linux: bump default version to 3.17.1

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

linux-headers: bump 3.{17, 16, 14, 10}.x series

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

openssl: security bump to version 1.0.1j

Fixes:
CVE-2014-3513 - SRTP memory leak
CVE-2014-3567 - Session ticket memory leak
CVE-2014-3568 - Build option no-ssl3 is incomplete
And adds SSL3 fallback protection against POODLE.

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

whois: bump to version 5.2.1

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

php: security bump to version 5.5.18

Fixes:
CVE-2014-3669 - Integer overflow in unserialize() (32-bits only)
CVE-2014-3670 - Heap corruption in exif_thumbnail()
CVE-2014-3668 - Global buffer overflow in mkgmtime() function

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Acked-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

package/libnfs: bump to version 1.9.5

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

manual: Add note about uses of github helper

Add a little comment regarding the uses of the github helper for the releases.

[Arnout: extend the comment a little]

Signed-off-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Samuel Martin <s.martin49@gmail.com>
Acked-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

pkg-utils: describe the case conversion macros in more detail.

After the refactoring of the case conversion macros in dd5e620ac, the
macro became even harder to understand. Improve this explanation a
little so future generations can make sense of what we're doing here.

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Thomas De Schampheleire <patrickdepinguin@gmail.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

libgtk3: ensure host version of pkg-config is used for native compilation

This already was in Eric's submission, but I removed it when committing
while cleaning up.  The package infrastructure already passes the correct
pkg-config to use for cross compilation, but NOT for building native tools,
so we ended up using the target variant and got wrong libraries/headers.

Fixes:

http://autobuild.buildroot.net/results/e0c/e0c2623e2d400ce66963eae23f136c0158556b36/
http://autobuild.buildroot.net/results/5bf/5bf66b8d16a3ddded24827bf278ad24b55ded6e7/
http://autobuild.buildroot.net/results/75f/75f93cea8ab59f052ddefa51c0bfc14c3df47a60/

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

libgtk3: ensure cups-config can be found

Fixes:

http://autobuild.buildroot.net/results/db6/db6cd8922444490056bf5157fc0c9349e9bf383b/
http://autobuild.buildroot.net/results/4bd/4bdcd642eaa880dbbfbed8f4ed5ffac0c52edaa8/
http://autobuild.buildroot.net/results/8d3/8d30687f92178de15d7697a7b8a1b7e2d6e93deb/

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

getent: look for glibc's getent in more places

External toolchains don't always install the executable in /usr/bin -
God knows why. So check in a few more places.

We can use make's $(wildcard ...) here, because the variable only gets
expanded when the install step is executed, and by that time the
STAGING_DIR has already been installed.

Probably fixes most getent autobuilder errors, at least if fixes
http://autobuild.buildroot.net/results/615b9d17f713b4a53192efb00188560a76b9efa3/

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

xz: add hash file

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

bzip2: add hash file

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

samba: fix config symbols

Some of the configure cache symbols changed name because some routines
were moved to libreplace instead of samba itself.
So change the naming for those affected so that samba can work nicely
again with some configuration directives (like interfaces = X).

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

kmod: add hash file

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

crda: add hash file

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

tftpd: add hash file

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

libqmi: bump to 1.10.4

Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

musl: bump to version 1.1.5

Patch upstream, add hash file.

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

libgtk3: new package

This package provides Gtk+ 3.0, a graphical toolkit.

[Peter: fix Config.in deps, ensure atleast 1 backend, remove redundant info]
Signed-off-by: Hadrien Boutteville <hadrien.boutteville@gmail.com>
Signed-off-by: Eric Le Bihan <eric.le.bihan.dev@free.fr>
[Eric: added backend support, use of pkgconf for host-libgtk3]
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

hicolor-icon-theme: new package

This new package provides Freedesktop Hicolor icon theme.

