Some ZynqMP device trees in the kernel have a dot in their name,
for example "zynqmp-zcu102-rev1.0". The post image script
left out the part after the dot and created an invalid symlink.
Signed-off-by: Francois Beerten <fbeerten.git@colabti.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 4f56655b2b)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since commit 916b21a7fb ("package/tar:
Depends on MMU"), BR2_PACKAGE_TAR depends on BR2_USE_MMU. However, the
Config.in comment does not take into account this dependency, an
inconsistency which is fixed by this commit.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 2a3ef05520)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The new version, among other changes, includes important fixes
for unaligned memory access on ARM (both for 32 and 64-bit), as well
as performance improvements and build fixes.
Patch "0001-Tell-CMake-to-not-check-for-a-C-compiler.patch" is not
needed due to the issue being fixed upstream, and therefore is removed.
Patch "0001-CMake-Allow-using-BUILD_SHARED_LIBS-to-choose-static.patch"
is rebased against the latest upstream changes.
Signed-off-by: Adrian Perez de Castro <aperez@igalia.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit be733b54a7)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When the runtime tests were first introduced, they still ran pretty
quickly. Nowadays, however, there are a lot of runtime tests, and some
of them take a really long time. So running them on every push is
really too much.
Just like we do for the defconfigs, run them on explicit trigger only.
The explicit trigger is now done every week, but it can be increased
to e.g. twice or three times per week.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 96123c1c4f)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When we do a release, we know only of a set of gcc versions that the
host may have. But in the future, distributions with newer gcc versions
may show up.
Currently, we do not recognise those versions, and thus we do as if they
were older than the oldest we know of. This means that a set of packages
become unselectable, when they should be.
We fix that by capping the detected version to the highest we know of.
Reported-by: gargar_ on IRC
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Peter Korsgaard <peter@korsgaard.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 3950e69dad)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Qt KMS support should not explicitly depend on the availability of opengl.
Don't explicitly disable KMS if opengl is not available and fallback to
detecting if libdrm is available before disabling kms.
The scenario where this is necessary involves using the Qt linuxfb backend
"dumb buffer" support via the DRM API. This is new in Qt 5.9 [1] and only
requires KMS, but not opengl. Although on Qt 5.6, only eglfs actually
uses libdrm/kms, it doesn't hurt to add the dependency and the -kms
option there as well, and doing so keeps the logic in the .mk file
simple.
[1] http://doc.qt.io/qt-5/embedded-linux.html#linuxfb
Cc: Peter Seiderer <ps.report@gmx.net>
Cc: Julien Corjon <corjon.j@ecagroup.com>
Signed-off-by: Joshua Henderson <joshua.henderson@microchip.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
(cherry picked from commit f91ea94a6f)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The utils/diffconfig script works only on variables with the BR2_
prefix. This is OK for Buildroot [def]configs since this is the prefix
for all user-facing variables, but it prevents using the same script
to compare configs from kconfig-based packages.
Remove the BR2_ restriction, allowing usage such as:
./utils/diffconfig \
board/qemu/xtensa-lx60/linux.config \
board/qemu/xtensa-lx60/linux-nommu.config
Signed-off-by: Marcel Patzlaff <m.patzlaff@pilz.de>
Reviewed-by: Luca Ceresoli <luca@lucaceresoli.net>
Tested-by: Luca Ceresoli <luca@lucaceresoli.net>
Reviewed-by: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit bf9ccfc37b)
Signed-off-by: Peter Korsgaard <peter@korsgaard.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>
(cherry picked from commit 076fd27da7)
Signed-off-by: Peter Korsgaard <peter@korsgaard.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>
(cherry picked from commit c61b49e5b5)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
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>
(cherry picked from commit 42b10634c6)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
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>
(cherry picked from commit d07ddd8e4e)
Signed-off-by: Peter Korsgaard <peter@korsgaard.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>
(cherry picked from commit 7007dc2bc9)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes CVE-2018-10873: A vulnerability was discovered in SPICE before version
0.14.1 where the generated code used for demarshalling messages lacked
sufficient bounds checks. A malicious client or server, after
authentication, could send specially crafted messages to its peer which
would result in a crash or, potentially, other impacts.
Drop patches as they are now upstream.
Add host-pkgconf as the configure script uses pkg-config. Drop removed
--disable-automated-tests configure flag.
Add optional opus support, as that is now supported and needs to be
explicitly disabled to not use. Explicitly disable optional gstreamer
support for now as the dependency tree is fairly complicated.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit f33f7a4f64)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Currently, when a custom user table and a package define the same user,
the settings from the package takes precedence over the ones from the
custom user table.
