gstreamer can optionally build against libxml2, so account for it in
the dependencies.
Functionality seems somewhat reduced without it, an example being
misdetecting subtitle (.srt) files as subrip mime type instead of
subtitle itself.
A failed scenario would be:
$ gst-typefind-0.10 some.srt
some.srt - application/x-subrip
When built with libxml2:
$ gst-typefind-0.10 some.srt
some.srt - application/x-subtitle
This doesn't affect gstreamer 1.x since it dropped any use of libxml2.
Signed-off-by: Gustavo Zacarias <gustavo.zacarias@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
It is general safe to use -fPIC for all architectures.
-fpic breaks sparc64 compile.
Generally gcc just optimize position independent code
for m68k, powerpc and sparc with -fpic.
The size differences are minimal, f.e. for powerpc:
text data bss dec hex filename
235983 5336 684 242003 3b153 output/target/usr/lib/libjim.0.75
236255 8456 684 245395 3be93 output/target/usr/lib/libjim.so.0.75
So instead of keeping special settings for sparc64 it would be better
to always use -fPIC in general.
Runtime tested on Qemu ARM, MIPS, PPC and SPARC64.
Fixes:
http://autobuild.buildroot.net/results/98f391ad13f22828c022f185c0166daabdb4c1ad/
[Peter: tweak comment as suggested by Thomas]
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The created symlinks are wrong.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
With OVERRIDE_SRCDIR we don't apply any of the qt5base patches, but the
custom specs files are needed to be able to build - So install these in the
configure step instead of having them as a patch.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Fixes only one connection regresssion in 5.3.4, see
https://wiki.strongswan.org/issues/1213
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
IDN can be picked up from the distro installation so use a prefix when
it's available or otherwise just disable it, fixes:
http://autobuild.buildroot.org/results/5d9/5d9baf528c47c5167ec8d2a6941cb06b4ca761ca/
Also disable libgcrypt support which wasn't accounted for and doesn't
seem to work either to possibly avoid the same problem.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This commit adds a patch to quagga's configure.ac to fix a library
ordering problem that is causing a build failure in static linking
scenario.
Fixes:
http://autobuild.buildroot.org/results/c08/c086f3128232af70d24fd5faf9282ce91d6488ef/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Static linking userspace programs such as MPD against libaudiofile fails if
FLAC is available, because libaudiofile is linked against FLAC, but this isn't
expressed in the pkg-config file:
[..]
arm-buildroot-linux-uclibcgnueabi/sysroot/usr/lib/libaudiofile.a(FLAC.o): In function `FLACDecoder::reset2()':
FLAC.cpp:(.text+0x58): undefined reference to `FLAC__stream_decoder_seek_absolute'
/home/buildroot/build/instance-1/output/host/usr/arm-buildroot-linux-uclibcgnueabi/sysroot/usr/lib/libaudiofile.a(FLAC.o): In function `FLACEncoder::sync2()':
FLAC.cpp:(.text+0x88): undefined reference to `FLAC__stream_encoder_finish'
/home/buildroot/build/instance-1/output/host/usr/arm-buildroot-linux-uclibcgnueabi/sysroot/usr/lib/libaudiofile.a(FLAC.o): In function `FLACDecoder::~FLACDecoder()':
FLAC.cpp:(.text+0xc4): undefined reference to `FLAC__stream_decoder_delete'
/home/buildroot/build/instance-1/output/host/usr/arm-buildroot-linux-uclibcgnueabi/sysroot/usr/lib/libaudiofile.a(FLAC.o): In function `FLACEncoder::~FLACEncoder()':
FLAC.cpp:(.text+0x164): undefined reference to `FLAC__stream_encoder_delete'
/home/buildroot/build/instance-1/output/host/usr/arm-buildroot-linux-uclibcgnueabi/sysroot/usr/lib/libaudiofile.a(FLAC.o): In function `FLACDecoder::runPull()':
[..]
The Libs.private field is specifically designed for such usage:
>From pkg-config documentation:
Libs.private:
This line should list any private libraries in use. Private
libraries are libraries which are not exposed through your
library, but are needed in the case of static linking.
Therefore, this patch adds a reference to FLAC as well as to lcov in the
Libs.private field of the pkg-config file.
