package/mono: bump to version 5.0.1.1 and monolite to version 1050001000
This patch bumps to mono 5.0.1.1 and updates the unit tests patch
accordingly.
The latest version of mono carries a bit of changes in the monolite
package: the version string changes and from now on, monolite should
be installed in a subdirectory with the exact version string as a name.
Therefore, mono and monolite are bumped in a single commit.
Patch 0004, although applied in upstream master, is not yet in this
(stable) release.
Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com>
[Arnout: squashed the patches, removed /usr part from HOST_DIR] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Sergio Prado [Sun, 6 Aug 2017 14:52:28 +0000 (11:52 -0300)]
package/stella: bump version to 5.0.1
Removed 0003-Use-gnu-11-standard-to-prevent-errors-on-PPC.patch, since
stella now uses -std=c++14 that builds fine when using PPC altivec
vectorization.
Removed 0004-gcc7.patch, backported from upstream.
Update toolchain dependencies to gcc 4.9 since codebase now uses C++14 features.
Bernd Kuhls [Sat, 5 Aug 2017 09:49:35 +0000 (11:49 +0200)]
package/mdadm: bump version to 4.0
This version of mdadm can optionally use the corosync and libdlm
packages. However, it checks for their presence by directly probing the
host /usr/include, which fails for cross-compilation:
Yann E. MORIN [Sun, 6 Aug 2017 15:41:02 +0000 (17:41 +0200)]
fs/squashfs: honour the number of parallel jobs
By default, mksquashfs uses all the CPUs available, like us with
BR2_JLEVEL=0.
However, by setting BR21_JLEVEL to non-zero, a user volutarily limits
the number of jobbs allowed to run concurrently (for various reasons,
like this is not a priority task on the machine; or like the
autobuilders, to spread the load accross multiple runs without
overloading the machine).
So, propagate tthe parallelisn settings down to mksquashfs, to repect
the user's will.
Francois Perrad [Sun, 6 Aug 2017 10:56:08 +0000 (12:56 +0200)]
ficl: fix build with gcc 7
gcc 7 with -Os (optimize for size) takes the liberty to remove the code
of inline functions entirely - which is indeed allowed according to the
C99 standard, cfr. this [1] very complete StackOverflow answer. This
leads to undefined function references at link time.
The proper approach is to patch the code to use 'static inline'
instead. However, for various reasons such a patch is not entirely
trivial, and upstream has not moved since 7 years, so instead take the
simpler approach of restoring the gcc original inline behaviour with
-fgnu89-inline to fix this issue.
The underlying reason for doing the checks is to ensure that a
merged-usr setting is properly enforced, even for custom skeletons.
Before that patch, a custom skeleton where both /bin and /usr/bin were
missing was accepted; but then the first package that intalled something
in /bin would create it as a directory, thus breaking the merged-usr
situation. Ditto sbin and lib, of course.
cf6d321 was created to detect that situation, and the fix was to require
that both directories do exist in the custom skeleton, so that we new
the check for consistency were OK, in all cases.
However, that broke existing skeletons which where missing both
directories, and some people are shinning about it...
The crux of the problem is that Buildroot should be responsible for
creating the directories or the symliks when they are missing, and only
fail if the existing ones are incorrect, but not impose the burden on
the user.
A situation where the problem arises is when a skeleton is shared
between various builds, some using a merged-usr while other do not.
We fix that by reverting the offending changes change, back to the
previous behaviour. We keep the stderr redirection in stat calls, and
the variable renaming.
A proper fix to create the missing directories can be added later.
gdb 7.12+ is compiled and linked with C++, therefore the thread_db
headers need to use extern "C", to be usable for C++ code.
Fixes:
Issue when building gdbserver for m68k-coldfire:
thread-db.o: In function `thread_db_init()':
thread-db.c:(.text+0x5b6): undefined reference to `td_ta_new(ps_prochandle*, td_thragent**)'
thread-db.c:(.text+0x61e): undefined reference to `td_thr_get_info(td_thrhandle const*, td_thrinfo*)'
thread-db.c:(.text+0x632): undefined reference to `td_symbol_list()'
...
Samuel Martin [Sat, 5 Aug 2017 18:18:53 +0000 (20:18 +0200)]
pkg-cmake.mk: set pkg-config env. vars. in the host configure environment
This change is fixing the bug reported in [1].
