busybox: Make it possible to override LOGIN_FAIL_DELAY in CONFIG_EXTRA_CFLAGS

Signed-off-by: Peter Kümmel <syntheticpp@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

opkg: bump to 0.2.4 and fix build failure

This commit bumps opkg to 0.2.4, and adds a new patch to fix a build
failure apparently caused by the recent bump of the libtool version:

  http://autobuild.buildroot.org/results/5fc/5fc9fa24563213d1ad77e55ab52c8e59bf21055f/

It also refreshes the existing patch, turns it into a Git formatted
patch, and fixes a typo in its title.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

liblinear: requires C++ support

Fixes:

  http://autobuild.buildroot.org/results/1be/1be890ece519ebc0d0ff4f3f31abc06eb0b7356c/

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package: indentation cleanup

Signed-off-by: Jerzy Grzegorek <jerzy.grzegorek@trzebnica.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/mysql: do not force a static build for internal programs

mysql wants to statically link some of the programs it uses internally,
like conf_to_src (whatever it is used for) which links against -lz.

This does not work for a shared-only build, as we would be missing
libz.a in that case.

Fix that by not forcing static builds for internal programs.

Fixes a slew of autobuild failures:
    http://autobuild.buildroot.net/results/018/018a579f37c9c8eca5a36fc76e93508dcf9bba35/
    http://autobuild.buildroot.net/results/2cc/2cc481b0ae43713a6fc50a1ca720f91384892af7/
    http://autobuild.buildroot.net/results/d98/d985c5824f9012374d64e6085d6c8ec9b3cd932b/
    ...

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/mysql: rename patches to the new convention

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

pinentry: remove non-existent configure options for Qt3 and GTK+1

Qt3 and GTK+1 support were removed by this commit:
  http://git.gnupg.org/cgi-bin/gitweb.cgi?p=pinentry.git;a=commit;h=f2ed42297528f2a126d189c6181d3105ceecca22

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

rsyslog: install systemd services links

Signed-off-by: Thierry Bultel <tbultel@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

boot/uboot: bump to version 2014.10

Signed-off-by: Jörg Krause <jkrause@posteo.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

boot/uboot/uboot.mk: Strip quotes from BR2_TARGET_UBOOT_SPL_NAME

Remove the quotes from the BR2_TARGET_UBOOT_SPL_NAME variable.

Signed-off-by: Jörg Krause <jkrause@posteo.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

boot/uboot/Config.in: Fix BR2_TARGET_UBOOT_SPL_NAME

Point out that the SPL name is an image name.

Signed-off-by: Jörg Krause <jkrause@posteo.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

boot/uboot/Config.in: Fix BR2_TARGET_UBOOT_SPL

Building an SPL image depends on the board configuration. This option
does not enable the SPL build, but only copies the built SPL image to
the binary images folder. The current help text is misleading.

Signed-off-by: Jörg Krause <jkrause@posteo.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/guile: rename patch to new convention

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

samba: Swap libnss_win*.so* files and symlinks

The libnss_wins and libnss_winbind libraries were installed in this way:
/lib/libnss_<service>.so
/lib/libnss_<service>.so.<major> -> /lib/libnss_<service>.so

This had been done like this in order to get something similar to the
other NSS libraries, which are usually installed in this way:
/lib/libnss_<service>-<version>.so
/lib/libnss_<service>.so.<major> -> /lib/libnss_<service>-<version>.so

However, besides these files, these other NSS libraries usually come
installed with:
/usr/lib/libnss_<service>.so -> /lib/libnss_<service>.so.<major>

This means that the NSS libraries follow the usual library installation
practice, i.e. that the non-versioned .so is a symlink to the versioned
.so, so that switching versions is easy. In the case of the NSS
libraries, the versioned .so is just also a symlink to a .so with a more
accurate version.

