gst1-plugins-base: add missing indirect dependency on atomics for pango
Also removed the redundant indirect dependencies on wchar, threads, mmu
(gstreamer depends on libglib2 so this indirect dependency is obvious).
[Peter: drop wchar+threads from comment as suggested by Jerzy Grzegorek] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
gst-plugins-base: add missing indirect dependency on atomics for pango
Also removed the redundant indirect dependencies on wchar, threads, mmu
(gstreamer depends on libglib2 so this indirect dependency is obvious).
[Peter: drop wchar+threads from comment as suggested by Jerzy Grzegorek] Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Yann E. MORIN [Sat, 7 Feb 2015 18:28:21 +0000 (19:28 +0100)]
br2-external: make it available to post-{build, image} scripts
Currently, BR2_EXTERNAL is not always exported to sub-processes that we
spawn, like post-build or post-image scripts. This all depends on how
the user passes BR2_EXTERNAL; consider the following:
- make BR2_EXTERNAL=/path/to/br2-ext
- BR2_EXTERNAL=/path/to/br2-ext make
In the first case, it is just a make variable, not an environment
variable, and thus not exported, while in the second case it is an
environment variable and gets exported to all sub-processes make may
spawn.
Explicitly export it using EXTRA_ENV.
Reported-by: Julian Scheel <julian@jusst.de> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Cc: Julian Scheel <julian@jusst.de> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Mono 3.12.0 doesn't build properly on PowerPC, but there are already
patches upstream to fix those issues. This commit backports the two
necessary patches to get the PowerPC build working.
Since the bump to version 3.12 of Mono, it fails to build on MIPS. The
patch added by this commit fixes this build failure, simply caused by
a missing include.
qt5/qt5base: Fix build on MIPS without DSP support
The qt_memfill32_asm_mips_dsp function is only declared if
QT_COMPILER_SUPPORTS_MIPS_DSP is defined, so we can't reference it
unless the same macro is defined.
Backporting an upstream patch to fix this issue.
This is the error message:
qt-everywhere-enterprise-src-5.4.0-alpha/qtbase/src/gui/painting/qdrawhelper.cpp:
In function 'void qInitDrawhelperAsm()':
qt-everywhere-enterprise-src-5.4.0-alpha/qtbase/src/gui/painting/qdrawhelper.cpp:6843:
error: 'qt_memfill32_asm_mips_dsp' was not declared in this scope
Peter Korsgaard [Fri, 6 Feb 2015 12:02:23 +0000 (13:02 +0100)]
qemu: correct kernel headers check when major numbers differ
commit f7add51c39 (qemu: add host/target Linux version check) added a
version check between the host kernel version and the version of kernel
headers used by the toolchain, but the logic would fail unless BOTH major
and minor versions were >=, which isn't true for E.G. host kernel = 3.2 and
toolchain 2.6.x.
Instead calculate a single version number (as major << 8 + minor) and
compare that.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The host-qemu package currently builds the user mode emulation, which
isn't available on all architectures, so this commit adds the
necessary architecture dependencies to the Config.in.host file.
Reported-by: Max Filippov <jcmvbkbc@gmail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Using --with-zlib=$(STAGING_DIR) causes the configure script to basically do
a find $(STAGING_DIR) -name zlib.h to figure out the correct include path
for zlib. That is normally ok, but if botan is enabled and built before
rpm, then it installs a zlib.h header in
$(STAGING_DIR)/usr/include/botan-1.10/botan which confuses this logic and
causes configure to fail.
Fix it by using --with-zlib=external instead which causes the configure
script to use pkg-config to get the CFLAGS/LDFLAGS. Also do it for popt so
we don't end up with a similar issue in the future. We unfortunately cannot
do it for beecrypt as it doesn't install a .pc file.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
libvips requires C++ support, otherwise it will fail with an error
message like this one:
error: ../libvips/.libs/libvips.so: No such file or directory
The '--disable-cxx' configure option seems to be broken, so in the
meantime a compiler with C++ support is required in order to build
libvips successfully.
