There was already a post-build hook to delete the /var/lib/dbus symlink
created by buildroot after the package's own installation, to prevent
a dbus installation error during "make dbus-rebuild". However, this
misses the case for when one might delete the .stamp_target_installed
file manually, outside of dbus-rebuild. This can be fixed by changing
the post-build hook to a pre-install hook. This seems appropriate,
since it is really addressing an installation issue, not a build issue.
Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This options was recently removed and made it to legacy (in 815f713773)
but miraclecast still selects it.
miraclecast builds fine without it, it seems, so just drop the select
altogether.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Cc: Ryan Barnett <rjbarnet@rockwellcollins.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
lutok is a lua module, so it depends on ! static libs. However, the
dependency is implicit, being done because the lua modules are sourced
globally under an if-block, and thus it is not easy to find that
dependency.
Propagate that dependency to kyua, which was missing it (because it is
not a lua module, so was missing the dependency).
[Peter: also update toolchain comment]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Sebastien Bourdelin <sebastien.bourdelin@savoirfairelinux.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This commit reverts cdf63517de "gmp: disable assembly for arc"
as in GMP v6.1.2 there's already a proper fix for ASM constraints, see
https://gmplib.org/repo/gmp/rev/58879634af3ci
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The midi profile was introduced in BlueZ 5.44.
Cc: Marcin Bis <marcin@bis.org.pl>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Disregard to increase just minor version BlueZ 5.44 is drastically different in
a way what plugins that used to be enabled by the --enable-experimental
configure option, now have their own option.
Extend Buildroot package to cover these plugins.
Cc: Marcin Bis <marcin@bis.org.pl>
Cc: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
In BlueZ 5.44 the --enable-experimental configure option only controls the
build of the Nokia OBEX PC Suite plugin. Correct the option prompt and help
text.
Cc: Marcin Bis <marcin@bis.org.pl>
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
binutils 2.27 triggers a segfault in elf2flt on ARM/noMMU. While Arnout
has identified a binutils 2.28 commit that can be backported on 2.27,
this commit is huge and we don't clearly understand the impact.
Since both binutils 2.26 and 2.28 are unaffected by this issue, we
simply disallow the selection of binutils 2.27 on ARM/noMMU, and default
to binutils 2.28.
Fixes:
http://autobuild.buildroot.net/results/e14cadb290b0b86cac12c4bfb681eb6eee9e6dea/
and lots of other similar ARM/Cortex-M4 failures
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
libunwind use sigreturn() and setcontext() while building for x86 [1]
but this function is not available with musl [2].
This throw a warning during libunwind build:
In file included from x86/Los-linux.c:4:0:
x86/Gos-linux.c: In function ‘_ULx86_local_resume’:
x86/Gos-linux.c:298:7: warning: implicit declaration of function ‘sigreturn’ [-Wimplicit-function-declaration]
sigreturn (sc);
^
But any program trying to link against libunwind-generic.so or
libunwind.so fail to build:
[...]/usr/lib/libunwind-generic.so: undefined reference to `sigreturn'
[...]/usr/lib/libunwind-generic.so: undefined reference to `setcontext'
collect2: error: ld returned 1 exit status
Disable libunwind for x86 target when musl is used.
Reported upstream by Waldemar [3]
[1] http://git.savannah.gnu.org/gitweb/?p=libunwind.git;a=blob;f=src/x86/Gos-linux.c;h=17aebc2974af50eb0bf8292689b2ed22a4c97866;hb=HEAD#l299
[2] http://openwall.com/lists/musl/2016/02/04/3
[3] http://lists.nongnu.org/archive/html/libunwind-devel/2017-04/msg00030.html
Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Waldemar Brodkorb <wbx@openadk.org>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since aubio 0.4.4, the libsamplerate support require fftw
single precision mode [1].
[1] 8be88e77f5
Signed-off-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
BR2_ARM_CPU_HAS_ARM already embeddeds the fact that this is an ARM chip,
so (BR2_arm || BR2_armeb) is forcibly true when BR2_ARM_CPU_HAS_ARM is.
