On machines supporting Riscv SV57 mode like Qemu, Go programs currently crash
with the following type of error:
runtime: lfstack.push invalid packing: node=0xffffff5908a940 cnt=0x1
packed=0xffff5908a9400001 -> node=0xffff5908a940
The upstream PR fixes this error, but has not yet been merged.
Upstream: https://go-review.googlesource.com/c/go/+/409055/4
Signed-off-by: Christian Stewart <christian@paral.in>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Unfortunately local patch is still pending upstream. I've pinged them
again after 3 months with no answer:
https://bugzilla.mozilla.org/show_bug.cgi?id=1765100
Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
imhttp support needs apr-util in addition to civetweb since bump to
version 8.2204.1 in commit bc469ee632
and
a8b8d6c1f9. Without
apr-util, the build fails with:
configure: error: Package requirements (apr-util-1 >= 1.0) were not met:
Package 'apr-util-1', required by 'virtual:world', not found
Instead of having the user needing to figure out that apr-util is
needed for imhttp support, have the top-level rsyslog option selects
apr-util when civetweb-lib is enabled
Fixes:
- http://autobuild.buildroot.org/results/4454ae5321df0c04c9dee19693f3908e69fbf23d
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Fix the following build failure without BUILD_COMPLEX raised since bump
to version 3.10.1 in commit b3e96b1fb0 and
bd1fba15d0:
CMake Error at LAPACKE/CMakeLists.txt:38 (message):
LAPACKE_BUILD_COMPLEX requires BUILD_COMPLEX to be ON
CMake Error at LAPACKE/CMakeLists.txt:38 (message):
LAPACKE_BUILD_COMPLEX16 requires BUILD_COMPLEX16 to be ON
Fixes:
- http://autobuild.buildroot.org/results/418d15f734c87a70edf0d670c0548d550e6e70ba
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Fix the following build failure without C++ raised since the addition of
the package in commit e27ef76582:
CMake Error at CMakeLists.txt:3 (project):
No CMAKE_CXX_COMPILER could be found.
Tell CMake where to find the compiler by setting either the environment
variable "CXX" or the CMake cache entry CMAKE_CXX_COMPILER to the full path
to the compiler, or to the compiler name if it is in the PATH.
Fixes:
- http://autobuild.buildroot.org/results/17e2d6e6d6ddf7845a37a8bbf733faf40d9faa61
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
dump1090 needs threads and gcc >= 4.9 since bump to version 7.2 in
commit e80c99e151 to avoid the following
build failures:
In file included from dump1090.c:50:
dump1090.h:69:10: fatal error: pthread.h: No such file or directory
69 | #include <pthread.h>
| ^~~~~~~~~~~
In file included from dump1090.c:50:0:
dump1090.h:68:23: fatal error: stdatomic.h: No such file or directory
#include <stdatomic.h>
^
Fixes:
- http://autobuild.buildroot.org/results/a8993d9a09c79231358f50d2c1eb2e08cf7d5fd6
- http://autobuild.buildroot.org/results/7367abae7dc8187e4a2aae5397618c6e1b7aa521
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
When doing a build with "make -s" (silent option), the build of Boost
fails. Indeed, in this case boost.mk is passing -d as a option to
Boost, but -d must be followed by a digit.
Fixes:
- https://bugs.buildroot.org/show_bug.cgi?id=14951
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
libfreeimage is a mixed CRLF-LF source code. The patch introduced by
commit d3ee0c781e (package/libfreeimage: fix musl build) was missing
the CRLF where they were needed (because patchwork mangles them?).
Fix the three hunks in two files that have the issue.
Reported-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
While bumping version I've fixed build failure due to missing host
asciidoctor by passing --disable-asciidoc to CRYPTSETUP_CONF_OPTS but that
is missing in HOST_CRYPTSETUP_CONF_OPTS and this leads the same build
failure of target building. So let's fix it by adding --disable-asciidoc
to host HOST_CRYPTSETUP_CONF_OPTS too.