Patch status: Pending
https://github.com/mpruett/audiofile/pull/26
Fixes:
http://autobuild.buildroot.net/results/e95/e959e703a8db81032da27bece295c121e53d830d/http://autobuild.buildroot.net/results/10b/10bd00209f098e8782016355cb2f46d1d3e7d3a3/
Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
As discussed on the list it requires a matching XDG runtime API level
which is currently not the case (wayland/weston newer than libgtk3 at
the moment). It doesn't normall exhibit build failures, but it fails at
runtime regardless.
Also fixes:
http://autobuild.buildroot.net/results/aae/aae4266c16189b295cc7ceb60000c504ebb60752/
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes a port-forwarding regression in 2015.68
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When cross-compiling it assumes there's none, however that's not true.
Specify the usual /dev/urandom for this usage, it's used for random key
generation on startup.
Signed-off-by: Gustavo Zacarias <gustavo.zacarias@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
autossh looks for 'ssh' in the PATH, which may not exist on a build
machine.
Fixes a private autobuild failure (soon to go public).
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Peter Korsgaard <jacmet@uclibc.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The iodine build system correctly appends the needed CFLAGS/LDFLAGS to what
was passed in the environment, so use TARGET_CONFIGURE_OPTS to ensure our
flags are passed.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
- Bump version to 4.3.0
- Update hash value
- Add new mandatory blkid dependency
- Remove upstreamed patch:
0001-xfsprogs-replaces-static-with-static-libtool-libs.patch
- Add new patch which has been sent upstream:
0001-mdrestore-do-not-do-dynamic-linking-of-libtool-libra.patch
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Adjust patch 0001 to trivial changes.
Adjust patch 0007 to remove new DES things.
Add patch 0008 directly from upstream to fix a late problem in the
1.0.1 release.
Signed-off-by: Peter Rosin <peda@axentia.se>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Upstream has reworked its build system, so it now supports PREFIX and 'make
install', so use those.
Notice that the tools have also gained a sunxi- prefix, so use that as well
when installing the target variant.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This patch bumps web2py to the latest released version (2.12.3)
and adds hash file.
Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Patches are upstream so drop them.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The problem looks like this:
CCLD ntpd
../compat/.libs/libcompat.a(libcompat_la-arc4random.o): In function
`arc4random':
/br/output/build/openntpd-5.7p4/compat/arc4random.c:183: undefined
reference to `pthread_mutex_lock'
/br/output/build/openntpd-5.7p4/compat/arc4random.c:185: undefined
reference to `pthread_mutex_unlock'
../compat/.libs/libcompat.a(libcompat_la-arc4random.o): In function
`arc4random_buf':
/br/output/build/openntpd-5.7p4/compat/arc4random.c:192: undefined
reference to `pthread_mutex_lock'
/br/output/build/openntpd-5.7p4/compat/arc4random.c:194: undefined
reference to `pthread_mutex_unlock'
/br/output/build/openntpd-5.7p4/compat/arc4random.c:194: undefined
reference to `pthread_mutex_unlock'
collect2: error: ld returned 1 exit status
Adding -pthread to CFLAGS fixes the problem.
Fixes:
http://autobuild.buildroot.net/results/b12/b120b6da87ae4eb355d06f2fd8f4fc8050410897/
[Peter: unconditionally pass in CFLAGS, add comment about why]
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The explanations given in the commit log of 7d6c79 (Compile static
versions of gcc libraries) do not explain why we have to provide custom
configure commands, instead of just adding --enable-static to the
configure options.
Add a comment in the code that explains why that is so. Add a pointer to
the ML archives with the explanations, too.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Jérôme Pouiller <jezz@sysmic.org>
Acked-by: Jérôme Pouiller <jezz@sysmic.org>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The host configure commands are not indented the same way as the target
ones. Besides, there are leading spaces (instead of TABs) on one line.
Replicate the same indentation as used for the target commands.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Quote from https://github.com/libass/libass/blob/master/Changelog
"* Drop ENCA support"
Therefore we remove the dependency to libenca.
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
HOSTCC_VERSION is no longer used since gcc switched to the package
infrastructure in e236fe481. It was in fact no longer needed since we
dropped support for gcc 4.3.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes build with glibc-2.22.
Also fixes hard-to-trigger deadlock in highly threaded case, and fix
libraries dependencies.
Misc minor improvements in test programs as well.
Fixes#8446.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: gamadril.dev@gmail.com
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When building with BR2_STATIC_LIBS=y, no shared library gets built, so
the installation of libopagent.so* fails. This commit gets rid of this
installation step when BR2_STATIC_LIBS=y.
