package/x11r7/xlib_libxkbfile: bump version to 1.1.0

Added all hashes provided by upstream and license hash.

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

package/x11r7/xlib_libXxf86dga: bump version to 1.1.5

Added all hashes provided by upstream and license hash.

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

package/x11r7/xlib_libXvMC: bump version to 1.0.11

Added all hashes provided by upstream and license hash.

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

package/x11r7/xlib_libXrandr: bump version to 1.5.2

Added all hashes provided by upstream and license hash.

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

package/x11r7/xlib_libXmu: bump version to 1.1.3

Added all hashes provided by upstream and license hash.

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

package/x11r7/xlib_libXft: bump version to 2.3.3

Added all hashes provided by upstream and license hash.

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

package/x11r7/xlib_libXext: bump version to 1.3.4

Added all hashes provided by upstream and license hash.

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

package/x11r7/xdriver_xf86-video-ati: bump version to 19.0.1

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

package/x11r7/xdriver_xf86-video-amdgpu: bump version to 19.0.1

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

package/x11r7/xapp_x11perf: bump version to 1.6.1

Added all hashes provided by upstream and license hash.

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

putty: security bump to version 0.71

CVE-2019-9894: A remotely triggerable memory overwrite in RSA key
exchange can occur before host key verification.

CVE-2019-9895: A remotely triggerable buffer overflow exists in any kind
of server-to-client forwarding.

CVE-2019-9897: Multiple denial-of-service attacks that can be triggered
by writing to the terminal.

CVE-2019-9898: Potential recycling of random numbers used in
cryptography.

Disable static build for now. When building statically configure defines
NO_GSSAPI. Build with NO_GSSAPI is currently broken. The issue has been
reported upstream.

Cc: Alexander Dahl <post@lespocky.de>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libcpprestsdk: add optional brotli dependency

Official brotli support has been added in version 2.10.8 and
https://github.com/Microsoft/cpprestsdk/commit/f3824aa8d01705656e7ee4c0c49ae6980c868b20

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

package/intel-gmmlib: needs dynamic library

Fixes:
 - http://autobuild.buildroot.org/results/0121befb9b4ba6cffc47ce7f861ed81077fc900f

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

package/dropbear: add upstream patch fixing TTY mode reset regression

2019.77 added a regression related to resetting TTY mode on exit.  Add an
upstream patch to fix it.

For more details, see the mailing list:
https://lists.ucc.gu.uwa.edu.au/pipermail/dropbear/2019q1/002157.html

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

package/moarvm: bump to version 2019.03

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

package/radvd: bump to version 2.18

Add hash for license file

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

package/dropbear: bump version to 2019.77

Drop patches as they are now upstream.  Add a hash for the license file.

Verified that runtime test still works:

./support/testing/run-tests -o tests.package.test_dropbear
20:42:44 TestDropbear                             Starting
20:42:45 TestDropbear                             Building
20:44:18 TestDropbear                             Building done
20:44:24 TestDropbear                             Cleaning up
.
----------------------------------------------------------------------
Ran 1 test in 100.727s

OK

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

package/x11r7/xlib_libXdmcp: security bump version to 1.1.3

Fixes CVE-2017-2625:
https://lists.x.org/archives/xorg-announce/2019-March/002974.html

Added all hashes provided by upstream and license hash.

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

package/asterisk: security bump to version 16.2.1

Fixes the following security issue:

AST-2019-001: Remote crash vulnerability with SDP protocol violation
When Asterisk makes an outgoing call, a very specific SDP protocol violation
by the remote party can cause Asterisk to crash (CVE-2019-7251)

https://downloads.asterisk.org/pub/security/AST-2019-001.html

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

configs/freescale_imx8mqevk: bump packages to imx_4.14.78_1.0.0_ga

Bump kernel, U-Boot and ATF packages to the NXP imx_4.14.78_1.0.0_ga
release.

Signed-off-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/supertux: fix build on powerpc64le

Fixes:
 - http://autobuild.buildroot.org/results/c484079b2736eb3c21adff257f3e3ab1acc67f9a

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

package/keepalived: fix build with kernel < 4.1

Fixes:
 - http://autobuild.buildroot.org/results/a33433abeb122cfb15f7f21ab777e84040bdcb8b

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

lvm2: Update to 2.02.183

In particular update to solve an issue where LVM would fail to
initialise LVM devices when asynchronous IO was not available and it
would fail to fall back to synchronous IO. [0][1]

[0] https://bugzilla.redhat.com/show_bug.cgi?id=1650652
[1] https://github.com/open-power/petitboot/issues/60

Signed-off-by: Samuel Mendoza-Jonas <sam@mendozajonas.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/eigen: bump to 3.3.7

Bumped the version of eigen to 3.3.7, which fixes issues when compiling with GCC7 (notably int-in-bool-context errors, fixed in 3.3.5), in addition to various other issues noted in the eigen change logs: http://eigen.tuxfamily.org/index.php?title=ChangeLog
Legal hashes are unchanged in 3.3.7

Signed-off-by: Reed Nightingale <reed.nightingale@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/qt5base: fix xkbcommon option

- Qt 5.6: according to configure --help the command line option to
  enable x11 xkbcommon support is '-system-xkbcommon-x11'

- Qt 5.12: according to configure --help the command line option to
  enable xkbcommon support is '-xkbcommon'

Fixes [1]:

  ERROR: Invalid value given for boolean command line option 'xkbcommon'.

