Python3 at runtime identifies the uuid backend to use to implement safe
time-based UUID generation functions. When the python _uuid extension
module, /usr/lib/python3.7/lib-dynload/_uuid.cpython-37m-<arch>.so
is not found, the fall back is to use a pure python implementation.
- If uuid.h from util-linux is available at build time, the _uuid
module is built and used. A Buildroot patch allows us to disable this
build-time check and never build the _uuid module.
With the bump to python 3.7, an optional dependency on util-linux was
added to build the _uuid module. However, this may lead to circular
dependencies. To break this circular dependency, we can disable the
build of the _uuid module and always use the pure python implementation.
Ref:
http://lists.busybox.net/pipermail/buildroot/2018-October/233113.htmlhttp://lists.busybox.net/pipermail/buildroot/2018-September/231060.htmlhttp://lists.busybox.net/pipermail/buildroot/2018-October/233079.html
CC: Adam Duskett <aduskett@gmail.com>
Signed-off-by: Matthew Weber <matthew.weber@rockwellcollins.com>
Tested-by: Carlos Santos <casantos@datacom.com.br>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This adds one column to the legal-info manifest table. It contains the
dependencies of the given package and their licenses. This information
is useful when assessing license compatibility of the packages and
their libraries.
An example of the content of the new column for the MPD package is
shown below:
"alsa-lib [LGPL-2.1+ (library), GPL-2.0+ (aserver)] boost
[BSL-1.0] libid3tag [GPL-2.0+] libmad [GPL-2.0+] libogg
[BSD-3-Clause] libvorbis [BSD-3-Clause] libzlib [Zlib]
skeleton-init-common [unknown] skeleton-init-sysv [unknown] sqlite
[Public domain] toolchain-external-linaro-arm [unknown]"
[Credits to Yann E. MORIN <yann.morin.1998@free.fr> for suggesting a
few simplifications.]
Signed-off-by: Michal Sojka <sojka@merica.cz>
Tested-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Luca Ceresoli <luca@lucaceresoli.net>
Tested-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The last parameter {HOST|TARGET} is now first. With this change,
adding new columns to the legal manifest file (as in the next commit)
will be slightly easier to review.
Signed-off-by: Michal Sojka <sojka@merica.cz>
Tested-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Luca Ceresoli <luca@lucaceresoli.net>
Tested-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Enable keyring if libsecret is available
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since bluez5_utils and bluez5_utils-headers would install the same set
of files (headers), we want to ensure the former is run after the
latter.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Grzegorz Blach <grzegorz@blach.pl>
Cc: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
This patch adds a simple way to change the linux bootup logo.
The patch was kept purposely simple to support only the use cause
where a user needs a color linux boot up logo.
Signed-off-by: Angelo Compagnucci <angelo@amarulasolutions.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Peter: clarify/reword option/help text, automatically enable framebuffer
support, fix convert path]
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This patch adds the host package for imagemagick.
It comes with a minimal selection of configure options to
compile fast cause it is used only to convert
the image files for the custom linux logo.
Signed-off-by: Angelo Compagnucci <angelo@amarulasolutions.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Peter: explicitly disable other configure options like it is done for the
target variant, add host-pkgconf]
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
For some reason, the hash of the Github-provided tarball for mongoose
6.7 has changed. The actual contents are the same between the old and
new tarballs, only the order of the files is changed.
However, changing the hash would break older Buildroot releases
(because sources.buildroot.net would pick up the tarball with the new
hash, and old Buildroot releases would no longer be able to pick up
the tarball with the old hash, neither from Github nor from the
Buildroot mirror).
So, we simply bump to a newer version of mongoose. This way, old
Buildroot releases will continue to fetch Mongoose 6.7 from the
Buildroot mirror, with the old hash matching. New Buildroot releases
will be using the new Mongoose version, with a correct hash that
allows to download from Github successfully.