This is the fallback icon theme, which only creates the standard
directory structure for storing icons of third-party applications (i.e.
the ones not available in usual icon themes).

Signed-off-by: Eric Le Bihan <eric.le.bihan.dev@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

kexec: bump to version 2.0.8

Also add hash file.

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

hostapd: add hash

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

perf: fix typo and syntax error

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

libnss: bump to version 3.17.2

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

libcdio: bump to version 0.93

Also add hash file.

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

make: bump to version 4.1

Also add hash file.

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

gnutls: bump to version 3.2.19

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

radvd: bump to version 2.8

Also add hash file.

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

uclibc: drop stray patches

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

binutils: drop stray patch

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

libv4l: bump version

And add tarball hashes.

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

package/mesa3d: Add hash file

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

package/mesa3d: Bump version to 10.3.1

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

openssh: place configuration files in /etc/ssh

pkg-autotools.mk fix --sysconfdir to "/etc". This patch restore
--sysconfdir to its default value (/etc/ssh)

Signed-off-by: Jérôme Pouiller <jezz@sysmic.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

connman: select dbus support in wpa_supplicant

... else connman cannot communicate with wpa_supplicant

Signed-off-by: Jérôme Pouiller <jezz@sysmic.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

hostapd: needs MMU

hostapd calls fork() since last version bump.

Fixes:
http://autobuild.buildroot.net/results/945/94543b865db2ff5da34434fdcdf15ea7db73f392/

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

uclibc: explicitly set HAVE_SHARED depending on BR2_PREFER_STATIC_LIB

Now with change of BR2_PREFER_STATIC_LIB meaning to "do not build dynamic libs
and build statically linked applications" it's possible to disable support of
shared libs in uClibc as well as builting of shared libc libs.

Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Cc: Peter Korsgaard <peter@korsgaard.com>
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

Config.in: update description of BR2_PREFER_STATIC_LIB

With time BR2_PREFER_STATIC_LIB option meaning changed from "prefer static
libs when possible" to "do not build dynamic libs and build statically linked
applications".

This patch changes the option description.

Note that variable name is kept unchanged on purpose - it will be modified later
when real "prefer static" mode will be introduced.

[Peter: tweak help text as suggested by Thomas and Yann]
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Cc: Peter Korsgaard <peter@korsgaard.com>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/upmpdcli: fix missing dependency of libcurl

Signed-off-by: Jörg Krause <jkrause@posteo.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

libva-intel-driver: indirectly depends on XORG7

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

libva-intel-driver: add indirect dependency on BR2_PREFER_STATIC_LIB

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Acked-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

valgrind: add patch to fix build on PowerPC uClibc

uClibc on PowerPC forgot to define some PTRACE_* definitions. Even
though the Buildroot uClibc package has a fix for this, it doesn't fix
external uClibc toolchains that are not patched. Therefore this commit
adds a patch to Valgrind to workaround the problem by defining
PTRACE_GETSIGINFO to the appropriate value on PowerPC when it's not
defined by the C library.

Fixes:

  http://autobuild.buildroot.org/results/bc6/bc6be9aa1ad39201a1045a65c0c3eb1e968371f2/

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

nginx: disallow AIO support on AArch64

The nginx AIO support cannot build on AArch64, because it hardcodes
the usage of SYS_eventfd(), while it should just use eventfd() from
the C library.

See https://bugs.launchpad.net/linaro-aarch64/+bug/1160013 for details.

Fixes:

  http://autobuild.buildroot.org/results/54b/54bb761f9e774d91066b78217efcd583fdd85ff3/

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

wpa_supplicant: uses fork(), not available on no-MMU platforms

The only package selecting wpa_supplicant is connman, and it is
already depends on BR2_USE_MMU.