[1] http://autobuild.buildroot.net/results/87076d03bb92a9940447ca80fc1f202e35ef39da

Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libopenssl: bump to version 1.1.1b

Drop patches 4..6 as they are now upstream.

Update the hash of the license file as the copyright dates changed.

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

package/libseccomp: security bump to version 2.4.0

>From the advisory:

Jann Horn  identified a problem in current versions of
libseccomp where the library did not correctly generate 64-bit syscall
argument comparisons using the arithmetic operators (LT, GT, LE, GE).
Jann has done a search using codesearch.debian.net and it would appear
that only systemd and Tor are using libseccomp in such a way as to
trigger the bad code.  In the case of systemd this appears to affect
the socket address family and scheduling class filters.  In the case
of Tor it appears that the bad filters could impact the memory
addresses passed to mprotect(2).

The libseccomp v2.4.0 release fixes this problem, and should be a
direct drop-in replacement for previous v2.x releases.

https://www.openwall.com/lists/oss-security/2019/03/15/1

v2.4.0 adds a new scmp_api_level utility, so update 0001-remove-static.patch
to match.

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

gummiboot: Add to DEVELOPERS file

Signed-off-by: Esben Haabendal <esben@haabendal.dk>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libssh2: security bump to latest git

Bump the version to latest git to fix the following security issues:

CVE-2019-3855
 Possible integer overflow in transport read allows out-of-bounds write
 URL: https://www.libssh2.org/CVE-2019-3855.html
 Patch: https://libssh2.org/1.8.0-CVE/CVE-2019-3855.patch

CVE-2019-3856
 Possible integer overflow in keyboard interactive handling allows
 out-of-bounds write
 URL: https://www.libssh2.org/CVE-2019-3856.html
 Patch: https://libssh2.org/1.8.0-CVE/CVE-2019-3856.patch

CVE-2019-3857
 Possible integer overflow leading to zero-byte allocation and out-of-bounds
 write
 URL: https://www.libssh2.org/CVE-2019-3857.html
 Patch: https://libssh2.org/1.8.0-CVE/CVE-2019-3857.patch

CVE-2019-3858
 Possible zero-byte allocation leading to an out-of-bounds read
 URL: https://www.libssh2.org/CVE-2019-3858.html
 Patch: https://libssh2.org/1.8.0-CVE/CVE-2019-3858.patch

CVE-2019-3859
 Out-of-bounds reads with specially crafted payloads due to unchecked use of
 `_libssh2_packet_require` and `_libssh2_packet_requirev`
 URL: https://www.libssh2.org/CVE-2019-3859.html
 Patch: https://libssh2.org/1.8.0-CVE/CVE-2019-3859.patch

CVE-2019-3860
 Out-of-bounds reads with specially crafted SFTP packets
 URL: https://www.libssh2.org/CVE-2019-3860.html
 Patch: https://libssh2.org/1.8.0-CVE/CVE-2019-3860.patch

CVE-2019-3861
 Out-of-bounds reads with specially crafted SSH packets
 URL: https://www.libssh2.org/CVE-2019-3861.html
 Patch: https://libssh2.org/1.8.0-CVE/CVE-2019-3861.patch

CVE-2019-3862
 Out-of-bounds memory comparison
 URL: https://www.libssh2.org/CVE-2019-3862.html
 Patch: https://libssh2.org/1.8.0-CVE/CVE-2019-3862.patch

CVE-2019-3863
 Integer overflow in user authenicate keyboard interactive allows
 out-of-bounds writes
 URL: https://www.libssh2.org/CVE-2019-3863.html
 Patch: https://libssh2.org/1.8.0-CVE/CVE-2019-3863.txt

Drop 0003-openssl-fix-dereferencing-ambiguity-potentially-caus.patch as that
is now upstream.

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

package/libfuse: Install udev rules and set permissions

This fixes some omissions from the installation.

Install the udev rules.

Tell buildroot about the fuse device.

Apply setuid permissions on the fusermount tool.