Fixes:
http://autobuild.buildroot.net/results/0c1a6a8e59653afaa09303550f74475b4a3898bf/
Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Commit
88dd3fcd72 ("package/pkg-meson.mk: add
support for fortran cross-compilation") incorrectly handled the
FCFLAGS value to create the cross-compilation.conf file. This commit
fixes that by making the PKG_MESON_CROSSCONFIG_SED macro take a 4th
argument containing the fcflags, and adjusting the call sites, as well
as the PKG_MESON_INSTALL_CROSS_CONF macro.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
python-scipy is moving to a meson build infrastructure, and requires
fortran cross-compilation. This commit extends the cross-compilation
configuration file generated for use by meson to add the "fortran",
"fortran_args" and "fortran_link_args" properties.
Signed-off-by: Guillaume W. Bres <guillaume.bressaix@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This is the third release of the 1.1.z series of runc, and contains various
minor improvements and bugfixes.
https://github.com/opencontainers/runc/releases/tag/v1.1.3
Signed-off-by: Christian Stewart <christian@paral.in>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
nanopi-neo no longer builds, as uboot needs python2 on the host:
https://gitlab.com/buildroot.org/buildroot/-/jobs/2812053540
I no longer have access to that board, so I can't test an update to
either uboot or the kernel anymore.
Drop the board.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
lshw is seldom released, the last one being more than two years ago.
Although the amount of changes is not huge, there have been some
significant changes and fixes, so it warrants a bump to the current
HEAD of the git repo.
lshw tries to build the translations with msgfmt, but as a development
and debugging aid, or as a backend to other scripts, translations are
not really useful. We disable that by overriding the LANGUAGES
variables to an empty list.
Patches are dropped: the basename() patch has been superseded by an
upstream change (introduced shortname() as a wrapper to it), and the
LONG_BIT fix has been applied. It was a bit difficult to find, as
upstream has not applied the patches, nor kept authorship, nor kept the
commit logs...
Of noteworthy attention, this fixes the JSON output format (at least for
the flaws we were hitting in our cases).
Signed-off-by: Yann E. MORIN <yann.morin@orange.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
lshw was introduced in early 2012, and since then our coding standards
have evolved substancially, and so lshw is lagging way behind nowadays.
Reformat it with the current styles:
- mandatory dependencies go before build variables
- one item per line in multi-line assignments
- space after the last item in a multi-line assignment
- don't hard-code the build target 'all' in the common make opts
- split long command lines
- format build and install command lines similarly for ease of
comparison
Signed-off-by: Yann E. MORIN <yann.morin@orange.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Commit 29bccabdeb (lshw: use data files provided by hwdata) removed
the hardware info DBs installed by lshw, in favour of those provided by
hwdata, as the former are way out-dated while the latter are more
up-to-date.
However, it forgot to also select the PNP ID database, to replace the
one lshw would install.
Signed-off-by: Yann E. MORIN <yann.morin@orange.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
pistache couldn't build on uClibc due to just one missing
definition. In addition, the Config.in comment was wrong as it said
that a glibc toolchain was needed, while in fact it could be glibc or
musl.
Cc: Thomas Ruschival <thomas@ruschival.de>
Cc: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Sometimes buildroot fails with:
../grub-core/kern/emu/hostfs.c:20:10: fatal error: config-util.h: No such file or directory
20 | #include <config-util.h>
| ^~~~~~~~~~~~~~~
Add a patch which fixes the Makefile to correctly generate config-util.h
first.
Note: This re-adds a workaround to avoid re-running autoconf. This has
previously been used to avoid having to run the rather complex build
file generation machinery of GRUB2. See 7e64a050fb ("boot/grub2: Fix
GRUB i386-pc build with Ubuntu gcc"), but now we just need to touch
Makefile.in.
Signed-off-by: Stefan Agner <stefan@agner.ch>
[yann.morin.1998@free.fr:
- add comment with patch name before hook
- slightly extend commit log that we only touch Makefile.in
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Enable --refclocks=all configure flag, to be able to use several
synchronization sources (hardware).