Fixes:

  http://autobuild.buildroot.org/results/9ac/9ac3c0de573f5266a889f1ceb6d43aa92f715c55/

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

package/qemu: add SDL frontends

[Peter: use BR2_PACKAGE_QEMU_HAS_EMULS, only add SDL_CONFIG once]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/qemu: add fine-grained target selection

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

package/qemu: add basic target selection

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

package/qemu: build for the target (i386 or x86_64 only)

Untested for other architectures, so only enabled for known-working ones.

We explicitly disable all features that depends on other packages,
for now. Such features will be added one by one in the next patches.
Features that do not rely on external packages are explicitly enabled
to catch the dependencies on toolchain features early.

[Peter: propagate dependencies from libglib2]
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/qemu: bump version

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

package/qemu: do not use autotools for qemu

Although QEMU has a ./configure script, it is not an autotools-package,
so switch to providing build and install commands, and to using
generic-package.

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

ecryptfs-utils: add runtime dependency on getent

Fixes bug #7142.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

getent: new package

The ecryptfs-utils scripts require the 'getent' program to be
installed to find the home directory of users. However, Buildroot
currently never installs this program, and therefore bug #7142 was
reported, explaining that ecryptfs-utils is not working properly.

In normal Linux systems, the getent program is provided by glibc, and
allows to query not only /etc/passwd, but also other NSS databases
such as LDAP and others.

In the context of Buildroot, this gives us several cases:

 1/ Internal toolchain

    a/ glibc/eglibc. In this case, the getent program is already built
       and installed by Buildroot in the staging directory, so the
       only thing missing is installing it in the target directory.

    b/ uclibc. uClibc provides a simple shell script that emulates the
       behavior of getent. It is located in extra/scripts/getent in
       the uClibc sources, but is currently never installed.

    c/ musl. There seems to be no getent implementation, and musl does
       not support NSS.

 2/ External toolchain

    a/ glibc/eglibc. In several external toolchains that we tested,
       there is a pre-built getent binary available in the sysroot,
       but Buildroot is not installing it to the target.

    b/ uclibc. The getent wrapper script is typically not part of any
       external uClibc toolchain.

    c/ musl. There is no getent implementation.

This patch proposes to solve this problem by introducing a getent
package, which has the following behavior:

 - When the toolchain is glibc based (either internal or external), it
   installs the getent program that was built and installed in the
   staging directory. This covers cases 1/ a/ and 2/ a/ above.

 - When the toolchain is uclibc or musl based, it installs a version
   of uclibc's getent wrapper script that is built into the getent
   package. This script is unlikely to change over time, so having it
   directly built into the package should not cause much issues moving
   forward. This covers all other cases above.

This solution allows to install a NSS-capable getent when glibc/eglibc
is used, and otherwise to rely on uClibc's wrapper script.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

beaglebone: better support for starting and creating images

- Added short readme file with description of first steps how to start.
- Added uEnv.txt so the board can start from SDCard
- Added post-image.sh script to copy included uEnv.txt to output/images

[Peter: strip trailing spaces]
Signed-off-by: Marcin Jabrzyk <marcin.jabrzyk@gmail.com>
Tested-by: Stephan Hoffmann <sho@relinux.de>
Tested-by: Marcin Jabrzyk <marcin.jabrzyk@gmail.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

uClibc: ldso/libdl: Also include dl-tls.h for for !SHARED.

Apply an upstream patch:

  http://git.uclibc.org/uClibc/commit/ldso/libdl/libdl.c?id=b57e9640db53166c88cdac66b79a046e46b8d728

On MIPS, several relocations that were original only resolved by the
dynamic linker were reused as static relocations.  Consequently the
macros TLS_DTPREL_VALUE and TLS_TPREL_VALUE defined in
libpthread/nptl/sysdeps/mips/dl-tls.h need to be available even for
!SHARED.

Relevant: http://www.linux-mips.org/wiki/NPTL#History

Fixes:

  http://autobuild.buildroot.net/results/85e/85e66962da9a6b80e1d1b721031b6ef0da45e3d5/

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/vlc: Fix typo

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

inotify-tools: Bump version

The release is a few years old but the repository is still going, so
uses the SHA1 and uses the github function to download it.