In other words, BR2_ARM_CPU_HAS_ARM can't be true unless either BR2_arm
or BR2_armeb itslef is true.
Simplify the logic accordingly.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Romain Naour <romain.naour@openwide.fr>
Acked-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
libnvidia-egl-wayland.so was renamed in upstream archive.
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since uClibc-ng v1.0.20 we can force libc to either use llock/scond
instructions for atomic ops or if those are missing use kernel-assisted
emulation. By default atomics are used so we disable that feature in
libc if we build for a target without LLSC.
This allows us to build and use NPTL on ARC750 for example which was not
possible before.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
spidev_test has further evolved after kernel version 3.15, which
is the most recent version that could be built in this package.
This change allows to build the latest version of spidev_test if
the toolchain's kernel header version allows it. In this way, when
building with a newer toolchain, we are able to benefit from the
added functionality offered by the newest version of the tool.
Signed-off-by: Joeri Barbarien <joeri.barbarien@nokia.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
We usually have empty lines around multi-line variable definitions.
Signed-off-by: Carlos Santos <casantos@datacom.ind.br>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
For dynamic library builds, it's not needed to pass it explicitly.
Signed-off-by: Carlos Santos <casantos@datacom.ind.br>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The fsck wrapper from util-linux is better maintained, so it should be
preferred. In particular, systemd relies on features in the util-linux
fsck wrapper that don't exist in the e2fsprogs one. However, since
e2fsprogs depends on util-linux, the one from e2fsprogs currently has
precedence. So in order to have e2fsck from e2fsprogs and the fsck
wrapper from util-linux, we disable e2fsprogs' fsck if the one from
util-linux is selected.
Signed-off-by: Carlos Santos <casantos@datacom.ind.br>
CC: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
[Arnout: don't add comment]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Carlos Santos <casantos@datacom.ind.br>
[Arnout: don't split up the fuse2fs CONF_OPTS and DEPENDENCIES
handling]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
We don't need e2initrd_helper or the test I/O manager for the host, just
like we don't need it for the target.
Signed-off-by: Carlos Santos <casantos@datacom.ind.br>
[Arnout: don't add --disable-rpath, since we *do* need rpath for the
host; otherwise we have to explicitly pass LD_LIBRARY_PATH to be able
to run tools linked with util-linux' libuuid/libblkid]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
So far we attempted to solve the conflicts between busybox and e2fsprogs
by removing busybox programs from /bin and /sbin, leaving the e2fsprogs
ones at /usr/bin and /usr/sbin. This fails with BR2_ROOTFS_MERGED_USR=y,
leading to situations like the one described in bug 9436.
We could provide a better solution by means of a fine-grained selection
of programs, like util-linux does, but this would require big changes in
e2fsprogs. So instead of resorting to dirty tricks we switch to a more
pragmatic approach:
- Drop all configs to select/deselect utilities without corresponding
enable/disable options to the configure script. In other words, we
always install the basic set of utilities.
- fsck has a configure option, so use it. Note that --enable-fsck is
only about the wrapper, not about e2fsck.
- Install e2fsprogs utilities at /bin and /sbin, overriding the ones
eventually installed by busybox.
- Let the e2fsprogs build system create the symlinks mkfs.ext2 -> mke2fs
etc. fsck.ext4dev and mkfs.ext4dev were dropped in version 1.43.4.
Notice that these changes do exactly the opposite of what is requested
in bug 9436. On the other hand the policy for e2fsprogs becomes coherent
with the one for util-linux: busybox never wins.
Fixes:
https://bugs.busybox.net/show_bug.cgi?id=9436 (no fix, in fact)
Signed-off-by: Carlos Santos <casantos@datacom.ind.br>
[Arnout:
- don't add 'default y' to resize2fs;
- don't install the host package in /bin instead of /usr/bin - we
install everything under /usr (until /usr will be removed, soon)]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Newer gcc requires even lower optimization, only with -O0 we
can successfully compile binutils for the target.
Fixes:
http://autobuild.buildroot.net/results/fb95cd7f7fcc532d036ed8f13853bc6f9a64d1b3
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reformat help text in Config.in to avoid warnings from
check-package. No content change intended.
