configs/qemu: update defconfigs to Linux 5.4.35

Linux version are changed to the new LTS kernel 5.4.35 for all qemu
defconfigs, except for riscv and csky. Thoses defconfigs are left
unchanged because they require either a custom Linux repository
or a specific kernel header version causing some difficulties when
upgrading to 5.4.35.

Update the nios2-10m50 linux.fragment to update the .dtb build directory
due to a change in kernel 4.20 [1]:

    nios2: build .dtb files in dts directory

    Align nios2 with other architectures which build the dtb files in the
    same directory as the dts files. This is also in line with most other
    build targets which are located in the same directory as the source.
    This move will help enable the 'dtbs' target which builds all the dtbs
    regardless of kernel config.

    This transition could break some scripts if they expect dtb files in
    the old location.

For x86 and x86_64 kernel, add the CONFIG_PCI symbol due a change in kernel
5.0 [2]. The option was previously enabled by default (default y).

    PCI: consolidate PCI config entry in drivers/pci

    There is no good reason to duplicate the PCI menu in every architecture.
    Instead provide a selectable HAVE_PCI symbol that indicates availability
    of PCI support, and a FORCE_PCI symbol to for PCI on and the handle the
    rest in drivers/pci.

Update the kernel of ppc-mac99 defconfig added in Buildroot 2019.08.

This version bump was tested on gitlab [4] using the newly introduced
boot-qemu-image.py script [5].

[1] https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=118864869805123bf82d666062542440a0fda5dd
[2] https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=eb01d42a77785ff96b6e66a2a2e7027fc6d78e4a
[3] a8fac3fcfc2c71c8a30d28c3d2e74646490502c2
[4] https://gitlab.com/kubu93/buildroot/pipelines/139819874
[5] 0c7935063851accf3db4c9eb2fc4f11c3da0c391

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

board/qemu/microblaze: convert kernel patch to git format

Convert the patch for microblaze kernel added for kernel 3.14 by
Waldemar to git format.

Note: the Waldemar Sob line is missing in the original patch:
fa2798548368d51998fad18d7f6d1ae0ed065b33

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/paho-mqtt-c: disable tests and cpack docs

paho-mqtt-c by default enables the building of test materials and
install of CPack documentation:

PAHO_ENABLE_TESTING - "Build tests and run"
PAHO_ENABLE_CPACK - "Enable CPack"

Let's disable these to save a couple megabytes and time.  This is
in keeping with the generic settings in pkg-cmake.mk.

Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/conntrack-tools: bump to version 1.4.6

Signed-off-by: Ryan Barnett <ryanbarnett3@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/exfatprogs: add patch fixing missing <sys/types.h>

This patch had been merged by upstream but appears to have been
accidentially overwritten with a force push. Let's add back the
rebased version until fixed by upstream.

Fixes:
http://autobuild.buildroot.net/results/667/667409f1d44a3f9be43aaff3f9a3426fe1386de2/build-end.log

Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/mtdev2tuio: fix build with musl 1.2.0

Fixes:
 - http://autobuild.buildroot.org/results/a30823f1388f37820d8a4f29b9b2c1de9b3992ac

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/wireguard-linux-compat: bump version to 1.0.20200413

For details, see the announcement:
https://lists.zx2c4.com/pipermail/wireguard/2020-April/005295.html

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

package/python-mbstrdecoder: needs python3

python2 has been dropped since version 1.0.0 and
https://github.com/thombashi/mbstrdecoder/commit/d8f6fd76452ef82d08d1633a0e484349fedc190b

Fixes:
 - http://autobuild.buildroot.org/results/324dc351752d412a315955faf69e8f5637e75123

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Reviewed-by: Asaf Kahlon <asafka7@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/rauc: bump to version 1.3

See: https://github.com/rauc/rauc/releases/tag/v1.3

And update hash file formatting (2 spaces).

Signed-off-by: Pierre-Jean Texier <pjtexier@koncepto.io>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/gflags: force Release mode

Force Release otherwise libraries will be suffixed by _debug which will raise
unexpected build failures with packages that use gflags (e.g. rocksdb)

Fixes:
 - http://autobuild.buildroot.org/results/3545774258babc3584f69e7d523efdf7fff1acb5

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/{libuv,uvw}: bump to versions 1.37.0, 2.5.0_libuv-v1.37

Signed-off-by: Asaf Kahlon <asafka7@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/libarchive: remove double handling of zstd

When 4bcc34446485d27efd41de70a86355a0e3798838 was applied, it was not
noticed that 96dc6701af877a4aa4da20a9e0d5b849cc5c8356 (from another
contributor) had already been
applied. 4bcc34446485d27efd41de70a86355a0e3798838 essentially did the
same thing as 96dc6701af877a4aa4da20a9e0d5b849cc5c8356, except it also
disable zstd support explicitly in the host-libarchive package.