Hence, follow the same rules for libnss_win*.so*, and install these
libraries the other way around:
/lib/libnss_<service>.so -> /lib/libnss_<service>.so.<major>
/lib/libnss_<service>.so.<major>

This is also how these libraries are installed by a major OS like Ubuntu
14.10:
/lib/x86_64-linux-gnu/libnss_winbind.so -> libnss_winbind.so.2
/lib/x86_64-linux-gnu/libnss_winbind.so.2
/lib/x86_64-linux-gnu/libnss_wins.so -> libnss_wins.so.2
/lib/x86_64-linux-gnu/libnss_wins.so.2

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau.dev@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

liblinear: new package

Signed-off-by: Romain Naour <romain.naour@openwide.fr>
Cc: "Yann E. MORIN" <yann.morin.1998@free.fr>
[yann.morin.1998@free.fr: add ability to build a static lib; handle
static-only and static+shared cases]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/pkg-cmake.mk: add ccache support for host packages

This ccache support for host-cmake-package uses the same logic as the
one done for the target packages in the toolchainfile.cmake file.

[Thomas: fix lots of mistakes.]

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/pinentry: allow building without Qt accessibility

Backport a patch applied upstream to allow building without Qt
accessibility.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/pinentry: fix include paths

The pinentry tarball contains generated .moc files that contains
incorrect include paths: they are probably specific to the machine of
the user who generated the tarball.

Fix those paths.

Fixes numerous autobuild failures:
    http://autobuild.buildroot.org/results/96c/96c68e9a6c528e15016cb10f0f73e52f58815bec/
    http://autobuild.buildroot.org/results/50e/50ecb6b347ab4970c701b1aed7b42518fef9a52c/
    http://autobuild.buildroot.org/results/fcb/fcbe8b971ef37c0e7c458bc32295a89be9872b61/
    ...

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libiio: new package

Libiio is a library to ease the development of software interfacing
Linux Industrial I/O (IIO) devices.

http://wiki.analog.com/resources/tools-software/linux-software/libiio

[Thomas:
 - Remove changelog from the commit log.
 - Rename init script to S99iiod instead of S99iiod.sh
 - Remove "status" command from init script, and implement "restart"
   instead.
 - Add dependency on thread support in toolchain.
 - Fixup indentation in the .mk file.
 - Fixup wrapping in the Config.in help text.]

Signed-off-by: Paul Cercueil <paul.cercueil@analog.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/gdb: fix shared-only build

The shared build only of gdb is actually not supported.

When --disable-static is given to the gdb's main configure script,
it also disable the static build of bundled libraries like bfd, readline
and opcodes, which must be build statically.

Force the static build of bundled libraries by adding --enable-static in
GDB_CONF_OPTS.

Fixes:
http://autobuild.buildroot.net/results/219/21979e730dca1fce5bdda9a4c7fad4485e788866/
http://autobuild.buildroot.net/results/c10/c1096d0bd22de5c6feba848f743601ad0416a944/
http://autobuild.buildroot.net/results/d8a/d8a5bcc7fa374fb0c916a9d0f33ef283109cb404/
http://autobuild.buildroot.net/results/22a/22a86d0f1df0fc4698c0f734f3d659e6317404a4/
http://autobuild.buildroot.net/results/404/404e61e5c30040ee5756f4b5839149dca38660d4/

And many more.

Signed-off-by: Romain Naour <romain.naour@openwide.fr>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

toolchain: get rid of -pipe from optimisations

-pipe is causing some build failures in Linux kernel >= 3.17.

Also, nowadays, using -pipe does not gain as much as it used to back in
the days:

Measurements made with a 3.16.7 Linux kernel:
    make linux-depends
    time sh -c 'make linux-build >/dev/null 2>&1'

Without -pipe:
    716.32user 54.44system 3:42.12elapsed 346%CPU
    721.22user 54.47system 3:41.81elapsed 349%CPU
    722.44user 54.00system 3:42.13elapsed 349%CPU
    721.03user 53.81system 3:41.92elapsed 349%CPU
    713.21user 53.63system 3:40.51elapsed 347%CPU
    706.67user 52.42system 3:38.40elapsed 347%CPU
    714.40user 53.18system 3:40.16elapsed 348%CPU
    706.01user 53.09system 3:37.87elapsed 348%CPU
    705.98user 53.01system 3:38.03elapsed 348%CPU
    714.17user 53.55system 3:39.98elapsed 348%CPU
Average:                   3:40.29elapsed

With -pipe:
    720.13user 53.90system 3:41.98elapsed 348%CPU
    713.38user 53.69system 3:40.44elapsed 347%CPU
    711.60user 52.81system 3:39.06elapsed 348%CPU
    708.66user 53.09system 3:38.59elapsed 348%CPU
    711.76user 53.00system 3:38.48elapsed 350%CPU
    717.85user 53.97system 3:41.77elapsed 348%CPU
    716.77user 53.77system 3:40.91elapsed 348%CPU
    717.48user 53.65system 3:41.24elapsed 348%CPU
    721.44user 55.67system 3:43.45elapsed 347%CPU
    724.61user 55.63system 3:43.35elapsed 349%CPU
Average:                   3:40.93elapsed

The delta is well in the measurement noise.