Refclocks are all disabled by default, and they can only be enabled,
either one by one by name, or globally. Except for PPS support that
needs pps-tools, the other refcloks do not have external dependencies,
so we choose to enable all or none, without fine-grained selection.
Signed-off-by: Guillaume W. Bres <guillaume.bressaix@gmail.com>
[yann.morin.1998@free.fr:
- change comment and commit to explain refclock can't be disabled
- don't add a menu-endmenu
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Since version 10, the weston-launch command has been deprecated, and
can only be built with -Ddeprecated-weston-launch=true, which
Buildroot does not do.
So the only launcher currently available is logind, provided by
systemd.
But weston provides an alternate way, based on the seatd daemon. This
commit enables this possibility by adding an optional dependency on
BR2_PACKAGE_SEATD and using -Dlauncher-libseat=true.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
All .pc files have a Version field set to #VERSION#, which needs to be
replaced, otherwise packages checking the version of OpenGL ES, EGL,
GBM, etc. will fail. For example:
Dependency gbm found: NO unknown version, but need: ['>=17.1.0']
Such problems have been seen with both libepoxy and SDL2.
The version 21.1.1 is chosen because it matches what the Yocto recipe
is doing for this package:
https://github.com/STMicroelectronics/meta-st-stm32mp/blob/kirkstone/recipes-graphics/gcnano-userland/gcnano-userland.inc#L42
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
At least libepoxy, a user of OpenGL ES, expects libraries to be
available as libGLESv2.so.2 and libGLESv1_CM.so.1. While other OpenGL
implementations comply with this, gcnano-binaries does not, and
installs its libraries directly as .so, without any ABI version,
causing runtime failures with libepoxy or SDL2, as it tries to
dlopen() the OpenGL ES libraries with their ABI suffix.
See for example the libepoxy code:
https://github.com/anholt/libepoxy/blob/master/src/dispatch_common.c#L191
Fix this issue by adding the relevant symlinks.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
rapidjson has been added as a dependency of pistache as part of commit
2f32185679 ("package/pistache: bump
version & switch to meson build"), using a "depends on". However, as
it is a library dependency, a "select" is normally used in Buildroot.
rapidjson's only dependency is on C++ support, which is already a
dependency of pistache.
In addition, use = instead of += for the rapidjson dependency in the
.mk file, as it's an unconditional dependency.
Cc: Thomas Ruschival <thomas@ruschival.de>
Cc: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
The readme.txt contains a make target that does not match the actual
defconfig file name, fix that.
Fixes: 1500b7d5c8 ("configs/broadcom_northstar: new defconfig")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
USe $(error) to simplify the code (drop "exit 1") and sned the message
to stderr.
Reported-by: David Laight <David.Laight@ACULAB.COM>
Reported-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Quentin Schulz <quentin.schulz@theobroma-systems.com>
Broadcom Northstar family of SoCs is most commonly used for home
routers. It's an ARM platform with Cortex-A9 CPU(s).
All known Northstar devices come with CFE bootloader which almost
always expects a TRX firmware format (with exception for D-Link). Some
vendors (like Luxul and Netgear) wrap TRX in their own containers.
This board code provides:
1. Minimal kernel with support for on-SoC blocks. It enables Linux
drivers for SoC, watchdog, Ethernet, switch, USB, PCIe, LEDs).
2. Post image script building firmware images. In uses Buildroot
packages tools (lzma_alone, otrx, lxlfw) to build
bootloader-compatible images that can be flashed.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Reviewed-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
When one is applying patches, it is pretty common to end up with .orig
and/or .rej files lying around. Unfortunately, our 'Config.*' match in
check-package ends up matching those files, causing false positives
when running "make check-package". To avoid this, this commit
excludes *.orig and *.rej files for the find logic used in the
check-package target.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
protobuf changed it versioning scheme [1]. The Release version is now only the
Minor.Patch version of the language specific version.
C++ stays on Major version 3, but python got bumped to 4. So the handling in
the .mk files changed to reflect that.
python-protobuf changed it version to 4 because of some breaking changes,
see[1].
