but we don't have a
board anymore to test it.

Since updating U-Boot is generally not needed anyway, we just remove
the U-Boot support.

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

rapidxml: add patch to fix usage with gcc 4.7+

The official `rapidxml` package release only supports up to GCC 4.6.x
for all capabilities. Implementers attempting to include the file
"rapidxml_print.hpp" will all result in a compilation failure when using
a currently supported Buildroot GCC version (all 4.7+ at this time).
With the provided patch, dependent packages will be able to use all of
rapidxml's capabilities using a generated toolchain from Buildroot (let
alone external toolchains).

Signed-off-by: James Knight <james.knight@rockwellcollins.com>
[Thomas: rework patch so that it contains the proper line endings, and
applies properly without a pre-patch hook to convert the line endings
beforehand.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

configs/lego_ev3: bump kernel to 4.4 series

Bump kernel revision used by lego_ev3_defconfig to the latest tag from
ev3dev, as suggested by Peter. This is a 4.4 series kernel, which has the
added benefit of not necessitating a specific gcc to boot.

We need a patch to create an empty drivers/lego/Kconfig file for the build
to succeed though, as the ev3dev kernel normally expects a drivers tree to
be checked out there.

Signed-off-by: Vincent Stehlé <vincent.stehle@laposte.net>
Suggested-by: Peter Korsgaard <peter@korsgaard.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

qt5: bump version to 5.6.2

- qt5base: remove obsolete '-no-nis' configure parameter ([1])
- qt5base: remove  upstream committed patch 0007-Remove-libudev-dependency-from-kms.pro.patch ([2])
- qt5imageformats: remove upstream committed  patch 0001-tvOS-support.patch ([3])

[1] https://code.qt.io/cgit/qt/qtbase.git/commit/?id=644d0e75240811212ed9fd0c21e3bd05db3ba5db
[2] https://code.qt.io/cgit/qt/qtbase.git/commit/?id=72492735b7b7770808fcc9fe067e8f03fab827fc
[3] https://git.buildroot.net/buildroot/commit/package/qt5?id=9a55b996651fdfbc40e202cb4dd725dce624ca91

Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Tested-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

configs/arm_juno: use official kernel tarball

Since commit
0924208839336f1a1bb29eb1a627005ba6b9065d ("configs/arm_juno: use 4.6
mainline kernel"), we're using a mainline kernel for the ARM Juno
board. But we're still fetching it from a Git repository, which is not
really efficient, so let's switch to use a tarball instead.

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

package/kodi: needs .py modules

Kodi segfaults as soon as it tries to load a python module:

    Could not find platform independent libraries <prefix>
    Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
    ImportError: No module named site
    Segmentation fault

Turns out that keeping .py modules (with or without .pyc) fixes the
issue.

Currently, Kodi selects python, but since the format of modules is a
choice, we can not select it.

Fix that by inverting the dependency on python from Kodi:
  - turn it into a depends rather than a select,
  - add the dependency to ! pyc-only
  - update the comment accordingly.

In addition, the !MMU and !static dependencies are updated since they
are no longer needed for Python (we now depend on it rather than
selecting it), but for other packages. The !MMU dependency is moved to
the _ARCH_SUPPORTS symbol as well.

Also, we can no longer "select BR2_PACKAGE_SAMBA4", because samba4
selects python, and we can no longer select python. Therefore, we switch
to a "depends on" dependency for samba4, which is fine as a user willing
to use Kodi with Samba will certainly realize that Samba should be
enabled.

Fixes bug #9221.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Thomas:
 - update on latest master
 - take care of the !MMU/!static dependencies.
 - take care of the samba4 problem.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

configs/zynq_*_defconfig: enable VFP support

The A9 cores inside the Xilinx Zynq have a VFP unit, so enable it.

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

configs: add defconfig for Digilent Zybo

This configuration allows out-of-the box Digilent Zybo
support. It uses mainline U-boot and Linux kernel.
This configuration generates a SD card image named sdcard.img.

[Peter: rename genimage.cfg, cleanup post-image script, enable VFP support,
  use 4.6 kernel headers and add tools needed by genimage]
Signed-off-by: Sebastien Van Cauwenberghe <svancau@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

sqlite: bump version to 3150000 (3.15.0)

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

configs/beaglebone_qt5: new defconfig

Add a target configuration for the beaglebone that supports the SGX530
graphics acceleration and builds the qt5 demos with the eglfs backend.
The imagination technologies demos and fbv are also included. The Linux
kernel is based on the omap2plus defconfig, a config fragment adds the
required settings for SGX accelerator support.

Signed-off-by: Lothar Felten <lothar.felten@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

configs/beaglebone: update kernel/uboot, add AM335x EVM support

This patch updates the target config for the beaglebone based on the
AM335x CPU. It also supports the beagleboneblack, the TI evm and evmsk.
Device tree blobs for am335x-evm, am335x-evmsk, am335x-bone,
am335x-boneblack will be built and a sd card image is created with a
post-image script.  The kernel and driver versions match the
ti-processor-sdk 02.00.00.00.  Patches allow the use of old AM335x evm
revisions (1 bit sd card) and JTAG debugging.

Kernel JTAG patch: By default the kernel will disable the JTAG clock,
access via the JTAG port will not work after the kernel booted. The
AM335x-evm has a JTAG header fitted, this patch keeps the JTAG clock
alive.

