ninja is faster at building cmake packages than make, and according to reports
also more reliable at handling parallel builds
This commit includes a patch that adds GNU make jobserver support, in order to
allow more precise control over the number of parallel tasks
Enable parallel build by default for packages using ninja
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Rosen Penev <rosenp@gmail.com>
This old ELF library dating to 2009 used to be necessary on MacOS but
is not required for building the kernel or tools since [1]. On Linux
systems, libelf is already an OpenWRT build-system prerequisite [2].
Presence of the older library can mask or conflict with the system libelf
and lead to build errors, as seen compiling Linux kernels since v5.8 or
host tools such as dwarves (e.g. pahole).
Remove the unnecessary tools/libelf library and avoid the related issues.
[1] 5f8e587240 ("build: force disable stack validation during kernel build
on non-linux systems")
[2] https://openwrt.org/docs/guide-developer/build-system/install-buildsystem#prerequisites
Tested-by: Rosen Penev <rosenp@gmail.com> (Linux)
Tested-by: Georgi Valkov <gvalkov@abv.bg> (MacOS)
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
U-Boot requires xxd to create the default environment from an external
file as done in uboot-mediatek.
Build xxd (only, not the rest of vim) as part of tools to make sure it
is present on the buildhost.
Reported-by: David Bauer <mail@david-bauer.net>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
mediatek-mt7622 as well as mediatek-mt7623 require CPIO to create their
initramfs images. So build CPIO as part of the host toolchain.
Signed-off-by: David Bauer <mail@david-bauer.net>
Instead of using an ancient qemu version in-tree the building machine
should just have qemu-utils installed.
Signed-off-by: Paul Spooren <mail@aparcar.org>
use PKG_FIXUP:=autoreconf to generate configure
200-hide-dlsym-error.patch deleted due to fixed upstream in another way
other patches refreshed to reflect latest changes
Signed-off-by: Syrone Wong <wong.syrone@gmail.com>
Upstream switched to building with CMake. Adjust accordingly.
Reapplied patch as upstream changed the file format.
Added HOST_BUILD_PARALLEL for faster compilation.
Added cmake tool dependency and removed circular dependencies as a
result.
Adjusted dependent tools to use NOCACHE as they are needed to build
ccache.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
rules.mk always passes these as -I/-L to the toolchain.
Fixes rare errors like:
cc1: error: staging_dir/target-aarch64_cortex-a53_musl/usr/include: No such file or directory [-Werror=missing-include-dirs]
Signed-off-by: Andre Heider <a.heider@gmail.com>
Acked-by: Paul Spooren <mail@aparcar.org>
Acked-by: Rosen Penev <rosenp@gmail.com>
This sorts the added tools and builddir dependencies alphabetically
to make it easier to find something in the Makefile.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
These tools have been used by the orion target which has been
removed in Jan 2020 [1].
Both were specifically meant for the WRT350Nv2, which is not
supported anymore.
So, let's remove them as well.
[1] 89f2deb372 ("orion: remove unmaintained target")
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
lzma-old host package is required for building ath79 images, as
mksquashfs-lzma is required for generating netgear images.
Fixes commit 4e4ee46495 ("ar71xx: drop target")
Signed-off-by: David Bauer <mail@david-bauer.net>
This target has been mostly replaced by ath79 and won't be included
in the upcoming release anymore. Finally put it to rest.
This also removes all references in packages, tools, etc. as well as
the uboot-ar71xx and vsc73x5-ucode packages.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
SELinux support requires setting the appropriate SELinux security context
to files and directories, which needs to happen at build time in order
to support read-only root filesystem scenarios. In order to create these
security contexts, we will have to run some SELinux-specific tools on
the host machine, but that requires root access. This adds support for
fakeroot, which the build process will use to run the SELinux security
context creation and the image creation.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Apply to current master, and adjust commit message
Thomas' original work is available at
http://lists.infradead.org/pipermail/openwrt-devel/2019-November/025976.html.