Just get rid of it.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Romain Naour <romain.naour@openwide.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/ncurses: fix builds with both shared and static libs

When both shared and static libs are enabled, we have two shell
constructs to run, but they are not properly separated, leading to build
failures like:

    ln -sf libncursesw.a /home/idnc_sk/IOLINUX/builds/micro-x86_64-corei7/
    output/host/usr/x86_64-buildroot-linux-uclibc/sysroot/usr/lib/libcurse
    s.a for lib in libncurses libmenu libpanel libform; do ln -sf ${lib}w.
    so /home/idnc_sk/IOLINUX/builds/micro-x86_64-corei7/output/host/usr/x8
    6_64-buildroot-linux-uclibc/sysroot/usr/lib/${lib}.so; done
    /bin/bash: -c: line 0: syntax error near unexpected token `do'
    /bin/bash: -c: line 0: `ln -sf libncursesw.a /home/idnc_sk/IOLINUX/bui
    lds/micro-x86_64-corei7/output/host/usr/x86_64-buildroot-linux-uclibc/
    sysroot/usr/lib/libcurses.a for lib in libncurses libmenu libpanel lib
    form; do ln -sf ${lib}w.so /home/idnc_sk/IOLINUX/builds/micro-x86_64-c
    orei7/output/host/usr/x86_64-buildroot-linux-uclibc/sysroot/usr/lib/${
    lib}.so; done'

As can be seen, there is a missing semi-colon ';' between the symlink
command and the for loop:

    ln -sf libncursesw.a [...]/libcurses.a for lib in [...]

Fix that by adding a semi-colon after each first shell constructs, to
properly separate the two. If the second one is not enabled (i.e. for a
static-only build), there is a trailing semi-colon, but that's perfectly
valid shell syntax.

Reported-by: idnc_sk on IRC
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

python-cheetah: new package

[Thomas:
 - Remove dependency on host-python-markdown, apparently not needed
   (it builds fine in a minimal chroot, which doesn't have
   python-markdown installed)
 - Remove runtime dependency on markdown, I can run Cheetah basic
   examples without python-markdown installed.
 - Add dependency on Python 2 only, since it doesn't build for Python
   3.]

Signed-off-by: Gwenhael Goavec-Merou <gwenhael.goavec-merou@trabucayre.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

python-markdown: new package

[Thomas:
 - Remove host variant of python-markdown, as it is not needed.
 - Add runtime dependency on Python XML module.]

Signed-off-by: Gwenhael Goavec-Merou <gwenhael.goavec-merou@trabucayre.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

iperf3: new package

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/iucode-tool: bump version

- updated upstream location
  http://metadata.ftp-master.debian.org/changelogs/contrib/i/iucode-tool/unstable_changelog
  "iucode-tool (1.0.2-1) unstable; urgency=low
   [...]
   * debian/control: update for new upstream location at Gitorious"

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/intel-microcode: bump version

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libtool: fix help2man related failure

Libtool sometimes requires help2man to build manpages, but since we
don't use/need them and want to avoid pulling in and/or building
dependencies for that we just disable it. Fixes:
http://autobuild.buildroot.net/results/bfb/bfb9e298196d877b623580f882b6fe74b30bec9b/

Also add host-m4 as dependency, it's not strictly required but silences
many build warnings.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

tinc: new package

[Thomas:
 - fix license from GPLv2 to 'GPLv2+ with OpenSSL exception'
 - add -std=c99 to CFLAGS, otherwise it doesn't build with a minimal
   toolchain configuration
 - add BR2_USE_MMU dependency since fork() is used.]