Signed-off-by: Norbert Lange <norbert.lange@andritz.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

Revert "package/meson: bump version to 0.50.0"

This reverts commit 114e9dcd28e1001b74689215ec669b8940dc3ea9.
Indeed, there is major issue with this version because c_std is not
passed anymore when cross-compiling:
https://github.com/mesonbuild/meson/issues/5097

As a result, some meson packages (at least systemd, libmpdclient,
glib-networking) fail to build

Fixes:
 - http://autobuild.buildroot.org/results/9eae2181fb3fcfe12481e5496b7d87c0dcd109eb
 - http://autobuild.buildroot.org/results/e828c7925d9247e14d8a8694febfee8ce6c86e81
 - http://autobuild.buildroot.org/results/031aee7cf12ad882727eb2da5e953fd8813b52e2

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

system/skeleton: remove password expiration from shadow

The fields in /etc/shadow were set as follows:

root::10933:0:99999:7:::

This sets the date of last password change to Jan 1, 2000, the minimum
password age to 0 days, the maximum password age to near-infinity, and a
warning period of 7 days. In practice, this means the password never
expires. So all of this is quite useless.

On the other hand, mkusers creates lines without all of these options.
It just sets ::::: which disables password expiration completely.

To make things consistent, do the same for the skeleton entries.

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

package/optee-test: fix dependency in TAs build

Fix TA build command to build TA libraries before TAs.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

boot/arm-trusted-firmware: support debug mode

When the trusted firmware is built with debug support (DEBUG defined),
the generated images are located in a different path compared to a
build without debug support. The non debug images are located in
generated directory build/<platform>/release/ while the debug images
are located in generated directory build/<platform>/debug/.

This change introduces the boolean option
BR2_TARGET_ARM_TRUSTED_FIRMWARE_DEBUG to define whether the release or
debug configuration is used to build trusted firmware.

Note that enabling trusted firmware debug support using
BR2_TARGET_ARM_TRUSTED_FIRMWARE_ADDITIONAL_VARIABLES="... DEBUG=1 ..."
would not work since Buildroot will try to copy the generated files
from the wrong path.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

boot/arm-trusted-firmware: add support for using OP-TEE as BL32

This change allows one to build trusted firmware (TF-A) with OP-TEE as
BL32 secure payload.

When BR2_TARGET_ARM_TRUSTED_FIRMWARE_BL32_DEFAULT is enabled TF-A
builds a BL32 stage according the TF-A configuration directive.  If
these specify no BL32 stage then TF-A will build without BL32
support. This is the default configuration and reflects TF-A legacy
integration in BR.

When BR2_TARGET_ARM_TRUSTED_FIRMWARE_BL32_OPTEE is enabled TF-A builds
with support for the OP-TEE OS as BL32.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
[Thomas: rename config options, simplify option prompts, and rework
option description.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

boot/arm-trusted-firmware: support 32bit Arm targets

This change allows one to build the TF-A (Trusted Firmware A) for
32bit architectures Armv7 and Armv8 platforms. TF-A supports these
configuration since release v1.5.

Change dependency on BR2_aarch64 into a dependency on either
BR2_ARM_CPU_ARMV8A/BR2_ARM_CPU_ARMV7A to better define
the target architectures supported by the package.

Set package configuration directive ARM_ARCH_MAJOR according to
BR2_ARM_CPU_ARMV8A/BR2_ARM_CPU_ARMV7A. Set package configuration
directive ARCH based on BR2_aarch64/BR2_arm.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

gummiboot: fix compatibility with newer glibc versions

Signed-off-by: Esben Haabendal <esben@haabendal.dk>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

gummiboot: upgrade to last commit before being removed

This upgrades to version 48 plus the last commits done after that.
Note, that after that, gummiboot is officially dead, so there will
be no more upstream commits after this.

For various reasons, there are still users of gummiboot, so let's base our
work on the latest gummiboot "version".

Signed-off-by: Esben Haabendal <esben@haabendal.dk>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

gummiboot: use new official upstream git repository

Signed-off-by: Esben Haabendal <esben@haabendal.dk>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/jq: security bump to version 1.6

- Fix CVE-2015-8863 and  CVE-2016-4074:
  https://github.com/stedolan/jq/issues/1406
- Add hash for license file
- Disable oniguruma (enabled by default)

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

package/libuv: bump to version 1.27.0

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

package/uftp: openssl is optional, not mandatory

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

package/uftp: fix openssl static linking

Fixes:
 - http://autobuild.buildroot.org/results/051cfd993b65830c34e675d797e241c272b6f35a

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

package/mariadb: security bump to version 10.3.13

Release notes:
https://mariadb.com/kb/en/library/mariadb-10313-release-notes/

Changelog:
https://mariadb.com/kb/en/mariadb-10313-changelog/

Fixes the following security vulnerabilities:

CVE-2019-2510 - Vulnerability in the MySQL Server component of Oracle MySQL
(subcomponent: InnoDB). Supported versions that are affected are 5.7.24 and
prior and 8.0.13 and prior. Easily exploitable vulnerability allows high
privileged attacker with network access via multiple protocols to compromise
MySQL Server. Successful attacks of this vulnerability can result in
unauthorized ability to cause a hang or frequently repeatable crash
(complete DOS) of MySQL Server.