Cmake may run pkg-config to find the dependencies when configuring a
package. Because of the value of PATH, and it will use the Buildroot's
pkg-config wrapper, which, by default, is configured (via some
environment variables) to find the target dependencies.
When configuring a host package using cmake, to prevent cmake from
wrongly solving dependencies from the target tree (when the
*-config.cmake files are using pkg-config) instead of looking for them
in the Buildroot's host tree or in the host system itself, we need to
set the environment variables altering the pkg-config behavior in the
cmake configure environment.
So, this change is fixing the cmake host-packages configuration step,
by properly setting the pkg-config environment variables to their values
for finding host dependencies.
Bernd Kuhls [Fri, 4 Aug 2017 04:02:48 +0000 (06:02 +0200)]
package/snappy: remove pthread handling for static builds
Commit 0e2ee0e6e8076 (package.snappy: fix static builds with threads)
added LIBS=-pthread to configure options to fix static build with
threads on some architectures. This works for autotools packages. Then,
commit d26282b3aa14cb (package/snappy: bump version, switch to cmake)
switched to cmake that does not understand the 'LIBS' parameter.
As pointed out by Thomas[1] the special handling of libpthread for
static builds is not needed anymore: it was needed only to build the
tests, and these are no longer built. Note that packages that link
statically with snappy will still need to link with -pthread. Only
two in-tree packages currently depend on snappy: lua-csnappy depends
on !BR2_STATIC_LIBS and leveldb adds -pthread from its build system.
Peter Seiderer [Tue, 8 Aug 2017 16:57:58 +0000 (18:57 +0200)]
bind: fix configure in case lmdb devel files are present on the host
Fix configure failure in case lmdb devel files are present on the host
by adding --without-lmdb option (reported [1] and fix tested [2],[3] by
grunpferd@netscape.net).
Fixes:
checking for lmdb library... yes
checking for library containing mdb_env_create... no
configure: error: found lmdb include but not library.
0006-fix-ftbfs-with-clang.patch
fixes the subsequent build error:
common/memops.c.31.o: In function `sample_move_dither_rect_d16_sSs':
memops.c:(.text+0x4dc): undefined reference to `fast_rand'
In file included from ../sysdeps/x86_64/multiarch/varshift.c:19:0,
from ../sysdeps/i386/i686/multiarch/varshift.c:1:
../sysdeps/x86_64/multiarch/varshift.h: In function '__m128i_shift_right':
../sysdeps/x86_64/multiarch/varshift.h:26:1: error: SSE vector return without SSE enabled changes the ABI [-Werror=psabi]
{
[...]
.../lib/gcc/i686-buildroot-linux-gnu/7.1.0/include/tmmintrin.h:136:1: error: inlining failed in call to always_inline '_mm_shuffle_epi8': target specific option mismatch
The patch has been taken from glibc's patchwork at
https://patchwork.sourceware.org/patch/21003/.
Versions older than Linux v3.7 do not support the aarch64 architecture
so disable them, for reference see https://kernelnewbies.org/Linux_3.7
Without this patch these defconfigs fail to build
BR2_aarch64=y
BR2_KERNEL_HEADERS_3_2=y
BR2_aarch64=y
BR2_KERNEL_HEADERS_3_4=y
with error messages like this:
make[1]: Entering directory '/home/buildroot/br5_next/output/build/linux-headers-3.2.89'
Makefile:567: /home/buildroot/br5_next/output/build/linux-headers-3.2.89/arch/arm64/Makefile: No such file or directory
Baruch Siach [Mon, 7 Aug 2017 17:54:14 +0000 (20:54 +0300)]
mutt: fix build with BR2_PACKAGE_SKELETON_NONE
BR2_PACKAGE_SKELETON_NONE does not provide the /var subdirectory which
breaks symlink creation in that directory. Make sure that /var exists
before creating the symlink.
