When building a toolchain with upstream gcc 9.x the build
fail due to several issues.
Note: The upstream Binutils support csky target since
release 2.32 but the support was never enabled in the
Buildroot packaging. So the latest version (2.33.1) was
tested here.
[upstream gcc 9.x w/ glibc csky fork with binutils csky for or binutils 2.33.1]
In file included from <command-line>:
./../include/libc-symbols.h:534:26: error: '__EI___errno_location' specifies less restrictive attributes than its target '__errno_location': 'const', 'nothrow' [-Werror=missing-attributes]
534 | extern __typeof (name) __EI_##name \
[upstream gcc 9.x w/ glibc 2.30 w/ binutils csky fork]
/tmp/ccThLRhb.s: Assembler messages:
/tmp/ccThLRhb.s:10: Error: invalid or unsupported encoding in .cfi_personality
/tmp/ccThLRhb.s:11: Error: invalid or unsupported encoding in .cfi_lsda
[upstream gcc 9.x w/ glibc 2.30 w/ binutils 2.33.1]
build/elf/librtld.os: in function `__sync_fetch_and_add_2':
libgcc/config/csky/linux-atomic.c:116: undefined reference to `__kernel_cmpxchg'
Currenlty, only the toolchain using binutils, gcc, glibc
fork produce a working toolchain. So disable gcc 9.x for
csky.
Signed-off-by: Romain Naour <romain.naour@smile.fr>
Cc: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Guo Ren <guoren@kernel.org>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
When the csky cpu support was added [1], the gcc download url was selected
depending on the csky cpu architecture (BR2_csky) rather than the csky gcc
fork version (BR2_GCC_VERSION_CSKY)[2].
When adding gcc 9.x version [3], we forgot to update the condition in order
to use the url to the gcc csky fork only when BR2_GCC_VERSION_CSKY=y.
Due to this error, the toolchain build with the upstream gcc 9.x for csky
cpu is broken due a download error.
Fix this by using BR2_GCC_VERSION_CSKY instead of BR2_csky.
Fixes:
https://gitlab.com/kubu93/buildroot/-/jobs/470072924
[1] 7873a5bd5e
[2] https://git.buildroot.net/buildroot/tree/package/gcc/gcc.mk?id=7873a5bd5ebbeb1674293dae6b06b50f0a1f2184#n19
[3] 089000eccf
Signed-off-by: Romain Naour <romain.naour@smile.fr>
Cc: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Guo Ren <guoren@kernel.org>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
- Update site to get latest release
- License is GPL-3.0+ since version 4.3.0 and
2bc924c184
- Switch to cmake-infrastructure
- Remove most of the existing dependencies (alsa, speex, libupnp, ...)
as they are now only managed in mediatreamer
- Add external dependencies: belr, belle-sip, libxml2, mediastreamer,
sqlite
- This bump is needed as current linphone does not build with latest
libeXosip2
Fixes:
- http://autobuild.buildroot.org/results/ed8bdcbfb7d453ed7d436bada35417ee305e6ac9
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
- Retrieve official tarball and so remove autoreconf
- xlib_Xcursor is a mandatory dependency since
6dd85edc6f
- libtasn1 is a mandatory dependency since
673b267e66
- nettle is a mandatory dependency since
00d9e0c4c8
- gnutls is a mandatory dependency instead of openssl since
166d1bc14d
- pulseaudio is an optional dependency since
9b98c3b229
- iconv is mandatory since
2ea3f69873
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This comment is not needed since bump to version 1.5.2 and
83a2b4e9bf
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
The new version comes with some new features and minor fixes. The two
previously included patches were merged upstream. A new patch was added
to fix build on 32-bit systems. There's a new dependency to the pcap
library.
Signed-off-by: Alexander Dahl <post@lespocky.de>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Other changes:
- Convert the package to a meson package.
- Remove 0001-add-PYTHON_INCLUDES-override.patch as it no longer applies.
- Add gobject-introspection as a dependency.
- Add the package under myself in the DEVELOPERS file.
Because gobject-introspection is now a dependency of python-gobject, the test
must be updated at the same time.
- Change TestPythonPy2Gobject to TestPythonPy3Gobject as
gobject-introspection requires python3.
- Refactor test_python_gobject.py to no longer inherit the
TestPythonPackageBase class, as this class uses a base config that does not
support gobject-introspection.
- Update sample_python_gobject to use Glib to find the path of sh.
Signed-off-by: Adam Duskett <Aduskett@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Other changes:
- Remove upstream patches.