CVE-2019-2537 - Vulnerability in the MySQL Server component of Oracle MySQL
(subcomponent: Server: DDL). Supported versions that are affected are 5.6.42
and prior, 5.7.24 and prior and 8.0.13 and prior. Easily exploitable
vulnerability allows high privileged attacker with network access via
multiple protocols to compromise MySQL Server. Successful attacks of this
vulnerability can result in unauthorized ability to cause a hang or
frequently repeatable crash (complete DOS) of MySQL Server.

Note that the hash for README.md changed due to Travis CI and Appveyor CI
updates.

Signed-off-by: Ryan Coe <bluemrp9@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/gesftpserver: bump version to latest stable

By default, gesftpserver wants to treat al warnings as errors, but there
is an actual warning in the release, so we quiesce that by not treating
warnings as errors.

We also backport a patch from upstream, to fix the ordering of some
fields when sending file attributes.

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

package/libpcap: fix bluez5_utils-headers dependency

Commit c46afc37dc877f1e94ffde0d77585290711b27ec changed bluez5-utils
dependency by bluez5_utils-headers without replacing the test on
BR2_PACKAGE_BLUEZ5_UTILS by BR2_PACKAGE_BLUEZ5_UTILS_HEADERS

Fix this mistake and also add a select on
BR2_PACKAGE_BLUEZ5_UTILS_HEADERS if BR2_PACKAGE_BLUEZ5_UTILS is set
so the user does not have to do it

Fixes:
 - http://autobuild.buildroot.org/results/c6828df1f3782564451ddd4187ff026679bf37d8

Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Acked-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/angularjs: fix license file hash

The update of angularjs to 1.7.8 in commit
00dcde3eb12d194d31356fd6e54c0328acd371a1 broke legal-info for this
package, because the hash of the license file changed.

In fact, we are using angular.js as the license file and obviously it
is likely to change between each release. The new angular.js still
specifies a MIT license, so we can update the license file hash.

Fixes:

  http://autobuild.buildroot.net/results/0b8ad8d8384d605c2230e862548ccaba1f06d9b0/

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

package/libglib2: add optional dependency in libselinux

If libselinux is selected, explicitly set --enable-selinux in the
configure options and build the library first.

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

linux: allow BR2_LINUX_KERNEL_IMAGE on RISC-V

We will need to build Image files for OpenSBI so allow that now.

Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/uftp: needs MMU, threads and wchar

Fixes:
 - http://autobuild.buildroot.org/results/47dd067dfdcaae99d50216f59ea6af115d8f1e1e

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

package/qt5: bump latest version to 5.12.2

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

package/angularjs: bump version to 1.7.8

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

package/meson: bump version to 0.50.0

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

package/keepalived: drop autoreconf

Commit 78d4ddbf3b416885d2e99b32363b174755f9bb30 removed all patches so
autoreconf is not needed anymore

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

configs/qemu_riscv64_virt: re-order options

Fix the config layout, no functional change.

Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

configs/qemu_riscv32_virt: upgrade to 4.20 kernel

Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/glibc: update glibc version for RISC-V 32-bit

Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/gst1-plugins-bad: add webrtc option

- remove old webrtc Config.in.legacy entry introduced by [1] (misnamed
  webrtc option was introduced with 2017.02, renamed to webrtcdsp for
  2017.08 and although backported to 2017.02.4)

[1] https://git.buildroot.net/buildroot/commit/?id=4c06d2490a07f0b88f42c56c7409899fd2f5608a

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

Revert "package/binutils: install libiberty for host build"

This reverts commit 7ec7ba54055919d89afc93606794d161505fa2d8, as it
causes build failures of host-gdb:

/usr/bin/ld: /home/buildroot/autobuild/instance-0/output/host/lib/libiberty.a(cplus-dem.o): relocation R_X86_64_PC32 against symbol `_sch_istable' can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: final link failed: Bad value

Indeed, the host-gdb build picks up the libiberty installed in
$(HOST_DIR) instead of using its own internal version. This needs to
be addressed before we can make host-binutils install libiberty in
$(HOST_DIR).

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

Revert "package/prelink-cross: new package"

This reverts commit 0bb0f2ba849dac9208d3a5a4107b8edd5ee46a1e, because
using "host-binutils" is not correct as it should only be used with
internal toolchains, and not external toolchains.

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

package/kf5-modemmanager-qt: link with libatomic when needed

On some architectures, atomic binutils are provided by the libatomic
library from gcc. Linking with libatomic is therefore necessary,
otherwise the build fails with:

sparc-buildroot-linux-uclibc/sysroot/lib/libatomic.so.1: error adding symbols: DSO missing from command line

This is often for example the case on sparcv8 32 bit.