Also, it now needs AUTORECONF.

Signed-off-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

docs/manual: slightly improve faq entry about compiler on target

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

scripts/graph-build-time: properly warn about missing modules

Currently the graph-build-time script prints a python exception if a
needed module cannot be imported. Catch the exception and tell the user
which packages are missing, as we do for other missing dependencies.

Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

scripts/graph-build-time: clarify backend selection

This instruction in the middle of 'import' lines looks very strange.

Also, it was not obvious to me what the 'Agg' backend is.

Both things are actually correct, but it took a while to find out why.
So clarify with a comment to save someone else's time.

[Peter: fix s/soe/some/ typo]
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Sascha Arthur <sascha.arthur@gmail.com>
Cc: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

xserver_xorg-server: Add optional package xlib_libXcomposite as dependency

Make sure that xlib_libXcomposite is build before xserver_xorg-server.

With this defconfig

BR2_TOOLCHAIN_EXTERNAL=y
BR2_PACKAGE_XORG7=y
BR2_PACKAGE_XSERVER_XORG_SERVER=y
BR2_PACKAGE_XSERVER_XORG_SERVER_MODULAR=y
BR2_PACKAGE_XLIB_LIBXCOMPOSITE=y

and "make xserver_xorg-server" the package xlib_libXcomposite was ignored.

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

am335x-pru-package: new package

am335x-pru-package provides an assembler and program loader for Texas
Instrument's AM335x programmable real-time units.

[Peter: fix typo in patch description]
Signed-off-by: Frank Hunleth <fhunleth@troodon-software.com>
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>

qlibc : new package

[Thomas:
 - add dependency on threads, wchar and dynamic librayr
 - add dependency on libiconv when locale support is not enabled
 - replace patch hacking includedir by a patch that lets the Makefile
   obey to $(DESTDIR)
 - remove optional OpenSSL and MySQL support which simply cannot work
   due to the usage of AC_CHECK_FILE() in configure.ac, this required
   an additional patch to fix the bogus AC_ARG_WITH() calls.
 - move from 'Miscellaneous' to 'Libraries' -> 'Other'.
 - fixup the licensing information.]

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

pkg-generic.mk: strip leading/trailing spaces

The manual gives this example for using the github macro:

  FOO_VERSION = v1.0 # tag or full commit ID

Unfortunately, people copy/pasting this example will face weird make errors,
because it leads the FOO_VERSION variable to end with a space.  Similar
problems can happen when testing a version bump or similar, so strip
leading/trailing spaces before the version is used to construct the build
directory path.

Reported-by: Edd Robbins <edd.robbins@gmail.com>
Cc: Edd Robbins <edd.robbins@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/opencv: fix build failure on x86

This change adds a patch fixing some opencv asm causing build failures on
i*86 with PIC.

This patch has been sent upstream [1].

Fixes:
  http://autobuild.buildroot.org/results/483/4838285b25d6293a5cf0bb9eadd5040a7c75d766/
  http://autobuild.buildroot.org/?reason=opencv-2.4.10&&arch=i686

[1] https://github.com/Itseez/opencv/pull/3331

Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

uclibc: build uclibc unstripped by default

uClibc can be built stripped (implemented as link-time option in uClibc).
This could be useful for those who build root fs manually.

In Buildroot there is a global strip stage wich strips most of shared libs
in "target" folder (wexcept kernel modules known to not work properly being
stripped and libpthread required to be non-stripped for correct debugging with
gdb of multi-threaded apps).

So there're few problems with current implementation:
 1. uClibc is being stripped 2 times (first on its build stage, second on
    global Buildroot strip stage)
 2. uClibc libs in "staging" folder are also always stripped except if
    "no strip" (BR2_STRIP_none) is explicitly is selected in Buildroot config.
    That makes it possible to remote debug uClibc libs on target only if target
    rootfs was not stripped (which might not be possible due to huge libs
    like Qt)

This patch disables embedded strip in uClibc (still users may modify uClibc
config and explicitly set "DOSTRIP=yes" if really needed).

