busybox: add another upstream patch to fix (g)unzip
0002-unzip.patch was added in 69516e0 to fix a segmentation fault in the
gunzip applet. However, it introduced a new issue that made the unzipping
of some files fail.
Add an upstream patch that fixes this new issue.
Fixes #8501.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Cc: Jason Rush <rush0033@hotmail.com> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Yann E. MORIN [Tue, 1 Dec 2015 22:19:07 +0000 (23:19 +0100)]
package/gawk: do not use readeline ormpfr in host variant
gawk has an optional dependency on mpfr (and thus gmp) and readline,
and will probe for them. If they are present, they are used; if they
are missing, that's not an error. mpfr (and gmp) is used for "BIGNUM"
support on gawk; readline is used by the gawk debugger.
However, mpfr (bringing gmp) are also host-packages in Buildroot, but in
the standard build order (i.e. a plain 'make'), they are built after
gawk. Ditto readline (from ncurses).
If the user has the development files for gmp and mpfr, then gawk is
linked to them. Ditto readline.
Now, further on in the build, we build gmp and mpfr (for gcci or guile),
so we install them in the host dir. Ditto readline (for gdb, ncurses
itself and a few other packages...)
But because we forcibly set an RPATH tag on all our host binaries, our
host gawk will now dynamically link with our versions, when it was in
fact built against the host ones.
This did not seem to cause any harm so far, but is far from ideal.
Since we do not really need BIGNUM or the debugger in our host gawk, we
just forcibly disable them and configure gawk without readline or mpfr
(there's no switch for gmp, but it's not a direct dependency, it comes
just with mpfr).
[Adjust comment as suggested by Thomas/Yann] 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> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Yann E. MORIN [Tue, 1 Dec 2015 22:19:06 +0000 (23:19 +0100)]
support/check-host-rpath: remove trailing '/' in host dir
Despite the comment saying so, the trailing '/' in the host directory is
not removed. Note however that it is properly removed from extracted
RPATH tags.
This is not visible when the host directory is our default $(O)/host
location, but breaks for user-supplied external host directory, when
the user leaves a trailing slash in the path.
Fix that.
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> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Gustavo Zacarias [Mon, 30 Nov 2015 13:16:35 +0000 (10:16 -0300)]
gnutls: bump to version 3.4.7
The 3.4 series has been promoted to stable.
Handle autodeps more concisely (idn, p11-kit, zlib).
libtasn1 is now mandatory, since otherwise gnutls uses the bundled
version it makes no sense to try that because of target duplication.
Disable tpm support since we've got no trousers package.
Disable libdane support since we've got no dane package.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
- Bump version to 2.4.43
- Use md5 and sha1 upstream provided hashes. This combination is
stronger than a sha256 hash.
- Switch to an http download site to avoid problems with firewalls
blocking the ftp.
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Doug Kehn [Sun, 24 May 2015 19:12:32 +0000 (14:12 -0500)]
package/nss-pam-ldapd: new package
Name Service Switch (NSS) module that allows your LDAP server to provide
user account, group, host name, alias, netgroup, and basically any other
information that you would normally get from /etc flat files or NIS. It
also provides a Pluggable Authentication Module (PAM) to do
authentication to an LDAP server.
[Thomas:
- bump to version 0.9.6
- use --disable-<foo> rather than --enable-<foo>=no.
- fix license information: there is no LICENSE file, it is named
COPYING.]
Signed-off-by: Doug Kehn <rdkehn@yahoo.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Ryan Barnett [Sun, 5 Apr 2015 22:10:11 +0000 (17:10 -0500)]
miraclecast: new package
[Thomas:
- add hash file
- fix license information: use comma as a separator in
<pkg>_LICENSE, and htable is under LGPLv2.1+ and not GPLv2.1+
(which doesn't exist)]
Signed-off-by: Ryan Barnett <ryanbarnett3@gmail.com> CC: Baruch Siach <baruch@tkos.co.il> CC: Yann E. MORIN <yann.morin.1998@free.fr> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Rodrigo Rebello [Mon, 30 Nov 2015 15:40:39 +0000 (13:40 -0200)]
conntrack-tools: work around build issue with musl
Building conntrack-tools with kernel headers >= 4.2 + musl fails due to
a well-known symbol clash that occurs when userspace and kernel headers
are included simultaneously (see [1], question 7, for details).
In the case of conntrack-tools, the inclusion of both 'netinet/in.h' and
'linux/in.h' occurs inside the C helper files (src/helpers/*.c)
indirectly via e.g. 'libnetfilter_conntrack/libnetfilter_conntrack.h',
which itself includes 'netinet/in.h', and 'linux/netfilter.h', which
includes 'linux/in.h' in kernel headers >= 4.2.