Fixes:
 - http://autobuild.buildroot.org/results/b941a3deaa57cac79f1686d47ca6ababf2f0d5e4

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

package/qpdf: bump to version 8.4.0

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

package/python-pylibftdi: bump to version 0.17.0

Add hash for license file

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

Revert "package/python-ipython: bump to version 7.3.0"

This reverts commit bc516052593947a067c98956931df9df31ee3195.
which adds a dependency on python3 as well as adddtional python runtime
modules. So revert this patch to polish a better bump.

Fixes:
 - http://autobuild.buildroot.org/results/7d765d2b15bfe72b05de62e86a01d57f1696c837
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/binutils: upstream fixes for 2.31.1

Combining musl and binutils 2.31.1 will produce static applications
that crash immediately. This commit picks up 3 upstream commits to
remedy this.

See https://sourceware.org/bugzilla/show_bug.cgi?id=23428

Signed-off-by: Norbert Lange <nolange79@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/nodejs: link with libatomic when needed

Fixes:
 - http://autobuild.buildroot.net/results/804951467d3f5b212bd33b95dc8fe9c6f5f63838

Signed-off-by: Martin Bark <martin@barkynet.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/nodejs: correct license file hash

The LICENSE file in nodejs 10.15.3 updated to include additional
3rd party licenses.

Fixes:
 - http://autobuild.buildroot.net/results/4bf66b9f05e9d04171156e44c7dd87619e429747
 - http://autobuild.buildroot.net/results/a65776ba1045f12263a686933eb6a4b5d6ff7333
 - http://autobuild.buildroot.net/results/aab619106a7d763f45e95a7ae3d85816fb4a6512
 - http://autobuild.buildroot.net/results/bb6cb93f323be22e4c0efc05011042526c01f62c

Signed-off-by: Martin Bark <martin@barkynet.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/python-aiorwlock: bump to version 0.6.0

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

package/python-aiojobs: drop aiohttp dependency

aiohttp isn't a required dependency for aiojobs

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

package/libpcap: fix bluez circular dependency

The optional bluez_utils dependency of libpcap creates a circular
dependency:

$ make dbus-show-recursive-depends

Recursion detected for  : systemd
which is a dependency of: dbus
which is a dependency of: bluez_utils
which is a dependency of: libpcap
which is a dependency of: iptables
which is a dependency of: systemd
make: *** [package/dbus/dbus.mk:121: dbus-show-recursive-depends] Error 1

Drop support for bluez_utils. For bluez5_utils, which also depends on
dbus, we only need the headers in the bluez5_utils-headers package. Use
that to break the circular dependency.

Fixes:
http://autobuild.buildroot.net/results/9c3/9c3ee798fa6bb501a20a7892c0b085d2b279b664/

Suggested-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/zlog: bump to version ccf8c36e918d6f3c7e6fe9e36eedf8583014c5c9

- Remove patches (already in version)
- Add hash for license file

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

package/grpc: update to 1.18.0

Bump version.  Freshen a patch that had a conflict.

Remove duplicate env setting.

Add comment about the hack being used to get gRPC to use the buildroot
c-ares library.  Otherwise it looks like the cmake env settings are out
of date vs what the gRPC build documentation says to use.

Signed-off-by: Trent Piepho <tpiepho@impinj.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/python-terminaltables: wrap Config.in help text

Fixes the following check-package error:

package/python-terminaltables/Config.in:4: help text: <tab><2 spaces><62 chars> (http://nightly.buildroot.org/#writing-rules-config-in)

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

configs/imx8mpico: new defconfig

Signed-off-by: Julien Olivain <juju@cotds.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/python-ipython: bump to version 7.3.0

- Update hash for COPYING.rst, it has been updated to remove
  BSD-3-Clause text:
  https://github.com/ipython/ipython/commit/5a6ef3e35db9ac04bf47a17d96484e03370f03dd
- Add LICENSE to license files as this is now the file that contain
  BSD-3-Clause text

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

package/jack2: bump to version 1.9.12

Remove patches (already in version)

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

package/jhead: bump to version 3.0.3

- Remove patch (already in version)
- Add hash for license file

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

package/keepalived: bump to version 2.0.13

Remove patches (already in version)

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

package/wpan-tools: bump to version 0.9

- Switch site to github
- Remove patch (already in version)
- Add hash for license file

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

package/mender-artifact: new package

mender-artifact is a host tool to generate update images
in the Mender artifact file format.
This package uses the binary archive from github because it bundles the
external dependencies.

Example usage:

    $ mender-artifact write rootfs-image \
         --update rootfs.ext4 \
         --output-path rootfs.mender \
         --artifact-name "release-v1.0.0" \
         --device-type "beaglebone"

Above will generate a Mender artifact called "rootfs.mender"
containing the "rootfs.ext4" image along with meta-data.

