{linux, linux-headers}: bump 4.{9, 14, 19}.x / 5.{0, 1}.x series

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

package/v4l2loopback: bump to version 0.12.2 to fix Linux 5.1 compat

This fixes a kernel 5.1.x compatibility issue. The only changes
between 0.12.1 and 0.12.2 are:

d3b198ef6f57ca512fb25147c9d85b922fd4651a Released v0.12.2
376c2c28bd7d4470cd92ff646d6087ca70cd9d2e fixed typo
6edc4b164b1f05bee74cb507a4f50776a65ceb73 mentioned support for 5.0.0
0b8feb80fdef9a415d8250bca1790b3ff23e8391 Replace v4l2_get_timestamp with ktime_get_ts(64)
541e3bc7aaf46dc9a21f92c7f527397fce03dfd8 Update README.md

So the only functional change is the actual ktime_get_ts() fix, which
is needed for Linux 5.1 compatibility. Therefore, bumping is pretty
much the same as backporting just this commit.

Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
[Thomas: extend the commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/wpewebkit: fix build with musl

Fixes:
 - http://autobuild.buildroot.org/results/c703d45ab691641beabc3440f081a8bc195b4a23

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

package/netsurf: disallow on archs requiring ABI specific CFLAGS

Fixes:
http://autobuild.buildroot.net/results/67ef520d82ea529a9fe593d83a3aeae5f8b0ee5d/
http://autobuild.buildroot.net/results/eafc3e4be571d5ecee549a11530ac4e508f31782/
http://autobuild.buildroot.net/results/ba7f30833fef54162a82f4b336a72d6599594526/

The netsurf build system mixes up host and target CFLAGS, so it isn't
compatible with architectures where we pass ABI specific compiler flags (in
TARGET_ABI).

Add a _ARCH_SUPPORTS kconfig variable matching the TARGET_ABI logic we have
in package/Makefile.in, and use it to disallow netsurf for those
architectures.

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

package/wireshark: security bump to version 3.0.2

The following vulnerabilities have been fixed:
 - wnpa-sec-2019-19 Wireshark dissection engine crash. Bug 15778.

Update patches to use the ones merged upstream

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libopenssl: fix static build

no-dso option has been removed with
https://github.com/openssl/openssl/commit/31b6ed76dfd53529b74e79830c81372d0b756929

To fix this error, use "gcc" target in static builds. This target is
very minimalistic, we need to manually pass -lpthread and
-DOPENSSL_THREADS however we can also remove libdl workarounds

Fixes:
 - http://autobuild.buildroot.org/results/96d6b89d20980e8f7fa450b832474a81d492b315

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/openmpi: fix build failure due to Gcc Bug 68485

Package openmpi manifests Microblaze Gcc Bug 68485 resulting in a build
failure due to an Internal Compiler Error.

As done for other packages in Buildroot work around this Gcc Bug by
setting optimization to -O0 if BR2_TOOLCHAIN_HAS_GCC_BUG_68485=y.

Fixes:
http://autobuild.buildroot.net/results/8f3/8f334427e7475154d69469f8ee4efab6df80e403/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/gstreamer1/gst1-plugins-bad: require dynamic library for decklink plugin

The decklink plugin uses <dlfcn.h> functions: dlopen(), dlsym(), etc.

Signed-off-by: Adam Duskett <Aduskett@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/opencv3: fix build due to gcc bug 68485

With Microblaze Gcc version <= 9.x the build fails due to gcc bug 68485:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68485. The bug show up when
building opencv3 with optimization but not when building with -O0. To
work around this, if BR2_TOOLCHAIN_HAS_GCC_BUG_68458=y, we force using
-O0.

Fixes:
 - http://autobuild.buildroot.org/results/c78eac84d1c5a6702e7759cd5364da1c3e399b4b

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Reviewed-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/dosfstools: introduce custom install routine

We can't use dosfstools' install target, because it'll install *all*
binaries, even the disabled ones. Also, we can't just delete dosfstools
binaries from the target directory after installing them, because other
packages (specifically Busybox) may provide tools of the same name, and
we may end up deleting those instead.

To avoid any issues, we create our own install routines, which only
copy the enabled binaries into the target location.

Signed-off-by: Markus Mayer <mmayer@broadcom.com>
[Thomas: use full destination path for INSTALL commands.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

DEVELOPERS: add Mirza Krak to mender-artifact package

Signed-off-by: Mirza Krak <mirza.krak@northern.tech>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/mender: update readme.txt

Provide additional details on how Mender works within Buildroot.