- Change cross-compile options to upstream equivilants
- Add -Ddoctool=disabled to GOBJECT_INTROSPECTION_CONF_OPTS
- Change -Dcairo from a bool to a feature.
Signed-off-by: Adam Duskett <Aduskett@gmail.com>
[yann.morin.1998@free.fr: keep patches ordering]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Other changes:
- Change -Dlibmount from a bool to a feature.
Signed-off-by: Adam Duskett <Aduskett@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
If multiple entries are specified for _MESON_EXTRA_BINARIES, the current
sed expression will only replace the first one.
Specifically, from GNU sed 4.8 the manual says:
/regexp/
Match lines matching the regular expression regexp. Matching
is performed on the current pattern space, which can be modified
with commands such as ``s///''.
so after the first binary has been added, the next entry no longer
matches since the pattern space has been modifed.
Instead of adding a script for each value, apply the match once and add
a subsitution for all entries at once.
Signed-off-by: John Keeping <john@metanate.com>
Tested-by: Peter Seiderer <ps.report@gmx.net>
[yann.morin.1998@free.fr: do a single substitution]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Keep custom install logic because qt5cinex has no built-in install
target.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Convert but keep the slightly complicated custom install step for now.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Mostly straightforward conversion but keeping custom config/build env
to apply the python2 workaround.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Straightforward conversion. Only conditional dependencies have to
remain.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The conversion is mostly straightforward. The jsc binary was not
installed by the qt5webkit build system, so we're keeping some custom
hook to install it.
It's worth mentioning that the package build system installs
QtWebProcess in /usr/libexec/ instead of /usr/bin.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The installation of the Javascript code is custom and thus needs to be
kept.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The values in the cross-compilation file are expected to be quoted with
single quotes, which we have in our template.
However, the endian value we are injecting comes from Kconfig, so it is
double-quoted, and those quotes end up in the cross-compilation files we
generate (the internal one, and the SDK one):
endian = '"little"'
So qstrip the value before we inject it.
Propagate the fix to the two generated files by using the same variable
HOST_MESON_TARGET_ENDIAN in both cases, rather than replicating the
(flawed) logic.
While at it, also use the common GCC_TARGET_CPU variable for the SDK
file too.
Signed-off-by: Gleb Mazovetskiy <glex.spb@gmail.com>
Reviewed-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Straightforward conversion except for the wayland-compositor configure
option.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
It seems the old method was missing some files from the install
step since after the conversion the following additional files
appear in target after a build with all config options selected:
usr/lib/qt/plugins/lipi_toolkit/libactivedtw.so
usr/lib/qt/plugins/lipi_toolkit/libboxfld.so
usr/lib/qt/plugins/lipi_toolkit/libl7.so
usr/lib/qt/plugins/lipi_toolkit/liblipiengine.so
usr/lib/qt/plugins/lipi_toolkit/liblogger.so
usr/lib/qt/plugins/lipi_toolkit/libneuralnet.so
usr/lib/qt/plugins/lipi_toolkit/libnn.so
usr/lib/qt/plugins/lipi_toolkit/libnpen.so
usr/lib/qt/plugins/lipi_toolkit/libpointfloat.so
usr/lib/qt/plugins/lipi_toolkit/libpreproc.so
usr/lib/qt/plugins/lipi_toolkit/libsubstroke.so
usr/lib/qt/plugins/virtualkeyboard/libqtvirtualkeyboard_lipi.so
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Use generic configure command but for now keep custom build and
install logic, as we only build some parts of qt5tools depending on
the selected sub-options.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
$$(STAGING_DIR) -> $(STAGING_DIR) in PKG_MESON_INSTALL_CROSS_CONF.
$$ resulted in `$(STAGING_DIR)` in the file instead of the expanded
value.
Note that this change only affects the etc config at:
host/etc/meson/cross-compilation.conf
Per-package cross-compilation.conf files are already correct.
Reviewed-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Gleb Mazovetskiy <glex.spb@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
The "network" script allows configuring an interface by means of two
/etc/ifup.<if_name> and /etc/ifdown.<if_name> files with the commands
required to bring it up and down, respectively. So we can support the
BR2_SYSTEM_DHCP config if netifrc is not selected.
- Replace the dependency on !BR2_PACKAGE_OPENRC in Config.in by a
dependency on !BR2_PACKAGE_NETIFRC.
- Add an IFUPDOWN_SCRIPTS_DHCP_OPENRC macro to create the ifup.<if_name>
and ifdown.<if_name> files.