The approach to solving this type of conflict with musl usually involves
removing the inclusion of kernel headers or refactoring the code so as
to avoid the mentioned simultaneous inclusion. This is unfortunately
non-trivial in the case of conntrack-tools since the clashing headers
get included indirectly by headers that are strictly necessary (because
of definitions used in some helper callbacks).
Work around the issue by defining __GLIBC__ when musl is used. This
eliminates the conflicts as the kernel headers avoid redefining certain
symbols when they see __GLIBC__ defined (linux/libc-compat.h). Note that
other glibc-compatible libraries, like uClibc, already do that
internally.
Compile tested for sparc/glibc and sparc64/glibc.
Fixes following autobuilder problem:
http://autobuild.buildroot.net/results/1d8e0c22594aef53f7a4de3491454aa78311c815/
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
valgrind: add a patch to fix MIPS R6 by replacing ADDI with ADDIU
The ADDI instruction has been removed for MIPS R6, so valgrind fails to
build for this revision level because it uses that assembly
instruction. Using ADDIU instead fixes the problem.
Peter Korsgaard [Mon, 30 Nov 2015 07:19:59 +0000 (08:19 +0100)]
libglib2: force arm mode when building with thumb1
Fixes #8491
glib/valgrind.h contains inline asm not compatible with thumb1, so build it
in arm mode:
CC libglib_2_0_la-gstringchunk.lo
/tmp/ccSVEXME.s: Assembler messages:
/tmp/ccSVEXME.s:587: Error: shifts in CMP/MOV instructions are only supported
in unified syntax -- `mov r12,r12,ror#3'
/tmp/ccSVEXME.s:587: Error: shifts in CMP/MOV instructions are only supported
in unified syntax -- `mov r12,r12,ror#13'
/tmp/ccSVEXME.s:588: Error: shifts in CMP/MOV instructions are only supported
in unified syntax -- `mov r12,r12,ror#29'
/tmp/ccSVEXME.s:588: Error: shifts in CMP/MOV instructions are only supported
in unified syntax -- `mov r12,r12,ror#19'
/tmp/ccSVEXME.s:589: Error: lo register required -- `orr r10,r10,r10'
Makefile:2087: recipe for target 'libglib_2_0_la-gslice.lo' failed
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Yann E. MORIN [Wed, 4 Nov 2015 21:42:39 +0000 (22:42 +0100)]
core/printvars: allow dumping a set of variables
Dumping our 176164 variables can take quite some time (~12s here). What
takes the most time is sorting the variables (~9s), followed by the
parsing of our Makefiles (~3s), with the actual printing in the noise.
However, sometimes only one or a few variables are needed. For example,
one may want to retrieve the Linux build dir from a post-build hook (to
get the Linux' actual .config after our fixups and check for various
features).
Add the possibility to only dump the variables listed in $(VAR) which
must be passed as a make argument, like so:
This is much faster (the time is just about the time it takes to parse
our Makefiles, 3s here) and easier to parse.
[Thomas: improve comment above the printvars target.]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Vivien Didelot [Sat, 7 Nov 2015 04:39:24 +0000 (23:39 -0500)]
support/download: fetch all refs on full git clone
When specifying BR2_LINUX_KERNEL_CUSTOM_REPO_VERSION, a user may want to
specify the SHA of a reference different than a branch or tag.
For instance, Gerrit stores the patchsets under refs/changes/xx/xxx, and
Github stores the pull requests under refs/pull/xxx/head.
When cloning a repository with --bare, you don't fetch these references.
This patch uses --mirror for a full clone, in order to give the user
access to all references of the Git repository.
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com> Reviewed-by: "Maxime Hadjinlian" <maxime.hadjinlian@gmail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Erico Nunes [Sat, 21 Nov 2015 02:14:29 +0000 (00:14 -0200)]
spi-tools: new package
This package contains some simple command line tools to help using Linux
spidev devices.
https://github.com/cpb-/spi-tools
The included version is based in a sha1 and not a tagged version as the
current master branch contains autotools support for build but there is
no tagged version with autotools support yet.
The only other quirk with the recipe is related to the VERSION define
which is normally obtained in build time via git for this package, but
that would fail within Buildroot as the package build is normally not
performed inside a git repository.
To work around that, I have added a hook which will substitute change
the VERSION to match that of the Buildroot package.
I have tested this package in an ARM target with a custom SPI device.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Gustavo Zacarias [Fri, 20 Nov 2015 13:34:15 +0000 (10:34 -0300)]
squashfs: always enable xattr support
In 785fef10 xattr support was disabled, probably because the old version
did require the xattr package to be present to support it.
