libuuid needs to be linked with -lintl when gettext is needed (with
locales).
For a shared build, this is solved with the DT_NEEDED ELF tags, but for
a static link, it is not pulled in automatically.
Use pkg-config to find libuuid, instead of AC_CHECK_LIB().
Second part at fixing:
http://autobuild.buildroot.net/results/c46/c467aeec258909bb82eda77123803944f97d8df8/
[Peter: add host-pkgconf to _DEPENDENCIES]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Peter Korsgaard <jacmet@uclibc.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When gettext is needed (with locales), libuuid needs to be linked
against -lintl.
In dynamic link, this is solved via the DT_NEEDED ELF tag, but for
static links, it does not get pulled in automatically.
Fix that by adding a Libs.private section to uuid.pc, but only if
it needed.
First part at fixing:
http://autobuild.buildroot.net/results/c46/c467aeec258909bb82eda77123803944f97d8df8/
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Because shairport-sync uses AC_CHECK_LIB to find some of the needed
libraries (opensl, alsa, popt...), it fails to bring in the dependencies
of those libraries; for example, -lssl requires -lz.
When doing a dynamically-linked build, those dependent libraries are
pulled in thanks to the DT_NEEDED ELF tags, but those are not available
in static libraries.
Use PKG_CHECK_MODULES instead of AC_CHECK_LIB, and emulate the results
of AC_CJECK_LIB (i.e. set the HAVE_LIBxxx and the LIBS variable
appropriately) to minimise the amount of code touched.
This should fix:
http://autobuild.buildroot.net/results/1a5/1a56dbff3583bed0d693508dfc16859086846ecd/http://autobuild.buildroot.net/results/6d5/6d5a8157592f7a2c9469bf71b0453796ef982ab1/
At the same time, remove a comment at the end of a variable assignment,
since this has proven to cause some problems.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Jörg Krause <jkrause@posteo.de>
Cc: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When we set LD_LIBRARY_PATH when building our host tools, we append any
pre-existing value to our custom path:
LD_LIBRARY_PATH="$(HOST_DIR)/usr/lib:$(LD_LIBRARY_PATH)"
But then if LD_LIBRARY_PATH was previously empty, we end up with an
LD_LIBRARY_PATH that ends with a colon.
Also, when we check that an existing LD_LIBRARY_PATH does not contain
CWD, we previously did not look for a zero-length prefix.
Since 'man ld.so' says of LD_LIBRARY_PATH:
A colon-separated list of directories in which to search for ELF
libraries at execution-time. Similar to the PATH environment
variable.
And POSIX states about PATH:
A zero-length prefix is a legacy feature that indicates the current
working directory.
And bash also recognises a zero-length prefix to search in CWD:
A zero-length (null) directory name in the value of PATH indicates
the current directory.
We may thus end up on a system where a zero-length prefix in
LD_LIBRARY_PATH is interpreted as CWD.
Do not append the previous LD_LIBRARY_PATH if it was empty, and check
for a zero-length prefix when checking dependencies.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Peter Korsgaard <jacmet@uclibc.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Due to a bug in makefile logic, monolite libraries path will not
be selected when EXTERNAL_MCS parameter is specified.
Setting EXTERNAL_MCS to a non existent binary, makes the internal logic
to fallback to monolite, setting correctly both binary and
library paths.
This has the benefit to force host-mono to always use monolite as
bootstrap compiler also in case a mono compiler is already installed
on the system.
Fixes:
http://autobuild.buildroot.net/results/fc00952bef03cf3b4785be1a0d9c08dc84aa3f54/
[Peter: add comment explaining why this is done]
Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com>
Reported-by: Rainer Berns <rainer.berns@berns-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
First of two patches to fix
http://autobuild.buildroot.net/results/238/2386edb7f95920e84a35811a33f4333ee0a7a860/
gd links against libiconv if it is already built, depend on libiconv
to get reproducable builds.
readelf output without libiconv present:
$ output/host/opt/ext-toolchain/bfin-linux-uclibc/bin/bfin-linux-uclibc-readelf \
-a output/staging/usr/lib/libgd.a | grep iconv
15: 00000000 12 FUNC GLOBAL HIDDEN 1 _iconv_open
16: 0000000c 12 FUNC GLOBAL HIDDEN 1 _iconv
17: 00000018 12 FUNC GLOBAL HIDDEN 1 _iconv_close
readelf output with libiconv present:
$ output/host/opt/ext-toolchain/bfin-linux-uclibc/bin/bfin-linux-uclibc-readelf \
-a output/staging/usr/lib/libgd.a | grep iconv
000000e4 0000100a R_BFIN_PCREL24 00000000 _libiconv_open + 0
00000140 0000140a R_BFIN_PCREL24 00000000 _libiconv + 0
0000019a 0000160a R_BFIN_PCREL24 00000000 _libiconv_close + 0
16: 00000000 0 NOTYPE GLOBAL DEFAULT UND _libiconv_open
20: 00000000 0 NOTYPE GLOBAL DEFAULT UND _libiconv
22: 00000000 0 NOTYPE GLOBAL DEFAULT UND _libiconv_close
[Peter: also add to LIBS so it ends up in gdlib-config --libs output]
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The luajit package is a provider of a lua interpreter, but does not install a
lua executable. This is fine for scripts that explicitly invoke luajit, but
not so good for scripts that just need a lua interpreter and call lua. This
mod creates a lua symlink so that the non-jit-specific scripts will still work.
Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This patch forces the use of internal compiler for building mono,
for more information please refer to mono official documentation
(README.md)
Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The previous no-op dhclient-script is useless and in fact can lead
users to believe things will work.
dhclient-script is supposed to, among many other things, bring the
interface up for dhclient to pump an address, set the interface
address, set resolv.conf appropiately and set the default gateway.
It's a complex script since it has to deal with a lot of details so
let's just use the best fit which is the OpenWRT one since it uses the
"legacy" tools (ifconfig, route, ...) which can be easily provided by
busybox and/or net-tools.
There are newer and more feature-complete versions around but they
require full iproute2 ip and wouldn't fit as well with the busybox-only
approach.
License: likely GPLv2+
Source:
http://git.openwrt.org/?p=packages.git;a=blob;f=net/isc-dhcp/files/dhclient-script;h=4afebc0ad20ebac51c5baae5ed01c6713e3a0fd0;hb=HEAD
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes:
CVE-2014-8767 - denial of service in verbose mode using malformed OLSR
payload
OLSR payload
CVE-2014-8768 - denial of service in verbose mode using malformed Geonet
payload
CVE-2014-8769 - unreliable output using malformed AOVD payload
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Qt Webkit Module uses pthread_getattr_np() which is implemented in
toolchains with NPTL. Otherwise the compilation will fail showing this
error message:
wtf/StackBounds.cpp: In member function ‘void
WTF::StackBounds::initialize()’:
wtf/StackBounds.cpp:172:38: error: ‘pthread_getattr_np’ was not declared
in this scope
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Qt Script Module uses pthread_getattr_np() which is implemented in
toolchains with NPTL. Otherwise the compilation will fail showing this
error message:
../3rdparty/javascriptcore/JavaScriptCore/runtime/Collector.cpp: In
function ‘void* QTJSC::currentThreadStackBase()’:
../3rdparty/javascriptcore/JavaScriptCore/runtime/Collector.cpp:692:42:
error: ‘pthread_getattr_np’ was not declared in this scope
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Somehow busybox thinks debian standards are the rule and installs lspci
to /usr/bin instead of /usr/sbin where upstream says it belongs.
So install pciutils binaries there as well to really win over busybox
PATH-wise and overwrite the little bugger symlink.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Some crazy folks use MIPS machines as build machines. ;-)
On MIPS, the only acceptable hash-style is 'sysv', because the MIPS ABI
defines that the GOT ordering to be the same as the symbols ordering,
while GNU hash requires symbols to be sorted by their hash.
Looking at binutils' code, it seems that only MIPS suffers from that
limitation.
Currently, we force the toolchain wrapper to be linked with both hash
styles, which breaks on MIPS.
So, fix that by singling out MIPS, and use sysv in that case, and both
otherwise.
Reported-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
NM only currently supports isc-dhcpc and dhcpcd as clients because NM
communicates with the client through DBus, and support for clients needs to
be written in. (as per files in src/dhcp-manager)
Buildroot's default, udhcpc is not supported.
Add the dependencies to Config.in, favoring dhcpcd as
default due to size.
[Peter: drop references to dhclient, isc-dhcpc is dhclient]
Signed-off-by: Nathaniel Roach <nroach44@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This fixes the following build errors seen when assembling huge files
produced by gcc:
Error: jump target out of range; no usable trampoline found
Error: operand 1 of 'j' has out of range value '307307'
Fixes:
http://autobuild.buildroot.net/results/545/545168d9caf3bdb3dd3eb3bae58ba9db8a33384a/
Backported from: d92b6eece424f0ad35d96fdd85bf207295e8c4c3
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Not all arm (E.G. non-VFP) toolchains define the control word macros, so
disable debug handling for arm(eb) as well.
As the debug handling has limited value (it simply forces CFLAGS which we
don't want and enables exceptions on FP errors) don't try to be clever about
what toolchain variants to disable it for, but just disable it for all arm(eb).
Fixes http://autobuild.buildroot.net/results/672/672a31c0bd329dec566ba74e0e1be43abd0f16c2/
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes a buffer overflow in ksba_oid_to_str.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
clamav unconditionally includes dlfcn.h which is missing on a uClibc
that is configured as a pure-static C library.
Thus, the build fails.
But the including file does not even makes use of any function from the
dlopen() familly, so it does not need to include dlfcn.h to start with.
Add a patch to clamav to not include dlfcn.h where not needed.
Fixes:
http://autobuild.buildroot.net/results/b49/b491f4e5e1760248adb8d21b404e8aa15f7dbdd1/
[Peter: fix typo in patch description]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes an out-of-bounds memory access in png_user_version_check().
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
libshairplay uses dlopen to open libdnssd at runtime (unless dlopen() isn't
available, then static linking is used - But avahi depends on
!BR2_PREFER_STATIC_LIB so that isn't possible), so ensure avahi is available.
Fixes http://autobuild.buildroot.net/results/54d/54de48b23439cef4cbea3b8a65b20e2a8b049f67/
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Modern architectures such as ARC do not provide the sysctl() system
call, since it is deprecated. However, uClibc still installs
<sys/sysctl.h> in such cases, which defeats radvd check for the
availability of sysctl(). This commit adds a patch to radvd which
improves the sysctl() checking.
Fixes:
http://autobuild.buildroot.org/results/458/4581c4220adeaebbf6761e3b923088d8de8522d5/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
This commit renames the radvd patches to follow the new naming
convention, and reformats them to be Git patches.
Since the patches are renamed, we adjust the comment in the .mk file
explaining why we autoreconf the package.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
The ushare configure script is a custom shell script, and it is not
designed to understand that the library order is important when doing
static linking.
Fixes:
http://autobuild.buildroot.org/results/32e/32ed2a521043f929cc290145c8a651e69042e104/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