Signed-off-by: Mirza Krak <mirza.krak@northern.tech>
[Thomas: remove duplicate "Default configuration files" title, rewrap
text]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/libnss: security bump to version 3.42.1

Fixes the following security issues:

(3.41) CVE-2018-12404: Cache side-channel variant of the Bleichenbacher
attack

(3.42.1) CVE-2018-18508: Add additional null checks to several CMS functions
to fix a rare CMS crash.  Thanks to Hanno Böck and Damian Poddebniak for the
discovery and fixes

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

package/libopenssl: security bump to version 1.1.1c

Fixes the following security issues:

Prevent over long nonces in ChaCha20-Poly1305 (CVE-2019-1543)

ChaCha20-Poly1305 is an AEAD cipher, and requires a unique nonce input for
every encryption operation.  RFC 7539 specifies that the nonce value (IV)
should be 96 bits (12 bytes).  OpenSSL allows a variable nonce length and
front pads the nonce with 0 bytes if it is less than 12 bytes.  However it
also incorrectly allows a nonce to be set of up to 16 bytes.  In this case
only the last 12 bytes are significant and any additional leading bytes are
ignored.

It is a requirement of using this cipher that nonce values are unique.
Messages encrypted using a reused nonce value are susceptible to serious
confidentiality and integrity attacks.  If an application changes the
default nonce length to be longer than 12 bytes and then makes a change to
the leading bytes of the nonce expecting the new value to be a new unique
nonce then such an application could inadvertently encrypt messages with a
reused nonce.

Additionally the ignored bytes in a long nonce are not covered by the
integrity guarantee of this cipher.  Any application that relies on the
integrity of these ignored leading bytes of a long nonce may be further
affected.  Any OpenSSL internal use of this cipher, including in SSL/TLS, is
safe because no such use sets such a long nonce value.  However user
applications that use this cipher directly and set a non-default nonce
length to be longer than 12 bytes may be vulnerable.

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

infra/pkg-kconfig: do not rely on package's .config as a timestamp

Since linux-4.19, the kernel's build system internally touches its
.config file.

However, we currently used that file as a timestamp to detect whether
our kconfig fixups were to be (re)applied or not, which in turn is used
to decide whether we should (re)build the package or not.

But with latest kernel versions, this timestamp heuristic is now broken,
and we always rebuild the kernel on subsequent builds.

We fix that by introducing a separate timestamp file of our own, which
we know the kernel (or the kconfig-based packages, for that matters)
does not use.

Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Thomas De Schampheleire <patrickdepinguin@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/matchbox-panel: fix build with NLS

Fixes:
 - http://autobuild.buildroot.org/results/3e72c2c2baae3a70fb5145b1b09a1f3716e4bce2

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/dropbear: rework license information

According to the LICENSE file curve25519-donna is licensed under
BSD-3-Clause license.

There is only BSD-2-Clause license mentioned so remove
BSD-2-Clause-like.

Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Reviewed-by: Thomas Huth <huth@tuxfamily.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

toolchain/toolchain-external/toolchain-external-andes-nds32: add missing dependencies/select

This external toolchain is pre-built for x86, so it can only work on
x86 and x86-64, and for the latter, the ia32 libraries are necessary.

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

package/python-cython: bump to version 0.29.9

Fixes a number of crashes / bugs. For details, see:
https://github.com/cython/cython/blob/master/CHANGES.rst

Signed-off-by: Asaf Kahlon <asafka7@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/php: security bump version to 7.3.6

Release notes: https://www.php.net/archive/2019.php#id2019-05-30-1

Fixes
CVE 2019-11038: https://bugs.php.net/bug.php?id=77973
CVE 2019-11039: https://bugs.php.net/bug.php?id=78069
CVE 2019-11040: https://bugs.php.net/bug.php?id=77988

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/flare-engine: fix check-package failure

Check-package gives error on flare-engine due to trailing whitespace in
flare-engine.mk comment inserted with commit:
https://git.buildroot.net/buildroot/commit/?id=1203cca1b51eefb4ee236d998ab6ec32bdd087d8

Remove trailing whitespace in flare-engine.mk comment.

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/mender: install additional inventory scripts

There are two additional inventory scripts provided in
the Mender client repository, let's install them.

- mender-inventory-os
    - will push content of /etc/os-release
- mender-inventory-rootfs-type
    - will push filesystem type of rootfs

Signed-off-by: Mirza Krak <mirza.krak@northern.tech>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/mender: update rootfs parts in mender.conf

The configuration options (RootfsPartA/RootfsPartB) must
provide a valid path to a block devices.

Signed-off-by: Mirza Krak <mirza.krak@northern.tech>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/suricata: needs autoreconf

Commit 1ce2db1090c494f57019af4f29691e0f86b48330 was the second patch of
a third patch serie, it was applied without the first one so
AUTORECONF=YES was missing and patch number was wrong. Fix these two
errors.