Fixes:
http://autobuild.buildroot.net/results/2a6604622392a3f127fb61505e2751a88a90af7b/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Add a patch from upstream to fix build error:
error: no match for 'operator<<'
Fixes:
http://autobuild.buildroot.net/results/ed4508a00663b03632a4eb9411ba05852d02fd88/
Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This patch bumps web2py version to 2.11.3
Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
In case libpng is not found apitrace its own version of libpng. However
current bundled version (1.1.16) is impacted by
https://sourceforge.net/p/libpng/bugs/230/. We simplify problem by
forcing use of external libpng.
Notice this patch also impact use of zlib. Apitrace also use its own
version of zlib if none was found. Since we select libpng, external
zlib is also selected.
Fix:
http://autobuild.buildroot.net/results/ac790191127cf4d8b19d0c7c640981264772f503/
Signed-off-by: Jérôme Pouiller <jezz@sysmic.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Remove two patches applied upstream:
0e2be0c21724ccb6ebc5
Due to the removed patches autoreconf is not necessary anymore.
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Fix (works around) bug #6872https://bugs.busybox.net/show_bug.cgi?id=6872
The gpsd package was disabled due to an internal compiler error for
the microblaze architecture in the following commit:
http://git.buildroot.net/buildroot/commit/?id=d517415bcda380336f4a27bf248cef4d81791c9e
The problem is due to an specific GCC optimization flag
(-fexpensive-optimizations), so disabling it we can work around this
problem and compile gpsd for the microblaze architecture.
Signed-off-by: Sergio Prado <sergio.prado@e-labworks.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
- Fixes:
- CVE-2015-5312 - Another entity expansion issue
- CVE-2015-7497 - Avoid an heap buffer overflow in xmlDictComputeFastQKey
- CVE-2015-7500 - Fix memory access error due to incorrect entities boundaries
- CVE-2015-8242 - Buffer overead with HTML parser in push mode
- Incorporates upstreamed patches as well, which also fixed:
- CVE-2015-1819 - The xmlreader in libxml allows remote attackers to cause
a denial of service (memory consumption) via crafted XML data, related
to an XML Entity Expansion (XEE) attack.
- CVE-2015-7941 - out-of-bounds memory access.
- CVE-2015-7942 - heap-buffer-overflow in xmlParseConditionalSections.
- CVE-2015-8035 - DoS via crafted xz file.
Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This patch fixes the following error messages:
ustr-cmp-code-so-dbg.o: In function `ustr_pool_make_subpool':
ustr-cmp-dbg-code.c:(.text+0x0): multiple definition of `ustr_pool_make_subpool'
ustr-b-code-so-dbg.o:ustr-b-dbg-code.c:(.text+0x0): first defined here
ustr-cmp-code-so-dbg.o: In function `ustr_pool_clear':
ustr-cmp-dbg-code.c:(.text+0x44): multiple definition of `ustr_pool_clear'
ustr-b-code-so-dbg.o:ustr-b-dbg-code.c:(.text+0x44): first defined here
ustr-cmp-code-so-dbg.o: In function `ustr_pool_free':
...
That's because the application assumes gnu89 behaviour of "inline" and
"extern inline", but gcc5 defaults to gnu11. So we need to force gnu89
in the compilation flags.
Fixes:
http://autobuild.buildroot.org/results/fde02ae2883a058b98e57ba087466d820ab1a5eb/http://autobuild.buildroot.org/results/e04849c31e2a2528282ce6cdcf69aa10076d721d/http://autobuild.buildroot.org/results/7b3295e711fc51907c311adb63f0b7d06a90a7b9/
...and many others.
Signed-off-by: Sergio Prado <sergio.prado@e-labworks.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Disable gtests since they require C++ support.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This feature was already discuted here:
http://lists.busybox.net/pipermail/buildroot/2015-April/125419.html
Personally, I have a big central directory with all tarballs I have
ever downloaded. I use this feature to isolate tarballs necessary to
build a configuration:
make project_defconfig
make BR2_DL_DIR=/tmp/mirror-project BR2_PRIMARY_SITE=file:///home/user/dl source
tar -C /tmp -czvf mirror-project.tgz mirror-project
Signed-off-by: Jérôme Pouiller <jezz@sysmic.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The sdl libpth autodetection gets confused in case a host
version is present. Overwrite libpth autodection for the
target build dependent on BR2_PACKAGE_LIBPTHSEM_COMPAT and
use pth-config from staging directory.