Let's drop the part of 4bcc34446485d27efd41de70a86355a0e3798838 that
duplicates 96dc6701af877a4aa4da20a9e0d5b849cc5c8356.

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

package/libarchive: add zstd optional dependency

zstd is available since version 3.3.3 and
https://github.com/libarchive/libarchive/commit/26838cf5c17642f57192753cc5c3880b16b65ba3

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/libarchive: add lz4 optional dependency

lz4 is available since version 3.2.0 and
https://github.com/libarchive/libarchive/commit/724f3f918e509ddca60ce62fad696d7621d3623d

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/libiio: fix install of C# bindings

libiio-sharp.dll.mdb is not installed since version 0.19.0 and
https://github.com/analogdevicesinc/libiio/commit/8b571969fa245fb25c6431340f23ff157e34841b

Indeed, EXISTS "${LIBIIO_CS_DLL}.mdb" will always return false because
the file will not exist when cmake will create the makefiles.

Fixes:
 - http://autobuild.buildroot.org/results/42ffbd07d0e998c75d1afde6a5db7d6418c10f45

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/docker-engine: bump version to 19.03.7

Signed-off-by: Christian Stewart <christian@paral.in>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/docker-containerd: bump version to 1.2.13

Signed-off-by: Christian Stewart <christian@paral.in>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/libcpprestsdk: zlib is optional, not mandatory

The CMakeLists.txt contains:

  if(CPPREST_EXCLUDE_COMPRESSION)
  if(NOT CPPREST_EXCLUDE_BROTLI)
  message(FATAL_ERROR "Use of Brotli requires compression to be enabled")
  endif()

libcpprestsdk requires openssl, so if the libopenssl provider is
chosen, it anyway pulls in zlib. But if libressl is used as a
provider, it doesn't pull in zlib, and building libcpprestsdk becomes
possible without zlib thanks to this patch.

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/python-tornado: bump version to 6.0.4

Convert the hash file to using the two space format for hashes. The
has for the LICENSE file has been updated since version 6.0.4 now
includes DOS line endings (\r\n).

Signed-off-by: Ryan Barnett <ryanbarnett3@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/{apparmor,libapparmor}: bump to version 2.13.4

Signed-off-by: Angelo Compagnucci <angelo@amarulasolutions.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/pkg-utils: show-info: report the package build directory

Some people want to be able to inspect the current state of the build,
and report information about it, like:
  - report the number of packages already built out of the total,
  - list the packages being actually built (e.g. for TLPB)
  - etc...

However, the location where a package is built is inherently an internal
detail, so expose that to the user in the output of show-info. We only
expose the location relative to the base directory (basically, either
output/ or the user-suppiled $(O)), so that show-info does not contain
local information (the output of show-info can be shared).

Interested parties will be able to poke in there to identify the stamp
files and deduce the package's state.

Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Vadim Kochan <vadim4j@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: eeppeliteloop@gmail.com
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/pkg-utils: show-info: virtual packages can also install things

Although most virtual packages do not install anything, some of them
do. For example, the virtual package 'toolchain' does tweak the musl
headers during a post staging hook.

So, it makes sense for "show-info" to report if a virtual package has
its target, staging and/or image installation steps enabled or not.

This commit changes show-info to report the type of install steps for
virtual packages too, which means for all packages.

Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Vadim Kochan <vadim4j@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: eeppeliteloop@gmail.com
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/pkg-generic.mk: don't set INSTALL_{TARGET, STAGING, IMAGES} for host packages

By their very nature, host packages have no target, staging, or images
install steps; they have a single install step, that is always
performed.

As such, setting the corresponding _INSTALL_{TARGET,STAGING,IMAGES}
variables does not make sense for host packages.

However, people (and scripts) may get confused when they process the
output of printvars, e.g.:

    $ make printvars VARS=HOST_LIBTOOL_INSTALL_TARGET
    HOST_LIBTOOL_INSTALL_TARGET=YES

Only set those variables for target packages. There is no
corresponding variable for host packages, as they are always installed
(and only once).

Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Vadim Kochan <vadim4j@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: eeppeliteloop@gmail.com
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/pkg-utils: show-info: do not show install types for host packages

By their very nature, host packages have no target, staging, or
images install steps; they have a single install step, that is
always performed.

However, we currently report those three types of install, which
is misleading at best, and really incorrect.

If we were to report any install type for host package, that would
be a single one, and it would always be true.

So, do not report any install type for host packages, as it does
not make sense to report anything that is always true.

Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Vadim Kochan <vadim4j@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: eeppeliteloop@gmail.com
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/perl-type-tiny: bump to version 1.010001

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

package/perl-net-ssh2: bump to version 0.71

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

package/perl-net-dns: bump to version 1.23

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

package/moo: bump to version 2.004000

diff LICENSE:
-This software is Copyright (c) 2019 by mst - Matt S. Trout (cpan:MSTROUT) <mst@shadowcat.co.uk>.
+This software is Copyright (c) 2020 by mst - Matt S. Trout (cpan:MSTROUT) <mst@shadowcat.co.uk>.