However, it makes sense to allow the settings from the custom user table
take precedence. For example, it would allow redirecting the user's
home directory to an alternate location (e.g. away from tmp and into a
partition that is persistent).
The support/scripts/mkusers script will only retain settings from the
latest definition it finds.
Thus, by passing the custom user table after the package defined users,
it is possible to override the package provided user definitions.
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Matthew Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
(cherry picked from commit c3edec0018)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The license heading in source files includes the "or any later"
language.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit cfa3447a78)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes CVE-2017-14501: An out-of-bounds read flaw exists in
parse_file_info in archive_read_support_format_iso9660.c in libarchive
3.3.2 when extracting a specially crafted iso9660 iso file, related to
archive_read_format_iso9660_read_header.
Drop upstream patches.
Use upstream provided tarball hash.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 946f136fe1)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
U-Boot has a copy of dtc in-tree. However, it has a bug in its build
system which could result in both one of the in-tree dtc include files
and the same host-installed include file to be #included.
Normally, that wouldn't be a problem, because (a) the two include files
are compatible, so it doesn't matter which one you include, and (b) the
include guards are the same in both, so only one of them really does
get included. However, upstream dtc has changed the include guards,
removing the leading underscore. Therefore, now the header file does
get included twice, which leads to multiple definitions like:
/builds/buildroot.org/buildroot/output/host/include/libfdt.h:1790:19: error: redefinition of 'fdt_appendprop_cell'
static inline int fdt_appendprop_cell(void *fdt, int nodeoffset,
^~~~~~~~~~~~~~~~~~~
In file included from tools/fdt_host.h:11:0,
from tools/imagetool.h:24,
from tools/atmelimage.c:8:
tools/../include/libfdt.h:1656:19: note: previous definition of 'fdt_appendprop_cell' was here
static inline int fdt_appendprop_cell(void *fdt, int nodeoffset,
^~~~~~~~~~~~~~~~~~~
To fix this, patch (host) dtc to accept the old include guard as well,
which restores the old behaviour. This patch is probably not
upstreamable, since it's really a hack to work around an issue in
U-Boot. Note that it has been fixed upstream, but Buildroot supports
building older versions of U-Boot as well.
Note that the problem may still occur if you have libdtc-dev installed
on the host. However, now there is a simple workaround: enable
BR2_TARGET_UBOOT_NEEDS_DTC.
Note that a similar problem also occurs with the beaglebone fork of the
kernel. It's not clear if it has been fixed there.
Signed-off-by: Lothar Felten <lothar.felten@gmail.com>
[Arnout: rewrite commit message, rewrap patch commit message]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
(cherry picked from commit c7ffd8a75d)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When two Buildroot builds run in parallel, and they both happen to call
npm at roughly the same time, the two npm instances may conflict when
accessing the npm cache, which is by default ~/.npm
Although npm is supposed to lock access to the cache, it seems it does
sometimes fail to do so properly, bailling out in error, when it would
never ever crash at all when not running in parallel. We suspect that
the sequence leading to such failures are something like:
npm-1 npm-2
lock(retry=few, sleep=short) .
does-stuff() .
. lock(retry=few, sleep=short)
. # can't lock local cache
. download-module()
. # can't download
. exit(1)
unlock()
As per the docs [0], few = 10, short = 10. So if the first npm (npm-1)
takes more than 100s (which can happen behind slow links and/or big
modules that contain native code that is compiled), then the second npm
(npm-2) will bail out (the download would fail if there is no network
access, for example, and only local modules are used).
Point npm to use a per-build cache directory, so they no longer compete
across builds.
That would still need some care when we do top-level parallel builds,
though.
Note also that the conflicts are not totally eliminated: two or more npm
instances may still compete for some other resource that has not yet
been identified.
But, at least, the conflict window has been drastically shortened now,
to the point where it now seldom occurs.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 4a16182d5f)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
While Erlang includes a version of zlib, it's intended for Windows and
there's an expectation that non-Windows platforms provide it. It's also
not as regularly updated as the one in Buildroot. This change makes
Erlang always use a Buildroot-provided zlib.
Fixes this compile error:
CC /home/buildroot/autobuild/run/instance-0/output/build/erlang-21.0/erts/emulator/zlib/obj/x86_64-buildroot-linux-musl/opt/adler32.o
In file included from zlib/adler32.c:11:0:
zlib/zutil.h:172:39: error: "_LFS64_LARGEFILE" is not defined [-Werror=undef]
(!defined(_LARGEFILE64_SOURCE) || _LFS64_LARGEFILE-0 == 0)
^~~~~~~~~~~~~~~~
See http://autobuild.buildroot.net/results/fc633f80c7c36a90e641487f5a888fbb767c2a54/.