[1] https://developers.google.com/protocol-buffers/docs/news/2022-05-06
Signed-off-by: Michael Nosthoff <buildroot@heine.tech>
Tested-by: Julien Olivain <ju.o@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
construct is a Python library for declarative serialization/
deserialization of structured binary data.
Signed-off-by: Martin Povišer <povik+lin@cutebit.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
uClibc-ng 1.0.41 added ns_t_tkey, but since external toolchains may
have a version older than 1.0.41, we introduce a compatibility patch
that avoids the use of ns_t_tkey. This compatibility patch can be
dropped in the future when uClibc-ng 1.0.41 is sufficiently deployed.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
printvars returns nothing when VARS is not passed or empty. This is done
on purpose, see commit fd5bd12379 ("Makefile: printvars: don't print
anything when VARS is not set").
An error message making explicit what is required from the user in order
to use printvars is however better than silently doing nothing.
This adds a check for a non-empty VARS variable.
Cc: Quentin Schulz <foss+buildroot@0leil.net>
Signed-off-by: Quentin Schulz <quentin.schulz@theobroma-systems.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Commit 5c54c3ef3d (Makefile: workaround make 4.3 issue for 'printvars
and 'show-vars') did not fully fix the show-vars case, which still
segfaults.
Overall, show-vars generates a JSON blurb. That is supposed to be
machine-readable, so we do not care that the variables are sorted, so
we get rid of it to (slightly) simplify the code.
Then, we currently iterate twice on the list of variables: the first one
to filter-out the 'internal' variables, and the second one to filter
only the variables matching the pattern. We can do away by iterating
only once, and applying both filters at once.
Since we now have an 'and' condition, we can take advantage of it: when
none of the items in $(and) are empty, $(and) evaluates to the last
item, while it evaluates to empty if any of the items is empty. So we
can coalesce the $(if) and $(and) together: $(if $(and a,b),c) is
equivalent to: $(and a,b,c) ; this gains us one parentheses depth.
Finally, the cause for the segfault is an overly-long call to $(info).
Reducing that is not easy: we want to call clean-json on the whole of
the JSON blurb, so we can't emit the individual variables one by one, or
the trailing comma would not be trimmed away.
So, we go crazy: we just output each word from clean-json with $(info).
We can do that, because mk-json-str transforms all spaces in a string
to an escaped UTF-8 sequence, so we will never have spaces in values;
the keys are the variables, so they won't have spaces either; spaces in
the rest of the JSON blurb are totally optional, so we don't care how
many there are. We know there are spaces, because we explicitly
introduce some (after "expanded" or "raw", for example), so we should
never hit a too-big word for $(info) to print.
Thanks to Henri for the suggestion to push $(info) further inside the
macro.
Reported-by: Quentin Schulz <quentin.schulz@theobroma-systems.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Roosen Henri <Henri.Roosen@ginzinger.com>
Cc: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Tested-by: Quentin Schulz <quentin.schulz@theobroma-systems.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Test the scipy::io module, to demonstrate that runtime dependencies
are correct.
Signed-off-by: Guillaume W. Bres <guillaume.bressaix@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
We need python built with the --zlib feature, for the io
module to import properly and be usable
Signed-off-by: Guillaume W. Bres <guillaume.bressaix@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Importing the scipy module in the Python interpreter running in Qemu
takes quite a while on slow machine, so a timeout extension to 30
seconds is needed to make sure the test has the time to run.
Signed-off-by: Guillaume W. Bres <guillaume.bressaix@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
python-attrs package is required since version 2.6.0.
Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This adds an optional linebreak at the vendor attribute in the CPE ID.
It should be noted this is purely for formatting/layout purposes
and does not actually insert any additional characters
(newline or otherwise) into the rendered text.
This means that even though the text renders across two lines,
copy-pasting will still yield one line of text.
example: https://sen-h.github.io/pkg-stats/c245575.html
see also: https://developer.mozilla.org/en-US/docs/Web/HTML/Element/wbr
Signed-off-by: Sen Hastings <sen@phobosdpl.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