Fixes [1]:
checking for pthreads... no
checking for pth-config... /usr/bin/pth-config
checking pth... yes
[...]
./src/SDL.c:33:17: fatal error: pth.h: No such file or directory
#include <pth.h>
[1] http://autobuild.buildroot.net/results/733/73330e51b3c099d179004db0b9522ac13358b8f7
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This commit adds two patches to the iproute2 package that are
necessary to build it against the musl C library. Both patches have
been submitted upstream.
Fixes:
http://autobuild.buildroot.org/results/5d3/5d3f8c2242e9a89bb545166fc7d70429d2cddd42/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
With the new version, documentation is not generated part of the main
build. This removes the need to alter the Makefile to remove the invoke
on the `chrony.txt` target. Cleanup also explicitly disabling seccomp
(since its experimental) and tomcrypt (Buildroot does not have
libtomcrypt at this time).
[Thomas: use tabs for indentation in the definition of
CHRONY_CONF_OPTS.]
Signed-off-by: James Knight <james.knight@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The host variant was added by us (DATACOM) because it is useful to
build some third-party packages that we use in our firmware. We humbly
request you to keep it alive.
Declare that it requires host-perl-module-build.
Signed-off-by: Carlos Santos <casantos@datacom.ind.br>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Like c-periphery and lua-periphery, libserial fails to build because
it tries to use some baud rate definitions that are not available on
sparc and sparc64 (the highest baud rates).
This commit fixes that by introducing a patch to libserial to make the
use of these high baud rates conditional on their availability.
Fixes:
http://autobuild.buildroot.org/results/f9b/f9bbb8a6636cd3e3203b059f627aac7b1d511eb2/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
libvorbis.pc does not exist, it is named vorbis.pc.
Also, sqlite3 was missing when querying pkg-config.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This is the only package where we select directfb. Change it to use depends
on to match what we do for the X11 backend.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
If system tools are selected, the host's lib/ directory may shadow
libraries from the system which are configured differently and do not
have all of the symbols required by the system tool.
Since buildroot now uses rpath everywhere, LD_LIBRARY_PATH should not
be necessary anyways.
Signed-off-by: Ben Boeckel <mathstuf@gmail.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When we build our host programs, and they depend on a host library we
also build, we want to ensure that program actually uses that library at
runtime, and not the one from the system.
We currently ensure that in two ways:
- we add a RPATH tag that points to our host library directory,
- we export LD_LIBRARY_PATH to point to that same directory.
With these two in place, we're pretty much confident that our host
libraries will be used by our host programs.
However, it turns our that not all the host programs we build end up
with an RPATH tag:
- some packages do not use our $(HOST_LDFLAGS)
- some packages' build system are oblivious to those LDFLAGS
In this case, there are two situations:
- the program is not linked to one of our host libraries: it in fact
does not need an RPATH tag [0]
- the program actually uses one of our host libraries: in that case it
should have had an RPATH tag pointing to the host directory.
For libraries, they only need an RPATH if they depend on another library
that is not installed in the standard library path. However, any system
library will already be in the standard library path, and any library we
install ourselves is in $(HOST_DIR)/usr/lib so already in RPATH.
We add a new support script that checks that all ELF executables have
a proper DT_RPATH (or DT_RUNPATH) tag when they link to our host
libraries, and reports those file that are missing an RPATH. If a file
missing an RPATH is an executable, the script aborts; if only libraries
are are missing an RPATH, the script does not abort.
[0] Except if it were to dlopen() it, of course, but the only program
I'm aware of that does that is openssl, and it has a correct RPATH tag.
[Peter: reworded as suggested by Arnout, fix HOT_DIR typo in comment]
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>
Cc: Peter Korsgaard <jacmet@uclibc.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes:
CVE-2015-1819 - The xmlreader in libxml allows remote attackers to cause
a denial of service (memory consumption) via crafted XML data, related
to an XML Entity Expansion (XEE) attack.
CVE-2015-7941 - out-of-bounds memory access.
CVE-2015-7942 - heap-buffer-overflow in xmlParseConditionalSections.
CVE-2015-8035 - DoS via crafted xz file.
All patches upstream.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Configuring the network interface with DHCP via
/etc/network/interfaces generally does not work when NFS booting. The
DHCP configuration will initially bring the interface down and system
then hangs at that point as the root filesystem is no longer
accessable.