Signed-off-by: Zoltan Gyarmati <mr.zoltan.gyarmati@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

sudo: adjust libtool comment

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libfribidi: adjust libtool comment

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

lzo: adjust libtool comment

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

subversion: security bump to version 1.7.19

Fixes:
CVE-2014-3580: mod_dav_svn DoS from invalid REPORT requests.
CVE-2014-8108: mod_dav_svn DoS from use of invalid transaction names.

Also add hash file.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libtool: bump to version 2.4.4

Patch is upstream so remove it.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

support/libtool: add patch for newer versions

Add patch to support newer (>2.4.2) versions.
Adjust patch logic to check for patchlevel greater than 2 (apply new patch) or
not (apply current patch).
Some people/distributions used unreleased versions, with the string being
2.4.2.x, this packages are AUTORECONFed and have to be kept like this since
the up-to-2.4.2 patch doesn't work, neither does the from-2.4.3 version patch.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/perl-gd: rename patches to follow the new name convention

Signed-off-by: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

python-enum: new package

This is an alternative to python-enum34 with a somewhat different API.

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/zlib: allow shared-only builds

Currently, we can build a static-only zlib, but not a shared-only one.

This is because zlib's build-system is a custom ./configure (not using
autotools), and does not allow building/installing only the shared
library.

Simply remove the .a as a post-staging install hook. We don't care
removing it from target, since it is not used at link time to build
other packages, and it is anyway removed later before assembling the
filesystem images anyway.

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>

package/oprofile: Bump version to 1.0.0

Remove opcontrol and oprofiled from the install list, they are no longer
present in the package.

Remove all patches, they have been integrated upstream.

Signed-off-by: Ilya Lipnitskiy <ilya.lipnitskiy@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package: indentation cleanup

Signed-off-by: Jerzy Grzegorek <jerzy.grzegorek@trzebnica.net>
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/weston: do not select rpi-userland

rpi-userland is a provider for some virtual packages, so we can not
select it, as instructed in the manual:

    http://nightly.buildroot.org/#_infrastructure_for_virtual_packages

    ---8<---
    If your package really requires a specific provider, then you’ll
    have to make your package depends on this provider; you can not select a
    provider.
    ---8<---

Instead, just depend on it. Remove the comment as well.

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>

xfsprogs: fix shared-only build

The shared-only build is not supported by xfsprogs because without
--enable-static, the builtin libraries (libxfs, libxcmd ...)  are only
built as shared libraries and xfs tools requires the static ones.

Also, we needs to replace -static with -static-libtool-libs for the
shared-only case, in order to fallback to linking with the shared
version of libuuid library.

Fixes:
http://autobuild.buildroot.net/results/139/139fa30756d35a24209136068c818802074908de/
http://autobuild.buildroot.net/results/e2e/e2eef5dbadec555756a912a24978c78c0013ee70/
http://autobuild.buildroot.net/results/08c/08c4bc5037379a8ca312e3e7574067235da74030/

And many more.

[Thomas: fix misc typos in the patch description and commit title.]

Signed-off-by: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/hostapd: use correct include path for libnl-3.0

Add a dependancy on host-pkgconf.

Signed-off-by: Romain Naour <romain.naour@openwide.fr>
Reported-by: Jörg Krause <jkrause@posteo.de>
Cc: Jörg Krause <jkrause@posteo.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/wpa_supplicant: use correct include path for libnl-3.0

Fixes:
ERROR: unsafe header/library path used in cross-compilation: '/usr/include/libnl3'

Add a dependancy on host-pkgconf when BR2_PACKAGE_LIBNL is enabled.

Signed-off-by: Romain Naour <romain.naour@openwide.fr>
Reported-by: Jörg Krause <jkrause@posteo.de>
Cc: Jörg Krause <jkrause@posteo.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

json-c: change upstream location to Github

json-c website and download locations have changed, the project is now
hosted on Github.

Signed-off-by: Sagaert Johan <sagaert.johan@skynet.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/ffmpeg: bump version to 2.5.1

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

manual: remove the extra whitespaces

Signed-off-by: Jerzy Grzegorek <jerzy.grzegorek@trzebnica.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

xserver_xorg-server: security bump to 1.16.3

- for a list of fixed CVE´s check
  http://lists.x.org/archives/xorg-announce/2014-December/002506.html
- added hash file

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/vlc: Update SITE url

old SITE is now password-protected

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/libdvbcsa: Update SITE url

old SITE is now password-protected

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

orc: bump version

And add hash file.