- Add an IFUPDOWN_SCRIPTS_INSTALL_INIT_OPENRC hook that includes the
IFUPDOWN_SCRIPTS_{PREAMBLE,DHCP,DHCP_OPENRC} macros. LOCALHOST is not
required, since "lo" is configured by the "loopback" script.
Signed-off-by: Carlos Santos <unixmania@gmail.com>
[yann.morin.1998@free.fr:
- adapt after the preamble was spit to itw own patch
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Currently, we always suppose that localhost settings will always come
before the (optional) DHCP ones. This is always true when using
sysvinit/busybox or systemd.
However, with openrc, this will no longer be the case: only the DHCP
settings will be present (OpenRC handles lo on his own).
So, prepare the macros to be independent one from the other, and for
symetry, in both:
- use append-redirection >> to populate the interfaces file,
- prepend a leading empty line at the beginning of each section.
The origianl singe '>' redirection would ensure the file would not grow
on a reinstall, but that's no longer the case, so reset the file prior
to filling it, using the canned preamble.
Signed-off-by: Carlos Santos <unixmania@gmail.com>
[yann.morin.1998@free.fr: split off to its own patch]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Generate it in the IFUPDOWN_SCRIPTS_INSTALL_INIT_<FOO> hook. This will
allow creating a file to configure an interface via DHCP for the openrc
network service, in a forthcoming change.
Signed-off-by: Carlos Santos <unixmania@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Use the qmake infrastructure for its generic configure and build steps
but keep the manipulation of qwtconfig.pri as a PRE_CONFIGURE_HOOK as
well as the custom install steps.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The script is minuscle and can be useful to configure an interface via
DHCP at run-time.
Signed-off-by: Carlos Santos <unixmania@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Two simple tests to ensure that openrc boots without any services crashing
with a read only and a read write filing system.
Signed-off-by: Adam Duskett <Aduskett@gmail.com>
Signed-off-by: Carlos Santos <unixmania@gmail.com>
[yann.morin.1998@free.fr: really check the init process]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Add a patch to prevent starting agetty on tty[1-6]. Buildroot starts a
single getty, according to the system configuration. Also tty[1-6] may
not even exist (e.g. embedded devices with serial consoles only).
Signed-off-by: Adam Duskett <Aduskett@gmail.com>
Signed-off-by: Carlos Santos <unixmania@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
In Buildroot getty is provided by BusyBox or util-linux (agetty). Both
create a /sbin/getty symlink pointing to the actual program, so make the
agetty service use that path instead of /sbin/agetty.
Also start agetty after all other services, to prevent mixing startup
messages with the getty prompt on devices with a single serial console,
as the custom unit we're dopping was doing already.
This allows us to drop the custom getty service, which causes an error
message from openrc 0.42.1 (in a forthcoming patch):
Error: getty is the name of a real and virtual service.
The patches are required because it's not possible to override the
command or startup order by means of a configuration file. Anyway it's
still better to patch something maintained upstream than depending on a
custom piece that may easily become obsolete.
Signed-off-by: Carlos Santos <unixmania@gmail.com>
[yann.morin.1998@free.fr:
- expand commit log to explain why we ve 'after *' in the same patch
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Convert build, staging and target install to the qmake infra commands.
This removes a lot of conditional code because we solely rely on all install
rules being properly generated according to the configured options.
Of course we keep the configure command/logic as is, because qmake does not
yet exist.
Handling of the custom qt.conf is now done in the post staging hook.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
As we are about to convert qt5base to the qmake-package
infrastructure, we need to avoid a circular dependency: the qt5base
package itself should not depend on qt5base.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This commit converts the qextserialport to the qmake infra, which is a
straightforward conversion as it doesn't require any custom
environment option or additional hook.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This converts all qmake packages to the new infra for which no custom
environments, option or additional hooks are needed.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This provides generic functions for Qt5 qmake based packages. It will
make it possible to remove lots of redefinition of
QT5_xxx_{CONFIGURE|BUILD|INSTALL_STAGING}_CMDS. Additionally it
provides a generic target install method which will make most of the
package specific commands obsolete.
This is done by re-running the install step of the qmake generated
Makefile with the package build directory prepended (to the
staging/host path). Even though this does create lengthy pathes it
allows for easy separation of the staging files from the host destined
files by just omitting the resulting BUILD_DIR+HOST_DIR path from the
following rsync call to the real target folder. The cleanup of many
files we dont want in target is deferred to the target-finalize
step. In addition to what's being removed already, we also have to
cleanup some Qt5 specific files (prl) and the documentation directory.