Peter Seiderer [Mon, 7 Aug 2017 21:15:44 +0000 (23:15 +0200)]
qt5base: fix examples compile without opengl support
Fixes [1]:
make[1]: Entering directory '.../build/qt5base-5.9.1/examples/opengl'
cd hellowindow/ && ( test -e Makefile || .../build/qt5base-5.9.1/bin/qmake -o Makefile .../build/qt5base-5.9.1/examples/opengl/hellowindow/hellowindow.pro ) && make -f Makefile
make[2]: Entering directory '.../build/qt5base-5.9.1/examples/opengl/hellowindow'
.../host/bin/sparc-linux-g++ -c -pipe -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -Os --sysroot=.../host/sparc-buildroot-linux-uclibc/sysroot -O2 -O3 -fno-exceptions -Wall -W -D_REENTRANT -fPIC -DQT_NO_EXCEPTIONS -D_LARGEFILE64_SOURCE -D_LARGEFILE_SOURCE -DQT_NO_DEBUG -DQT_GUI_LIB -DQT_CORE_LIB -I. -I../../../include/QtGui/5.9.1 -I../../../include/QtGui/5.9.1/QtGui -I../../../include -I../../../include/QtGui -I../../../include/QtCore/5.9.1 -I../../../include/QtCore/5.9.1/QtCore -I../../../include/QtCore -I.moc -I../../../mkspecs/devices/linux-buildroot-g++ -o .obj/hellowindow.o hellowindow.cpp
In file included from hellowindow.cpp:51:0:
hellowindow.h:96:5: error: ‘QOpenGLShaderProgram’ does not name a type
QOpenGLShaderProgram *m_program;
^~~~~~~~~~~~~~~~~~~~
hellowindow.h:97:5: error: ‘QOpenGLBuffer’ does not name a type
QOpenGLBuffer m_vbo;
^~~~~~~~~~~~~
hellowindow.cpp: In constructor ‘Renderer::Renderer(const QSurfaceFormat&, Renderer*, QScreen*)’:
hellowindow.cpp:62:40: error: invalid use of incomplete type ‘class QOpenGLContext’
m_context = new QOpenGLContext(this);
^
In file included from ../../../include/QtGui/qsurfaceformat.h:1:0,
from ../../../include/QtGui/../../src/gui/kernel/qsurface.h:45,
from ../../../include/QtGui/qsurface.h:1,
from ../../../include/QtGui/../../src/gui/kernel/qwindow.h:51,
from ../../../include/QtGui/qwindow.h:1,
from ../../../include/QtGui/QWindow:1,
from hellowindow.h:51,
from hellowindow.cpp:51:
glibc: add patch fixing the build with binutils 2.29
Fixes the following build failure, which occurs on all architectures:
/tmp/cc6APNmN.s: Assembler messages:
/tmp/cc6APNmN.s: Error: `loc1@GLIBC_2.17' can't be versioned to common symbol 'loc1'
/tmp/cc6APNmN.s: Error: `loc2@GLIBC_2.17' can't be versioned to common symbol 'loc2'
/tmp/cc6APNmN.s: Error: `locs@GLIBC_2.17' can't be versioned to common symbol 'locs'
../o-iterator.mk:9: recipe for target '/tmp/output/build/glibc-2.25/build/misc/regexp.os' failed
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Microblaze support was never upstreamed in gdb. We currently use a gdb
from the Xilinx Github repository, but this gdb is based on 7.6, and
has never been updated in the last 4 years. There are no other active
branches at https://github.com/Xilinx/gdb/branches.
Xilinx has a slightly newer gdb, based on 7.7, available at
https://github.com/Xilinx/meta-xilinx/tree/master/recipes-microblaze/gdb. However,
it's apparently only available in the form of stack of big patches
(https://github.com/Xilinx/meta-xilinx/tree/master/recipes-microblaze/gdb/files).
Currently, gdb on Microblaze fails to build in various conditions:
* Against glibc, with "error: conflicting types for 'ps_lgetfpregs'"
* Against musl, with "error: unknown type name 'elf_gregset_t'"
While those issues can probably be fixed, the fact that there is no
active upstream significantly reduces the incentive to fix those
problems.
Therefore, let's drop support for gdb on Microblaze entirely.
Marcus Hoffmann [Fri, 4 Aug 2017 16:58:32 +0000 (18:58 +0200)]
dbus: add upstream patch to fix startup hang with with expat >= 2.2.1
After c0ad6ded018ffbc33f7f5 expat: security bump to version 2.2.1
the system can hang on startup under certain circumstances.
This happens when:
* we use systemd as init system
* the random nonblocking pool takes a while to initialize
* this apparently doesn't happen on qemu, so this would not have
been caught by the runtime testing infrastructure
* it also doesn't seem to happen when network booting
For a more detailed description of the bug see here:
https://bugs.freedesktop.org/show_bug.cgi?id=101858
The patch should be in next dbus version 1.10.24
Set DBUS_AUTORECONF = YES because configure.ac is changed.