Signed-off-by: Lothar Felten <lothar.felten@gmail.com>
[Thomas: remove 1-bit MMC patch, it's causing a significant performance
regression for users of BeagleBone, which are 99.99% of the users of
this defconfig.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

DEVELOPPERS: add packages I'm interesting in

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

gcc: add patches to fix PR70473

gcc 5.x introduced a regression in the ARM build, which causes the
s-automata program to consume a very significant amount of RAM during
the gcc build. This causes numerous failures with our Travis-CI based
testing of defconfigs.

In order to address this, this commit backports a commit from the gcc
master branch, to both our gcc 5.x and gcc 6.x support.

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

toolchain-buildroot: enable Musl for MIPS R6

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

gcc: add support for MIPS R6 Musl

Backport upstream patch to add support for MIPS R6 Musl:

  https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=83717065090bb8b954556d1216dd9dc397dc0243

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

DEVELOPERS: change developer for wpa_supplicant

The wpa_supplicant was in fact no introduced by Peter, but by Gustavo,
and Gustavo is much more active in maintaining this package, so let's
mention Gustavo as the developer for this package.

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

DEVELOPERS: update entry for gcc

Refering to package/gcc/gcc-initial and package/gcc/gcc-final was not a
good idea, as it doesn't notify when gcc patches are
added/removed/modified, so change this to package/gcc so that it catches
all changes.

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

openblas: use MIPS specific cores for P5600 and I6400

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

qt53d, qt5serialbus: move to technology preview section

As explained in
https://wiki.qt.io/New_Features_in_Qt_5.6#Technology_Preview_Modules,
not only qt5quickcontrols2 is a technology preview component in Qt 5.6,
but also qt53d, qt5serialbus.

Signed-off-by: Matt Kraai <kraai@ftbfs.org>
[Thomas: improve commit log with a reference.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

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

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

aufs-util: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

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

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

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

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

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

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

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

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

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

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

mtdev2tuio: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

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

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

pptp-linux: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

live555: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

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

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

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

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

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

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

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

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

jack2: use $(TARGET_MAKE_ENV) when calling $(MAKE)

In this case in particular it's $(TARGET_CONFIGURE_OPTS), which includes
$(TARGET_MAKE_ENV).

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

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

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

gadgetfs-test: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

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

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

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

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

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

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

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

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

ti-utils: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

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

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

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

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

open2300: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

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

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

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

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

smstools3: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

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

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

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

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

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

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

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

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

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

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

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

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

irda-utils: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

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

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

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

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

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

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

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

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

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

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

lz4: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

p910nd: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

wf111: use $(TARGET_MAKE_ENV) when calling $(MAKE)

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

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

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

configs/arm_juno: add the bootloader build

Thanks to the recent addition of the vexpress-firmware and
arm-trusted-firmware packages, this commit improves the ARM Juno
defconfig to also build a bootloader.

Signed-off-by: Joao Pinto <jpinto@synopsys.com>
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>

arm-trusted-firmware: new package

Signed-off-by: Joao Pinto <jpinto@synopsys.com>
[Thomas:
 - Rename the package from atfirmware to arm-trusted-firmware, in order
   to match upstream.
 - Remove option to apply custom patches. We no longer add such options,
   and use the global patch directory instead.
 - Rename the repo URL/version options, in order to not be Git specific,
   in case support for fetching from other VCS is added later. This is
   consistent with how other bootloaders handle this.
 - Add license information.
 - Do not add a weird dependency on the vexpress-firmware package in
   Config.in. Instead, simply use it if it's available. Of course, some
   configurations (such as the juno configuration) will fail to build if
   vexpress-image is not enabled, but it's the responsibility of the
   user to create a config that builds.
 - Simplify misc aspects in the .mk file.]
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>

MIPS: add support for P6600 cores

-march=p6600 is not yet supported in GCC upstream, so disabling all
versions when selecting this core.

Note that P6600 implies a MIPS R6 CPU, and some GCC versions are already
disabled for R6, so we don't need to disable those ones for P6600 as
well.

The external Codescape IMG GNU Linux Toolchain has support for this
core.

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

MIPS: add support for P5600 cores

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

MIPS: add support for I6400 cores

-march=i6400 support starts from GCC-6, so disable previous versions
when selecting this core.

Note that I6400 implies a MIPS R6 CPU, and some GCC versions are already
disabled for R6, so we don't need to disable those ones for I6400 as
well.

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

MIPS: add support for M5101 cores

-march=m5101 support starts from GCC-6, so disable previous versions
when selecting this core.

Note that M5101 implies a MIPS R5 CPU, and some GCC versions are already
disabled for R5, so we don't need to disable those ones for M5101 as
well.

Also disable external toolchains that don't support this core.

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

MIPS: add support for M5100 cores

-march=m5100 support starts from GCC-6, so disable previous versions
when selecting this core.

Note that M5100 implies a MIPS R5 CPU, and some GCC versions are already
disabled for R5, so we don't need to disable those ones for M5100 as
well.

Also disable external toolchains that don't support this core.

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

MIPS: add support for interAptiv cores

-march=interaptiv support starts from GCC-6, so disable previous
versions when selecting this core.

Also disable external toolchains that don't support this core.

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

MIPS: enable Codescape MTI toolchain for R5

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

MIPS: add mips64r5 generic architecture variant

-march=mips64r5 support started from GCC-5, so disable previous versions
when the CPU is R5.

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

MIPS: add mips32r5 generic architecture variant

-march=mips32r5 support started from GCC-5, so disable previous versions
when the CPU is R5.

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