Fixes:
 - http://autobuild.buildroot.org/results/a26d3493399c43faa37d2d67d772e0833971a9de

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

toolchain: introduce BR2_TOOLCHAIN_HAS_GCC_BUG_68485

GCC hangs while building brotli for the Microblaze Arch:
http://autobuild.buildroot.net/results/d86/d86251974a0a348a64d9a1d1fd7d02dd4aff0792/

Originally reported for gpsd:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68485

Still not fixed. Every Microblaze Gcc version up to and including 9.1
is affected.

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/libhtp: add -lz to htp.pc

This will fix static build of suricata.

Fixes:
 - No autobuilder failures yet

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/suricata: fix pcap static build

Fixes:
 - no autobuilder failures yet

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/brotli: work around gcc bug 68485

With Microblaze Gcc version <= 9.x the build fails due to gcc bug 68485:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68485. The bug show up when
building brotli with optimization but not when building with -O0. To
work around this, if BR2_TOOLCHAIN_HAS_GCC_BUG_68458=y, we force using
-O0.

Fixes:

  http://autobuild.buildroot.net/results/d86/d86251974a0a348a64d9a1d1fd7d02dd4aff0792/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/flare-engine: fix typo in comment

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/supertux: fix debug build on uclibc/musl

Always set CMAKE_BUILD_TYPE to Release otherwise supertux will be built
with -pg since:
https://github.com/SuperTux/supertux/commit/afd5f1b33c110104a7f286d6d7854157919a0a52

This will result in the following build failure on uclibc or musl:
[ 77%] Linking C executable sq_static
CMakeFiles/sq_static.dir/sq.c.o: In function `quit':
/home/buildroot/autobuild/instance-3/output/build/supertux-0.6.0/external/squirrel/sq/sq.c:42: undefined reference to `__gnu_mcount_nc'

Fixes:
 - http://autobuild.buildroot.org/results/b0b2e25af198d01713d1e2bcf38c77ae8ffbd7de

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/flare-engine: re-enable package on microblaze

With Microblaze Gcc version < 8.x the build hangs due to bug 85180:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180
To avoid this, the flare-engine package has a
!BR2_TOOLCHAIN_HAS_GCC_BUG_85180 dependency. However, gcc bug 85180 only
triggers when optimization is enabled, so we can work around the issue
by passing -O0, which is what we do in other Buildroot packages to work
around this bug.

So, this commit passes -O0 when BR2_TOOLCHAIN_HAS_GCC_BUG_85180, and
re-enables flare-engine on Microblaze. It is not enough to set
CMAKE_CXX_FLAGS, because flare-engine's CMakeLists.txt sets
CMAKE_CXX_FLAGS_<BUILD_TYPE> (depending on the value of
CMAKE_BUILD_TYPE), and the build-type-specific flags come after the
generic flags, so our -O0 gets overridden again. Therefore, also set
CMAKE_BUILD_TYPE to the dummy value Buildroot.

Fixes:
http://autobuild.buildroot.net/results/706/7065e14917a8bbc0faf21b29183ac55b6c800ee3/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
[Arnout: extend explanatory comment and update commit log]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

support/gnuconfig/config.sub: add C-SKY support

Modify config.sub so that it knows about the C-SKY
architecture. Without this, all autotools projects fail to build on
C-SKY.

Signed-off-by: Guo Ren <ren_guo@c-sky.com>
[Thomas: improved commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/tpm2-totp: fix error due to variable used uninitialized

Some inline declarations of strtok_r (specifically in Sourcery CodeBench
Lite 2016.11-19) contain code where an '__s' local variable can be used
uninitialized.

When GCC expands that declaration in tpm2-totp, __s becomes an alias to
a variable which fact is not initialized, but this is not relevant since
the execution path leading to the uninitialized use is never followed.

Anyway, apply a patch already submitted upstream to fix the compilation
error.

Upstream bug report:
  https://github.com/tpm2-software/tpm2-totp/issues/32

Fixes:
  http://autobuild.buildroot.net/results/5693a35e4d6bc76a1f46fe0e217abc49f7188aad/

Signed-off-by: Carlos Santos <unixmania@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/linuxptp: fix build with new headers

net_tstamp.h in recent kernel versions requires time.h for clockid_t.

Fixes:
  http://autobuild.buildroot.net/results/93903e8abfeaa7df0b6525033fa63e29f2c9e7ae

Signed-off-by: Artem Panfilov <panfilov.artyom@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/qt5/qt5wayland: install compositor QML files

Fixes using QtWayland.Compositor QML import.