In a409d76e xattr support was made optional via (lib)attr, however the
squashfs package does in fact include all of the required bits to
support xattr without the need for the attr package.
So enable xattr support by default for both the host and target
variants.
Signed-off-by: Gustavo Zacarias <gustavo.zacarias@free-electrons.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
- Bump version to 3.1.16
- Update the hash value
- Remove upstreamed patches:
0001-configure.ac-remove-manual-compiler-check-with-AC_TR.patch
0003-Makefile.in-add-LDFLAGS-to-linking-stage.patch
- Tweak patches for the new version:
0004-Makefile.in-replace-IROOT-by-DESTDIR.patch
0005-Makefile.in-make-install-fix.patch
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Jörg Krause [Mon, 23 Nov 2015 22:23:21 +0000 (23:23 +0100)]
package/zmqpp: fix static build issues
The Makefile for zmqpp builds both shared and static libraries and the client
binary. This leads to several build issues in a pure static library context:
* R_ARM_TLS_LE32 relocation not permitted in shared object
* relocation R_ARC_32_ME against `_ZSt7nothrow' can not be used when making
a shared object; recompile with -fPIC
We add a minimal patch to add some basic handling of building a shared or a
static library only.
Additionally, disable the client for static only builds as it depends on
building the shared library in zmqpp's Makefile. As there is already version
4.1.2 available which provides a CMake build file which solves this issue
(probably, not tested) we don't care for now.
Gustavo Zacarias [Sun, 29 Nov 2015 11:39:50 +0000 (08:39 -0300)]
yad: needs gtk3-x11
Doesn't build with broadway or wayland since it wants the X-variant of
gdk. Fixes:
http://autobuild.buildroot.net/results/91d/91dc6ab749f967ae046f856f3f592d5a548499c6/
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Thu, 26 Nov 2015 22:13:41 +0000 (23:13 +0100)]
package/skeleton: also handle merged /usr in the staging
When the target uses a merged /usr setup, gdbserver will only report
paths in /lib to the remote gdb, which in turn will only look for
libraries in staging/lib and never in staging/usr/lib.
So. the merged (or non-merged) /usr setup must be replicated in the
staging.
The best solution where to do so is in the skeleton package, since it
is guaranteed to come before any package that installs things in the
staging, and even before the (internal or external) toolchain as well.
Reported-by: Pieterjan Camerlynck <pieterjan.camerlynck@gmail.com> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Cc: Pieterjan Camerlynck <pieterjan.camerlynck@gmail.com> Cc: Maxime Hadjinlian <maxime.hadjinlian@gmail.com> 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>
Yann E. MORIN [Thu, 26 Nov 2015 22:13:40 +0000 (23:13 +0100)]
package/skeleton: parameterise the symlinks-or-mkdir macro
currently, we only make the merged symlinks or non-merged mkdirs in the
target directory. This is fine, as long as one does not run gdb.
However, as soon as one wants to run gdbserver on the target,
gdb will only search for libraries in /lib in the sysroot.
In preparation for a patch to fix that, make the symlink-or-mkdir macro
a function that takes as parameter the base directory where to create
symlinks or mkdir in.
Move the function declarations out of the custom/non-custom skeleton
conditional block, so they are always available.
Reported-by: Pieterjan Camerlynck <pieterjan.camerlynck@gmail.com> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Cc: Pieterjan Camerlynck <pieterjan.camerlynck@gmail.com> Cc: Maxime Hadjinlian <maxime.hadjinlian@gmail.com> 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>
opencv: use the same logic as opencv3 for gstreamer support
The logic for enabling gstreamer support in opencv is not correct. We
select gstreamer-0.10 packages but then we enable gstreamer1 support in
the opencv.mk file. opencv3 has the correct logic, so let's use it in
opencv as well.
package/minidlna/S60minidlnad: add force-reload to rebuild DB
Add a force-reload operation that restarts minidlnad and makes it
rebuild its database. This is what Debian does, and this is useful when
media_dir is changed in the configuration file or when inotify can not
detect changes inside the media directories (e.g. in case of a mount).
James Knight [Fri, 27 Nov 2015 16:25:57 +0000 (11:25 -0500)]
gpsd: enable kpps support
GPSD has the ability to support KPPS. Its enablement is determined
primarily from the existence of `sys/timepps.h`. This file is provided
from the pps-tools package. Adjusting GPSD's Makefile to depend on the
pps-tools package if a developer has included pps-tools as part of a
build.
Signed-off-by: James Knight <james.knight@rockwellcollins.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Fri, 27 Nov 2015 22:20:03 +0000 (23:20 +0100)]
package/autossh: really fix the path to ssh
The ./configure script is not up-to-date with what is in configure.ac,
which prevents --with-ssh from working properly.