If --zmq isn't passed then python-pymq falls back to using 'pkg-config
--variable=prefix libzmq' which returns /usr, causing it to look in /usr/lib
for libraries breaking the build.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Peter Korsgaard [Tue, 3 Feb 2015 14:58:07 +0000 (15:58 +0100)]
iprutils: drop unneeded ncurses5-config handling
ncurses nowadays installs the headers directly into $prefix/usr/include, so
ncurses5-config doesn't return anything important and the logic wasn't
taking the wchar variant into consideration - So just drop it.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Gustavo Zacarias [Thu, 27 Nov 2014 22:41:49 +0000 (19:41 -0300)]
skeleton: make /run a proper directory/filesystem
Making /var/run and /run symlinks to /tmp is bad since the underlying
tmpfs filesystem is mode 1777 which leads to possible security attack
vectors via badly owned/mask-mode pidfiles and state files residing there.
So make /run a proper directory with /var/run symlinked to it.
Eventually all startup scripts and state info should be pointed to /run
directly as per the linux fhs and good practice.
Add a tmpfs filesystem entry for /run so that busybox inittab, systemv
inittab and systemd automount mounts it there to avoid breaking the
system.
While at it fix inconsistent spacing in /etc/fstab by using tabs and
drop the "static file system information" header whatever that means.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since the tarball we download is generated from GitHub, there's no
saying that the tarball won't differ in a month from now.
So the hash has no value there.
An update to the manual should be done.
Signed-off-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Samuel Martin [Fri, 16 Jan 2015 17:58:35 +0000 (18:58 +0100)]
bootlader/grub2: disable x86-64-efi choice on !BR2_ARCH_IS_64
When x86-64-efi platform is selected, grub2 automatically adds -m64 to the
CFLAGS. This makes the configure script failed when the toolchain does not
have multilib support (like the Buildroot ones).
Reported-by: Noe Rubinstein <nrubinstein@aldebaran-robotics.com> Signed-off-by: Samuel Martin <s.martin49@gmail.com> Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
gnutls links to a number of libraries depending on configuration (pthread,
iconv, ..), so we need to link against those as well when statically
linking.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Johan Oudinet [Mon, 12 Jan 2015 09:32:20 +0000 (10:32 +0100)]
ejabberd: new package.
[Thomas:
- Adjust the comment about the dependency on erlang
- Fix license to be 'GPLv2+ with OpenSSL exception' and not just 'GPLv2+'
- Use double quotes instead of simple quotes in the .mk file.
- Don't use the EJABBERD_MAKE_ENV variable, since it's not defined
anywhere.
- Remove the 0007-fix-init.patch patch, since we're not using the
init script provided by ejabberd, and rename
0008-fix-install-permissions.patch to
0007-fix-install-permissions.patch.]
Signed-off-by: Johan Oudinet <johan.oudinet@gmail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Mon, 12 Jan 2015 09:32:07 +0000 (10:32 +0100)]
docs/manual: add documentation for the pkg-rebar infrastructure
Signed-off-by: Johan Oudinet <johan.oudinet@gmail.com>
[yann.morin.1998@free.fr: split the doc into its own patch] Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Johan Oudinet [Mon, 12 Jan 2015 09:32:06 +0000 (10:32 +0100)]
package/pkg-rebar: new infrastructure
Ease the development of packages that use the erlang rebar tool as
their build system.
Signed-off-by: Johan Oudinet <johan.oudinet@gmail.com>
[yann.morin.1998@free.fr: split the patch into semantically separated
patches; large rewrites of the rest] Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
[Thomas, with help from Yann and Arnout:
- Fix the comment about the symlink used to make sure rebar does not
download dependencies. The comment was not up-to-date with where
the symlink is actually created.
- Make <pkg>_USE_BUNDLED_REBAR and <pkg>_USE_AUTOCONF be inherited by
host packages from their corresponding target package.
- Make sure host dependencies are inherited from the corresponding
target packages dependencies. This requires copying some logic from
inner-autotools-package and inner-generic-package, just like
inner-autotools-package duplicates some logic from
inner-generic-package.
- Fix host variant of $(2)_BUILD_CMDS indentation, use double quotes
instead of simple quotes. So that it matches the target
$(2)_BUILD_CMDS, and what we do elsewhere in Buildroot.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Yann E. MORIN [Mon, 12 Jan 2015 09:32:05 +0000 (10:32 +0100)]
package/erlang: export EI_VSN so other packages can use it
Packages that depend on Erlang may need the Erlang Interface Version
(aka EI_VSN), so just export it.
We do export a static value instead of looking in the file, to avoid
spawning a shell every time the variable is dereferenced.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[johan.oudinet@gmail.com: update EI_VSN since the erlang update] Signed-off-by: Johan Oudinet <johan.oudinet@gmail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>