Signed-off-by: Marcus Hoffmann <m.hoffmann@cartelsol.com>
[Arnout: add upstream commit sha + Marcus's Sob to the patch] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Stefan Becker [Thu, 3 Aug 2017 09:05:45 +0000 (12:05 +0300)]
package/skeleton-common: fix recursive variable
This fixes a recursive variable definition introduced by the refactoring
in commit 120307520f69f801bf05df1d2fd3433c6d498551. Simply rename the
hook to remove the recursion.
The azmq library provides Boost Asio style bindings for ZeroMQ. This
library is built on top of ZeroMQ's standard C interface and is intended
to work well with C++ applications which use the Boost libraries in
general, and Asio in particular.
instrumentation: extend packages-file-list.txt with symlinks and directories
Using the package instrumentation hooks, a file packages-file-list.txt is
created containing the list of files installed by a package with their size.
Due to the use of 'find -type f', symbolic links to files are not included
in this file list. Since the original purpose of this file was to calculate
the total size of a package and symbolic links have file size 0, this was
not a problem.
However, if packages-file-list.txt is reused for other purposes, for example
to get a complete list of files installed by a package regardless of size,
symbolic links to files are important too.
Likewise, to get a complete view of what a package installs, directories
should be included too.
Update the instrumentation hook accordingly.
Although for files an md5sum is taken, we cannot do this for directories.
Instead, mimic the output of md5sum with a fake hash string.
Note: for directories that are used by several packages, e.g. /etc/init.d/,
the package that created the directory will be treated as the 'owner' of
that directory. This gives a somewhat distorted view of ownership.
Similarly, the package size reported by 'make graph-size' (based on
packages-file-list.txt) will include the 'size' of a directory for that
owner, which is reported as e.g. 4096 bytes.
Signed-off-by: Thomas De Schampheleire <thomas.de_schampheleire@nokia.com> Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
instrumentation: extract duplication to get list of installed files
Before and after the building of each package, the instrumentation hooks are
run. One of these hooks obtains the list of files installed by a package.
The code to obtain this list is currently duplicated in the start and end
part of the hook. While the amount of duplication is currently small, a
subsequent patch will make more changes to this code, increasing the
duplication.
Therefore, split off into a helper function.
Signed-off-by: Thomas De Schampheleire <thomas.de_schampheleire@nokia.com> Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Fabio Estevam [Wed, 21 Jun 2017 20:49:47 +0000 (17:49 -0300)]
imx6q-sabresd: Add a qt5 defconfig variant
Introduce imx6q-sabresd_qt5_defconfig that supports the opensource
Etnaviv graphical stack.
This defconfig provides a way to quickly test some graphical applications,
such as kmscube, qt5, glmark2.
Currently kernel mainline exhibits issues when running cpufreq as ondemand
governor on mx6, so add a linux fragment that disables such option for the
time being.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
[Arnout:
- Add Fabio to DEVELOPERS
- Rename linux config fragment to linux_qt5.fragment
- Add comment to linux_qt5.fragment why it is needed
- Add explicit CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE=y to it
- Reorder defconfig entries to they match imx6q-sabresd_defconfig
- Remove strace and host-uboot-tools, neither seems to be needed
- Drop BR2_PACKAGE_QT5BASE_LICENSE_APPROVED, it no longer exists
- Replace BR2_TARGET_ROOTFS_EXT2_BLOCKS with BR2_TARGET_ROOTFS_EXT2_SIZE
- Regenerate .gitlab-ci.yml] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Bernd Kuhls [Wed, 2 Aug 2017 16:29:42 +0000 (18:29 +0200)]
package/kodi: add support for more architectures
For the autoconf-based build system it was necessary to allow each
supported architecture separatly:
https://github.com/xbmc/xbmc/blob/Krypton/m4/xbmc_arch.m4
For the CMake build system this is not necessary anymore so we only
blacklist archs in Config.in known to be broken.
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
[Thomas: slightly tweak the BR2_PACKAGE_KODI_ARCH_SUPPORTS definition.] Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Adam Duskett [Mon, 31 Jul 2017 14:22:09 +0000 (10:22 -0400)]
systemd: bump version to 234
Other changes:
- Removed the following patches due to being merged upstream:
- resolved-bugfix-of-null-pointer-p-question-dereferencing.patch
- resolved-simplify-alloc-size-calculation.patch
- resolved-do-not-allocate-packets-with-minimum-size.patch
- Updated the following patches to work with 234:
- fix-getty-unit.patch
- build-check-for-ln-relative.patch
- fix-am-path-libgcrypt-no-found.patch
- Updated ordering of remaining patches.