Signed-off-by: David Rosca <nowrep@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/wireshark: fix static build with libxml2 and lzma

Fixes:
 - http://autobuild.buildroot.org/results/533c75836ba6281000b89a648fc653b9ff4685ee

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/botan: fix static build with atomic

Fixes:
 - http://autobuild.buildroot.org/results/9db39b209a0f4be1e09b7e4ddb5e9a63da4fbbfa

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>

package/mutt: fix static build without unistring

Fixes:
 - http://autobuild.buildroot.org/results/9f3dd4e81ba9e82db0e858f91de13ef23bc2bf14

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/qt5webkit-examples: fix build with qt5webkit

Fixes:
 - http://autobuild.buildroot.org/results/68dd6f1ffed49f8e2570f7364e5e635573807d3e

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

package/qt5multimedia: fix install with gstreamer 0.10 and QT 5.6

QT 5.6 fails to correctly detect gstreamer support with gstreamer 0.10
as a result libqgsttools_p.so is not built and build fails on:
cp: cannot stat '/home/buildroot/autobuild/run/instance-1/output/host/xtensa-buildroot-linux-uclibc/sysroot/usr/lib/libqgsttools*.so.*': No such file or directory

Fixes:
 - http://autobuild.buildroot.org/results/c88afcd365418e29c89f247d9d887b5f786b0ec8

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

Update for 2019.05-rc3

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

package/gpsd: fix BR2_PACKAGE_GPSD_MAX_{CLIENT, DEV} options

limited_max_clients and limited_max_devices have been renamed in 3.12
as max_clients and max_devices, respectively.

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Reviewed-by: Benoît Thébaudeau <benoit.thebaudeau.dev@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

linux/linux.hash: fix typo in 5.1.4 entry

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

package/keepalived: fix build with -Os

Fixes:
 - http://autobuild.buildroot.org/results/955aef1c25b977d6697fdf8cd1e710f40c7e60a3

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

{linux, linux-headers}: bump 4.{9, 14, 19}.x / 5.{0, 1}.x series

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

linux: default to kernel 5.1.x

To match the default kernel headers we use.

[Peter: drop 5.1.x bump, tweak commit message]
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/intel-microcode: make target installation optional

For early microcode loading, there is no need to install the individual
microcode files to /lib/firmware - So make that optional.

Let the option default to y for backwards compatibility, and select it from
iucode-tool as the init script relies on the /lib/firmware files.

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

linux: build after intel-microcode if enabled for early loading support

To support building in (a subset of) the intel-microcode files into the
kernel using the CONFIG_EXTRA_FIRMWARE option, we need to ensure that the
microcode files are installed before the Linux kernel is built.

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

package/intel-microcode: install into images for early loading support

Microcode based security mitigation (E.G.  MDS) requires that the microcode
gets loaded very early. This can be handled by one of:

- Concatenating (a subset of) the intel-microcode files and write to
  kernel/x86/microcode/GenuineIntel.bin in the initrd.  Requires that the
  (first) initrd is external from the kernel and NOT compressed.

- Build (a subset of) the intel-microcode files into the kernel using the
  CONFIG_EXTRA_FIRMWARE option.

Install the microcode files into images to support these use cases (E.G.
through a post-build script for the initrd, or by pointing
CONFIG_EXTRA_FIRMWARE_DIR to ${BR_BINARIES_DIR}, similar to how we include
the .cpio image inside the kernel).

Notice that there may be licensing concerns when embedded non-GPL firmware
in the kernel.

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

package/assimp: re-enable package on Microblaze

With Microblaze ccc version < 8.x the build hangs due to bug 85180:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. To avoid this, the
assimp package has a !BR2_microblaze dependency. However, gcc bug
85180 only triggers when optimization is enabled, so we can work
around the issue by passing -O0, which is what we do in other
Buildroot packages to work around this bug.

So, this commit passes -O0 when BR2_TOOLCHAIN_HAS_GCC_BUG_85180, and
re-enables assimp on Microblaze.

Note that the comment was talking about gcc bug 71124, but this gcc
bug is a duplicate of 85180. Since all Buildroot packages now use the
reference to gcc bug 85180 and the option is named
BR2_TOOLCHAIN_HAS_GCC_BUG_85180, we use this naming as well for
assimp.

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/ffmpeg: re-enable package if gcc bug 85180 is present

With Microblaze Gcc version < 8.x the build hangs due to gcc bug
85180: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. To avoid
this problem, until now, ffmpeg could not be selected on
Microblaze. However, this problem only happens due to optimization,
and can worked around by forcing -O0.

So, when BR2_TOOLCHAIN_HAS_GCC_BUG_85180=y, this commit:
- adds --disable-optimizations to FFMPEG_CONF_OPTS
- passes -O0 to CFLAGS in FFMPEG_CONF_ENV

Then, we remove 'depends on !BR2_microblaze' from Config.in to
re-enable the package.