This adds a script to be run via a pre-up line in
/etc/network/interfaces. It checks for a NFS root filesystem and if
the interface to be configured is used for access to the NFS server.
If this is the case, then it returns a failure code so that ifup will
not configure the interface. This works for DHCP and another config
methods (static, bootp, etc.). This system does detect if the
interface to be configured isn't the one used for NFS and doesn't skip
it when that is the case.
NFS filesystems that aren't the root fs aren't considered.
Fixes bug #4790.
Signed-off-by: Trent Piepho <tpiepho@kymetacorp.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The QEMU configure script incorrectly assumes SSP is supported by the
toolchain in some cases where the compiler accepts -fstack-protector*
flags but the C library does not provide the necessary __stack_chk_*()
functions.
Even though a full compile and link test is performed by the script,
this is done with a code fragment which does not actually meet any of
the conditions required to cause the compiler to emit canary code when
the -fstack-protector-strong variant is used. As no compile or link
failure occurs in this case, a false positive is generated and a
subsequent error is seen when the probe for pthreads is performed.
The fix consists in patching the configure script to use a more
appropriate test program for the SSP support checks.
Fixes:
http://autobuild.buildroot.net/results/efb/efbb4e940543894b8745bb405478a096c90a5ae2/http://autobuild.buildroot.net/results/32d/32d6d984febad2dee1f0d31c5fa0aea823297096/http://autobuild.buildroot.net/results/aa6/aa6e71c957fb6f07e7bded35a8e47be4dadd042c/
...and many others.
Signed-off-by: Rodrigo Rebello <rprebello@gmail.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
In all other package the condition for BR2_PACKAGE_LIBGUDEV selection is
BR2_INIT_SYSTEMD. Do the same for udisks.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Make the case of adding libgudev as a dependency more consistent with all
other dependency packages. The <pkg>_DEPENDENCIES variable sets the build
order which is conceptually independent of the underlying reason for that
dependency.
Also, remove the comments in hidapi.mk and udisks.mk. These comments explain
why BR2_PACKAGE_LIBGUDEV is selected, so the comments in corresponding
Config.in files are enough.
Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Packages built with automake use a `py-compile` helper to byte-compile
Python source files. This script uses the "py_compile" module from the
standard library. In turn, the compile() function in the "py_compile"
module invokes the cache_from_source() function provided by importlib.
This commit adds a new patch named "020-importlib-no-pep3147.patch"
that changes cache_from_source() and source_from_cache() in importlib
to get rid of the "__pycache__" directory.
This commit fixes the following import error in kmod when the module
is built for Python 3:
>>> from kmod import Kmod
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
ImportError: cannot import name 'Kmod'
Moreover, this commit removes two patches that are no longer necessary
since modifying cache_from_source() and source_from_cache() disables
PEP 3147 for the standard library and distutils / setuptools.
* 004-old-stdlib-cache.patch
* 016-distutils-no-pep3147.patch
Signed-off-by: Christophe Vu-Brugier <cvubrugier@fastmail.fm>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
/etc/inputrc is configuration file for readline. However, until now, it
was provided by skeleton. This patch install /etc/inputrc from readline
recipe.
Signed-off-by: Jérôme Pouiller <jezz@sysmic.org>
Acked-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Instead of using a patch to specify the architecture variant for
sparc, use some Boost configure variables.
[Thomas: reword commit log.]
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
[Thomas:
- use BR2_INIT_SYSTEMD instead of BR2_PACKAGE_SYSTEMD for the
libgudev handling. This is consistent with what we do in libmbim,
modem-manager and network-manager. Only udisks is diverging from
this by using BR2_PACKAGE_SYSTEMD.
- Fix the license info: it is GPLv3 *or* BSD-3c *or* the specific
HIDAPI license. Added LICENSE-orig.txt to LICENSE_FILES.]
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
They're unable to build liquid-dsp successfully resulting in an internal
compiler error. Fixes:
http://autobuild.buildroot.net/results/2ac/2acd745f4e5ec6edfd09afcd57b9ece216724c1b/
[Thomas: propagate the dependencies to the comment.]
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Currently Qt 5.5 only detects and build the eglfs_mali device
integration if the commercial Mali driver package from ARM is used.
This patch makes sure the Qt configure script also test for the
sunxi-mali driver package. It also removes the dependency of
the proprietary fbdev_window.h.