Interesting that DOSTRIP was not only selected in uClibc config but also was
force set with Buildroot on uClibc configuration step with UCLIBC_STRIP_CONFIG.

Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Cc: Peter Korsgaard <peter@korsgaard.com>
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Cc: Thomas De Schampheleire <patrickdepinguin@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Waldemar Brodkorb <wbx@openadk.org>
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

system: add option to choose what /bin/sh points to

Not all our shells do install a pointer to /bin/sh. Besides, between
those that do and multiple ones are enabled, the last one to install
wins the the symlink.

Add a new config choice in the system sub-menu that allows the user to
explicitly select the shell to provide /bin/sh. If busybox is not
enabled, default to using dash, a POSIX shell.

Remove the symlink creation from bash.mk at the same time.

Note: for every shell, we select them, except busybox, on which we
depend, on the assumption that we do not want to force busybox in case
the user decided not to enable it.

[Peter: remove redundant !BR2_PACKAGE_BUSYBOX dependency]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Cc: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/systemd: check BR2_TARGET_GENERIC_GETTY

Do not create a [serial-]getty link if BR2_TARGET_GENERIC_GETTY is not set.
Might be useful for custom setups (rootfs overlay).

[Peter: use BR2_TARGET_GENERIC_GETTY_PORT to match the code below]
Signed-off-by: André Erdmann <dywi@mailerd.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/systemd: fix getty setup

When trying to run a buildroot system configured with
BR2_TARGET_GENERIC_GETTY_PORT="tty1" (x86_64), the boot process hangs
with the following message:
  "A start job is running for dev-tty1.device (<time> / 1min 30s)"

Replacing /etc/systemd/system/getty.target.wants/serial-getty@tty1.service
(linking to serial-getty@) with getty@tty1.service (-> getty@) fixes the issue.

This patch adds a check that "detects" the tty type by removing digits at the
end of BR2_TARGET_GENERIC_GETTY_PORT and comparing the resulting base name.
An instance of getty@service gets created if the name matches "tty",
otherwise serial-getty@ gets instantiated (as before).

So, tty1,tty2,... are created as links getty@tty1.service -> getty@,
while ttyS0, ttyAMA0, ... are created as instances of serial-getty@.

[Peter: simplify logic]
Signed-off-by: André Erdmann <dywi@mailerd.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/Makefile.in: Fix dependency for selecting uclinux as TARGET_OS

Current setting only allows blackfin to select uclinux as TARGET_OS.
However, some noMMU ARM platforms that using FLAT binary format also need to
select uclinux as TARGET_OS. Fix the dependency.

Signed-off-by: Axel Lin <axel.lin@ingics.com>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

sysklogd: add initscript

Since we don't handle it in sysvinit inittab any longer let's add an
initscript to get things back into plug-and-play shape.

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

package/sysvinit: don't start logging

Don't try to start logging from /etc/inittab because busybox installs a
S01logging script that handles that and can lead to duplicate logging
processes.
Let's handle the logging in each package that provides it since we can
get multiple variants in the future like syslog-ng that have different
initialization routines.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

sunxi-cedarx: bump version, improve EABI support

This commit bumps the sunxi-cedarx package version to the latest git
commit, which also provides a better support for ARM EABI, since it
now has a non-EABIhf variant of avheap.so available.

Built tested on ARM EABI and ARM EABIhf.

[Thomas: compared to Spenser's original version, bumped to a newer
 version, and dropped support for the demo since it didn't build,
 and fixing it isn't trivial]

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

zsh: security bump to version 5.0.7

Fixes shellshock-alike exploits.

Install binary to /bin as all shells should be and add hash file.