One can read-out the meta-data with the following command:

    $ mender-artifact read rootfs.mender
      Mender artifact:
      Name: release-v1.0.0
      Format: mender
      Version: 2
      Signature: no signature
      Compatible devices: '[beaglebone]'

    Updates:
      0000:
        Type:   rootfs-image
        Files:
          name:     rootfs.ext4
          size:     52428800
          modified: 2018-08-27 09:10:55 +0200 CEST
          checksum: e70b113fb0964a810a3043586eb4fc1c48e684ba78b02ba65fead4aa3e540d87

Signed-off-by: Angelo Compagnucci <angelo@amarulasolutions.com>
Tested-by: Adam Duskett <aduskett@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/pkg-golang: add support for building host packages

With this you can add:

    $(eval $(host-golang-package))

to a package .mk file to build for host.

Signed-off-by: Mirza Krak <mirza.krak@northern.tech>
Acked-by: Angelo Compagnucci <angelo@amarulasolutions.com>
Tested-by: Angelo Compagnucci <angelo@amarulasolutions.com>
Signed-off-by: Angelo Compagnucci <angelo@amarulasolutions.com>
Tested-by: Adam Duskett <aduskett@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/go: rename BR2_PACKAGE_HOST_GO_{ARCH_SUPPORTS,CGO_LINKING_SUPPORTS}

The hidden Config.in option BR2_PACKAGE_HOST_GO_ARCH_SUPPORTS name is
not very clear as to whether it says whether Go is available for the
target architecture or the host architecture.

Until now, this was fine since there was support for host Go
packages. But as we are about to introduce support for building host
Go packages, we need to clarify the meaning of
BR2_PACKAGE_HOST_GO_ARCH_SUPPORTS. Since it says whether the target
architecture has support for Go or not, we rename it to
BR2_PACKAGE_HOST_GO_TARGET_ARCH_SUPPORTS.

And since BR2_PACKAGE_HOST_GO_CGO_LINKING_SUPPORTS is tightly related,
we rename it to BR2_PACKAGE_HOST_GO_TARGET_CGO_LINKING_SUPPORTS.

Signed-off-by: Angelo Compagnucci <angelo@amarulasolutions.com>
Tested-by: Adam Duskett <aduskett@gmail.com>
[Thomas: entirely rewrite commit log]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/fltk: add optional xlib_libXrender dependency

xlib_libXrender is enabled by default and has been added since version
1.3.4-1 and
https://github.com/fltk/fltk/commit/a6c4b29a184ce7708819f4706877eedcd99a30f5

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

package/pkg-generic: tweak only .la files that need it

Currently, when we tweak the .la files, we do so unconditionally on all
.la files, even those we already fixed in a previous run.

This has the nasty side effect that each .la file will be reported as
being touched by all packages that are installed after the package that
actually installed said .la file.

Since we can't easily know what files were installed by a package (that
is an instrumentation hook, and comes after the mangling), we use a
trick (like is done in libtool?): we do mangle all files, each into a
temporary location; if the result is identical to the input, we remove
the temporary, while if the result differs from the output, we move
the temporary to replace the input.

Reported-by: Nicolas Cavallari <nicolas.cavallari@green-communications.fr>
Reported-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Nicolas Cavallari <nicolas.cavallari@green-communications.fr>
Cc: Thomas De Schampheleire <patrickdepinguin@gmail.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/python-aiomonitor: new package

aiomonitor adds monitor and python REPL capabilities for
asyncio application.

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

package/python-aioconsole: new package

Asynchronous console and interfaces for asyncio.

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

package/python-terminaltables: new package

Generate simple tables in terminals from a nested list of strings.

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

core: add make-based full-dependency list

Currently, when we need to build the full dependency graph, we call make
to show the list of packages (make show-targets), and then call it again
and again iteratively while it returns new packages.

Since calling make will parse the whole set of our Makefiles, this takes
quite a bit of time (~4s each here), and the total can get pretty long.

However, make being make, already builds the whole dependency tree
information, so we can just ask for it.

Add a new top-level rule 'show-dependency-tree' that displays the whole
set of dependencies for all packages. For each package, its name, type
and version is displayed, then all the direct, first-level dependencies
are dumped. We choose a format that is not unlike the dot-graph format,
because it is both easy to read as a human, and easy to parse as a
machine:

    foo: target 1.2.3
    foo -> bar host-meh
    bar: target virtual
    bar -> buz
    buz: target 2.3.4
    buz ->
    host-meh: host virtual
    host-meh -> host-bleark
    host-bleark: host 3.4.5
    host-bleark ->
    rootfs-meh: host
    rootfs-meh -> host-bleark

To be noted: rootfs are currently reported as if they were 'host'
packages, to stay aligned with how graph-depends currently treats them.
Ideally, graph-depends could be enhanced to recognise them separately,
but that is another story.

