In particular, I am interested in a relatively recent cmake, which has
interesting features.

Should fix broken RPi CI build

Required for some of our drivers

This commit enables On-the-go Ethernet over the USB-C connector. This
allows the user to simply connect their laptop to the Pi using a USB-C
cable, get an IP address on the USB interface and immediately start
working on the Pi, reducing setup time.

The following changes were made:

- Enable dwc2, which is a USB driver handling the OTG events. This also
  required changing the way the config.txt file is generated.
- Enable g_ether, which is the USB gadget driver for USB Ethernet.
- Configure systemd to start a DHCP server on usb0.

It appears to have better performance and stability on that type of
devices

This commit backports an upstream patch made for gnupg2 into gnupg, in
order to fix build failures with gcc 10 due to the use of
-fno-common. Due to the code differences between upstream gnupg2 and
the old gnupg 1.x, the backport is in fact more a rewrite than an
actual backport.

Fixes:

  http://autobuild.buildroot.net/results/496a18833505dc589f7ae58f2c7e5fe80fe9af79/



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

Installing qt5declarative examples on fast/fast/multicore machines sometimes
failes with a variation of the following error messages:

 - Cannot touch [...]/chapter5-listproperties/app.qml: No such file or directory
 - Error copying [...]/chapter2-methods/app.qml: Destination file exists

Fix it by using OTHER_FILES instead of a seperate qml files install target
to fix the race between install_target, install_qml and install_sources.

Fixes:

 - https://gitlab.com/buildroot.org/buildroot/-/jobs/565470221



Signed-off-by: Romain Naour <romain.naour@gmail.com>
[Reworked patch and commit log]
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

Added upstream patch for fixing build failure when using GCC10 as a host
compiler (-fno-common is now default).

Fixes:
http://autobuild.buildroot.net/results/47f/47fcf9bceba029accdcf159236addea3cb03f12f/



Cc: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Heiko Thiery <heiko.thiery@gmail.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

Added upstream patch for fixing build failure when using GCC10 as a host
compiler (-fno-common is now default).

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

In version 5.6 a minor change was made to this file, stating tht "[a]ll
contributions to the Linux Kernel are subject to this COPYING file",
and hence the hash changed.

We can update the hash, because the licensing information is only
accounted for the "latest" version, so the hash change will not impact
older kernel versions as the user would have to switch to a non-latest
kernel.

Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

This patch was wrongly removed when bumping the version to 1.4.0 in
commit 6976f312

Fixes:
 - http://autobuild.buildroot.org/results/7a53a59dd08c043f371bea967c3b450a7bddcde8



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

The default iamge size is 32MiB, which is quite low by today's standards.
Besides, the AArch64 kernels are relatively big, which leaves not much
room, if at all, for users to experiment on the default image.

Increase the vfat size to a more reasonable 64MiB.

Note that users who derive an in-tree defconfig for their own case will
allways hit any arbitarary size we put here, so they will anyway have to
also derive this template for their own use-cases.

Signed-off-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

As reported by Nicolas Carrier on the Buildroot mailing list [1],
there is a new build issue while building a program which interacts with
the u-boot environment. This program uses the headers of the ubootenv
library provided by uboot-tools.

This is an upstream change from uboot [2] adding "#include <env.h>" to
fw_env.h. Adding env.h require a board configuration to build.

But only fw_env.h header is installed in the staging directory by
uboot-tools package, but since it now include env.h the build is broken
because env.h is missing from the staging directory.

It's seems an upstream bug since env_set() is not used in fw_env tool.
Nicolas removed env.h from fw_env tool and fixed it's build issue.

This problem is present since uboot v2019.10, so the uboot version
present in Buildroot 2020.02 is affected.

It's probably not a problem for upstream uboot but it's a problem
for uboot-tools package that build uboot tools without a board
configuration for the target.

[1] http://lists.busybox.net/pipermail/buildroot/2020-April/280307.html
[2] https://gitlab.denx.de/u-boot/u-boot/-/commit/9fb625ce05539fe6876a59ce1dcadb76b33c6f6e



Reported-by: Nicolas Carrier <nicolas.carrier@orolia.com>
Signed-off-by: Romain Naour <romain.naour@gmail.com>
[yann.morin.1998@free.fr: add URL to upstream commit]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

Added upstream patch for fixing build failure when using GCC10 as a host
compiler (-fno-common is now default).

Fixes:
http://autobuild.buildroot.net/results/c4b/c4bba80e9fc476247c7ba28850831c6a8edd559f/build-end.log



Cc: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Heiko Thiery <heiko.thiery@gmail.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>

Pull a patch pending in an upstream pull request to fix the detection
of the snappy library when we are in static linking configurations.

Fixes:

  https://bugs.busybox.net/show_bug.cgi?id=12671



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

snappy is not a mandatory dependency to build leveldb. Back when it
was introduced in Buildroot, as of version 1.18, the build logic
already made snappy an optional dependency.

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

Libdrm freedreno depends on BR2_arm || BR2_aarch64 || BR2_aarch64_be
as such we need to propagate those dependencies to mesa's gallium
freedreno driver.

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

According to https://prosody.im/doc/depends#bitop the correct bitop
package to use with prosody for Lua 5.1 is:
https://luarocks.org/modules/siffiejoe/bit32



As such replace BR2_PACKAGE_LUABITOP with BR2_PACKAGE_LUA_BIT32

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

So far in 2020, Logilin and Tap2Open made some financial donations to
the Buildroot Association, so let's thank them on our sponsors page.

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

Hash was not updated by commit 18079e20

Fixes:
 - http://autobuild.buildroot.org/results/0f7179ed4706f05551af330d7f12b3efaeffd278



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