[Peter: drop /bin/sh handling as we're going to handle it globally]
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

dash: bump to version 0.5.8-1

Add hash file and use $(INSTALL) instead of cp for proper mode handling.

[Peter: drop /bin/sh handling as we're going to handle it globally]
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/upmpdcli: bump to version 0.8.3

Since version 0.8.2 upmpdcli builds with gcc >= 4.5. This allows compilation
with the CodeSourcery toolchain versions 2011.03 and 2010.09 for the PowerPC
architecture.

Signed-off-by: Jörg Krause <jkrause@posteo.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

Makefile: make help: remove <package>- help, add link to online manual

We only documented a few of the <package>- targets and it's hard to
decide which ones are relevant for make help. Since the help is already
way too long, it's better to remove these advanced targets.

Instead, let's refer to the online manual.

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Acked-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

libuci: bump version

Signed-off-by: Alexey Mednyy <swexru@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

libubox: bump version

Signed-off-by: Alexey Mednyy <swexru@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

biosdevname: new package

Signed-off-by: Ryan Wilkins <ryan@deadfrog.net>
[yann.morin.1998@free.fr: bump to 0.6.0; fix dependencies (udev+zlib);
 add hashes; fix title]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

gcc: remove gcc snapshot option

As discussed during the dev days. It is broken for uClibc/musl and
architectures not using mainline gcc, so it has only very limited
usefulness.

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

kmod: needs to autoreconf

Fixes:
http://autobuild.buildroot.net/results/7a7/7a70b179f17070d0a0883f71c60030dd7d8a9102/
Didn't see widespread failure because it autoreconfed on its own just
fine if autotools infra was already present.
And didn't work before the automake patch with AUTORECONF=YES because of
an older gtk-doc.m4

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

automake: update gtk-doc.m4 to serial 2

Update gtk-doc.m4 infra to serial (version) 2.
Some packages start to need/ship with the new version and can't be
properly autoreconf'ed, like kmod 18+.
The file was picked up from kmod-18 itself actually.

[Peter: don't pull in host-pkgconf as it was already needed by v1]
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

cramfs: add host utilities menu entry

Signed-off-by: Eric Le Bihan <eric.le.bihan.dev@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

rt-tests: enable on MIPS uClibc again

We have fixes in the internal toolchain; an external toolchain is most
likely a buildroot-built toolchain anyway, so it will also have the
fix. But we still keep a comment to warn the user for potentially-broken
external toolchains.

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

uclibc: Make __SIGEV_PAD_SIZE to take __WORDSIZE into account

Applying an upstream patch to make __SIGEV_PAD_SIZE to take __WORDSIZE
into account for alpha, mips and ia64 arches.

Upstream patch URL:
   http://git.uclibc.org/uClibc/commit/libc/sysdeps/linux?id=b97b4b698b023f75b54f987859c856ab4861ea00

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

uclibc: add a missing function member to siginfo.h

Applying an upstream patch to add a missing function member on ia64,
mips and sparc arches.

Upstream patch URL:
   http://git.uclibc.org/uClibc/commit/libc/sysdeps/linux?id=b4e6e61e2f7c6fb4bf59f66efaa74591a2112912

Fixes:
   http://autobuild.buildroot.net/results/fa0/fa03ecc087a4b30df8b0366bb238be3d167a56d9/

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

vsftpd: Add build option to disable utmpx update code

This was modeled after a similar option for Dropbear.

The utmpx code is automatically disabled when compiling with musl,
to avoid a build error due to WTMPX_FILE being undefined. Note that
musl has an empty utmpx implementation, so no functionality is lost
by not calling it.

[Peter: use positive logic]
Signed-off-by: Maarten ter Huurne <maarten@treewalker.org>
Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

eudev: improve S10udev script for startup

This commit adjusts the S10udev script by:

 * Using the --action=add option to udevadm trigger. By default, only
   the "change" events are handled by 'udevadm trigger', which means
   it doesn't handle all the "device add" events that occured during
   the boot time, before eudev was started.