For just plain defconfig, which is about the smallest config we can have
with an internal toolchain, we already have a seven-fold improvement
(with the graph-depends rule modified to not run the pdf generation, to
be able to just compare the tree generation):

    $ time make graph-depends
    real    0m27.344s
    $ time make show-dependency-tree
    real    0m3.848s

>From defconfig, C++, wchar, locales, ssp, and allyespackageconfig,
tweaked for even more packages (qt5 not qt4, luajit to avoid multi
providers, etc...), the timings are (graph-depends still modified to
not generate the pdf):

    $ time make graph-depends
    real    1m56.459s
    $ time make show-dependency-tree
    real    0m5.748s

There. I don't think those numbers need any explanation whatsoever;
they do speak on their own. OK, for maths sake, the ratio is about
twenty-fold. So, "yeah", I guess... ;-)

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

infra/pkg-generic: use pure Makefile-based recursive dependencies

Calling to the graph-depends script is very costly, as it calls back to
'make' a lot of time.

It turns out that we already have the list of recursive dependencies, so
we can just print that.

As for the recursive reverse dependencies, we use the same memoisation
technique to cut-down on the expansion cost, which would otherwise be on
the order of 𝑶(𝑛²) (with 𝑛 enabled packages).

>From a defconfig, modified to use glibc, C++, wchar, locales, ssp, and
allyespackageconfig (tweaked to avoid multi providers, etc...), the
timings for X-show-recursive-rdepends are:

                    before      after       speedup     #rdeps
    libnss          0m22.932s   0m5.775s     3.97x      3
    qt5base         0m41.176s   0m5.781s     7.12x      67
    libjpeg         0m56.185s   0m5.749s     9.71x      228
    libxml2         0m54.964s   0m5.795s     9.48x      271
    freetype        0m46.754s   0m5.819s     8.07x      287
    libpng          0m53.577s   0m5.760s     9.30x      303
    sqlite          1m15.222s   0m5.807s    12.95x      801
    libopenssl      1m25.471s   0m5.844s    14.63x      931
    readline        1m13.805s   0m5.775s    12.78x      958
    libzlib         1m11.807s   0m5.820s    12.34x      1039
    toolchain       1m23.712s   0m6.080s    13.77x      2107
    skeleton        1m27.839s   0m6.293s    13.96x      2111 (+1)
    host-skeleton   1m27.405s   0m6.350s    13.76x      2172 (+2)

  - speedup: ratio before/after
  - #rdeps: number of recursive reverse dependencies, with the extra
            dependencies returned with this patch, see below for the
            reason.

So, for a low-level package with a lot of reverse dependencies, like
libzlibz, libopenssl or readline are, the timings are already very much
in favour of the change. This is less impressive with packages that
have few dependencies (libnss), but still much faster.

Also, remember that the config tested has as much packages enabled as
possible, so is in itself a degenerate case. With simpler and more
realistic configurations, the gains would probably be a bit lower than
reported above, but various tests still report good improvements
overall (note: coming up with a 'realistic' configuration is pretty
hard, as everyone and their dog have their notion of what is realistic
in their context, so nothing displayed here; timings are left as an
exercise for the interested parties to report aggravation in their
cases should they notice some regression).

Note that, more recursive reverse dependencies may be displayed now,
since we do not apply the exceptions applied in graph-depends. For
example, host-skeleton gains two new recursive reverse dependencies:
skeleton and toolchain, which are both exceptions in graph-depends.

As for direct (not reverse) dependencies: the gain is not as fantastic
as for reverse ones, but it is still noticeable, especially thanks to
a21212fb7cf (package/pkg-generic: speed up RECURSIVE_FINAL_DEPENDENCIES);
just a few examples for %-show-recursive-depends:

                    before      after       speedup     #deps
    libzlib         0m46.864s   0m5.902s     7.94x      17
    qt5base         0m57.590s   0m5.848s     9.85x      190
    sqlite          0m46.601s   0m5.816s     8.01x      24

Basically, displaying recursive dependencies, direct or reverse, is
almost a constant now: it only slightly varies by about 10% depending
on the complexity of the dependency chain, with the parsing of the
Makefiles still accounting for the large majority of the time.

(PS. Thanks to Joseph for suggesting a list of interesting packages
to test, and thanks to Trent for his example of memoisation!)

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Thomas De Schampheleire <patrickdepinguin@gmail.com>
Cc: Joseph Kogut <joseph.kogut@gmail.com>
Cc: Trent Piepho <tpiepho@impinj.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

support/graph-depends: also cut on host-gzip

When host-gzip is needed, it is a mandatory dependency of all packages.
As such, drawing the dependency lines toward host-gzip would uselessly
clutter the graph.

So, like for the skeleton, host-skeleton, and host-tar, we cut the
dependency chains toward host-gzip.