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

ympd: fix license typo

Signed-off-by: Jerzy Grzegorek <jerzy.grzegorek@trzebnica.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/libbluray: bump version to 0.6.2

- old SITE is now password-protected
- add hash

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

gcc: bump 4.8.x series to version 4.8.4

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

jasper: add patches to fix CVE-2014-8137 and CVE-2014-8138

Fixes:
CVE-2014-8137 - double-free in jas_iccattrval_destroy()
CVE-2014-8138 - heap overflow in jp2_decode()

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

git: security bump to version 2.2.1

Fixes:
CVE-2014-9390 - arbitrary command execution vulnerability on
case-insensitive file systems.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

erlang: fix typo in dependencies

If libatomic_ops was enabled, then the host-erlang dependency was lost.

Signed-off-by: Frank Hunleth <fhunleth@troodon-software.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/snappy: add dependency on host-pkgconf

Snappy doesn't configure without host pkg-config, causing this totally
unhelpful diagnostic from autoconf:

  configure.ac:42: error: possibly undefined macro: AC_DEFINE
        If this token and others are legitimate, please use m4_pattern_allow.
        See the Autoconf documentation.
  configure.ac:44: error: possibly undefined macro: AC_MSG_FAILURE

So add host-pkgconf to the package's DEPENDENCIES list.

Signed-off-by: Steve James <ste@junkomatic.net>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package: cleanup whitespace of assignments

Signed-off-by: Jerzy Grzegorek <jerzy.grzegorek@trzebnica.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

icu: bump to version 54.1

Also add hash file.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

boost: bump to version 1.57.0

Also add hash file.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

mpd: bump to version 0.19.7

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

tcpreplay: bump to version 4.1.0

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

linux-pam: add security patches

To fix:
CVE-2013-7041 - use case sensitive comparison in pam_userdb
CVE-2014-2583 - potential path traversal issue in pam_timestamp

Also add hash file (computed, the hash files upstream cover up to 1.1.7)

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

php: security bump to version 5.5.20

Fixes:
CVE-2014-8142 - Use after free vulnerability in unserialize()

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

jquery: bump to version 1.11.2

Also add hash file.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

nftables: bump to version 0.4

Also add hash file.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

libnftnl: bump to version 1.0.3

Also add hash file.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

libnss: security bump to version 3.17.3

Fixes CVE-2014-1569 - The definite_length_decoder function in
lib/util/quickder.c in Mozilla Network Security Services (NSS) before
3.16.2.4 and 3.17.x before 3.17.3 does not ensure that the DER encoding
of an ASN.1 length is properly formed, which allows remote attackers to
conduct data-smuggling attacks by using a long byte sequence for an
encoding.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

apr-util: add hash file

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

apr: add hash file

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

linux: bump default to version 3.18.1

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

linux-headers: bump 3.{18, 17, 14, 10}.x series

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/Config.in: fix alphabetical order

Signed-off-by: Jerzy Grzegorek <jerzy.grzegorek@trzebnica.net>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/Config.in: fix menu alphabetical order

Signed-off-by: Jerzy Grzegorek <jerzy.grzegorek@trzebnica.net>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

dbus: bump version

Regression fix for CVE-2014-3636, see release announcement:
http://lists.freedesktop.org/archives/dbus/2014-November/016430.html

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

configs/qemu: update infos for latest qemu version

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

stunnel: adjust initscript

Use proper status messages, make spacing standard instead of a mix of
spacing/tabbing, drop boringly obvious comment from the header.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

stunnel: bump to version 5.08

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

ncurses: add hash file

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

ncurses: fix gpm support

The CF_LIB_SONAME macro doesn't work when cross compiling so we need to
specify the lib name for libgpm explicitly. While at it make gpm support
explicit in the form of --without-gpm when it's not selected and adding
it to dependencies when it is. Fixes:
http://autobuild.buildroot.net/results/32a/32a5ba3905772a3f2f2ec9d1b290a109fe22d9f9/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

linux: fix recursively defined variable

When running 'make printvars', the output stops at the time we dump the
Linux related variables, with:

    linux/linux.mk:109: *** Recursive variable `LINUX_TARGET_NAME'
    references itself (eventually).  Stop.