Signed-off-by: W. Michael Petullo <mike@flyn.org>
[add rules.mk FAKEROOT variable]
Signed-off-by: Paul Spooren <mail@aparcar.org>
[update, fix macos build]
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Some autotools based build systems are using autoconf-archive scripts
and are expecting them to almost always be available. This is not
required for regular releases as tar balls generated for releases
commonly have existing configure script. This is rather intended to be
used with autotools.mk's autoreconf and in cases it is not always
possible to get release tar ball.
Including this adds little to no overhead in terms of build time as
those are just m4 scripts copied to an appropriate location.
Signed-off-by: Karel Kočí <karel.koci@nic.cz>
[fixed From: to match SoB]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
As part of the kernel build process there are utilities built to run on
the host system that expect linux kernel headers to be available.
Unfortunately macos/darwin doesn't have these headers.
vdso2c requires types.h so provide a minimal stub to satisfy it.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
pkgconf is a newer, actively maintained implementation of pkg-config that
supports more aspects of the pkg-config file specification and provides a
library interface that applications can use to incorporate intelligent
handling of pkg-config files into themselves (such as build file
generators, IDEs, and compilers). Through its pkg-config compatibility
interface (activated when it is run as "pkg-config"), it also can
completely replace the original implementation.
It is also lighterweight and does not require glib2, as pkg-config does.
On other distros, pkgconf is symlinked to pkg-config. For simplicity here,
it is renamed to pkg-config.real, as in the original package.
Initial results have been positive. As before, pkgconf works as long as
the pkg-config files point to the proper paths.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
[backported upstream fix for Meson]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Currently, if "make tools/install" is called after tools have already
been installed, a symbolic link named "lib" will be created inside
$(STAGING_DIR_HOST)/lib, pointing to "lib" (i.e. itself).
During tools/prepare, a "lib64" symlink is created inside
$(STAGING_DIR_HOST) that points to "lib" (also inside
$(STAGING_DIR_HOST)).
If tools/prepare is called and the "lib64" symlink already exists, then
ln will treat it as a directory and instead create a symlink named "lib"
inside of that directory.
This adds the -n option for ln so that $(STAGING_DIR_HOST)/lib64 is
always treated as a normal file (the link name), not as a directory.
Signed-off-by: Jeffery To <jeffery.to@gmail.com>
"$(STAGING_DIR)/include" was carried over from buildroot-ng to OpenWrt
in commit 60c1f0f64d. buildroot has
dropped this directory a long time ago.
In OpenWrt the directory is still created by the PrepareStaging macro
and is part of the default TARGET_CPPFLAGS. But nothing at all installs
headers into this directory, nor should anything be installed under this
path.
Removing this directory from TARGET_CPPFLAGS will cut down the log noise
a bit. Not only will CPPFLAGS be shorter, there will be less warnings
set off by "-Wmissing-include-dirs" (or even failures when paired with
"-Werror"). After all the directory does not even _exist_ in the SDKs,
which are used on the build bots when building packages (see [1] and
[2]).
make[8]: Entering directory '/builder/shared-workdir/build/sdk/build_dir/target-aarch64_generic_musl/libmbim-1.20.0/src/common'
CC libmbim_common_la-mbim-common.lo
cc1: error: /builder/shared-workdir/build/sdk/staging_dir/target-aarch64_generic_musl/include: No such file or directory [-Werror=missing-include-dirs]
cc1: all warnings being treated as errors
[1] https://github.com/openwrt/packages/issues/10377
[2] https://github.com/openwrt/packages/pull/10378
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Acked-by: Jo-Philipp Wich <jo@mein.io>
Acked-by: Rosen Penev <rosenp@gmail.com>
This patch removes scons host build tool, as commit 7087efd72a8d
("scons: move host build tool to a proper place") in the packages feed
has moved scons into the new home.
There are currently no packages in the master tree which would need
scons, yet scons is build always as part of host tools, just in order to
satisfy host build dependency of few packages in the packages feeds.
Ref: https://github.com/openwrt/packages/pull/9584
Acked-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
This provides board configuraion tables for various Tegra boards needed
by cbootimage tool to create flashable bootloader images.