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

support/graph-depends: also cut on host-tar

When host-tar is needed, it is a mandatory dependency of all packages.
As such, drawing the dependency lines toward host-tar would uselessly
clutter the graph.

So, like for the skeleton and host-skeleton, we cut the dependency chains
toward host-tar.

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

support/graph-depends: also cut on host-skeleton

host-skeleton is a dependency of almost all packages, except a very few.
As such, it clutters the dependency graph uselessly.

Do with it as we do for the skeleton: cut the dependency chains.

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

support/graph-depends: add option to exclude mandatory deps

Some times, multiple dependency graphs for a set of packages (mostly
the application-level packages for the project) are included in reports
(e.g. delivery notes). Repeating the mandatory dependencies on all
those graphs is useless and clutters the important dependencies.

When we had only two such mandatory dependencies (toolchain, skeleton),
it was manageable to list them as manual exclusions:
    -x toolchain -x skeleton

But we now have quite a few such dependencies, and it becomes a bit more
cumbersome to manage, not counting the ones we may add in the future.

Add an option to exclude all those mandatory dependencies, to generate
neat graphs.

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

support/graph-depends: make sure mandatory deps are displayed

The current graph-depends implementation filters out a number of
"mandatory" dependencies that all packages have: dependency on
"toolchain" and dependency on "skeleton".

Despite this filtering, in full graph dependencies, "toolchain" and
"skeleton" are still shown, because they are target packages, and
therefore appear in the result of "make show-targets". Thanks to this,
they will be visible as dependencies of the "ALL" node, which is the
root of the dependency tree.

However, as we are going to introduce host-skeleton as a "mandatory
dependency" to be filtered out, this is no longer going to work.

This commit adjusts the remove_extra_deps() function to ensure that
when a mandatory dependency is removed, this dependency exists between
the root of the dependency tree and the mandatory dependency.

This issue was noticed by Yann E. Morin, and this commit provides a
different implementation than what Yann proposed in
https://patchwork.ozlabs.org/patch/910453/.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
[yann.morin.1998@free.fr:
  - list mandatory deps before removing them
  - fix flake8 warnings
]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/openjdk: support zero hotspot variant

The OpenJDK package supports a non-assembler variant of the hotspot
virtual machine.  This allows archtectures which don't have hard-coded
support (which provides better performance) to still build a virtual
machine that can execute JAVA.

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

package/systemd: drop optiopns that are now conditional

Commit 6f28ce53222 (package/systemd: enable building of systemd-boot)
made the efi and gnu-efi optiona conditional upon whether systemd-boot
is built or not, but forgot to remove the non-conditional ones.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: James Hilliard <james.hilliard1@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

package/xenomai: drop empty line at end of .hash file

Fixes the following check-package warning:

package/xenomai/xenomai.hash:3: empty line at end of file

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

support/testing/tests/package/test_openjdk: new test

This test is a simple "Hello, World" integration test of the OpenJDK
package.

It compiles the Java app on the host, then runs it on an emulated
AARCH64 target and verifies "Hello, World" is printed.

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

package/openjdk: use AArch64 optimizations

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

package/openjdk: new package

OpenJDK is a free and open-source implementation of the Java Platform.
This package provides the option to build a client or a server JVM
interpreter.

The default option is the server option, as that is what the majority
of users use. This JVM interpreter loads more slowly, putting more
effort into JIT compilations to yield higher performance.

Unlike most autotools packages, OpenJDK is exceptionally different and
has many quirks, some of which are below:

- X11, alsa, and cups are required to build Java, even if it's a headless build.
  See
  http://hg.openjdk.java.net/jdk10/jdk10/raw-file/tip/common/doc/building.html#external-library-requirements
  for more information.

- host-zip is needed for the zip executable.

- There is no autogen.sh file, instead, a user must call "./configure
  autogen."

- OpenJDK ignores some variables unless passed via the environment.
  These variables are: PATH, LD, CC, CXX, and CPP.

- OpenJDK defaults ld to the ld binary but passes -Xlinker and -z as
  arguments during the linking process, which causes linking failures.
  To fix this issue, ld is set to gcc.

- Make -jn is unsupported. Instead, one must use the "--with-jobs="
  configure option, and use $(MAKE1).

Signed-off-by: Adam Duskett <Aduskett@gmail.com>
[Thomas:
 - drop explanations about CC, LD, CXX, etc. be set to their "actual
   binaries" instead of ccache: TARGET_CC/TARGET_LD/TARGET_CXX point
   to the compiler wrapper, so the usage of ccache is hidden
 - make sure at least one of the variants is enabled in Config.in
 - drop the submenu for variant selection
 - use system zlib instead of the bundled one. This works fine when
   BUILD_SYSROOT_CFLAGS and BUILD_SYSROOT_LDFLAGS are passed
 - fix minor nits in the Config.in comments]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>