nfs-utils: add support to use RPC support from libtirpc

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

lmbench: add support to use libtirpc when available

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

busybox: add support to link against libtirpc when available

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libtirpc: ensure getrpcby{number,name} are provided

Needed by E.G. xinetd.

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libtirpc: new package

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

toolchain-external: ARM 2012.03 has RPC support

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

pkg-stats: fix the broken "results" link

At the top of the output html page there is a dangling "results" link.

Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

auto{conf,make}: only make available on archs supported by qemu

Now that we're using full perl (which needs qemu), only make auto{conf,make}
available on the supported archs.

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

qemu: needs libglib2

Build error message is:
  glib-2.12 required to compile QEMU

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

qemu: needs zlib

Build error message is:
  Error: zlib check failed
  Make sure to have the zlib libs and headers installed.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

qemu: needs pkg-config

Fixes

  http://autobuild.buildroot.org/results/ccb7d6da503a81015275ff1e4ba9a564203932ec/build-end.log

Even though the build failure happens on host-qemu, adding
host-pkgconf to QEMU_DEPENDENCIES seems like the right thing to do
since that this dependency will also be needed for the target qemu
once it gets included.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

package: drop games/ subdir

As discussed during the dev day.

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

legal-info: split _LICENSE and _REDISTRIBUTE constants

Currently packages can be described in two ways: proprietary (tarball not
saved, license not described in further detail), and others (tarball
saved, license described).

Split the logic to allow the license to be always described whether or not
the source code can be redistributed.

Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

legal-info: move $(2)_LICENSE to a proper line

Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

ulogd: fix build failure

Fix Peter's reported build failure:
http://autobuild.buildroot.net/results/1dcf63799937f2216e3c712af8880ad431dbd47c/build-end.log
This is because the configure script uses absolute paths to detect dbi,
mysql and pgsql, so forcibly disable them.
Also add the optional sqlite dependency when it's selected.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

legal-info: support per-package hooks

Useful to produce extra warnings for packages that have special
licensing-related issues.

Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

QT: Fix build for architectures using generic atomics

The problem has to do with const-correctness. This has been resolved
for various architectures, but not for the generic case.

Signed-off-by: Mischa Jonker <mischa.jonker@synopsys.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

system: fixup default device table locations after system/ got added

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

Split target/Config.in.arch into multiple Config.in.* in arch/

target/Config.in.arch had become too long, and we want to remove the
target/ directory. So let's move it to arch/ and split it this way:

 * An initial Config.in that lists the top-level architecture, and
   sources the arch-specific Config.in.<arch> files, as well as
   Config.in.common (see below)

 * One Config.in.<arch> per architecture, listing the CPU families,
   ABI choices, etc.

 * One Config.in.common that defines the gcc mtune, march, mcpu values
   and other hidden options.

[Peter: space->tab fix, mipsel64 little endian, mips3 as noted by Arnout]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

New top-level directory: system

This directory groups the following elements:
 * the default root filesystem skeleton
 * the default device tables
 * the Config.in options for system configuration (UART port for
   getty, system hostname, etc.)
 * the make rules to apply the system configuration options

Even though the skeleton and device tables could have lived in fs/, it
would have been strange to have the UART, system hostname and other
related options into fs/. A new system/ directory makes more sense.

As a consequence, this patch also removes target/Makefile.in, which
has become useless in the process.

[Peter: fixup TARGET_SKELETON settings / documentation to match]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Yann E. MORIN <yann.morin.1998@free.fr>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

sqlite: Add new compile-time options

[Peter: whitespace/wording tweaks]
Signed-off-by: Stefan Froberg <stefan.froberg@petroprogram.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

cups: fix libdir handling

Cups uses /usr/lib64 if the HOST has it, no matter if it is cross compiling
for a 32bit arch, breaking the build for stuff looking in /usr/lib.

