As reported by Sébastien Szymanski [1], the apply-patches script
doesn't stop if a tar command can't extract an archive.
Use "set -e" to exit immediately if a command return an error.
[1] http://patchwork.ozlabs.org/patch/626196
Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: Sébastien Szymanski <sebastien.szymanski@armadeus.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This makes it possible to clone git repositories hosted on Github while
behind a firewall.
Signed-off-by: Paul Cercueil <paul.cercueil@analog.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
For some platforms, there are multiple generated spl images. Extend
BR2_TARGET_UBOOT_SPL_NAME to allow these multiple images to be installed
after uboot build completes.
For example, the NextThingCo C.H.I.P. uses two binaries from uboot,
spl/sunxi-spl.bin and spl/sunxi-spl-with-ecc.bin.
Signed-off-by: Jason Abele <jason@nextthing.co>
[Maxime:
- Add foreach loop for general case and mkpimage
- Use firstword for zynq case]
Signe-off-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Signed-off-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
[Thomas: minor tweaks to Config.in.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The two previous patches have been merged upstream, so they have been
deleted here. Another patch (picked from upstream) has been added to fix
the build with thread-less toolchains.
Libiio v0.7 provides two new backends, a USB backend (using libusb-1.0)
and a serial backend (using libserialport).
Additionally, it is now possible to compile libiio with thread-less
toolchains. In that case, thread safety is disabled.
Signed-off-by: Paul Cercueil <paul.cercueil@analog.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
We now have a real file in that directory, so we do not need a .empty
placeholder.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Register package-specific target-finalize hooks with the
newly-introduced <PKG>_TARGET_FINALIZE_HOOKS.
This incidentally fixes luarocks, which was registering target-finalize
hooks even when it was not enabled.
To be noted, the skeleton package is not converted, because it is not
optional, we always have it; so its hooks would always be registered
anyway. Besides, the followup patches would render this conversion moot
anyway, since those hooks would be spread across the various skeleton
packages.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Currently, packages using target finalize hooks must remember that they
need to register their hook in TARGET_FINALIZE_HOOKS
conditionally (otherwise their hook will be triggered even if the
package is disabled).
In order to avoid this potential mistake, this commit introduces a
per-package target-finalize hook variable, in which packages can
register their target-finalize hooks, with the guarantee that they will
only be triggered if the package is enabled.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
[Thomas: rework commit log.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The hostname does not look like it serves any useful purpose, except
maybe to set set perladmin email.
Which is undoubtfully useless on the target.
A followup commit will make the hostname depend on the default skeleton,
so it won't always be available. We can not rely on it to be set.
Besides, even today it is not guaranteed to be set; a user may well
leave it empty.
Use a dummy hostname.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Marcin Niestroj <m.niestroj@grinn-global.com>
Reviewed-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Supporting a R/O rootfs needs a bit more love to be fully operational.
This will come in later patches...
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Currently, our /etc/mtab points to /proc/mounts. This was all neat so
far, and was good for a sysv-like init system.
However, the way today is to point it at /proc/self/mounts, the
per-process mount tab.
Additionnally, that's what systemd expects. If /etc/mtab is not a
symlink to ../proc/self/mounts and the rootfs is readonly, systemd would
whine loudly (and a service unit would be marked failed).
Since it works well for sysv-like init systems too, just use that.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
In makefiles, variables have global scope. We traditionally ensure that
we get no name clashing by prefixing all variables with the package
name.
Currently, this is not the case in the skeleton package, for historical
reasons (code snippets copied over from the common Makefiles). We
currently have a mix of naming for the variables:
- some are indeed prefixed with SKELETON_
- some are prefixed with SYSTEM_
- some are prefixed with both
- some are not prefixed
Clean up these discrepancies, and prefix all variables with just
SKELETON_ and drop the SYSTEM_ prefix.
Also include SET_ in all variables that do set something (getty) for
consistency across all the variables.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The 2.26 binutils tarball has incorrect timestamps for the .info files,
so binutils build system tries to regenerate them with makeinfo. In
order to avoid depending on host-texinfo, we simply touch the .info
files so that their timestamp is newer than the corresponding source
files.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Tested-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
gdb versions fetched from git (such as the ARC version) will not build
with host-flex and host-bison.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Just like binutils, when gdb is fetched from git, it wants to
re-generate its documentation. We were trying to avoid that by patching
the Makefiles, but it wasn't working in all situations.