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

package/perl-mojolicious: bump to version 8.40

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

package/perl-libwww-perl: bump to version 6.44

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

package/perl-io-socket-ssl: bump to version 2.068

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

package/perl-data-uuid: bump to version 1.226

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

fs/cpio: make initramfs init script survive 'console=' kernel argument

When booting with 'console=<empty>' in the kernel command line (as e.g.
U-Boot does with silent flags in effect), opening /dev/console fails.

As per POSIX [0], when iany redirection fails, the shell running exec
shal exit in error. So, when 'console=<empty>' is specified.
/dev/console can't be opened, and the redirection fails, and /init is
killed.

That behaviour was fixed on the kernel side with commit 2bd3a997befc2
(Open /dev/console from rootfs), present since 2.6.34, released in May
2010, so any [dr]ecent kernel will have that fix.

Furthermore, busybox will fix things up anyway (in bb_sanitize_stdio()),
falling back to opening /dev/null if no console is availble. systemd
does a similar thing (in make_console_stdio()), and sysvinit again has
a similar approach (in console_init()).

The archealogy search turned up those relevant commits:

    2011-08-04 10a130f91e initramfs/init: make sure that 0, 1, 2 fds are available
        introduces the three exec redirections in initramfs

    2011-09-06 3fac21ef8d cpio: fix boot with dynamic /dev
        introduces the three exec redirections in cpio

    2011-09-06 13a3afc536 fs/initramfs: refactor with fs/cpio
        dropped the initramfs tweaks to reuse the cpio ones

    2012-11-04 e1ebae700a fs/common: Create initial console device
        introduces the /dev/console char,5,1 pseudo device creation in
        cpio

    2018-03-31 dec061adce fs/cpio: don't extend packages' permissions table
        switched from the permission-table to a manual mknod to create
        /dev/console

The redirections were added before we could guarantee there was a
/dev/console in the rootfs.

We're now guaranteed to have /dev/console in an initramfs, and any recent
kernel will automatically open /dev/console before spawning /init.

The three redirections are useless now, and cause harm under certain
conditions. Drop them.

[0] https://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html#tag_18_20_01

Signed-off-by: Timo Ketola <timo.ketola@exertus.fi>
Cc: Peter Korsgaard <peter@korsgaard.com>
[yann.morin.1998@free.fr:
  - extend commit log with the analysis done with Peter
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/wireguard-linux-compat: tweak kernel-headers < 5.6 logic

The reason why the external wireguard kernel module is not allowed with
kernel headers >= 5.6 is that wireguard is included in the upstream kernel
since 5.6 rather than some kind of (fixable) incompatibility issue.  Adjust
the comment to make that clear.

While we're at it, drop the redundant !5.6 dependency on the kernel headers
dependency comment.  If headers are older than 3.10, then they are also
older than 5.6, so the statement is redundant.

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

package/mongodb: bump to version 4.2.4

- Drop !BR2_PACKAGE_PYTHON3 dependency as SConstruct requires
  host-python3 since version 4.1.10 and:
  https://github.com/mongodb/mongo/commit/8dd6d4755734ed37c1b98dfdefce3ca6bc65f1f6
- host-python-psutil is needed since 4.1.8 and
  https://github.com/mongodb/mongo/commit/ff03811e31851984617d73739de8acc7640b19a2
- Drop unneeded host-python-typing dependency
- C++17 and so gcc 7 is required since 4.1.8 and
  https://github.com/mongodb/mongo/commit/01d84b2565fc9ea9626a55dcf1f3f817968088f2
- Set --disable-minimum-compiler-version-enforcement as mongodb enforces
  gcc >= 8.2 since verson 4.1.8 and
  https://github.com/mongodb/mongo/commit/9ac90b128ebeb1bb431ebe3fe9176cb6142818cb
- Explictly disable ssl if needed as ssl is enabled by default
- Add host-python-cheetah dependency if needed
- Manage new http-client option added in version 4.1.2 and
  https://github.com/mongodb/mongo/commit/4d7dcca91bc2bbde65843fe88c41fa040e0d1877

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
[Titouan: Fix patch conflicts with master]
Signed-off-by: Titouan Christophe <titouan.christophe@railnova.eu>
Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/python3-psutil: add a special host variant

Needed to build mongodb.

With this host variant there is no need to enforce host
python in the original package.

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

package/python3-pyyaml: add special host variant

Needed to build mongodb.

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

package/python3-requests: add special host variant

Needed to build mongodb.

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

package/python3-cheetah: add special host variant

Needed to build mongodb.

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

package/python3-regex: add special package variant for the host build

Needed to build mongodb.

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

package/python-regex: new package

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

package/wireguard-linux-compat: depend on headers < 5.6

Since wireguard is built into kernels 5.6 and later we can't build
wireguard-linux-compat on them, so we need to depend on
!BR2_TOOLCHAIN_HEADERS_AT_LEAST_5_6.

Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

configs/olpc_xo175: needs host-openssl for kernel build

Fixes:

  https://gitlab.com/buildroot.org/buildroot/-/jobs/517918177

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/apparmor: fixing installation os detection

Actually the install target of some tools like the parser tries to do
an os detection to understand what install and where.
Incidentally, when the install is invoked on SuSE, this will trig a bug
in parallel install which manifests as "target/lib/apparmor: File exists"
error. For this problem, a patch is already sent upstream.

For buildroot instead, the os detection is useless so we disable it
and use a generic install method.

Fixes:
http://autobuild.buildroot.net/results/b18c6a9ce67065dcb7968c6f473b3b403d2925d2

Signed-off-by: Angelo Compagnucci <angelo@amarulasolutions.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/libuhttpd: disable test plugin

test plugin needs dynamic library support

Fixes:
 - http://autobuild.buildroot.org/results/ce7c69ceac3390f6d22f777e4cf908188486c6b7

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/libtasn1: needs autoreconf

Commit 6e1fde98e4deda8506e694c6d3ffba3fbd4d1d0e forgot to set
LIBTASN1_AUTORECONF

Fixes:
 - http://autobuild.buildroot.org/results/383e3cba56894df0e3dd53bfe9cee3f4bb80d412

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
[yann.morin.1998@free.fr: add dependency to host-pkgconf]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/apache: fixing apxs path mangling

Latest version of Apache introduce a new apxs with a slightly modified
path handling logic. In order to simplify the crosscompilation, the
software removes the common prefix from bin install dir and build
install dir, but for this to work they both should have a common prefix.

So we introduce a new regexp to fix /usr/bin to staging dir, the regexps
are also fixed to replace only the exact path between double quotes, to
avoid replacing the she-bang line.

Fixes:
http://autobuild.buildroot.net/results/c41f31566974209897a3a1ec35afe2536fb248cc
http://autobuild.buildroot.net/results/b93f19976ce96e79ea159c25ed74a7377c78f334

Signed-off-by: Angelo Compagnucci <angelo@amarulasolutions.com>
[yann.morin.1998@free.fr:
  - add the last few words about the she-bang blurb
  - do not use quotes in the existing /sur/bin regexp
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/sqlite: bump version to 3.31.1

Changelog: https://sqlite.org/releaselog/3_31_1.html

Reformated hash file.

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/exfatprogs: bump to version 1.0.2

Drop patch that is now upstream.

Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/evtest: switch to Gitlab

Downloading from the cgit repo is now broken and it is a mirror of the
Gitlab repo so use the Gitlab repo.

Signed-off-by: Sébastien Szymanski <sebastien.szymanski@armadeus.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

configs/nitrogen*: bump u-boot revision

Latest bump introduced a regression for Nitrogen8MM, now fixed.

Changelog:
- Fix Nitrogen8MM SPL regression
- Add multi-rank lpddr4 training support
- Improve fastboot support
- Add Android AVB / BCB support

Fixes: https://gitlab.com/buildroot.org/buildroot/-/jobs/517918160
Signed-off-by: Gary Bisson <gary.bisson@boundarydevices.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/binutils: fix xtensa PR ld/25861

Handle link-time relaxation of full 16-bit difference values correctly
in xtensa linker. This fixes the following kind of build errors:

.../five-point.cpp:338:(.debug_line+0x3b12):
        dangerous relocation: overflow after relaxation

Backported from:
https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commitdiff;h=30ce8e47fad9b057b6d7af9e1d43061126d34d20
Fixes:
http://autobuild.buildroot.net/results/67a5545779bbe8ca2b21da1f3ff2002053710ce1/
http://autobuild.buildroot.net/results/87f38998d2d60b6bce3128589973187ef9596e28/

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

package/freescale-imx/imx-gpu-g2d: fix the file name for ARM

NXP apparently changed file naming convention for this particular file.

Fixes: 1d0ea8d433 (package/freescale-imx/imx-gpu-g2d: bump to version 6.4.0.p1.0)
Signed-off-by: Timo Ketola <timo.ketola@exertus.fi>
Reviewed-by: Gary Bisson <gary.bisson@boundarydevices.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/{mesa3d, mesa3d-headers}: bump version to 20.0.5

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

package/iwd: fix build with glibc < 2.24

Fixes:
 - http://autobuild.buildroot.org/results/63f983aa22ca5303a55bc55e004f05a94e5f7259

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/systemd: bump version to v245.5

Signed-off-by: Adam Duskett <Aduskett@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/paho-mqtt-c: fix bump to version 1.3.2

Commit e69b8ddd43292b03b8f30f748ee2279b8df64c47 bumped paho-mqtt-c to
version 1.3.2 but hash was wrong moreover licence has changed to EPL-2.0
since
https://github.com/eclipse/paho.mqtt.c/commit/1e91229cb6403e3a9b02dfdc11c71b04c0518cff