- Reformatted remaining patches as git style patches.
- Updated Upstream-Status comment in "build-check-for-ln-relative.patch" to
"Denied [No desire to support building on old distributions]"
Signed-off-by: Adam Duskett <aduskett@gmail.com> Tested-by: Marcus Hoffmann <m.hoffmann@caretelsol.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Max Filippov [Wed, 2 Aug 2017 18:39:59 +0000 (11:39 -0700)]
package/binutils: fix assertion in 2.29 xtensa ld
Section sorting code in xtensa ld was incorrectly cleaned up in release
2.29, resulting in the following assertion when --sort-section,alignment
option is used:
Max Filippov [Wed, 2 Aug 2017 18:40:20 +0000 (11:40 -0700)]
package/binutils: fix crash caused by buggy xtensa overlay
In some xtensa configurations there may be system/user registers in
xtensa-modules with negative index. ISA initialization for such config
may clobber heap and result in program termination.
Don't update lookup table entries for register with negative indices.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This host-only package was used to enable graphite support in GCC 4.9,
but Buildroot no longer supports Graphite optimizations in GCC 4.9
(only GCC >= 5.x is supported). Therefore, we can remove the cloog
package.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Andrey Yurovsky [Tue, 1 Aug 2017 23:43:01 +0000 (16:43 -0700)]
isl: bump to version 0.18
Now that we no longer support Graphite optimizations in gcc 4.9 (which
required ISL 0.14.1), we can bump ISL to a newer version, which allows
to support Graphite optimizations in GCC >= 5.x.
The comment about the version being limited to 0.14.1 because of cloog
is no longer relevant, so it is removed.
Our libtool patch now applies properly, so we can remove
ISL_LIBTOOL_PATCH = NO.
Signed-off-by: Andrey Yurovsky <yurovsky@gmail.com>
[Thomas:
- remove ISL_LIBTOOL_PATCH = NO
- improved commit log] Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Andrey Yurovsky [Tue, 1 Aug 2017 23:43:00 +0000 (16:43 -0700)]
gcc: support grahite only for GCC 5 or above
GCC 4.9.x requires an old version of ISL to support graphite, and this
old version of ISL is not sufficient to support graphite in modern GCC
versions.
Since GCC 4.9.x is getting older, and to keep things simple, we stop
supporting graphite on old GCC < 5.x. This way, we will be able to
bump ISL to a version that is suitable for modern GCC versions.
In addition, this allows to drop the dependency on cloog, which was
only needed for GCC 4.9.
The tests just verify what the /sbin/init binary is, and that we were
able to grab an IP address. More tests can be added later, for example
to check each systemd features (journal, tmpfiles...)
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
[Arnout: regenerate .gitlab-ci.yml] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Tue, 1 Aug 2017 22:52:23 +0000 (00:52 +0200)]
system: make systemd work on a read-only rootfs
When the rootfs is readonly, systemd will expect /var to be writable.
Because we do not really have a R/W filesystem to mount on /var, we make
it a tmpfs [*], and use the systemd-tmpfiles feature to populate it with
"factory" defaults.
We obtain those factory defaults by redirecting /var to that location at
build time, using a symlink /var -> /usr/share/factory which is the
location in which systemd-tmpfiles will look for when instructed to
"recursively copy" a directory.
With a line like:
C /var/something - - - -
it will look for /usr/share/factory/something and copy it (recursively
if it is a directory) to /var/something, but only if it does not already
exist there.
We also mark this copy with the exclamation mark, as it is only safe to
copy on boot, not when changing targets.
To be noted: the real format for such lines are:
C /var/something - - - - /from/where/to/copy/something
But if the source is not given, then it is implicitly taken from
/usr/share/factory (which in our case is as-good a location as whatever
else, so we use it, and thus we need not specify the source of the
copy).
Note that we treat symlinks a little bit specially, by creating symlinks
to the factory defaults rather than copying them.
Finally, /var at build time is a symlink, but at runtime, it must be a
directory (so we can mount the tmpfs over there). We can't change that
as a target-finalize hook, because:
- some packages may want to set ownership and/or access rights on
files or directories in /var, and that only happens while assembling
the filesystem images; changing /var from a symlink to a (then
empty) directory would break this;
- /var would be a directory on sub-sequent builds (until the next
"make clean").