The hash of the license file has changed, due to a change in the
copyright years.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
When python3 depends on bluez5_utils-headers,
python3 is build with bluetooth support.
Signed-off-by: Grzegorz Blach <grzegorz@blach.pl>
[Arnout: select bluez5_utils-headers is bluez5_utils is enabled]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Header files from bluez utils version 5.x.
Separated headers are required to break circular dependencies
when python3 is built with bluetooth support.
This is a blind option, it will be automatically selected by python3
when bluez5_utils is enabled.
Signed-off-by: Grzegorz Blach <grzegorz@blach.pl>
[Arnout: add comment to keep versions in sync; make it a blind option]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
By default clang is assuming the system linker /usr/bin/ld. This
patchset updates the default to point at the cross toolchain ld.
Previously clang had been used on target only, this update is
required for host clang/clang++ to be used as a frontend for
cross-compiling.
Example build command:
output/host/bin/clang++ -mcpu=cortex-a53 \
--sysroot /<abs path>/output/staging/ \
-B /<abs path>/output/host/opt/ext-toolchain/ \
foobar.cc -o foobar
Cc: Valentin Korenblit <valentinkorenblit@gmail.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Matthew Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Since commit 8f8e9162fa ("package/gcc:
do not mourn avr32 for too long..."), in which we dropped AVR32
support, the BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE is always
'y'. Therefore, it is totally useless to keep this option around, and
this commit removes it, making the corresponding code unconditional
along the way.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
libcilkrts was introduced in gcc 4.9, and removed in gcc
8.x. Therefore, it does not make sense to pass --disable-libcilkrts in
gcc 8.x and higher. This commit introduces BR2_GCC_SUPPORTS_LIBCILKRTS
that allows the makefile code to know if the selected gcc version
supports libcilkrts or not.
This new option mainly allows to annotate the fact that libcilkrts
exists in gcc [4.9, 7] and that once we have dropped support for those
gcc versions, the libcilkrts related logic can be removed.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The libcilkrts configure script errors out with "Pthreads are required
to build libcilkrts" if the C library doesn't have thread support. To
fix that, we disable libcilkrts when thread support is not available.
This issue was not noticed until now, because we only regularly build
a no-thread toolchain for ARM, and libcilkrts was enabled on ARM only
starting in gcc 7.x.
This fixes the build of no-thread toolchains on architectures where
libcilkrts is supported, i.e x86/x86-64, ARM and Sparc.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Utilize the predefined variable MENDER_PKGDIR which points to the
directory containing mender.mk and Config.in, along with the local
files.
One benefit is also that it works when BR2_EXTERNAL is used, which
makes it easier to "backport" this to an older external tree.
Signed-off-by: Mirza Krak <mirza.krak@northern.tech>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Sort the certificates into alphabetical order so the contents of
ca-certificates.crt can be built reproducibly.
Note: The certificates are sorted uppercase then lowercase filenames
so the contents of ca-certificates.crt matches the source debian package.
Signed-off-by: Martin Bark <martin@barkynet.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Rebuilding ca-certificates using make ca-certificates-rebuild
caused duplicate certificates to be installed in the target. Its build
system is broken: it doesn't detect that the output file already exists,
and instead of overwriting it, a duplicate is generated under a
different name. The net effect is that all certificates are installed
twice after rebuild.
Fix this by cleaning the build directory before building the package.
Signed-off-by: Martin Bark <martin@barkynet.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
c_rehash looks at all files in /etc/ssl/certs, generates the hash for
the certificates in them, and makes a symlink from the hash to the
certificate file.
However, ca-certificates.crt is also installed in /etc/ssl/certs and
it contains all the certificates. c_rehash will take one of them (the
first?) and create a symlink from that hash to ca-certificates.crt.
Usually, this results in an error like:
WARNING: Skipping duplicate certificate ca-certificates.crt
and all is well. However, depending on filesystem order,
ca-certificates.crt may come first, and the actual certificate is
not symlinked.