So fix that and add LICENSE to the list of license files

Fixes:
 - http://autobuild.buildroot.org/results/7ea1791778053613e9ef6b146dbd1992a0f63dc6

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/apcupsd: always depend on threads

Fixes:
http://autobuild.buildroot.net/results/aa2ad7bdb37da611a4ec863ed2e986f34ea5228b/build-end.log

Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/luafilesystem: bump to version 1.8.0

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

package/uacme: bump version to 1.2.3

This version fixes https://github.com/ndilieto/uacme/issues/22

Signed-off-by: Nicola Di Lieto <nicola.dilieto@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/libva: bump version to 2.7.1

Changelog:
https://github.com/intel/libva/blob/5cef34fdf55b5f8a08fbfdfe5c5f8cc13b1d77ed/NEWS

Added upstream sha1 hash.

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

package/libva-utils: add upstream sha1 hash

Reformat remaining hashes.

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

package/libtasn1: bump to version 4.16.0

- Add patch to fix build
- Update indentation of hash file (two spaces)

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/python: bump version to 2.7.18

Release notes:
https://pythoninsider.blogspot.com/2020/04/python-2718-last-release-of-python-2.html

Updated license hash due to upstream commit:
https://github.com/python/cpython/commit/aa5b196c180d4c1dbe5a8593e55066e630d38c09

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

package/protobuf-c: bump to 1.3.3

Signed-off-by: Matthew Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/python-colorama: bump to 0.4.3

Signed-off-by: Matthew Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/linknx: bump to version 0.0.1.38

- Drop first and second patches (already in version)
- Retrieve third patch from upstream
- Retrieve an additionnal patch from upstream
- host-gettext is needed for AM_ICONV since
  https://github.com/linknx/linknx/commit/0512f6d0a07d50df33bfdca22d48360d164922a8
- Update indentation of hash file (two spaces)

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/python-wcwidth: bump to version 0.1.9

Update indentation of hash file (two spaces)

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/{openjdk, openjdk-bin}: add support for building either lts or latest

As Java is used quite a bit in the enterprise world, having the option to
build the LTS version of OpenJDK is quite convenient and also a requirement
for many companies wanting to use Java.

As such, there are three options:
  1) Continue only to support the latest version of OpenJDK.
  2) Downgrade our existing OpenJDK package from 14 to 11.
  3) Add an option to support either OpenJDK 11 or 14.

OpenJDK 11 and 14 currently have:
  - The same configure options.
  - The same license files and hashes for those license files.
  - The same dependencies.
  - The same method to build and install.

As such, supporting both 11 and 14 is not only an easy option to add to
Buildroot, but also a nice feature for users who wish to use Java in an
embedded environment with a company that mandates the use of the LTS version.

To make it explicit that this choice really is about LTS vs. latest, and
not about 11 vs. 14, the options are really named with LTS and LATEST,
so that future defconfigs will not have to migrate when the versions
changes (e.g. we update from 14->15, or from 11 to the next LTS).

Signed-off-by: Adam Duskett <Aduskett@gmail.com>
[yann.morin.1998@free.fr:
  - keep latest as the default, for existing defconfigs
  - rename options: drop numbers, use LTS and LATEST
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/python-jedi: add missing license

python-jedi bundles its own copy of typeshed since version 0.14.0 and
https://github.com/davidhalter/jedi/commit/7d2b7bb3c1c4f79f325ccbde9d2b2734aa872312

So add it to the license files (and update indentation of hash file to
two spaces while at it)

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

{linux, linux-headers}: bump 4.{4, 9, 14, 19}.x / 5.{4, 5, 6}.x series

Notice: 5.5.x is now EOL, so should be dropped at the next version bump.

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

package/perl: bump perl-cross to version 1.3.3

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

package/fio: bump to 3.19

- add upstream patch to fix build on musl
- update indentation of hash file (two spaces)

Signed-off-by: Heiko Thiery <heiko.thiery@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/libspatialindex: bump to version 1.9.3

- Drop patch (already in version)
- Update indentation in hash file (two spaces)
- Add upstream to allow building static library

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/iproute2: bump to version 5.6.0

Signed-off-by: Petr Vorel <petr.vorel@gmail.com>
[yann.morin.1998@free.fr: two spaces in hash file]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

support/scripts/boot-qemu-image.py: wait before using expect

As reported by a gitlab runtime test [1] and on the mailing list
[2], some runtime tests are failing on slow host machines when
the qemu-system-<arch> is missing on the host.

The boot-qemu-image.py script need to wait some time after
calling pexpect.spawn() in order to make sure that the qemu
process has been executed in start-qemu.sh.

If start-qemu.sh failed due to missing qemu-system binary
an exception will be thrown by child.expect() and should be
catched by the error handling (pexpect.EOF).

After spending a lot of time to investigate with Yann E. MORIN
[3]. It seems that short-lived child processes are a corner-case
that is not very correctly handled...