Note that the comment was talking about gcc bug 71124, but this gcc
bug is a duplicate of 85180. Since all Buildroot packages now use the
reference to gcc bug 85180 and the option is named
BR2_TOOLCHAIN_HAS_GCC_BUG_85180, we use this naming as well for
ffmpeg.

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
[Thomas: tweak CFLAGS logic, improve commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/gstreamer/gst-ffmpeg: re-enable package when gcc bug 85180 is present

With Microblaze Gcc version < 8.x the build hangs to gcc bug 85180:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. This package has
been excluded from building if the toolchain is affected by gcc bug
85180. To be consistent with how we deal with this issue in other
packages, we re-enable the package and instead work around the issue
by building with -O0, since gcc bug 85180 manifests itself only when
optimization is enabled.

To achieve this:

- add --disable-optimizations to EXTRA_OPTS, which gets passed down to
  the built-in libav library
- pass -O0 to CFLAGS in CONF_ENV

Then we remove 'depends on !BR2_TOOLCHAIN_HAS_GCC_BUG_85180' and its
comment if not available from Config.in

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
[Thomas: tweak CFLAGS logic, improve commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/postgresql: work around gcc bug 85180

With Microblaze Gcc version < 8.x the build hangs due to gcc bug
85180: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. The bug
shows up when building postgresql with optimization but not when
building with -O0. To work around this, if
BR2_TOOLCHAIN_HAS_GCC_BUG_85180=y we force using -O0.

Fixes:

  http://autobuild.buildroot.net/results/53c270c696ee999d541da0b09bf6a215e5335e9a/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
[Thomas: tweak CFLAGS logic, improve commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/opus: work around gcc bug 85180

With Microblaze Gcc version < 8.x the build hangs due to gcc bug
85180: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. The bug
shows up when building opus with optimization but not when building
with -O0. To work around this, if BR2_TOOLCHAIN_HAS_GCC_BUG_85180=y we
force using -O0.

Fixes:

  http://autobuild.buildroot.net/results/73dc9610a13d6e14eec58d529617210d93d5dec4/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
[Thomas: tweak CFLAGS logic, improve commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/libcpprestsdk: work around gcc bug 85180

With Microblaze Gcc version < 8.x the build hangs due to gcc bug
85180: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. The bug
shows up when building libcpprestsdk with optimization but not when
building with -O0. To work around this, if
BR2_TOOLCHAIN_HAS_GCC_BUG_85180=y we force using -O0.

Fixes:

  http://autobuild.buildroot.net/results/7065e14917a8bbc0faf21b29183ac55b6c800ee3/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
[Thomas: tweak CFLAGS logic, improve commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/kismet: work around gcc bug 85180

With Microblaze Gcc version < 8.x the build hangs due to bug 85180:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. The bug shows up
when building kismet with optimization but not when building with
-O0. To work around this, if BR2_TOOLCHAIN_HAS_GCC_BUG_85180=y we
force using -O0.

Fixes:

  http://autobuild.buildroot.net/results/02f6e616631f05a4f15f31f2b2e67d65ba7afcb0/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
[Thomas: tweak CFLAGS logic, improve commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/glibmm: use GLIBMM_CXXFLAGS

Commit 8d8d4ecaf1bc22a89d0f373e9d1655f59f283480 ("package/glibmm: work
around gcc bug 85180") was incorrect reworked by me, and
TARGET_CXXFLAGS was used instead of GLIBMM_CXXFLAGS, causing the
Microblaze workaround to be ineffective. This new commit fixes this
typo.

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

package/jasper: work around gcc bug 85180

With Microblaze Gcc version < 8.x the build hangs due to gcc bug
85180: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. The bug
shows up when building jasper with optimization but not when building
with -O0. To work around this, if BR2_TOOLCHAIN_HAS_GCC_BUG_85180=y,
we force using -O0.

Fixes:

  http://autobuild.buildroot.net/results/5546087cdf684099525dc5824ed428cd0fd74283/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
[Thomas: tweak CFLAGS logic, improve commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/glibmm: work around gcc bug 85180

With Microblaze Gcc version < 8.x the build hangs due to gcc bug
85180: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. The bug
shows up when building glibmm with optimization but not when building
with -O0. To work around this, if BR2_TOOLCHAIN_HAS_GCC_BUG_85180=y,
we force using -O0.

Fixes:

  http://autobuild.buildroot.net/results/e196d77626b877dc3454d21febe20a04877c02a9/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
[Thomas: tweak CXXFLAGS logic, improve commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/ddrescue: work around gcc bug 85180

With Microblaze Gcc version < 8.x the build hangs due to gcc bug
85180: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. The bug
shows up when building ddrescue with optimization but not when
building with -O0. To work around this, if
BR2_TOOLCHAIN_HAS_GCC_BUG_85180=y we force using -O0.