And that's expected, since we have:

    109 LINUX_TARGET_NAME = $(LINUX_IMAGE_NAME)
    [...]
    112 ifeq ($(LINUX_IMAGE_NAME),)
    113 LINUX_IMAGE_NAME = $(LINUX_TARGET_NAME)
    114 endif

Even though they are defined in a way that ensures they are in fact not
recursively defined (the if-block ensures that), 'printvars' does dump
all our variables by evaluating all of them, which in that specific case
implies they are recursively defined.

Fix that by explicitly setting LINUX_IMAGE_NAME in each if-block.

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>

erlang: use libatomic_ops if available

Erlang can either provide its own atomic_ops implementation or use the
one in libatomic_ops. This patch ensures that it only uses the one from
libatomic_ops if it is available.

Fixes:

http://autobuild.buildroot.net/results/3eb/3ebfd36948d5a642cc83ce31ddf5902cb369bfbd/
http://autobuild.buildroot.net/results/04c/04cbf79ccf159e15f9ea3953b425faadafc8a1c0/
http://autobuild.buildroot.net/results/fc9/fc9904446553e67dd44954310fea26d751a0df5d/
http://autobuild.buildroot.net/results/501/501a12da25615071fb6197a898456cfe41a6451d/
http://autobuild.buildroot.net/results/34e/34e2fbb4d39737fd7defa5bd607726d52ef38568/
http://autobuild.buildroot.net/results/722/7220c3905ebe9e081f1e3b90641b0859d996c2ac/
http://autobuild.buildroot.net/results/ffd/ffdc531d8cf9adb43340d045aa88eb0b39fc9bb7/
http://autobuild.buildroot.net/results/b4f/b4f238bcb30b0191ce1a2be669fa792c8ccd3e69/

Signed-off-by: Frank Hunleth <fhunleth@troodon-software.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

pckage/ncurses: fix pkg-config files for ncursesw

Currently, ncurses creates symlinks from the non-'w' variants to the
equivalent 'w' variant, but forgets to do so for pkg-config files.

Do that for the pkg-config files too.

Should fix numerous autobuild failures:
    http://autobuild.buildroot.net/results/3c4/3c44dfc6e3f151bef751f39cac336fef3d6a0099/
    http://autobuild.buildroot.net/results/38c/38c81407bb8d0677e87b3a01f9a1fc1372ab4738/
    ...

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/ncurses: trim leading 'lib' from definition of libs to install

Currently, ncurses creates symlinks from the non-'w' variants to the
equivalent 'w' variant, but forgets to do so for pkg-config files.

To be able to share the same list between the libraries and the
pkg-config files to symlink, just trim the 'lib' prefix of libraries in
the definition, and just add it back at the time we need it.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/ncurses: slight code re-arrangement

Move the definition of libraries to install before it is actually used.

Also, in a coming changeset, it will also be used to know which
pkg-config files to symlink.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/ncurses: fix shared-only libs

For a shared-only build, do not create the symlinks to the static
libraries, since they do not exist.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/gpsd: remove duplicate dependency

No need to duplicate the host-pkgconf dependency on the Qt case, we
already depend on it in the general case.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

libyaml: add security patch for CVE-2014-9130

Fixes: assert failure when processing wrapped strings (CVE-2014-9130).
See https://bugzilla.redhat.com/show_bug.cgi?id=1169369

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/collectd: fix build without libgcrypt

In case the libgcrypt development files are present on the host system,
collectd's ./configure will mistakenly try to use them and will call the
host system's libgcrypt-config, thus leading it to use path to the host
system includes and libraries.

Fix that in two ways:
  - explicitly disable libgcrypt support when libgcrypt is not enabled;
  - pass the complete path to libgcrypt-config when libgcrypt is enabled.

However, collectd's configure.ac is utterly broken. The code in
configure.ac has special code to check for libgcrypt-config, and use
whatever is provided via --with-libgcrypt=/path/to/libgcrypt-config. But
that is promptly forgotten because they then call the AM_PATH_LIBGCRYPT
macro, that just does it all again from scratch, and does not use the
value previously found.

Instead, we set LIBGCRYPT_CONFIG in the environment and point it to our
own libgcrypt-config.