The fix of commit edd2716c didn't work, it would just force /usr/lib64 if
the target is 64 bit.  Instead, force installation in /usr/lib regardless
of the host.

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

links: Enable graphics for X11

Graphics can be enabled for X11 as well.
Do not select DirectFB for graphics,
set DirectFB or X11 dependencies instead.

Signed-off-by: Valentine Barshak <gvaxon@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libnl: bump to version 3.2.14

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

bison: bump to version 2.6.3

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

xapp_xsm: fix build against Xmu by bumping version

Fix a build issue similar to the one encountered on xapp_listres at
http://autobuild.buildroot.org/results/06d0ef38165769df38047c1f64d4cfae835e9e54/.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

xapp_xmag: fix build against Xmu by bumping version

Fix a build issue similar to the one encountered on xapp_listres at
http://autobuild.buildroot.org/results/06d0ef38165769df38047c1f64d4cfae835e9e54/.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

xapp_xgc: fix build against Xmu by bumping version

Fix a build issue similar to the one encountered on xapp_listres at
http://autobuild.buildroot.org/results/06d0ef38165769df38047c1f64d4cfae835e9e54/.

Note that this new version requires a minor cross-compilation fix in
configure.ac, which has already been committed upstream but is not yet
part of a release tarball.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

xapp_xfontsel: fix build against Xmu by bumping version

Fix a build issue similar to the one encountered on xapp_listres at
http://autobuild.buildroot.org/results/06d0ef38165769df38047c1f64d4cfae835e9e54/.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

xapp_xedit: fix build against Xmu by bumping version

Fix a build issue similar to the one encountered on xapp_listres at
http://autobuild.buildroot.org/results/06d0ef38165769df38047c1f64d4cfae835e9e54/.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

xapp_xbiff: fix build against Xmu by bumping version

Fix a build issue similar to the one encountered on xapp_listres at
http://autobuild.buildroot.org/results/06d0ef38165769df38047c1f64d4cfae835e9e54/.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

xapp_listres: fix build against Xmu by bumping the version

Fixes

  http://autobuild.buildroot.org/results/06d0ef38165769df38047c1f64d4cfae835e9e54

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

opus-tools: fix build on i386

Upstream patch.  Fixes e.g.
http://autobuild.buildroot.net/results/03cf0ee3c092960c37a6a78f2531b047a317d516/

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

toolchain-external: improve glibc support to test availability of RPC

Basically, the BR2_TOOLCHAIN_EXTERNAL_GLIBC option no longer
unconditionally selects BR2_TOOLCHAIN_HAS_NATIVE_RPC since there are
glibc toolchains that don't have RPC support. All the predefined
toolchain profiles are updated to take into account this change: for
the moment, all glibc toolchains that have pre-defined toolchains have
RPC support, but further patches in the series add pre-defined glibc
toolchains that don't have RPC support. In the case of custom glibc
toolchains, a question is asked to the user so that he can say whether
the external glibc toolchain has RPC support or not. The validity of
this configuration option is checked by the new
check_glibc_rpc_feature function in helpers.mk.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

netkittelnet: mark as deprecated

It is no longer being developed, the FTP is dead, and Debian has
removed it from the distribution.

Also it allows to not worry about adding RPC support through libtirpc
to this package.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

netkitbase: mark as deprecated

It is no longer being developed, the FTP is dead, and Debian has
removed it from the distribution.

Also it allows to not worry about adding RPC support through libtirpc
to this package.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

Rename BR2_INET_RPC to BR2_TOOLCHAIN_HAS_NATIVE_RPC

The BR2_INET_RPC has for a long time been a not very descriptive
configuration option name, and with the advent of non-RPC glibc
toolchains and the apparition of libtirpc, we really need to rename it
to something more sensible, BR2_TOOLCHAIN_HAS_NATIVE_RPC.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

toolchain-crosstool-ng: use the Crosstool-ng config options instead of the common hidden ones

Since we are some day going to finally rename the badly named common
toolchain options (BR2_USE_WCHAR, BR2_ENABLE_LOCALE, BR2_INET_RPC,
etc.) into something more logical, let's start using the Crosstool-NG
toolchain options in the Crosstool-NG code.