Fixes:

  http://autobuild.buildroot.net/results/2af69401acb76be3d3fd427d1c58fed3af33bf9b/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
[Thomas: tweak CFLAGS logic, improve commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/chocolate-doom: work around gcc bug 85180

With Microblaze Gcc version < 8.x the build hangs due to bug 85180:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. The bug shows up
when building chocolate-doom with optimization but not when building
with -O0. To work around this bug, if
BR2_TOOLCHAIN_HAS_GCC_BUG_85180=y, we force using -O0.

Fixes:

  http://autobuild.buildroot.net/results/e570839576be8963dc6bd36342e2f857da3c6146/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
[Thomas: tweak CFLAGS logic, improve commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/webkitgtk: security bump to version 2.24.2

This is a new major release which brings in many improvements and new
features. For a complete list, please refer to the release notes:

  https://webkitgtk.org/2019/03/13/webkitgtk2.24.0-released.html
  https://webkitgtk.org/2019/04/09/webkitgtk2.24.1-released.html
  https://webkitgtk.org/2019/05/17/webkitgtk2.24.2-released.html

Updating to version 2.24.2 also includes fixes for CVE-2019-6201,
CVE-2019-6251, CVE-2019-7285, CVE-2019-7292, CVE-2019-8503,
CVE-2019-8506, CVE-2019-8515, CVE-2019-8518, CVE-2019-8523,
CVE-2019-8524, CVE-2019-8535, CVE-2019-8536, CVE-2019-8544,
CVE-2019-8551, CVE-2019-8558, CVE-2019-8559, CVE-2019-8563,
CVE-2019-11070, CVE-2019-6237, CVE-2019-8571, CVE-2019-8583,
CVE-2019-8584, CVE-2019-8586, CVE-2019-8587, CVE-2019-8594,
CVE-2019-8595, CVE-2019-8596, CVE-2019-8597, CVE-2019-8601,
CVE-2019-8607, CVE-2019-8608, CVE-2019-8609, CVE-2019-8610,
CVE-2019-8615, CVE-2019-8611, CVE-2019-8619, CVE-2019-8622, and
CVE-2019-8623.

The detailed security advisories can be found at:

  https://webkitgtk.org/security/WSA-2019-0002.html
  https://webkitgtk.org/security/WSA-2019-0003.html

The BR2_PACKAGE_WEBKITGTK_ARCH_SUPPORTS_JIT configuration symbol is not
needed anymore, because the logic to decide whether the JavaScriptCore
JIT spport can be enabled has been improved upstream.

One of the new features in 2.24.x is the support for JPEG2000 images,
which is implemented using the OpenJPEG library. Therefore now
BR2_PACKAGE_OPENJPEG is selected.

This adds one small patch which did not make it to the 2.24.2 release
which solves a build issue when the building the GStreamer GL elements
is disabled.

Signed-off-by: Adrian Perez de Castro <aperez@igalia.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/mono-gtksharp3: fix build with latest mono

Fixes:
 - http://autobuild.buildroot.org/results/7cae5b32285fe0edf69eab3297be9c6a4e76b7b8

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/luarocks: bump to version 3.1.2

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

package/libcurl: security bump to version 7.65.0

Fixes the following security vulnerabilities:

- CVE-2019-5435: Integer overflows in curl_url_set()
  https://curl.haxx.se/docs/CVE-2019-5435.html

- CVE-2019-5436: TFTP receive buffer overflow
  https://curl.haxx.se/docs/CVE-2019-5436.html

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

package/atop: work around gcc bug 85180

With Microblaze Gcc version < 8.x the build hangs due to gcc bug
85180: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180. The bug
shows up when building atop with optimization but not when building
with -O0. To work around this, if BR2_TOOLCHAIN_HAS_GCC_BUG_85180=y
we force using -O0.

Fixes:

  http://autobuild.buildroot.net/results/94aa00f776b8a3491ac0b3212c95f5e666c5a29a/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

toolchain: gcc bug 85180 is fixed in gcc >= 8.x

Gcc bug 85180 (https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85180) has
been fixed on Gcc version >= 8.x, so this commit adjusts the
BR2_TOOLCHAIN_HAS_GCC_BUG_85180 option to no longer be true when the
gcc version is >= 8.x.

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
Reviewed-by: Petr Vorel <petr.vorel@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

tslib: update to 1.20

The release page has a changelog overview:
https://github.com/libts/tslib/releases

Signed-off-by: Martin Kepplinger <martink@posteo.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

support/scripts/check-bin-arch: ignore /usr/lib/grub

/lib/grub is already ignored, so add /usr/lib/grub to support
BR2_ROOTFS_MERGED_USR.