This issue is set to be fixed in upcoming Qt 5.6:
https://codereview.qt-project.org/#/c/125837/
[Thomas: renumber patch from 0010 to 0009.]
Signed-off-by: Daniel Nyström <daniel.nystrom@timeterminal.se>
Tested-by: Ariel D'Alessandro <ariel@vanguardiasur.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
[Thomas: build tested before the patch, verified that indeed the
eglfs-mali plugin doesn't get built, and that after the patch it gets
built as expected.]
This commit adds a patch to pulseaudio (already submitted upstream)
that avoids a structure redefinition build failure that occurs with
old gcc versions.
Fixes:
http://autobuild.buildroot.org/results/a1e/a1e7e59f9c65a6ce38800439c78b7808048ad708/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Trying to build anything at patch time will result in a broken
legal-info, as the needed host dependencies are not yet built.
Make that hook a pre-configure hook rather than a post-extract
hook.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Peter Korsgaard <jacmet@uclibc.org>
Cc: Romain Naour <romain.naour@openwide.fr>
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
To build libcurl, we need to override LD_LIBRARY and force it to a sane
value, otherwise libcurl is confused when target == host (see a51ce319,
libcurl: fix configure with openssl when target == host).
That is currently OK, since we always set LD_LIBRARY_PATH to a non-empty
value.
However, we're soon to stop setting it at all.
So, if the user has an empty (or no) LD_LIBRARY_PATH in his envirnment,
we'd end up adding the current working directory to LD_LIBRARY_PATH (as
an empty entry in a colon-separated list is most probably interpreted as
meaning the current working directory, which we do know can cause issue,
and which we expressely check against in support/dependencies/dependencies.sh
Fix that by only using an existing LD_LIBRARY_PATH if it is not empty.
Also use a Makefile construct as it is easier to read than a shell one
(we can do that, as all variables from the environment are available as
make variables).
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Peter Korsgaard <jacmet@uclibc.org>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
It does not build, and no one depends on it.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Francois Perrad <fperrad@gmail.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
To build the host-mysql, we only build parts of the source, just the
strictly minimum required to then cross-compile it.
However, the host variables (conf opts, build and install cmds) are only
defined when the mysql server is enabled in the configuration.
So, this breaks:
make defconfig; make host-mysql
Even though it is not much use to have that partial host-mysql on its
own, it is still very interesting to be able to build it, if at least
for testing changes in the core package infrastructures (like new step
hooks or the likes...)
Move the definitions of the host variant out of the server conditional
block.
[Peter: add comment about what we build and why as suggested by Arnout]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Marcelo Gutiérrez(UTN/FRH) <kuyurix@gmail.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
mkfs.axfs uses zlib, but does not have an rpath to our host dir.
That's because:
- we're not passing our host CFLAGS or LDFLAGS
- it is forcibly setting CFLAGS in the Makefile, overriding anything
specified by the user
- it is not using LDFLAGS at all
Add two patches so that CFLAGS and LDFLAGS from the environment are
used if present.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Tested-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Building with -mtune=e6500 led to build failures in glibc (probably in
uclibc as well) because gcc was built for a 32-bit target even though
the target tuple is powerpc64-*. This lead to a mix of 32-bit and
64-bit support and build errors like:
fatal error: gnu/lib-names-32.h: No such file or directory
The root cause is that the configure script is not handling e6500
correctly, because of stupid typo in the condition.
Change has been submitted upstream.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Alvaro Gamez <alvaro.gamez@hazent.com>
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
ARC is supported by uClibc-ng, cleanup old code.
Signed-off-by: Waldemar Brodkorb <wbx@uclibc-ng.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes a fanotify include bug where old kernels (2.6.x) fail to build.
Fixes recent failures on x86 with Linux 4.3.x kernel regarding
network sockets. Includes the removed patch.
Signed-off-by: Waldemar Brodkorb <wbx@uclibc-ng.org>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Force to use internal libatomic_ops of gauche, since the external
libatomic_ops packaged in Buildroot is not compatible with gauche.
Fixes:
http://autobuild.buildroot.org/results/a97/a979f6a58f0fbb1896830461d6c384f4a65c9429/
[Thomas: tweak commit log.]
Signed-off-by: Hiroshi Kawashima <kei-k@ca2.so-net.ne.jp>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This patch is ported from bdwgc package. Since Gauche uses an internal
copy of the boehm gc code, it is affected by the same problem.