Should fix numerous build issues:
    http://autobuild.buildroot.net/results/ad4/ad408aef5fb92fe9e031c7dbaf6999776b40ace4/
    http://autobuild.buildroot.net/results/967/96735bfa91bcf2e3dff89f69c0a12ed406e9efb9/
    ...
    http://autobuild.buildroot.net/results/3bd/3bdd9bdffb1d55414787d38fc2656d7a3391a957/
    ...

(the first two are with the paranoid wrapper, the third one was before
the paranoid wrapper.)

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Romain Naour <romain.naour@openwide.fr>
Cc: Peter Korsgaard <jacmet@uclibc.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

linux-headers: bump 3.{12, 2}.x series

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

libgpgme: bump to version 1.5.3

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

ethtool: bump to version 3.18

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

blktrace: really fix parallel build issue

Fixes:
http://autobuild.buildroot.net/results/dd4/dd47b3ca17bdb597d6b12de6b381cf5f25abaf71/
http://autobuild.buildroot.net/results/a84/a842234fbdc8f65d48089b90ee23523540a7091f/
http://autobuild.buildroot.net/results/e63/e6351692ad6aef397e68ffdf107f3a85e4490015/

And many more.

Commit a21163aed (package/blktrace: do not build in parallel) tried to fix
the parallel build issue, but the package doesn't use the autotools package
infrastructure so BLKTRACE_MAKE isn't used.  Instead replace the explicit
$(MAKE) calls with $(MAKE1).

Cc: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/tvheadend: rename patches to new convention

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/tvheadend: Bump version

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/tvheadend: Fix inclusion of DVB frequency tables in bundled build

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/mesa3d: Bump version to 10.4.0

- removed mesa3d-0002-execinfo_h.patch, applied upstream
- renamed remaining patch according to new convention

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/librtlsdr: fix build failure

Disable run test for pthread detection.

In case of build with only static library, CMake fails to detect the
pthread library (-lpthread), so it falls back trying to link using
-pthread and execute the test program.

<snip>
  [...]
  -- Looking for include file pthread.h
  -- Looking for include file pthread.h - found
  -- Looking for pthread_create in pthreads
  -- Looking for pthread_create in pthreads - not found
  -- Looking for pthread_create in pthread
  -- Looking for pthread_create in pthread - not found
  -- Check if compiler accepts -pthread
  CMake Error: TRY_RUN() invoked in cross-compiling mode, please set the following cache variables appropriately:
     THREADS_PTHREAD_ARG (advanced)
  For details see /home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/TryRunResults.cmake
  -- Check if compiler accepts -pthread - no
  -- Found Threads: TRUE
  -- Udev rules not being installed, install them with -DINSTALL_UDEV_RULES=ON
  -- Building with kernel driver detaching disabled, use -DDETACH_KERNEL_DRIVER=ON to enable
  -- Building for version: 0.5.3 / 0.5.3
  -- Using install prefix: /usr
  -- Configuring incomplete, errors occurred!
  See also "/home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeOutput.log".
  See also "/home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeError.log".
</snip>

Here it fails because, in the CMake FindThreads.cmake file, the
'try_run' function is unconditionally called [1].

Also, the CMakeError.log gives more details why linking with -lpthread
fails:

<snip>
  [...]
  Run Build Command:"/usr/bin/make" "cmTryCompileExec3012181287/fast"
  make[1]: Entering directory '/home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeTmp'
  /usr/bin/make -f CMakeFiles/cmTryCompileExec3012181287.dir/build.make CMakeFiles/cmTryCompileExec3012181287.dir/build
  make[2]: Entering directory '/home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeTmp'
  /home/buildroot/instance-1/output/host/usr/bin/cmake -E cmake_progress_report /home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeTmp/CMakeFiles 1
  Building C object CMakeFiles/cmTryCompileExec3012181287.dir/CheckFunctionExists.c.o
  /home/buildroot/instance-1/output/host/usr/bin/arm-linux-gcc   -pipe -Os -static  -DCHECK_FUNCTION_EXISTS=pthread_create   -o CMakeFiles/cmTryCompileExec3012181287.dir/CheckFunctionExists.c.o   -c /home/buildroot/instance-1/output/host/usr/share/cmake-3.0/Modules/CheckFunctionExists.c
  Linking C executable cmTryCompileExec3012181287
  /home/buildroot/instance-1/output/host/usr/bin/cmake -E cmake_link_script CMakeFiles/cmTryCompileExec3012181287.dir/link.txt --verbose=1
  /home/buildroot/instance-1/output/host/usr/bin/arm-linux-gcc  -pipe -Os -static  -DCHECK_FUNCTION_EXISTS=pthread_create  -static  CMakeFiles/cmTryCompileExec3012181287.dir/CheckFunctionExists.c.o  -o cmTryCompileExec3012181287 -rdynamic -lpthread -Wl,-Bstatic -lusb-1.0 -Wl,-Bdynamic
  /home/buildroot/instance-1/output/host/usr/arm-buildroot-linux-uclibcgnueabi/sysroot/usr/lib/libpthread.a(init.os): In function `__pthread_initialize_minimal':
  init.c:(.text+0x270): undefined reference to `__libc_setup_tls'
  collect2: error: ld returned 1 exit status
  CMakeFiles/cmTryCompileExec3012181287.dir/build.make:89: recipe for target 'cmTryCompileExec3012181287' failed
  make[2]: *** [cmTryCompileExec3012181287] Error 1
  make[2]: Leaving directory '/home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeTmp'
  Makefile:118: recipe for target 'cmTryCompileExec3012181287/fast' failed
  make[1]: *** [cmTryCompileExec3012181287/fast] Error 2
  make[1]: Leaving directory '/home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeTmp'

  Determining if compiler accepts -pthread returned PLEASE_FILL_OUT-FAILED_TO_RUN instead of 2. The compiler had the following output:
  Change Dir: /home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeTmp

  Run Build Command:"/usr/bin/make" "cmTryCompileExec1361818630/fast"
  make[1]: Entering directory '/home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeTmp'
  /usr/bin/make -f CMakeFiles/cmTryCompileExec1361818630.dir/build.make CMakeFiles/cmTryCompileExec1361818630.dir/build
  make[2]: Entering directory '/home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeTmp'
  /home/buildroot/instance-1/output/host/usr/bin/cmake -E cmake_progress_report /home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeTmp/CMakeFiles 1
  Building C object CMakeFiles/cmTryCompileExec1361818630.dir/CheckForPthreads.c.o
  /home/buildroot/instance-1/output/host/usr/bin/arm-linux-gcc   -pipe -Os -static     -o CMakeFiles/cmTryCompileExec1361818630.dir/CheckForPthreads.c.o   -c /home/buildroot/instance-1/output/host/usr/share/cmake-3.0/Modules/CheckForPthreads.c
  Linking C executable cmTryCompileExec1361818630
  /home/buildroot/instance-1/output/host/usr/bin/cmake -E cmake_link_script CMakeFiles/cmTryCompileExec1361818630.dir/link.txt --verbose=1
  /home/buildroot/instance-1/output/host/usr/bin/arm-linux-gcc  -pipe -Os -static    -static  CMakeFiles/cmTryCompileExec1361818630.dir/CheckForPthreads.c.o  -o cmTryCompileExec1361818630 -rdynamic -pthread
  make[2]: Leaving directory '/home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeTmp'
  make[1]: Leaving directory '/home/buildroot/instance-1/output/build/librtlsdr-v0.5.3/CMakeFiles/CMakeTmp'
</snip>

While falling back on -pthread is correct when using only static
libraries to get some tls-related macros defined [2,3], it is wrong to
try to run the test program in case of cross-compilation.

So, this patch only disable run test when pthread support is done using
-pthread, i.e. in case of the type of libraries is "static only".

Fixes:
  http://autobuild.buildroot.org/results/26c/26cb8c2e5f867d49451f61304bf0a1a6f7b825d9/

[1] http://www.cmake.org/gitweb?p=cmake.git;a=blob;f=Modules/FindThreads.cmake;h=a0bc4d133eb3f9d3ae096f0e03e419ef7be66ce8;hb=HEAD#l84
[2] http://lists.busybox.net/pipermail/buildroot/2014-June/100690.html
[3] http://stackoverflow.com/questions/23250863/difference-between-pthread-and-lpthread-while-compiling

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Cc: Baruch Siach <baruch@tkos.co.il>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

iprutils: bump to version 2.4.5

Rebase patch 0003 on top of 2.4.5 release.

Signed-off-by: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>