Signed-off-by: Alex Xu <alex_y_xu@yahoo.ca>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/intel-microcode: security bump to version 20190514a

Includes MDS mitigation (RIDL, Fallout, Zombieload), INTEL-SA-00223

Move to the Intel github repo as this release is not yet available on
downloadmirror.intel.com.

Update license hash because of copyright year and DOS/UNIX newlines change.

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

package/pcsc-lite: security bump to version 1.8.25

Contains the following security related fixes:
 - Race condition possibility; CWE-362: Concurrent Execution using
   Shared Resource with Improper Synchronization ('Race Condition'):
   https://github.com/LudovicRousseau/PCSC/commit/36c8eae890dc81f0e25ebf0be1c6a969ffd9543a
 - Fix realloc error handling:
   https://github.com/LudovicRousseau/PCSC/commit/430043842e05a300f1dda1b93fcfa8fceadd81e5

Update hash of COPYING and remove MIT from licenses because sd-daemon
files were removed:
https://github.com/LudovicRousseau/PCSC/commit/0984e0f4a03e5e29bb5725b6f149f56c9e9c73c2

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/pcsc-lite: fix license

pcsc-lite is not only licensed under BSD-3-Clause, so retrieve licenses
from COPYING (BSD-2-Clause, GPL-3.0+, MIT, ISC) and add GPL-3.0.txt to
license files

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/pcsc-lite: drop unneeded first patch

This patch is not needed as static build has been disabled since
commit ad8c327053f3736881e1ef6d7b394b3a500ed8fa

Moreover, pthread dependency should be correctly retrieved from
libusb-1.0.pc

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

utils/check-package: fix flake8 warning

Fix flake 8 warning (introduced with commit [1]):

  utils/checkpackagelib/lib.py:56:1: E302 expected 2 blank lines, found 1

[1] https://git.buildroot.net/buildroot/commit/?id=8e352c32b0beded97a8a5c1e9edc9d618514ee7b

Reported-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

utils/check-package: warn about utf-8 characters in .mk files

Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Tested-by: Titouan Christophe <titouan.christophe@railnova.eu>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

board/pandaboard: replace utf-8 graphic sympols by ascii characters

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

board/csky: replace utf-8 graphic sympols by ascii characters

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

board/beaglebone: replace utf-8 graphic sympols by ascii characters

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

board/beagleboardx15: replace utf-8 graphic sympols by ascii characters

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

board/armadeus: replace utf-8 graphic sympols by ascii characters

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

board/lemaker/bananapro/linux-wifi.fragment: add rfkill driver into kernel by default

This fixes following error:

# wpa_supplicant -i wlan0 -c /etc/wpa_supplicant.conf -B
Successfully initialized wpa_supplicant
rfkill: Cannot open RFKILL control device

Signed-off-by: Bartosz Bilas <b.bilas@grinn-global.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/libupnp18: add upstream patch to fix runtime crash with musl

Applications build with a musl 1.1.20+ toolchain and linked with libupnpp will
crash at runtime with `Illegal instruction` as musl is more strict with
trying to detach an already detached thread resulting in undefined
behaviour.

Upstream status:
https://github.com/mrjimenez/pupnp/issues/102

Backported from:
https://github.com/mrjimenez/pupnp/commit/04b454f693d0c71336252380d08f1d02967e133e

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

package/libupnp18: needs host-pkgconf

Otherwise build fails with:

```
configure.ac:630: error: possibly undefined macro: AC_MSG_ERROR
      If this token and others are legitimate, please use m4_pattern_allow.
```

Fixes:

  http://autobuild.buildroot.net/results/6d734e45ec923dfa7549e0480b3d8913fc29fddd/

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

support/testing/tests: drop Python 2.x test in test_ipython

Since commit 6ebaef3818d6e3dc821616a33391eb6d594a053e
("package/python-ipython: bump to version 7.4.0"), ipython is no
longer available for Python 2.x, as it requires Python 3.x.

However, the corresponding test case that was testing iPython under
Python 2.x was not removed at the same time, causing a failure of
TestIPythonPy2 test. Let's drop the test that is no longer relevant.