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 <jacmet@sunsite.dk>

busybox: don't force FEATURE_NFS_MOUNT

Instead of making the Busybox configuration more complicated by trying
to adjust it depending on whether RPC is available or not (which gets
complicated when RPC support can be provided by libtirpc), simplify
things by letting the user enable FEATURE_NFS_MOUNT or not depending
on whether RPC support is available or not.

Our default configuration do not enable FEATURE_NFS_MOUNT, so users
will not face any build problems by default. Only if they explicitly
enable FEATURE_NFS_MOUNT will they have to make sure that the
toolchain has RPC support, or that libtirpc is enabled (support for
this added in a followup patch).

[Peter: remove from CONFIGURE_CMDS as well]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

uClibc: use the Buildroot toolchain options instead of the hidden common options

Since we are some day going to finally rename the badly named common
toolchain options (BR2_USE_WCHAR, BR2_ENABLE_LOCALE, BR2_INET_RPC,
etc.) into something more logical, let's start using the Buildroot
toolchain options in the uClibc code.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libhid: disable on avr32

libhid uses timerfd, which is not yet available in the uClibc-0.9.31
that we have to rely on for avr32.  Since this is pretty much a corner
case, just disable libhid for avr32.

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

pkg-download: Make a shallow clone for git downloads

When downloading from git, we clone the whole repository and then only
use the latest commit.  That's a lot of redundant stuff.  So instead,
make a shallow clone.  Unfortunately that's only possible when
downloading a branch or tag, so fall back to the old method if git gives
an error.

This speeds up the cloning of a linux git from more than 2 hours to
20 minutes on a 200KB/s link).

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Acked-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

arptables: add new package

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

ulogd: add new package

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

conntrack-tools: bump to version 1.4.0

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

nfacct: add new package

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libnetfilter_log: add new package

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libmnl: add license info

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libnetfilter_acct: add new package

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libnetfilter_queue: add new package

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libnetfilter_cthelper: add new package

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libnfnetlink: bump to version 1.0.1

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libnetfilter_cttimeout: add license info

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libnetfilter_conntrack: bump to version 1.0.2

[Peter: adjust dnsmasq dependency]
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

cmake: bump version to 2.8.9

Signed-off-by: Thiago A. Correa <thiago.correa@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

qextserialport: new package

[Peter: adjust for BR changes]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

use avr32linux.org mirror

The avr32linux.org mirror has vanished and is replaced by a spam site
that returns HTML pages for whatever URL you request from it.  So the
download helper thinks that download has succeeded, while actually it
failed.

Fortunately, there is still a mirror of the site alive, so we can use
that one.

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

gdb: adjust behavior when thread debug is not available

When an external toolchain without thread debug is used, the gdb
package can be selected, but no version can be choosen, since none
match any of the requirements. This leads Buildroot to try to build
gdb for the target without a version being defined, as in the
following build log:

 http://autobuild.buildroot.org/results/84e8fd2df0cc22448052a572c2e9a6e03dd137eb/build-end.log

To fix this, we adjust the dependencies of the BR2_PACKAGE_GDB option
so that the package as a whole is not selectable when the required
conditions are not met. Basically, we have the choice of:

 * Having a toolchain that supports thread debugging, which is needed
   for gdb >= 7.x

 * Having BR2_DEPRECATED enabled, which allows gdb 6.8 to be selected,
   which doesn't require thread debugging

 * Using bfin, since this architectures has a special old gdb version

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

barebox: Allow to pass a custom configuration file

Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

cpanminus: new package

Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

iproute2: fix em_ipset build issue

Fixes http://autobuild.buildroot.net/results/be19e39c19b62946ea48dd6b3dfd55f5c3ccd7c9

em_ipset.c uses xtables.h from iptables, but the configure script fails
to to check for this, so forcibly disable it if not available.