To fix this install certificates.crt to /etc/ssl/certs *after* we run
c_rehash to prevent it getting hashed by mistake.
Note: $(TARGET_DIR)/etc/ssl/certs/ is already removed during install so
this fix also works for rebuilds.
Signed-off-by: Martin Bark <martin@barkynet.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
This patch provides support for the board Odroid XU4. It uses a recent
4.14 kernel released by Hardkernel, and the current u-boot 2017.05
also from Hardkernel.
Signed-off-by: Pierre CROKAERT <pct@crookies.net>
[Thomas:
- Update DEVELOPERS file
- Update .gitlab-ci.yml file
- Simplify the boot.ini script
- Add missing BR2_PACKAGE_HOST_LINUX_HEADERS_CUSTOM_4_14 option
- Add BR2_PACKAGE_HOST_DOSFSTOOLS and BR2_PACKAGE_HOST_MTOOLS since a
VFAT filesystem is generated
- Keep the default root filesystem size, since it is sufficient.
- Reduce the size of the VFAT filesystem]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
nss-myhostname is helpful to resolve local hostname without domain.
For instance, if we set hostname to myboard and ping myboard
without .local suffix, it fails if /etc/hosts stays unchanged.
nss-myhostname modifies /etc/nsswitch.conf file.
It set myhostname resolution just after files in hosts line,
for example:
hosts: files myhostname mdns4_minimal [NOTFOUND=return] dns
This choice comes from Debian commit [1] and Yocto recipe [2].
Also, nss-myhostname author suggests myhostname should be used as
a backup and appended at the end of hosts line. Both choices are valid.
Discussion on Debian bug tracker [3] gives some details on issue
that may occurs using nss-myhostname (see fqdn).
Note nss-myhostname is already included in systemd, so ensure is conflicts
with it.
[1] http://git.nomeata.de/?p=libnss-myhostname.git;a=commit;h=5104d7f0045df55ee8be526e8c84078750e1e0ed
[2] https://git.yoctoproject.org/cgit.cgi/poky/plain/meta/recipes-support/nss-myhostname/nss-myhostname_0.3.bb
[3] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=756224
[Peter: simplify sed invocation, adjust Config.in]
Signed-off-by: Vincent Prince <vincent.prince.fr@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
With the nsswitch.conf installation logic moved to the post-install-install
hook of the toolchain package, it is guaranteed to be available when
nss-mdns is run - So drop the logic.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
nsswitch.conf is processed both by the toolchain, nss-mdns and
nss-myhostname without any guaranteed ordering in between.
The toolchain package ensures that nsswitch.conf is available, and the two
nss-* packages tweaks the content, so the toolchain processing should run
before the nss-* ones. Toolchain is a dependency of all the packages, so
ensure this is done by moving the toolchain handling to a
post-target-install hook.
Also move the variable to toolchain/toolchain/toolchain.mk where the virtual
toolchain package is defined for clarity.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Both nss-mdns and nss-myhostname patch nsswitch.conf file so it needs to be
done at the final stage to support per-package host/target directories.
[Peter: reword]
Signed-off-by: Vincent Prince <vincent.prince.fr@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This reverts commit 15c00b9a23, because
the hashes do not match the upstream tarballs. We need to understand
what happened rather than blindly fixing the hashes.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
host-package that is able to shorten the load time of entry_points declared
in setup.py of python packages.