Instead, we use the newly-introduce pre- and post-rootfs command hooks,
to turn /var into a directory before assembling the image, and back to a
symlink after assembling the image.
[*] People who want the factory-defaults only on first boot will have
to tweak the fstab to mount something else than a tmpfs on /var.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Reviewed-by: Romain Naour <romain.naour@gmail.com> Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Tue, 1 Aug 2017 22:52:22 +0000 (00:52 +0200)]
fs: add pre- and post-command hooks
In some cases, the directory structure we want in the filesystem is not
exactly what we have in target/
For example, when systemd is used on a read-only rootfs, /var must be a
tmpfs. However, we may have packages that install stuff in there, and
set important rights (via the permission-table). So, at build time, we
need /var to be a symlink to the remanent location (/usr/share/factory)
while at runtime we need /var to be a directory.
One option would have been to have /var as a real directory even during
build time, and in a target-finalize hook, move everything out of there
and into the "factory" location. However, that's not possible because
it's too early: some packages may want to set ownership and/or acces
rights on directories or files in /var, and this is only done in the
fakeroot script, which is called only later during the assembling of the
filesystem images.
Also, there would have been no way to undo the tweak (i.e. we need to
restore the /var symlink so that subsequent builds continue to work) if
it were done as a target-finalize hook.
The only solution is to allow packages to register pre- and post-hooks
that are called right before and right after the rootfs commands are
executed, and inside in the fakeroot script.
We can however not re-use the BR2_ROOTFS_POST_FAKEROOT_SCRIPT feature
either because it is done before the filesystem command, but there is
nothing that is done after. Also, we don't want to add to, and modify a
user-supplied variable.
So, we introduce two new variables that packages can set to add the
commands they need to run to tweak the filesystem right at the last
moment.
Those hooks are not documented on-purpose; they are probably going to
only ever be used by systemd.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Reviewed-by: Romain Naour <romain.naour@gmail.com> Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Tue, 1 Aug 2017 22:52:21 +0000 (00:52 +0200)]
system: separate sysv and systemd parts of the skeleton
For systemd, we create a simple /etc/fstab with only an entry for /, as
systemd otherwise automatically mounts what it needs where it needs it.
systemd does not like that the content of /var be symlinks to /tmp,
especially journald that starts before /tmp is mounted, and thus the
journal files are hidden from view, which causes quite a bit of fuss...
Instead, move the current /var to a sysv-only skeleton.
systemd at install time will create the /var content it needs, so we
just create an empty /var for systemd.
systemd would create /home and /srv at runtime if they are missing, but
it is better to create them right now, to simplify supporting systemd on
a RO filesystem in the (near) future.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Cc: Maxime Hadjinlian <maxime.hadjinlian@gmail.com> Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Tue, 1 Aug 2017 22:52:20 +0000 (00:52 +0200)]
package/skeleton: introduce sysv, systemd and none specific skeletons
Currently, we use the same skeleton for sysv-like init systems and
systemd, even though systemd has some peculiarities that makes our
default skeleton unfit.
So, we'll need to provide different skeletons (really, only part of
it) for sysv-like and systemd. In addition, in order to support the
"no init system" (BR2_INIT_NONE) use case, we introduce a "none"
skeleton.
Introduce three new skeleton packages, aptly named skeleton-sysv,
skeleton-systemd and skeleton-none. All three are providers of the
skeleton virtual package, in lieu of the skeleton-common package,
which is now a simple dependency of all three new skeletons.
Those packages are empty for now. In followup changes:
- sysv-specific stuff will be moved out of skeleton-common and into
skeleton-sysv;
- systemd-specific stuff will be added to skeleton-systemd.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Arnout:
- merge with the patch that enables the BR2_INIT_NONE case
- simplify the BR2_PACKAGE_SKELETON_COMMON_ONLY select logic] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
[Thomas:
- remove the BR2_PACKAGE_SKELETON_COMMON_ONLY logic, and instead
introduce a separate skeleton-none package for the BR2_INIT_NONE] Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Tue, 1 Aug 2017 22:52:19 +0000 (00:52 +0200)]
package/skeleton-common: simplify staging install
We don't really care what's going in staging, as long as it is properly
setup for merged/non-merged /usr, especially for the lib/ directory.
So we can just copy the skeleton as-is.