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

package: don't hide webservers when !BR2_PACKAGE_BUSYBOX_SHOW_OTHERS

The busybox web server isn't providing the same binary name / cmdline
arguments as the "big" webservers, so they aren't equivalent.

As discussed on the dev day, don't hide them when _SHOW_OTHERS isn't
enabled.

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

perl: prepare the removal of microperl

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

perl: add option "custom install"

allows same behaviour (and footprint) as microperl

Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

perl: add GDBM_File

Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

gdbm: new package

Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

perl: add DB_File

Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

perl: new package

Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

host-qemu: new package

only user mode for the current target architecture

Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

lcdapi: fix autobuilder failures

Version 0.3 of the lcdapi package does not correctly handle CXXFLAGS passed
in from the environment/command line. In particular, the passing in of CXXFLAGS
clobbers the -fPIC flag which is set in the top-level lcdapi makefile.

The following autobuild logs highlight the problem.

  http://autobuild.buildroot.net/results/9a47b4a70d29308f668d65d983ba6146c0c56365/build-end.log

  http://autobuild.buildroot.net/results/4a4f6e2e7f80d5a23c9d26069094e9dafb1468e7/build-end.log

  http://autobuild.buildroot.net/results/0d96c207a9db44cf2791a66cad5162a739e7053a/build-end.log

This has been fixed upstream in version 0.4; this patch simply bumps the
version number in the Buildroot lcdapi package.

Signed-off-by: Simon Dawson <spdawson@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

iproute2: fix build with toolchains providing setns(2)

Fixes http://autobuild.buildroot.net/results/b2f28621bc9d6048057c406048b3f7d665cef165

ipnetns.c contains a replacement setns implementation, which is used
if the toolchain doesn't provide any (HAVE_SETNS not set).

We don't have any knowledge of toolchain setns support on buildroot
level, but the (handwritten) configure script contains a test for it,
so run the configure script before building.

The configure script isn't written for cross compilation, so it needs
to be massaged slightly to use the cross compiler / flags.

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

gdisk: close(2) needs unistd.h

Fixes http://autobuild.buildroot.net/results/c557ee7cd1247e024b5edccf5f824eb4d5212e25

Fixed upstream as well in the (big) 0.8.3 commit:
http://gptfdisk.git.sourceforge.net/git/gitweb.cgi?p=gptfdisk/gptfdisk;a=commit;h=3488294d718a0e8b7f312c80c9e5729671173f6a

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

dbus: bump version

Fixes security issues:
- CVE-2012-3524: Don't access environment variables (fd.o #52202)
  Thanks to work and input from Colin Walters, Simon McVittie,
  Geoffrey Thomas, and others.

- Be more careful about monotonic time vs. real time, fixing
  DBUS_COOKIE_SHA1 spec-compliance (fd.o #48580, David Zeuthen)

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

at91sam92*: update defconfigs to use latest u-boot

Signed-off-by: Alexandre Belloni <alexandre.belloni@piout.net>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

u-boot: update at91 patches to match mainline

The patch that was included in buildroot was not the same as the final
one submitted in the mainline.

Signed-off-by: Alexandre Belloni <alexandre.belloni@piout.net>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

at91bootstrap: update patch for latest u-boot

The u-boot patch got mainlined but not exactly as it was submitted to
buildroot. Update the at91bootstrap to match what is in the mainline.

Signed-off-by: Alexandre Belloni <alexandre.belloni@piout.net>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

linux: bump 3.6.x stable version

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

kernel-headers: bump 3.{0, 2, 4, 6}.x stable versions

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

Add default configuration for the AArch64 Foundation V8 simulator

Note that this configuration doesn't build entirely: the user must
manually run "make busybox-menuconfig" and disable the "Mount NFS
filesystems" option, because the toolchain does not have RPC support.