Both configure and configure.ac are modified because autoreconf fails
due to an incompatibility with the version of the autotools used by
Buildroot.
Fixes:
http://autobuild.buildroot.org/results/8c6/8c6ce526735a36b8c8b4c6047e3c69039c4527cb/http://autobuild.buildroot.org/results/894/89460829a737a4dab19103f7a46905636420b0a8/
[Thomas: adjust commit log and patch description.]
Signed-off-by: Hiroshi Kawashima <kei-k@ca2.so-net.ne.jp>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The boost context module does not build on sparc, even if sparc is
advertised as a supported architecture by the boost documentation. So
let's disallow this module for the time being.
Fixes:
http://autobuild.buildroot.org/results/206/2060e6e31c4d739947097faf6587e0a06681fee2/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Fixes:
CVE-2015-8023 - authentication bypass vulnerability in the eap-mschapv2
plugin that was caused by insufficient verification of the internal
state when handling EAP-MSCHAPv2 Success messages received by the
client.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
libcap neds headers >= 3.0, so propagate that down to fastd which
selects libcap.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Alexander Dahl <post@lespocky.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Changelog (see [1] for details):
- Added support when active framebuffer is not starting at offset 0
- Fixed broken re-ordering of colors, rgba to argb
- Fixed broken path for man page
[1] http://fbgrab.monells.se
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Adjust pulseview/libsigrok for the gcc >= 4.8 dependency.
Line wrap comment depends for pulseview.
Drop redundant libsigrok depend on comment for libsigrok++.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Changelog (see [1] for details):
Only a small fix, the suggested hwdb entry that the touchpad-edge-detector
printed used the wrong codes for ABS_X and ABS_Y. Fixed with this release,
the output is now suitable for copy/paste into the udev hwdb.
[1] http://lists.freedesktop.org/archives/input-tools/2015-November/001227.html
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes:
png_set_PLTE/png_get_PLTE functions failed to check for
an out-of-range palette when reading or writing PNG files with a bit_depth
less than 8.
CVE not yet assigned.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The configure script uses pkg-config to find the optional dependencies, so
make sure it is present:
checking pkg-config is at least version 0.9.0... yes
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
With SD, USB and PCI-E variants available for several chips,
bus-prefixed entries get mis-aligned badly.
Rearrange them to use bus-suffixes instead.
Config options and the files themselves remain bus-prefixed.
Signed-off-by: Alex Suykov <alex.suykov@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Firmware blobs for Marvell WiFi-Ex 8797, 8887, 8897 were already
there in linux-firmware tree but relevant options were not in Config.in.
Signed-off-by: Alex Suykov <alex.suykov@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes a potential heap corruption on Windows when
mbedtls_x509_crt_parse_path() is passed a path longer than 2GB. This
cannot be triggered remotely. Found by Guido Vranken, Intelworks.
Fixes a potential buffer overflow in some asn1_write_xxx() functions.
This cannot be triggered remotely unless you create X.509 certificates
based on untrusted input or write keys of untrusted origin. Found by
Guido Vranken, Intelworks.
The X509 max_pathlen constraint was not enforced on intermediate
certificates. Found by Nicholas Wilson, and fix and tests provided by
Janos Follath.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This was causing unexpected HTTP requests by the setup.py script for
python-protobuf. These "dependencies" aren't actually required for a successful
build, and are not staged into the target install directory.
Fixes:
http://autobuild.buildroot.org/results/e44/e44694f30e39266491a1040e284e504d6d37ef9e/
Signed-off-by: Steven Noonan <steven@uplinklabs.net>
Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
ffmpeg was already marked as not available for the NIOS2 Sourcery
toolchains, but it could still be built with the internal toolchain
backend or a custom external toolchain.
However, an inspection of the latest glibc source code indicates that
FE_INVALID, FE_OVERFLOW and FE_UNDERFLOW and indeed not available in
the nios2 variant of <fenv.h>.
Consequently, this patch makes ffmpeg not available on nios2, which
allows to simplify a bit the dependencies.
It propagates this dependency to:
- minidlna (and at the same time makes sure the minidlna comment is
not displayed on nios2, which wasn't properly taken into account
until now)
- mpd
- opencv
- opencv3
- squeezelite
- tovid
Even if it selects ffmpeg, Kodi does not need an update since Kodi is
only available on a limited number of architectures (which don't
include nios2, obviously). Other packages only make use of ffmpeg when
available.