This simplifies maintenance, should we ever need to tweak the layout:
we'd just have to do it once in the skeleton directory to have it
propagated to both target and staging.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
skeleton-common: rename SKELETON_ROOT_PASSWORD to SKELETON_COMMON_ROOT_PASSWORD
Package variables should always be prefixed with the package name. For
SKELETON_COMMON_ROOT_PASSWORD, this means we set the variable a second
time based on the value it previously had. Fortunately, this is fine
for make, even for recursively expanded variables.
This also allows to simplify the condition of the empty password - it's
not needed to set it again to empty if it already was empty.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Tue, 1 Aug 2017 22:52:16 +0000 (00:52 +0200)]
package/skeleton: split out into skeleton-common
Move all the handling of the default skeleton into a new package,
skeleton-common.
We don't name it skeleton-default, because it will be further split
later, into a skeleton for sysv and another for systemd, with some parts
still common between the two. So just name it skeleton-common right now;
this will save us a rename later.
While we're at it, also assign to SKELETON_COMMON_TARGET_FINALIZE_HOOKS
instead of directly to the global FINALIZE_HOOKS. Therefore, we don't
need to do all of that in a condition BR2_PACKAGE_SKELETON_COMMON==y.
Note: it would be technically sound to move the skeleton files together
within a sub-directory of the skeleton-common package. However, we refer
the user to those files, from various locations (manual, packages). It
will indeed be easier for the user to find those files in
system/skeleton/ rather than in package/skeleton-common/skeleton/
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
[Arnout: remove the mkdir $(STAGING_DIR)/usr/include which was removed
in skeleton.mk in master.] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
For target install, we don't need to check the merged-usr symlinks
any more, it was already checked before.
For staging, instead of creating directories, just copy the same
skeleton, which was already checked to be correct.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Arnout: split off in a separate patch + wrote commit message] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Tue, 1 Aug 2017 22:52:14 +0000 (00:52 +0200)]
package/skeleton-custom: also check for missing directories
In addition to checking the symlinks in the merged usr case, also check
whether /bin, /lib, /sbin and their /usr/... counterparts exist in the
non-merged case.
Missing directories are collected in the variable
SKELETON_CUSTOM_MISSING_DIRS. For symmetry and because it's more
logical, rename SKELETON_CUSTOM_NOT_MERGED_USR to
SKELETON_CUSTOM_NOT_MERGED_USR_DIRS.
When a directory is indeed missing, "stat" will print an error.
Buildroot *also* prints an error, which is clearer. So remove the error
from stat by redirecting it to /dev/null.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Arnout: split off in a separate patch + wrote commit message] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Tue, 1 Aug 2017 22:52:13 +0000 (00:52 +0200)]
package/skeleton-custom: rework the merged_usr and building conditions
skeleton-custom.mk contains a huge condition for BR_BUILDING and for
BR2_ROOTFS_MERGED_USR. This makes the code difficult to follow, so
rework the condition a bit so that only the parts that really need to
be conditional are indeed conditional.
Note that there is no overhead in needlessly evaluation the "stat"
commands. Indeed, the assignments use late evaluation so the "stat"
is only executed when evaluating the condition - when skeleton-custom
is not selected, stat is never called.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Arnout: split off in a separate patch + wrote commit message] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Tue, 1 Aug 2017 22:52:12 +0000 (00:52 +0200)]
package/skeleton: split out into skeleton-custom
For the custom skeleton, we practicaly do nothing, except ensure it
contains the basic, required directories, and that those are properly
setup wrt. merged /usr.
Furthermore, our current skeleton is not fit for systemd, and we'll
have to split things out into various skeletons.
So, off-load the custom skeleton into its own package.
Thus, the existing skeleton package is now limited to:
- when using our default skeleton, install and tweak it properly;
- when using a custom skeleton, do nothing except for depending on
the skeleton-custom package.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Arnout: split off in a separate patch doing only this] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The tests just verify what the /sbin/init binary is, and that we were
able to grab an IP address. More tests can be added later, for example
to check each systemd features (journal, tmpfiles...)