This issue will be fixed once the support for toolchain without RPC
will be integrated.

[Peter: fix readme typo]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

boot-wrapper-aarch64: new package

The boot wrapper contains both a minimal bootloader and a tool to
generate an image suitable for execution by the AArch64 software
simulator. The image generated embeds the minimal bootloader, the
kernel image, the Device Tree Blob and the kernel command line.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

Support KERNEL_ARCH for AArch64

The architecture tuple is 'aarch64', but the kernel people decided to
call it 'arm64', so we have to do some mungling to get the kernel
architecture name from the Buildroot architecture name.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

toolchain: make external toolchain the default for AArch64

Our internal toolchain backend does not yet have support for AArch64,
and Crosstool-NG also does not have support for AArch64 at the moment
(though it should be coming quickly since the Linaro AArch64 toolchain
is generated with a modified Crosstool-NG version).

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

toolchain-external: add aarch64 external toolchain

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

Add AArch64 to the list of architectures

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

mysql_client: bump version to 5.1.65

Signed-off-by: Sagaert Johan <sagaert.johan@skynet.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

liblog4c-localtime: fix underquoted path bug.

This patch fixes the autobuilder bug that shows up during the
libmbus build.

Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

sconeserver: fix bluetooth module build problem

This fixes a build problem in the sconeserver bluetooth module, as reported
by the Buildroot autobuild system:

  http://autobuild.buildroot.net/results/d1133da3f8d74c58fd271e0cab197a1108000e90/build-end.log

The build problem, due to the use of BDADDR_ANY, has been fixed upstream;
consequently, this patch simply bumps the sconeserver svn revision.

Further details of the underlying problem at the following URL.

  http://stackoverflow.com/questions/9751710/c-c-warning-address-of-temporary-with-bdaddr-any-bluetooth-library

Signed-off-by: Simon Dawson <spdawson@gmail.com>
Acked-by: Samuel Martin <s.martin49@gmail.com>
Tested-by: Samuel Martin <s.martin49@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

fluxbox: select xmodmap

Fluxbox uses xmodmap in its startup script.

Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

rtorrent: bump to version 0.9.3

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

libtorrent: bump to version 0.13.3

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

samba: bump to version 3.6.9

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

php: bump to version 5.3.18

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

lm-sensors: bump to 3.3.2

Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

xz: bump to 5.0.4

Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

strace: fix build for x86_64

strace fails to build on x86_64 because stat64 is not available.  This
is because the automatic detection of stat64 in configure is overridden
by buildroot, by setting ac_cv_type_stat64.  Just remove that override -
current strace seems to detect it correctly for non-largefile platforms.

Build-tested on x86_64 (with largefile), ARM (with and without largefile),
sh4, MIPS and ppc-32 (no largefile).

Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

qtuio: fix build/link issue when Qt is built with debug support

Fixes http://autobuild.buildroot.net/results/6dcbf4ca3fcc5ec911ac7a1680f7cb381fd18fcb

qtuio appends _d to the library name when Qt was built with debug support,
breaking linking step for examples and staging/target install.

There's no real advantage to the _d suffix, so simply fix it by removing
the logic appending _d.

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

squid: fix kerberos build failure

The configure script tries to autodetect kerberos 5 support and seems to
fail under some circumnstances.
Just disable it since we don't have kerberos support in buildroot.
Fixes:
http://autobuild.buildroot.net/results/c9e590a1034c3869cb7391984ce8281ec4c3ef9c/build-end.log

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

squid: bump to version 3.2.3

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

docs/manual: update to host-pkgconf

Update the manual to match now that we have moved to pkgconf instead of
pkg-config.

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>

iw: pkg-config cleanup

We no longer need to specify PKG_CONFIG_PATH since it's handled by the
host-pkgconf wrapper.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>