Fixes:
http://autobuild.buildroot.org/results/921/9212f5a6432c5e695ac0630695405cea05e28610/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Drop the patches since this is a major bump and they're no longer
relevant.
Introduce the new RRDTOOL_RRDGRAPH option, in previous versions the
graphing functionality was bundled with the stats/database backend,
however for newer versions this can be disabled.
Since it's normally expected functionality the default is yes, however
all of the buildroot rrd-using applications can work without this.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Bump trinity version to 1.6. Remove backported vt.c missing include patch
as it is now included in the upstream release.
Signed-off-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
[Thomas:
- add comment in Config.in to explain why this package depends on
glibc, and specifically why it doesn't build for uClibc and musl.
- add hash file.
- add 'COPYING' to the list of LICENSE_FILES
- add themissing <pkg>_DEPENDENCIES variable in the .mk file, to make
sure dependencies are built before netsniff-ng.
- add comment in the .mk file explain why we're using the
generic-package infrastructure even if the build procedure is
essentially ./configure, make, make install.
- fix indentation in NETSNIFF_NG_CONFIGURE_CMDS.]
Signed-off-by: Joris Lijssens <joris.lijssens@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
When using systemd, the policy in Buildroot is to use a merged /usr
(see c5bd8af6, "system: add options for /bin /sbin and /lib to be
symlinks into /usr" for more info). So, we apply a few tricks in some
packages to account for the merged /usr case.
However, when using a custom skeleton, we have no say in how that
skeleton is organised, so it may well have a split /usr. In that case,
our little tricks might not work as expected.
So, when the user uses a custom skeleton and wants systemd as an init
system, we must check that the custom skeleton is setup with a merged
/usr.
We do that by checking that each pair of {/lib,/usr/lib} {/bin,/usr/bin}
and {/sbin,/usr/sbin} have the same inode numbers, i.e. /lib must have
the same inode number as /usr/lib (and so on...). When a pair does not
share the same inode number, this is not a merged /usr and we abort.
We implement that check with make constructs, so it is done very early
in the build process, and we can abort early if need be.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Cc: Peter Korsgaard <jacmet@uclibc.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
In case the user wants to use a custom skeleton, we should not try to
handle the symlinks (resp. mkdir) to handle merged (resp. split) /usr.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Cc: Peter Korsgaard <jacmet@uclibc.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When a powerpc target is detected the build system tries to force
altivec support but that's wrong:
Not every powerpc processor has altivec instruction support.
It's using the removed/deprecated -faltivec compiler directive when it
should be using -maltivec instead.
Even if the above conditions are corrected the altivec codepath fails
to build properly.
So just disable altivec in general using the standard
--enable-simdoverride configure option. Fixes:
http://autobuild.buildroot.net/results/e1b/e1b6cf76a7a4ae754a928d211e779e9381198f48/
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The license is wrong, it's GPLv3+ rather than GPLv2+ now.
Also there's a specific clause for libintl that makes it LGPLv2.1+, see
gettext-runtime/COPYING.
Signed-off-by: Gustavo Zacarias <gustavo.zacarias@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The SSP support checks of the Ruby configure script do not do a full
link, and therefore don't properly detect if SSP support is fully
functional or not, which causes some build failures if the compiler
supports -fstack-protector but the C library does not provide the SSP
implementation.
To avoid this, we explicitly tell ruby if SSP support is available or
not.
Fixes:
http://autobuild.buildroot.net/results/001/00154b0142f51c197d5d53ae36f7adc4f8250d9d/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
In commit 53bf889cdc ("ruby: should not
use dl_iterate_phdr() on Blackfin") we added
ac_cv_func_dl_iterate_phdr=no to RUBY_CONF_ENV when building on the
Blackfin architecture.
However, while doing so, we overwrote the previous value of
RUBY_CONF_ENV, which was defined to have custom CFLAGS:
RUBY_CONF_ENV = CFLAGS="$(RUBY_CFLAGS)"
This has no visible effect because the custom CFLAGS are only used on
SuperH/SuperH 64, while the overwrite of RUBY_CONF_ENV takes place on
Blackfin.
However, it doesn't look very pretty, and future additions to
RUBY_CONF_ENV may break this subtle behavior.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Similar to how we do it for dhcpd.service. The file should really have been
named /etc/default/dhcpd instead, but changing it now would break existing
systemd setups so leave it as it is.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