So, we autoreconfigure autossh, which fixes the issue.
Note however that this is not really an autotools package, since it only
uses autoconf and not automake. However, making it a generic pacakge is
not trivial, since autossh really wants config.h as generated by
./configure. Using the autotools infra in this case helps us easily
autoreconfiguring without to manual depend on all the autotools-relates
host packages.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Bernd Kuhls [Sat, 28 Nov 2015 11:52:19 +0000 (12:52 +0100)]
package/swig: bump version to 3.0.7
This patch bumps the swig package to the current major release, the
comment about the CMake fix was also updated to reflect the version
bump. Swig 3.0.7 was positively tested using this defconfig:
Yann E. MORIN [Fri, 27 Nov 2015 13:58:56 +0000 (14:58 +0100)]
package/gstreamer1: add symlink to gstconfig.h in legacy location
After version 1.4 (the previous one packaged in Buildroot), gstreamer
has changed the location where it installs its gstconfig.h, because it
is architecture-dependent, see upstream commit:
http://cgit.freedesktop.org/gstreamer/gstreamer/commit/?id=444ceb068424122ceba46d8cc48a53b8b32c3cf5
However, this causes build failures in some pcakges (at least FreeRDP
is known to break) because either they do not use pkg-config (bad!) or,
like FreeRDP, have a broken buildsystem and have not yet fixed their
build failures [0].
In Buildroot, we do not really care about architecture-dependent
locatios, because we only build for one architecture. So, the legacy
location of that header is not problematic to us.
Create a symbolic link to gstconfig.h in its legacy location.
[0] FreeRDP does have a commit supposedly fixing the build with "newer"
gstreamer versions:
https://github.com/FreeRDP/FreeRDP/commit/bea27fd919b64ee8d97996409e279e1e83d13594
but it makes the whole build completely fail, as it does not even
detect gstreamer anymore. Sigh... :-(
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This patch bumps to version 4.2.1.102 and updates
accompaining patches accordingly.
* Removed upstreamed patch 0003-fix-parallel-install.patch
* Adding patch to fix eglib linking option (upstreamed here
https://github.com/mono/mono/pull/1971 but not included in
this release)
* Adding patch to remove compilation of unit-tests
Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Is not yet supported and the compilation will fail like this:
[ 4%] Building C object library/CMakeFiles/polarssl.dir/bignum.c.o
/tmp/ccLDxl9G.s: Assembler messages:
/tmp/ccLDxl9G.s:92: Error: opcode not supported on this processor:
mips32r6 (mips32r6) `multu $13,$14'
/tmp/ccLDxl9G.s:93: Error: opcode not supported on this processor:
mips32r6 (mips32r6) `addi $10,$10,4'
/tmp/ccLDxl9G.s:94: Error: opcode not supported on this processor:
mips32r6 (mips32r6) `mflo $14'
/tmp/ccLDxl9G.s:95: Error: opcode not supported on this processor:
mips32r6 (mips32r6) `mfhi $9'
[...]
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Philippe Proulx [Thu, 26 Nov 2015 19:42:05 +0000 (14:42 -0500)]
lttng-modules: bump to version 2.7.0
[Peter: drop extra whitespace in .hash file] Signed-off-by: Philippe Proulx <eeppeliteloop@gmail.com> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Lada Trimasova [Thu, 26 Nov 2015 17:21:38 +0000 (20:21 +0300)]
gcc: fix incorrect handling of the comparison arguments
This patch fixes compiler error during libbroadvoice build.
The comparison arguments where not correctly handled.
The fix is done in development tree:
https://github.com/foss-for-synopsys-dwc-arc-processors/gcc/commit/b4035128ba8f8bbbf9527f54f261a87b304ca4c5
and will be a part of the next release of ARC GNU tools.
Once that new release happens this patch must be removed.
Gustavo Zacarias [Fri, 20 Nov 2015 13:06:51 +0000 (10:06 -0300)]
gstreamer: can optionally use libxml2
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.
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.
Alex Suykov [Mon, 23 Nov 2015 11:12:42 +0000 (13:12 +0200)]
collectd: fix musl builds (ipc module)
src/ipc.c uses shm_info.used_ids which musl only provides when
_GNU_SOURCE is defined.
The issue has been fixed upstream, but the fix is not in their
latest 5.5.0 release. Adding their patch as a temporary fix
to be removed on the next version bump.
Peter Korsgaard [Tue, 24 Nov 2015 21:21:49 +0000 (22:21 +0100)]
qt5base: support OVERRIDE_SRCDIR
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>
projects / coffee / buildroot.git / log