Signed-off-by: Asaf Kahlon <asafka7@gmail.com>
Reviewed-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
gcc does not build when the srcdir path contains a '@', because that
path is then substitued in a texi file as argument to an @include
directive. But then, the '@' in the path will start a command evaluation
of its own, thus breaking the build. For example, with a $(O) path set
to /home/ymorin/dev/buildroot/O/to@ti :
perl ../../gcc/../contrib/texi2pod.pl ../../gcc/doc/invoke.texi > gcc.pod
../../gcc/doc/invoke.texi:1678: unknown command `ti'
../../gcc/doc/invoke.texi:1678: @include: could not find /home/ymorin/dev/buildroot/O/to/build/host-gcc-initial-7.3.0/build/gcc/../../gcc/../libiberty/at-file.texi
[Peter: use findstring instead of subst/compare]
Reported-by: c32 on IRC
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
If a custom external toolchain is used, we can't enable the fortran
support. Add a new option for that.
Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The download link was broken, former qt versions are stored into a
distinct location.
Signed-off-by: Francois Gerin <francois.gerin@essensium.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The download infra now knows to look for per-version hash files, so we
can now drop the unversioned hash files.
Instead of removing them, add a comment that redirects the developper to
update the per-version hash files instead (so they do not re-add a
unversioned one in the future).
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Gaël Portay <gael.portay@savoirfairelinux.com>
Cc: Peter Seiderer <ps.report@gmx.net>
Cc: Julien Corjon <corjon.j@ecagroup.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Now that packges may have hashes for their downloaded files in a
versioned subdir or in the unversioned hash file, we can no longer
hard-code the path to the package hash file anymore.
Instead, we now make use of the per-package variable, that points to the
package hash file.
Note: of the packages for which we offer a version choice, almost none,
but some of the qt5 ones, have a per-version hash file, so we still use
the unversioned hash file for them. As for the few qt5 packages that do
have a per-version hash file for their licensing terms, they've already
been updated to duplicate their download hashes in both the unversioned
and per-version hash files. So, one way or the other, no hash check
would go missing with this change.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Baruch Siach <baruch@tkos.co.il>
Reviewed-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The Qt5 packages may have different licensing terms between the two
versions we support, and in some cases, those different terms are
expressed in similarly named files, like files named plain 'LICENSE' for
example.
Similarly, glibc also has different license files, especially since the
arc version still has libidn, which got dropped from upstream.
This is problematic, because, in a .hash file, we can't store two
different hashes for the same file. We've started to handle this case by
moving the licenses hashes to the per-version sub directories.
However, the hashes for the downloads are still stored inside the non-
versioned hash file of the package, which is not totally coherent: if we
have a per-version hash file, it should list all the hases for that
version, downloads included, and there should be no unversioned hash
file.
In preparation for this, we duplicate the downloads hashes from the main
hash files, and into the versioned ones. Once the download infra learns
to look for those hashes in these per-version subdirs, we'll remove the
unversioned hash files.
Note that, now that we have versioned hash files, the main hash files
will not be used to check license files, so we can already drop the
hashes for license files from the main hash files.
Note also that there are a few other packages for which we support
different versions (binutils, gcc, gdb, lua, xserver_xorg-server,
uboot), but none of those have different licensing terms due to the
version. Qt5 and glibc are alone in this case.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Gaël Portay <gael.portay@savoirfairelinux.com>
Cc: Peter Seiderer <ps.report@gmx.net>
Cc: Julien Corjon <corjon.j@ecagroup.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
When a package has a version selection (e.g. Qt5), the licensing terms
may be different across versions, but lie in similarly named files (e.g.
'LICENSE').
However, when we check a file, all the hashes for it must match. So, we
can't have the hashes for two different content of the same file. We
overcame that limitation in the legal-license-file macro, which checks
whether a package has a .hash file in a versioned subdir.
For consistency, we would like to also store the source hashes in that
per-version subdir.
Rather than reconstruct the path to the hash file everywhere we need it,
add a variable that points to it.
Existing users will be converted over in followup patches.
Note: the check for a missing hash file is done in the check-hash helper
script, so this variable must always yield a filename, even of a missing
file, thus we do not use $(wildcard...) to resolve the hash file path;
we use $(wildcard...) only to check if the versioned .hash file exists.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Baruch Siach <baruch@tkos.co.il>
Reviewed-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