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Arnout Vandecappelle <arnout@mind.be>
[Arnout: update .gitlab-ci.yml] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
package/samba4: allow to use libtirpc instead of internal C implementation
This patch allows building samba4 with uClibc again after rpc support
was removed from uClibc. Building with musl is still broken:
The first error
../nsswitch/wins.c: In function ‘_nss_wins_gethostbyname_r’:
../nsswitch/wins.c:272:15: error: ‘NETDB_INTERNAL’ undeclared (first use in this function)
*h_errnop = NETDB_INTERNAL;
^~~~~~~~~~~~~~
../nsswitch/wins.c:272:15: note: each undeclared identifier is reported only once for each function it appears in
../nsswitch/wins.c:349:14: error: ‘NETDB_SUCCESS’ undeclared (first use in this function)
*h_errnop = NETDB_SUCCESS;
^~~~~~~~~~~~~
could be solved by applying
https://git.alpinelinux.org/cgit/aports/tree/main/samba/netdb-defines.patch
but then the build fails at
../ctdb/tests/src/test_mutex_raw.c: In function ‘main’:
../ctdb/tests/src/test_mutex_raw.c:170:28: error: ‘pthread_mutex_t {aka struct <anonymous>}’ has no member named ‘__data’
printf("pid=%u\n", mutex->__data.__owner);
^~
Therefore we disable musl support until a fix can be found.
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
[Thomas:
- instead of setting SAMBA4_CONF_ENV within the libtirpc condition,
introduce SAMBA4_CFLAGS and SAMBA4_LDFLAGS variables and use them
- add missing dependency on host-pkgconf, needed because we call
pkg-config to get the flags from libtirpc] Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Thomas Petazzoni [Sat, 29 Jul 2017 20:06:26 +0000 (22:06 +0200)]
gdb: force to use ncurses and not termcap for host-gdb
Both our target and host gdb depend on ncurses (host-ncurses for
host-gdb, of course). However, while for the target we passs
--with-curses, we are not doing this for the host variant. Due to
this, host-gdb default to using the termcap library: if such a library
is available on the build system, it will be used instead of the
host-ncurses we have built. This causes the host gdb binary to depend
on a library that we do not provide in $(HOST_DIR), breaking the
principle of a standalone SDK (which should only depend on the C
library).
To solve this, we simply pass --with-curses in HOST_GDB_CONF_OPTS,
which forces host-gdb to use the host-ncurses library.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Reviewed-by: Romain Naour <romain.naour@gmail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yegor Yefremov [Tue, 1 Aug 2017 08:20:13 +0000 (10:20 +0200)]
python-hpack: fix build breakage due to UTF-8 formatted file
UTF-8 formatted file breaks installations on some systems:
Traceback (most recent call last):
File "setup.py", line 34, in <module>
long_description=open('README.rst').read() + '\n\n' + open('HISTORY.rst').read(),
File "/home/peko/autobuild/instance-1/output/target/usr/lib/python3.6/encodings/ascii.py", line 26, in decode
return codecs.ascii_decode(input, self.errors)[0]
UnicodeDecodeError: 'ascii' codec can't decode byte 0xc2 in position 1125: ordinal not in range(128)
This patch replaces '§' with 'S' meaning 'Section' and hence converts file to plain ASCII.
package/skeleton: select it rather than default to y
Our current skeleton is tailored to sysv-like init systems; it is not
fit for systemd-based systems. So, in upcoming changes, we'll add
another skeleton for systemd.
This means we can no longer have the current skeleton default to 'y', or
it would be enabled also for systemd, which would be incorrect.
So, we remove the default to 'y' but have it selected by the default
skeleton choice.
However, we do not yet have a way to directly build (really, install)
the custom skeleton, it is built (really, installed) as a dependency of
the default skeleton. So we must also forcibly select the default
skeleton when using a custom one.
Until we have the means to do only one or the other; i.e. when we have a
virtual skeleton.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
package/skeleton: drop dependency on host-mkpasswd
Setting the root pasword is done in a target-finalize hook, so we do not
need to enforce a dependency from the skeleton onto host-mkpasswd.
Dropping that dependency will simplify making skeleton a virtual
package (in up-coming changes).
Instead, it is now selected as any other package. As such, it is
guaranteed to be built before target-finalize.
This however introduces a slight change in behaviour: previously,
host-mkpasswd would only be built if we needed to hash the root password
from its plain-text value. Now, host-mkpasswd is always built as soon as
the root password is non-empty, even if already pre-hashed.
Since host-mkpasswd is a really tiny weeny package bundled in Buildroot,
with only two C files, built as a single unit with a single gcc call,
the overhead is really minimal. Compared to the simplifications this
will allow in the skeleton packages (plural: common, sysv, systemd,
custom) to come, this overhead is acceptable.
Yet another simplification, even if small, to ease providing multiple
skeletons.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
projects / coffee / buildroot.git / log