 * Adds a call to 'udevadm settle' to make sure we wait for udev to
   handle all the events that occured before eudev was started.

Both of these change match what the Ubuntu and Debian udev startup
scripts are doing.

Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Signed-off-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>

package/wpa_supplicant: security bump to version 2.3

Fix CVE-2014-3686: wpa_cli and hostapd_cli action script execution vulnerability
(http://w1.fi/security/2014-1/wpacli-action-scripts.txt)

Signed-off-by: Jörg Krause <jkrause@posteo.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

package/hostapd: security bump to version 2.3

Fix CVE-2014-3686: wpa_cli and hostapd_cli action script execution vulnerability
(http://w1.fi/security/2014-1/wpacli-action-scripts.txt)

Signed-off-by: Jörg Krause <jkrause@posteo.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

mpg132: bump to version 1.21.0

Also add hash file.

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

file: bump to version 5.20

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

libxslt: remove configure patch

It dates back to the initial libxslt submission, has no documentation and
doesn't seem to be needed.

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

libxslt: add upstream patch to fix --maxvars handling

Needed to build the manual in PDF format if the host doesn't have a fixed
xsltproc.

(by make host-libxsltproc; PATH=output/host/usr/bin:$PATH make manual-pdf)

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

docs/manual: document the asciidoc infra

[Peter: move periods outside paranthesis as suggested by Thomas De Schampheleire]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Samuel Martin <s.martin49@gmail.com>
Cc: Thomas De Schampheleire <patrickdepinguin@gmail.com>
Reviewed-by: Samuel Martin <s.martin49@gmail.com>
Reviewed-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>

Makefile: enable generating documents from br2-external

Currently, the only way br2-external may generate a document is by
including that document's recipe from within external.mk.

But external.mk is only parsed when the tree is configured.

This is unlike our internal document (the manual) which can be generated
from within an unconfigured tree.

So, include the documents from br2-external at the same time we include
our own document:
  - expect the same layout as we have:      docs/DOC_NAME/doc-name.mk
  - do not fail if there is no document:    use "-include", not "include"

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

doc/asciidoc: always use UTC

Currently, the manual is rendered with the timezone of the user running
the rendering. This timezone can fluctuate, depending on the date, due
to DST (Daylight Saving Time). Currently, the manual is rendered in
either CET or CEST (Central European Time, or its DST variant.)

So, a manual rendered during the summer or the winter would refer to a
non-constant timezone. If the machine and/or user doing the rendering
also changes, there is no guarantee the timezone would still be CET/CEST.
This is not a hard issue, since the user can still deduce the time in
UTC, but is just incoherent.

Just force the timezone to be UTC when doing the rendering of the
manual, so we are not dependent on the machine or user doing the
rendering.

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

doc/asciidoc: always apply Buildroot's AsciiDoc config

As suggested by Thomas: the AsciiDoc options we use ensure we get a sane
output of the document. We want that configuration to be applied to
other documents as well.

Up until now, it was implicit that the configuration was applied to
our manual, becasue we only supported document-specific configuration,
and the configuration we had was in our manual dir, so we got to use it.

But now, we can render other documents, especially ones from
br2-external, and we want those to also use the default configuration
from Buildroot, but still be able to provide their own customisation.

So, always add Buildroot's configuration first, if available, before we
append the document's configuration.

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

doc/asciidoc: allow documents to pass a global asciidoc configuration

Currently, a document can specify an output-specific configuration, like
the text output, that hides images references, and formats hyperlinks.

But sometimes it is required that a specific configuration is used for
all the output formats, to always apply the same markup rules in all
documents (like using the [specialwords] section.)

Also look for a file named 'asciidoc.conf' in the document's directory.
If that file exists, add it to the asciidoc options (before the
output-specific config, so the latter takes precedence over the former.)

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