Without adding a sleep(1), child.expect() can trigger an
exception before setting the exitstatus of the spawned
process. This issue can be reproduced on a gitlab runner or
by adding "exit 1" in the first line of start-qemu.sh
(after the shebang).

There is even the same workaround in some pexpect examples [4].

Thanks to Yann for the help while investigating the issue.

Tested:
https://gitlab.com/kubu93/buildroot/pipelines/138472925

[1] https://gitlab.com/kubu93/buildroot/pipelines/135487475
[2] http://lists.busybox.net/pipermail/buildroot/2020-April/280037.html
[3] http://patchwork.ozlabs.org/project/buildroot/patch/20200418161023.1221799-1-romain.naour@gmail.com/
[4] https://github.com/pexpect/pexpect/blob/master/examples/ssh_tunnel.py#L80

Fixes:
https://gitlab.com/kubu93/buildroot/-/jobs/509053135

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: Yann E. MORIN <yann.morin.1998@free.fr>
[yann.morin.1998@free.fr: reorder imports]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/ltp-testsuite: fix build with musl 1.2.0

Fixes:
 - http://autobuild.buildroot.org/results/57e8d2a53f99dc804d66482f71653957fd1a6e47

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Acked-by: Petr Vorel <petr.vorel@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/{openjdk, openjdk-bin}: bump version to 14.0.1+7

Signed-off-by: Adam Duskett <Aduskett@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

packageJopenjdk: add support for building the full jdk

Some users may require the full JDK on the target to debug their programs.
This change is relatively trivial to add.

While the full JDK does have programs used for compiling on a target,
which is against Buildroot policy, the JDK also has several utilities used for
debugging purposes, which the JRE target does not build, and Buildroot supports
applications used for debugging purposes such as GDB.

As such, JDK support should be available for debugging purposes, and a note in
the Config.in file has been added under the JDK section, which informs the user
that JDK support is for debugging purposes only and that developing on a
target is not supported by Buildroot.

Signed-off-by: Adam Duskett <Aduskett@gmail.com>
Reviewed-by: Ryan Barnett <ryan.barnett@rockwellcollins.com>
Tested-by: Ryan Barnett <ryan.barnett@rockwellcollins.com>
[yann.morin.1998@free.fr:
  - s/OPENJDK_INSTALL_DIR/OPENJDK_VARIANT/
  - slightly rewrap help text
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/openjdk: Remove sparc support

Sparc support is deprecated and may be removed in future releases. There are
two choices to fix this issue:

1) Set --enable-deprecated-ports=yes in the CONF_OPTS to supress the error.
2) Remove support for Sparc.

Because this port is deprecated, it's safer to remove support alltogether.

Fixes:
    http://autobuild.buildroot.net/results/692820b4b6d4da42cd557fa7badbbd11806bbeba/

Signed-off-by: Adam Duskett <Aduskett@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/openjdk: copy all directories and files when installing

Several directories and files are currently not installed during the
target installation, these include:
  - conf
    Several configuration files, including security configuration files which
    may be necessary for running various java applications.

  - legal
    This directory contains legal notices that some java applications may
    require, as they may print legal information and will throw exceptions at
    runtime if the legal files are not present on the system.

  - release
    This file contains a list of modules included in the image.

Because these directories take up less than of megabyte extra, it is not an
issue to install all of them.

Signed-off-by: Adam Duskett <Aduskett@gmail.com>
Reviewed-by: Ryan Barnett <ryan.barnett@rockwellcollins.com>
Tested-by: Ryan Barnett <ryan.barnett@rockwellcollins.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

configs/qemu_riscv{32, 64}_virt: remove BR2_TARGET_GENERIC_GETTY_PORT

Giving an explicit getty port is not really needed, as we already
spawn a getty on the "console" device, which will match the serial
port passed as console= argument on the kernel command line.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/openjdk: fix installation with merged usr directories

Currently, Buildroot installs the jre libraries using
cp -dprf /build/linux-*-release/images/jre/lib/* $(TARGET_DIR)/usr/lib/

However, if a system has a merged /usr directory, and there is a built kernel
before installing OpenJDK, the installation fails because jre/lib has binary
modules file, which causes the following error: cp: cannot overwrite directory
'/usr/lib/modules with non-directory

The obvious fix is to install the modules to /usr/lib/jvm/ and set the
appropriate rpaths via the --with-extra-ldflags conf option. However, this fix
does not work because the built binaries themselves do not link against
libjava.so

Indeed, running readelf on the built java binary reports the following:
"(RUNPATH) Library runpath: [/usr/lib/jvm]" and /usr/lib/jvm/libjava.so exists.
However, when running the Java binary on the target, the following error
occurs: "Error: could not find libjava.so."