Signed-off-by: Frank Hunleth <fhunleth@troodon-software.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit ec5378038f)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
As reported by Jeff Wittrock in bug #11396, the U-Boot environment
image checksum is invalid for big endian targets, because the test on
the BR2_ENDIAN Config.in option doesn't take into account that it is
double quoted.
The fix was provided by Jeff himself on bugzilla.
Fixes bug #11396.
Reported-by: Jeff Wittrock <jwittrock@faultrecorder.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit d6fcf044a7)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
For some reason, ustr installs its own source code, which means we end
up with 448 KB of source code in /usr/share in the target filesystem:
$ tree output/target/usr/share/
output/target/usr/share/
└── ustr-1.0.4
├── malloc-check.h
├── ustr-b-code.h
├── ustr-b-dbg-code.c
├── ustr-b-opt-code.c
├── ustr-cmp-code.h
├── ustr-cmp-dbg-code.c
├── ustr-cmp-internal.h
├── ustr-cmp-opt-code.c
├── ustr-cntl-code.h
├── ustr-fmt-code.h
├── ustr-fmt-dbg-code.c
├── ustr-fmt-internal.h
[...]
$ du -sh output/target/usr/share/ustr-1.0.4/
448K output/target/usr/share/ustr-1.0.4/
So let's drop this source code in a post-install target hook.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit c27484b2ef)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes CVE-2018-10933: authentication bypass vulnerability in the server
code. By presenting the server an SSH2_MSG_USERAUTH_SUCCESS message in
place of the SSH2_MSG_USERAUTH_REQUEST message which the server would
expect to initiate authentication, the attacker could successfully
authenticate without any credentials.
https://www.libssh.org/security/advisories/CVE-2018-10933.txt
Drop an upstream patch.
Cc: Scott Fan <fancp2007@gmail.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit de24e47d90)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Pass TARGET_LDFLAGS to EXTRA_LDFLAGS to fix following issue:
/home/buildroot/autobuild/run/instance-3/output/build/host-gcc-final-7.3.0/build/arm-buildroot-linux-musleabihf/libgcc/../../../libgcc/config/arm/lib1funcs.S:1545: undefined reference to `raise'
Also pass TARGET_CFLAGS to EXTRA_CFLAGS and TARGET_CXXFLAGS to
EXTRA_CXXFLAGS and move all these variables to
OPEN_PLC_UTILS_MAKE_OPTS for readability
Fixes:
- http://autobuild.buildroot.org/results/67bc5e7ac8ae1c49c035b022a394d2f746705cf2
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit d8738d3b97)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
In commit a31a66802a ("freetype:
security bump to version 2.5.3"), the freetype package was changed to
call ./autogen.sh to regenerate the autotools stuff, because the
ltmain.sh provided by upstream freetype was not compatible with
Buildroot libtool-patching logic.
Since then, freetype has been bumped several times, and the current
version packaged in Buildroot has an ltmain.sh that is compatible with
our libtool-patching logic.
Therefore, this commit drops the no longer needed autogen stuff.
This autogen stuff was badly breaking per-package host/target
directory, because the autogen happened at the post-patch hook step,
at which point the host-automake/host-autoconf/host-libtool
dependencies have not yet been copied into this package host
directory.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 88c6329521)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Use system liblinear instead of using included liblinear.
liblinear in buildroot is at version 2.20 released on December 2017
whereas liblinear in nmap has not been updated since 7 years (except for
liblinear.vcxproj which has been updated 2 years ago)
Do not use --with-liblinear option as otherwise nmap will forget to add
-llinear to LIBS due to the following line in configure.ac:
if test $have_liblinear != yes; then
AC_CHECK_HEADERS([linear.h],
AC_CHECK_LIB(linear, predict, [have_liblinear=yes; LIBLINEAR_LIBS="-llinear"; break],, [-lm])
)
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit c0d9ba562c)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When used without spdx_lookup the BSD licence cannot be
detected correctly because many Python packages just specify
BSD without the exact version in their metadata. So add a
special message warning the user instead of the licence id.
Bonus: fix typo.
Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit d05e41eb1a)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Commit 87d759ced5 (ntp: fix build for no-MMU) added a patch to make MMU
dependent code hidden behind HAVE_WORKING_FORK. It turns out that the
patch covers too much code. When libcap is enabled we pass
--enable-linuxcaps, which in turn enables HAVE_DROPROOT. This adds calls
to code that is covered by HAVE_WORKING_FORK.
Update the no-MMU fix so that HAVE_WORKING_FORK only covers the no-MMU
incompatible routine.
Fixes:
http://autobuild.buildroot.net/results/c5c/c5cf28bb969fec7c07864cdd094dedfa4d5439d2/
Cc: Artem Panfilov <apanfilov@spectracom.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 7cf152852d)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