Signed-off-by: Tomasz Maciej Nowak <tomek_n@o2.pl>
Tegra BCT and bootable flash image generator/compiler
>From documentation:
This project provides a tool which compiles BCT (Boot Configuration
Table) images to place into the boot flash of a Tegra-based device.
The tool will either:
a) Compile a textual representation of a BCT into a binary image.
b) Generate an entire boot image from a previously compiled BCT and a
bootloader binary.
Signed-off-by: Tomasz Maciej Nowak <tomek_n@o2.pl>
gnu date is present in our staging_dir which means squashfskit will fail
to build. Prevent also a race condition if coreutils is build after
squashfskit.
To prevent a race condition, depend on coreutils.
Acked-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
squashfskit is a fork of the squashfs-tools.
squashfskit creates reproducible filesystems and includes
many of the distro patches.
Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
Tested-by: Paul Spooren <mail@aparcar.org>
This makes OpenWrt build an Ext2+Ext4 partitioned image
for the D-Link DNS-313 with two blank partitions, a boot
partition on /dev/sda3 and a rootfs partition on
/dev/sda4.
This uses the methods already used in apm821xx including
the tricks to convert the generated Ext2 filesystem
partition to version 1 so that the firmware bootloader
will properly recognize it.
We patch a bit around the build files to make sure we
get the rootfs size set and that genext2fs is properly
built.
Tested on the D-Link DNS-313.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
[Fixing and tricksing]
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
One image requires a zip compressed image, so add the zip util found in
the packages feed, and extend it with some useful debian patches.
Signed-off-by: Mathias Kresin <dev@kresin.me>
A new test case was adding in one of the patches fixing a problem, this
also included a change in the test/Makefile.am to add this test case.
The build system detected a change in the Makefile.am and wants to
regenerate the Makefile.in, but this fails because automake-1.15 is not
installed yet. As automake depends on patch being build first, make sure
we do not modify the Makefile.am.
This fixes build problem seen by the build bots.
Fixes: 4797dddfde ("patch: apply upstream cve fixes")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The Makefile.am changed and now patch wants to use automake to
regenerate the Makefile.in. Make sure automake was build before we build
patch.
This fixes build problem seen by the build bots.
Fixes: 4797dddfde ("patch: apply upstream cve fixes")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Certain Netgear and AVM devices use BE squashfs for the kernel image. As
squashfs4 only supports creating LE images, add squashfs (v3) into
the tools to be built for ath79.
Trying to use an LE squashfs (thus trying to use squashfs4 only for
building the image) for the kernel image results in the bootloader
barfing and stopping.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Zoltan HERPAI <wigyori@uid0.hu>
This version now uses autotools to configure the build system. They are
also using the newly added zlib package.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Tested-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
This allows us to link the other tools against our libz and we do not
need the system zlib any more.
Only the static linked library is copied to the staging directory so we
have a statically linked library on all systems and not only on Linux.
This also adds the new dependencies of the packages which are depending
on zlib.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Tested-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Remove upx from repo, it hasn't been used for a very long time and serves little to no purpose.
Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
A lot of packages (especially those using libtool) depend on sed being
available. xz fails on an incremental build pretty early if it's not
there.
Clean up handling if essential core tools (patch, tar, xz), build them
in parallel and make them depend on sed
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Ensure that pkg-config is available before building libressl as we're going
to need it for setting proper link flags in utilities requiring libcrypto.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
graphite is disabled by default and it's not worth carrying ppl and
cloog for supporting this feature on the old gcc 4.8, which is only used
for ARC
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This is intended to be used for a wide array of package sanity checks.
The first check that is implemented is for the hash of downloaded files.
It checks:
- Missing hash
- Use of SHA256 instead of MD5
- dl/<file> hash not matching hash in makefile
- deprecated MD5SUM variable
The deprecated MD5SUM variable check is skipped for feeds/ until OpenWrt
is updated as well
Signed-off-by: Felix Fietkau <nbd@nbd.name>