This approach was chosen over copying all files recorded in the pkg-files-list
after some discussion which Thomas Petazzoni summed up:
"We don't yet use pkg-files-list really as part of the build
process anywhere, I feel a bit more comfortable at this point with what
Andreas is proposing."
Thanks to this infrastructure, it will be possible to get rid of the
many conditional install commands because qmake already takes care of
this when generating the Makefile install targets with the given or
autodetected configure options of each package.
However, custom install steps may have to remain in cases where a
particular Buildroot option has no corresponding setting in the
packages configuration options.
Signed-off-by: Andreas Naumann <anaumann@ultratronik.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
host-python3 is a mandatory dependency since bump to version 4.10.0 and
5ddff307b4
because buildtools/bin/waf shebang is set to "/usr/bin/env python3"
Fixes:
- http://autobuild.buildroot.org/results/2634eb7824beb34f485cf40670f6959515f008ad
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Belr is Belledonne Communications' language recognition library, written
in C++11. It parses text inputs formatted according to a language
defined by an ABNF grammar, such as the protocols standardized at IETF.
https://gitlab.linphone.org/BC/public/belr
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Belle-sip is a modern library implementing SIP (RFC 3261) transport,
transaction and dialog layers.
http://www.linphone.org/technical-corner/belle-sip
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
[Thomas: select mbedtls as support for crypto is needed in bctoolbox]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
On master during the stabilisation phase, a new patch was added to
systemd; in parallel, on the next branch, systemd was bumped. Then, when
next was merged into master, there was no merge conflict, so the uneeded
patch was left unnoticed.
That patch was applied upstream, and is present in the version we now
have, so it no longer applies.
Drop that patch.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
- Drop first patch (already in version)
- Update second patch
- License is GPL-3.0+ since version 4.3.0 and
85094197cf
- Switch to cmake-package and so drop third patch as autotools is not
updated anymore (and fails to build due to missing po/Makefile.in.in)
- Add a mandatory dependency to mbedtls (to enable crypto support in
bctoolbox)
- Add bcg729, jpegturbo, libgsm, libpcap, libsrtp and zxing-cpp optional
dependencies
- Add a dependency on dynamic library as no pkg-config calls are done in
cmake (static build with ffmepg and opus will fail for example)
- Drop libupnp optional dependency (not available anymore)
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Add openssl linker flags via LIBS to fix configure gensio
library detection.
Fixes:
http://autobuild.buildroot.net/results/66e0d3e0a2a8dc5a62c267d16a53216f0f2ce8dd
checking gensio/gensio.h usability... yes
checking gensio/gensio.h presence... yes
checking for gensio/gensio.h... yes
checking for str_to_gensio in -lgensio... no
configure: error: libgensio won't link, please install gensio dev package
The build/ser2net-4.1.1/config.log files states:
.../arm-buildroot-linux-uclibcgnueabi/bin/ld: .../host/arm-buildroot-linux-uclibcgnueabi/sysroot/usr/lib/libgensio.a(gensio_filter_ssl.o): in function `gensio_do_ssl_init':
gensio_filter_ssl.c:(.text+0x34): undefined reference to `OPENSSL_init_ssl'
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The Buildroot's gitlab testing infra reported a build issue
with the qemu_arm_vexpress_tz_defconfig due to host-python3
modules issues [1]. Thoses issues has been fixed by the
previous patch.
But the defconfig doesn't boot with the current master
(2020.02-rc3).
It turn out that is an Qemu 4.2.0 regression that was
fixed upstream by [2]. This issue was found by using
git bisect old/new.
Fixes:
$ ../host/bin/qemu-system-arm -machine virt -machine secure=on -cpu cortex-a15 -smp 1 -s -m 1024 -d unimp -serial stdio -netdev user,id=vmnic -device virtio-net-device,netdev=vmnic -semihosting-config enable,target=native -bios bl1.bin
NOTICE: Booting Trusted Firmware
NOTICE: BL1: v2.0(release):2020.02-rc3-43-g9abf171ea6
NOTICE: BL1: Built : 12:44:52, Mar 8 2020
ERROR: Failed to load BL2 firmware.
After fixing host-python3 issue from [1]
[1] https://gitlab.com/buildroot.org/buildroot/-/jobs/456818689
[2] 21bf9b06cb
Signed-off-by: Adrien Grassein <adrien.grassein@smile.fr>
[Romain:
- improve commit log
- add upstream link
]
Signed-off-by: Romain Naour <romain.naour@smile.fr>
Cc: Etienne Carriere <etienne.carriere@linaro.org>
Cc: Gerome Burlats <gerome.burlats@smile.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