The following is the result of "strace java" ran on the target:
faccessat(AT_FDCWD, "/usr/lib/libjava.so", F_OK) = -1 ENOENT
faccessat(AT_FDCWD, "/usr/jre/lib/libjava.so", F_OK) = -1 ENOENT
newfstatat(AT_FDCWD, "/usr/lib/libjava.so", 0x7ffe7b4af8, 0) = -1 ENOENT
newfstatat(AT_FDCWD, "/usr/lib/jvm/libjli.so", [sic] AT_SYMLINK_NOFOLLOW) = 0

As seen above, the java binary searches for libjli.so in /usr/lib/jvm,
which demonstrates that the java binary searches for some of the
DT_NEEDED libraries using the correct rpath. But libjava.so is not
searched from the rpath; it is instead dl-opened manually, looked for in
the search paths hardcoded to the following directories:
  - /usr/lib/
  - /usr/jre/lib/
  - $(dirname $0)/../lib/

The reason behind the hardcoded paths given by the maintainers is due to
historical purposes for the need to support several java versions at the
same time on a single system, and that changing the above behavior is not
likely to ever happen.

As such, most distributions such as Redhat do the following:
  - Create the directory /usr/lib/jvm/java-$(JAVA_VERSION)/
  - Install all directories and files found in images/jre to that directory.
  - Symlink the binaries to in /usr/lib/jvm/java-$(JAVA_VERSION)/bin to
    /usr/bin.

However, because Buildroot does not need to support multiple versions of java
concurrently, there is no need for the additional java-$(JAVA_VERSION)
directory.

To fix the above issue, the following changes are performed:
  - Introduce the variable "OPENJDK_INSTALL_BASE" which points to /usr/lib/jvm
  - Set the --with-extra-ldflags conf_opt to
      "-Wl,-rpath,$(OPENJDK_INSTALL_BASE)/lib,-rpath,
      $(OPENJDK_INSTALL_BASE)/lib/$(OPENJDK_JVM_VARIANT)"
  - Run "mkdir -p $(TARGET_DIR)/usr/lib/jvm/" in the INSTALL_TARGET_CMDS step.
  - Copy both the lib and bin directories to /usr/lib/jvm/
  - Symlink the binaries in /usr/lib/jvm/bin/ to /usr/bin.

Fixes: https://bugs.busybox.net/show_bug.cgi?id=12751
Signed-off-by: Adam Duskett <Aduskett@gmail.com>
Reviewed-by: Ryan Barnett <ryan.barnett@rockwellcollins.com>
Tested-by: Ryan Barnett <ryan.barnett@rockwellcollins.com>
[yann.morin.1998@free.fr: fix two remaining mis-placed '/']
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

package/python-markdown: fix Python version dependency

Since upstream commit
https://github.com/Python-Markdown/markdown/commit/dab931f597cdc5b57d4e2458840fb991606506bc,
merged since version 3.2 of python-markdown, the Python 2.x support
has been dropped.

So our commit 6e538bf642d3584032d63bb54d24b88ab57e4283
("package/python-markdown: bump to version 3.2.1") should have made
python-markdown available only for Python 3.x.

Fixes:
http://autobuild.buildroot.net/results/6d1e796f6658accb3eafef265935db5e464638fe

Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Reviewed-by: Asaf Kahlon <asafka7@gmail.com>
Tested-by: Gwenhael Goavec-Merou <gwenhael.goavec-merou@trabucayre.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/libglib2: restore use of system printf

Prior to commit 4102db0f7a ("package/libglib2: bump to version 2.60.3")
which converted libglib2 to meson, Buildroot used to set a range of
autoconf options to bypass tests that require running binaries.

The meson version of libglib2's build system has many fewer of these
checks, but there are still some and these can be fed the "correct"
answer by adding properties to cross-compilation.conf.

Add the necessary properties to indicate that we have C99 compliant
print functions to avoid pulling in the gnulib fallback.

Signed-off-by: John Keeping <john@metanate.com>
Reviewed-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/uacme: bump version to 1.2.2

This version includes a new binary named "ualpn", a proxying
ACMEv2 tls-alpn-01 responder.

Signed-off-by: Nicola Di Lieto <nicola.dilieto@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/python-crossbar: needs cbor2

Since upstream commit 71f81bf83d3a1d4c564131ed7408f4d7ffe79242, cbor2
is needed by crossbar. This commit first appeared in crossbar release
v19.3.1, which means it is needed since the version bump that took
place in Buildroot commit 44a9f390efc282f41052b39716eab04cd76b1680.

Fixes:

  https://gitlab.com/buildroot.org/buildroot/-/jobs/517918425

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

package/libssh: add patch to fix regression

The libssh server is used by libnetconf2. With libssh version 0.9.4 a
regression was introduced that wrongly leads to session closed after the
poll timeout.