So, we simply add a dependency on host-texinfo when gdb is fetched from
git, and remove our hackhish work-around.
Fixes:
http://autobuild.buildroot.net/results/80e63edc009d23cb939880a6f5ae2f220f07a96a/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This commit adds the dependency on host-texinfo to binutils, when the
source is fetched from Git. Thanks to this, we can remove the
0800-Docs-Prevent-build-failures-when-makeinfo-is-missing.patch patch
from the ARC-specific version of binutils.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
We have tried multiple solutions to solve the "makeinfo" problem in
binutils and gdb, without finding anything really convincing. So it
seems like the easiest solution is to add a texinfo host package, and
use it as a dependency of binutils/gdb when they are fetched from
Git. This commit introduces just the host-texinfo package.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Gallium driver also provides DRI-like infrastructure and OpenGL.
This change prevents from removing pkg-configs of them also when
installing the driver.
Signed-off-by: Akihiko Odaki <akihiko.odaki.4i@stu.hosei.ac.jp>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This ensures the correct AR is used for some of the iptables build
process. This is the same fix as
<https://bugs.gentoo.org/show_bug.cgi?id=444282>
Signed-off-by: Jordan Yelloz <jordan@yelloz.me>
[Thomas: use upstream patch instead of passing AR in the environment.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Commit 7b17bafc5d by Davide Viti has a
detailed explanation of some unusual techniques used for building
host-cmake and (target-)cmake. This is useful information for whoever
starts hacking on it, so copy it in the makefile, where it will be
easily noticed.
Also remove the sentence about host-cmake having a runtime dependency
on host-pkgconfig (not true anymore: it's the specific cmake-packages
that depend on it) and fix typos.
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Samuel Martin <s.martin49@gmail.com>
Cc: Davide Viti <zinosat@tiscali.it>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
In 3d475ee0ba a dependency on
host-pkgconf was added to host-cmake. It is a workaround to fix build
failures for packages that use pkgconf through a cmake module, but do
not depend on host-pkgconf as they should.
Since it is the package that needs host-pkgconf and not host-cmake
itself, move the dependency to the proper place, in pkg-cmake.mk.
Also copy the explanation on the mentioned commit as a comment in
order to clarify why we do this.
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Samuel Martin <s.martin49@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Davide Viti <zinosat@tiscali.it>
Cc: Arnout Vandecappelle <arnout@mind.be>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Thomas:
- update on top of master
- drop empty HOST_CMAKE_DEPENDENCIES, no longer needed since host
dependencies are no longer derived from target dependencies.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Enable Linuxthreads, a small binutils patch from upstream is
required.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This is required to build uClibc-ng with threading support and FDPIC
binary format.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since commit 0dca644e44 ("boot/uboot: fix
missing host-openssl for i.MX28 target"), we pass HOSTCFLAGS/HOSTLDFLAGS
in the environment of U-Boot when building. This is needed to allow
U-Boot to find the OpenSSL headers/libraries when they are
needed. Unfortunately, this breaks the build with old U-Boot versions as
the U-Boot Makefiles were not designed to have HOSTCFLAGS passed in,
which causes some important CFLAGS from the U-Boot build system to be
ignored.
As suggested by Arnout, we pass the HOST_CFLAGS inside HOSTCC directly,
which allows to pass the CFLAGS without overriding the internal U-Boot
CFLAGS.
Tested with an old U-Boot (2012.10), and a modern U-Boot in a
configuration that needs OpenSSL.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This new package provides jemalloc, a malloc(3) implementation that
emphasizes fragmentation avoidance and scalable concurrency support.
A host variant is added as it will be used by rust.