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Add 'decorator'[1] package to Buildroot. Needed by 'traitlets'
package, which is added in the next patch.
[1] https://pypi.python.org/pypi/decorator
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
[Thomas: download from PyPi, add hash file, add entry in DEVELOPERS file.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The patch added in commit c78b65c4f1 was
bogus: it adds some code conditionally on HAVE_CANONICALIZE_FILE_NAME
being undefined, but there is nothing that defines
HAVE_CANONICALIZE_FILE_NAME if canonicalize_file_name() is already
provided by the C library.
This commit fixes that by adding the appropriate autoconf check.
Fixes:
http://autobuild.buildroot.net/results/4f0eb2af09c92ff4cd7448211ffda432754d4e60/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Adds an inline equivalent of canonicalize_file_name
using realpath().
Bug report (origin of this patch):
https://bugs.freedesktop.org/show_bug.cgi?id=99944
Resolves:
http://autobuild.buildroot.net/results/afa/afa4b97e012586585b11be1e70ed3c63a7c48a4d/
CCLD test-generated
CCLD test-utils
CCLD test-message
../../../src/libqmi-glib/.libs/libqmi-glib.so: undefined reference to `canonicalize_file_name'
collect2: error: ld returned 1 exit status
Makefile:440: recipe for target 'test-generated' failed
Signed-off-by: Matthew Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The current tools build assumes the host system when trying to detect if
swig/python are present. It then uses those tools from the path.
The upstream RFC included this commit set's patch but offered up
discussion on how to cleanly introduce a better method for detecting
swig and using the tools. The tools build really needs to be sysroot
and prefix/host dir tools aware.
Upsteam submission for RFC:
https://lists.denx.de/pipermail/u-boot/2017-May/289520.html
Workaround for:
http://autobuild.buildroot.net/results/6d52ac8bb71012aea6fc4c679691b31a3366728b
Signed-off-by: Matthew Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
tinyxml2 can build both static and shared libraries. By default, only
shared library is built. Shared/static builds are controlled using
separate cmake flags BUILD_SHARED_LIBS and BUILD_STATIC_LIBS.
In static libs configuration, we internally pass -DBUILD_SHARED_LIBS=OFF
cmake flag to build system which disables both shared and static builds
of library, resulting in failures while linking executable with library.
So pass -DBUILD_STATIC_LIBS=ON cmake flag in case of static libs
configuration.
fixes:
http://autobuild.buildroot.net/results/d30/d301bcbe5db26068b35eaa94bd816ae8cf8ef2e1
Signed-off-by: Rahul Bedarkar <rahulbedarkar89@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This configuration builds a 64bit image for RaspberryPi 3.
The size of the ext4 filesystem has to be extended to fit the large amount
(~60MB) of kernel modules enabled in the defconfig.
[Peter: drop gcc version selection / glibc, explain ext4 size requirement]
Signed-off-by: Gaël PORTAY <gael.portay@savoirfairelinux.com>
Tested-by: Ricardo Martincoski <ricardo.martincoski@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Enable support for OpenSSH to use a hardware SSL engine if
cryptodev-linux is included. Without this, OpenSSH uses only
OpenSSL software crypto implementation.
Signed-off-by: Gilad Ben-Yossef <gilad@benyossef.com>
CC: Baruch Siach <baruch@tkos.co.il>
CC: Arnout Vandecappelle <arnout@mind.be>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Add firmware file for the Raspberry Pi 3 and Zero W Bluetooth module.
Note that to successfully download the firmware bluez5_utils needs a patch[1].
Even though the firmware file name (BCM43430A1.hcd) only mentions the
Bluetooth module, it is most likely tailored for the Raspberry Pi boards.
Firmware for the same BT module on other boards look different. For this
reason the name of the package is rpi-bt-firmware.
[1] http://www.spinics.net/lists/linux-bluetooth/msg69470.html
[Peter: correct Config.in ordering]
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This allows a static build of picocom when BR2_STATIC_LIBS=y but the toolchain
provides static and shared libraries.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