The patch comes from upstrem:
https://git.libssh.org/projects/libssh.git/commit/?id=6417f5a3cac8537ac6f6ff7fc1642dfaa0917fb4

Signed-off-by: Heiko Thiery <heiko.thiery@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/midori: needs gcc >= 7

Commit ec1ff802df9a0f17dd2b734ba536a5e206aa5aa4 forgot to propagate
gcc >= 7 dependency to midori

Fixes:
 - http://autobuild.buildroot.org/results/736a99dcaa72fba3f06775d08c4395f506fce944

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/collectd: fix dependency of network-gcrypt

Commit c8b8744a1aeebeba4035fdba9556226696ac2abd forgot to propagate
libgcrypt reverse dependency

Fixes:
 - http://autobuild.buildroot.org/results/98c7734dccb4e01c02f4680cfbdf82ee59e17935

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/libopenssl: security bump to v1.1.1g

This fixes CVE-2020-1967:
Server or client applications that call the SSL_check_chain() function during
or after a TLS 1.3 handshake may crash due to a NULL pointer dereference as a
result of incorrect handling of the "signature_algorithms_cert" TLS extension.
The crash occurs if an invalid or unrecognised signature algorithm is received
from the peer. This could be exploited by a malicious peer in a Denial of
Service attack. OpenSSL version 1.1.1d, 1.1.1e, and 1.1.1f are affected by this
issue. This issue did not affect OpenSSL versions prior to 1.1.1d.

See https://www.openssl.org/news/secadv/20200421.txt

Also update the hash file to the new two spaces convention

Signed-off-by: Titouan Christophe <titouan.christophe@railnova.eu>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

configs/qemu_riscv*: remove BR2_TARGET_OPENSBI_USE_PLAT

This option never existed in opensbi package.

This fixes the new defconfig check.

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

configs/qemu_ppc_virtex_ml507_defconfig: select BR2_POWERPC_SOFT_FLOAT

The BR2_SOFT_FLOAT option is lost while loading the defconfig with:
make qemu_ppc_virtex_ml507_defconfig

On powerpc, BR2_POWERPC_SOFT_FLOAT must be used to enable soft
floating point support.

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

configs/olimex_a20_olinuxino_lime{, 2}_defconfig: use a glibc toolchain

Since 2e71b396a171594d5b913d8b8c11079d376198ed, this defconfig needs
a glibc toolchain to select sunxi-mali-mainline package.

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

configs/engicam_imx6qdl_icore_qt5_defconfig: needs udev to select glmark2

The BR2_PACKAGE_GLMARK2 is lost while loading the defconfig with:
make engicam_imx6qdl_icore_qt5_defconfig

In order to select gmark2 package, BR2_PACKAGE_GLMARK2_FLAVOR_ANY option
must be set.

Based on the defconfig without X11 and wayland package, the only missing
option to select BR2_PACKAGE_GLMARK2_FLAVOR_ANY is BR2_PACKAGE_HAS_UDEV.
The only possible option is to enable one of the udev provider
(eudev or systemd). Select eudev package for /dev management.

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

configs: remove BR2_PACKAGE_QT5BASE_LICENSE_APPROVED

This option has been removed since 6836f2a70acd259a846c88968924d60afef8cc16.

This fixes the new defconfig check.

Signed-off-by: Romain Naour <romain.naour@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@bootlin.com>

configs/nanopi_r1_defconfig: remove BR2_TARGET_UBOOT_BOARD_DEFCONFIG

This defconfig use a custom defconfig file located
in board/friendlyarm/nanopi-r1/uboot/nanopi_r1_defconfig.

BR2_TARGET_UBOOT_BOARD_DEFCONFIG is used to provide the name
of a in-tree defconfig. Since a custom defconfig is used,
this option is lost while loading the defconfig with:
make nanopi_r1_defconfig

This fixes the new defconfig check.

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

configs:minnowboard_max-graphical_defconfig: re-enable GLX support

This defconfig loses mesa3d-demo and glmark2 package since commit
5cb821d5635626b7327d5d704555c412e5ed5a1f that introduced an
explicit option to enable GLX support.

This fixes the new defconfig check.

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

configs: fix typo BR2_TARGET_ROOTFS_EXT2_4

The ext4 option is BR2_TARGET_ROOTFS_EXT2_4 not
BR2_TARGET_ROOTFS_EXT_4.

This fixes the new defconfig check.

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: Fabio Estevam <fabio.estevam@nxp.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

configs/amarula_a64_relic: add missing BR2_PACKAGE_HOST_ANDROID_TOOLS=y

board/amarula/a64-relic/readme.txt makes use the host fastboot utility
to flash the board. However, BR2_PACKAGE_HOST_ANDROID_TOOLS_FASTBOOT
(which is enabled in the defconfig) has a dependency on
BR2_PACKAGE_HOST_ANDROID_TOOLS, which is not enabled in the
defconfig. Due to this, BR2_PACKAGE_HOST_ANDROID_TOOLS_FASTBOOT=y is
lost when loading the defconfig.

Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: Jagan Teki <jagan@amarulasolutions.com>
[Thomas: change to add BR2_PACKAGE_HOST_ANDROID_TOOLS=y]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>