Signed-off-by: Eric Le Bihan <eric.le.bihan.dev@free.fr>
[Thomas:
- remove trailing slash (noticed by Romain Naour)
- add !BR2_STATIC_LIBS dependency (noticed by Romain Naour).]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
docker-containerd is a daemon and API for controlling and managing runC
containers.
https://containerd.tools/
Signed-off-by: Christian Stewart <christian@paral.in>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
[Thomas:
- simplify a bit the comments about the runtime dependencies
- add missing comment on wchar, and take into account the BR2_USE_MMU
dependency in the comment
- factorize the build step with a foreach loop.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
runC is a CLI tool for spawning and running containers according to the
OCP specification.
runC is used by the latest Docker engine versions, however, runc itself
is standalone and has no other dependencies.
https://runc.io/
Signed-off-by: Christian Stewart <christian@paral.in>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Due to compiler deficiences, gcc on ARC and Blackfin doesn't define
_REENTRANT, which the latest version of the ax_pthread.m4
autoconf-archive macros looks at to determine whether thread support is
available or not.
Fixing the compilers is definitely the right thing to do, but it's not
doable for pre-built external toolchains like the Blackfin one, so for
now, we simply define _REENTRANT manually on ARC and Blackfin. If more
packages are affected by this problem, a more general solution will have
to be found.
Fixes
http://autobuild.buildroot.net/results/593/593b5050473a83a9ddcada9de39f8f14ab38d554/http://autobuild.buildroot.net/results/ff0/ff058a378b91740b5b399f32a6b375fbc3c8df06/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The ACPI Component Architecture (ACPICA) project provides an operating
system (OS)-independent reference implementation of the Advanced
Configuration and Power Interface Specification (ACPI).
It contains tools such as acpidump, iasl, acpixtract, etc.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Cc: Romain Naour <romain.naour@gmail.com>
Cc: Peter Korsgaard <peter@korsgaard.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
[Thomas: adjust install flags to -m755.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The situation looks like following for elf2flt and binfmt FLAT:
* Only gcc for bfin/m68k implements
-msep-data (BR2_BINFMT_FLAT_SEP_DATA) and
-mid-shared-library (BR2_BINFMT_FLAT_SHARED), so the corresponding
options are made only visible on those architectures.
* When the default of BR2_BINFMT_FLAT_ONE is used on m68k, broken
binaries are produced, which mainly end up in SIGILL, so do not use
it for m68k.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
[Thomas:
- also add the dependencies on m68k/bfin to BR2_BINFMT_FLAT_SHARED
- rework commit log.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Library and some tools (dbus-lanuch, dbus-run-session) are dual
licensed under GPLv2 or AFLv2.1 and others (dbus-monitor, dbus-send,
dbus-cleanup-sockets, dbus-uuidgen) are licensed under GPLv2+
only.
Reviewed-by: Marcin Nowakowski <marcin.nowakowski@imgtec.com>
Signed-off-by: Rahul Bedarkar <rahul.bedarkar@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since version 6, GMP is distributed under the dual licenses LGPLv3+ or
GPLv2+
Reviewed-by: Marcin Nowakowski <marcin.nowakowski@imgtec.com>
Signed-off-by: Rahul Bedarkar <rahul.bedarkar@imgtec.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
CLI tool and lua bindings are released under GPLv2.
Reviewed-by: Marcin Nowakowski <marcin.nowakowski@imgtec.com>
Signed-off-by: Rahul Bedarkar <rahul.bedarkar@imgtec.com>
[Thomas: use proper construct to concatenate new licenses to
<pkg>_LICENSE, as suggested by Yann E. Morin.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
GnuTLS core library is licensed under LGPLv2.1+ while gnutls-openssl
library is licensed under GPLv3+. Annotate the license with components.
Reviewed-by: Marcin Nowakowski <marcin.nowakowski@imgtec.com>
Signed-off-by: Rahul Bedarkar <rahul.bedarkar@imgtec.com>
[Thomas:
- add comment about the license of the "core library" since some files
seems to be mistakenly under LGPLv3+ even though the library is
licensed under LGPLv2.1+.
- add the README file in the license information, since it contains a
lot of useful details.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
It's really not necessary to have each option for the different file
formats named "enable XYZ file format support". Saying "XYZ support" is
sufficient.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since there were multiple sub-menus, their prompts are turned into
comments.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The only menu was "Module Selection", even though it contained
options (like "NPM for target") which are not related to selecting
modules. This commit therefore removes the menu ... endmenu.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