Fixes:

  https://gitlab.com/buildroot.org/buildroot/-/jobs/210208754

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

package/libssh2: fix build with openssl and atomic

Fixes:
 - http://autobuild.buildroot.org/results/e381eefa39307f1c8eb22f872df438b3fe0352dd

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

package/wpewebkit: add BR2_PACKAGE_HAS_LIBEGL_WAYLAND dependency

Commit 5b3c7a3e7828107eaa9dad49ba2650466277d891 added a
BR2_PACKAGE_HAS_LIBEGL_WAYLAND dependency to wpebackend-fdo but forget
to add it to wpewebkit

Fixes:
 - http://autobuild.buildroot.net/results/49e04166de68358e69bce580b29ba3a25b313acf

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Reviewed-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/wpewebkit: needs gcc >= 6

gcc requirement has been bumped from gcc 5 to gcc 6 since
https://trac.webkit.org/browser/webkit/trunk/CMakeLists.txt?rev=231152

Fixes:
 - http://autobuild.buildroot.org/results/304956dda8e580cb39107e57a139cd1ae08ee655

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Reviewed-by: Adrian Perez de Castro <aperez@igalia.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/pcsc-lite: disable documentation

Fixes:
 - http://autobuild.buildroot.org/results/a6cd53b060c00e916b40d7bd6c99bd8232d0f978

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

package/tslib: update upstream source address

The project has been moved from the kergoth user account to the
libts organisation on github. While github seems to maintain redirects,
we should use the new location directly.

Signed-off-by: Martin Kepplinger <martin.kepplinger@ginzinger.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/cjson: security bump to version 1.7.12

Fix infinite loop in cJSON_Minify (potential Denial of Service), see
https://github.com/DaveGamble/cJSON/issues/354

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.{0, 1}.x series

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

package/gerbera: fix pidfile name in start-stop-script

`NAME` is not set resulting in a pidfile without a basename:
`/var/run/.pid`. Use the correct variable `DAEMON` instead.

Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/dhcp: fix upstream URL in Config.in

The current URL no longer exists, so update it to what looks like the
main DHCP upstream site.

This issue was noticed by the upstream URL check added by Matt Weber
in the pkg-stats script, whose results are visible at
http://autobuild.buildroot.net/stats/.

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

package/python: security bump to version 2.7.16

Fixes the following security issues:

- CVE-2013-1752: Change use of readline() in :class:`imaplib.IMAP4_SSL` to limit line length

- CVE-2018-14647: The C accelerated _elementtree module now initializes hash
  randomization salt from _Py_HashSecret instead of libexpat's default
  CSPRNG.

For more details, see the NEWS file:
https://github.com/python/cpython/blob/v2.7.16/Misc/NEWS.d/2.7.16rc1.rst

Refresh patches, drop now upstream
package/python/0035-bpo-35746-Fix-segfault-in-ssl-s-cert-parser-GH-11569.patch
and adjust hash of LICENSE file for a change of copyright years.

run-tests results:
16:05:41 TestPython2                              Starting
16:05:42 TestPython2                              Building
16:11:26 TestPython2                              Building done
16:11:32 TestPython2                              Cleaning up
.
----------------------------------------------------------------------
Ran 1 test in 351.905s

OK

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

package/wpebackend-fdo: package needs wayland libegl

Now wpebackend-fdo depends only on HAS_LIBEGL but if libegl is not of
wayland type, build breaks due to different EGLNativeWindowType types
depending on backends(fb,x11,wl).

Modify:
'depend on BR2_PACKAGE_HAS_LIBEGL'
to:
'depend on BR2_PACKAGE_HAS_LIBEGL_WAYLAND'
in Config.in to avoid building if there is not a wayland egl backend.
Modify also comment in Config.in when package is not selectable
mentioning the need to have an OpenEGL-capable Wayland backend.

Fixes:
http://autobuild.buildroot.net/results/4f02b91f6ffffd194e09ed18c917b4f678b1a52d/

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
Reviewed-by: Adrian Perez de Castro <aperez@igalia.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

Update for 2019.05-rc2

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

DEVELOPERS: add Giulio Benetti to Minicom package

Signed-off-by: Giulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libtorrent-rasterbar: fix build on m68k_cf

An internal compiler error is raised on m68k_cf at dwarf2cfi.c:2802 in
connect_traces. Error can be fixed by adding -fno-defer-pop, see
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58864

Fixes:
 - http://autobuild.buildroot.org/results/746f60fadb40288341c1e5642cfcf6373b70583a

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/lynx: remove duplicate dependency for host-pkg-conf

host-pkg-conf is already in default lynx's dependencies so remove it
from openssl condition block.

Signed-off-by: Vadim Kochan <vadim4j@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/lynx: fix build with libidn

In case of BR2_STATIC_LIBS=y the linking with libidn fails because
linker can't resolve libiconv symbols needed by libidn. Fix it by
required LIBS for libidn generated by pkg-conf.

The issue can be reproduced with the following defconfig:

BR2_x86_i686=y
BR2_STATIC_LIBS=y
BR2_PACKAGE_LIBICONV=y
BR2_PACKAGE_LIBIDN=y
BR2_PACKAGE_LYNX=y

Fixes:
http://autobuild.buildroot.net/results/23a421e15c32b17ff2f69f183a2e8620ecb93316/

Signed-off-by: Vadim Kochan <vadim4j@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>