Some architectures may need updates to the configure helper scripts
before it is possible to safely configure.
Signed-off-by: Will Newton <will.newton@gmail.com>
Add a new target for libuuid so it can be built without e2fsprogs.
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Recent glibc versions introduced a futimens() function, that gzip
already implemented. So waiting for a proper usage of this function in
gzip, we rename the gzip internal function to gz_futimens().
Patch inspired by
http://lists.pld-linux.org/mailman/pipermail/pld-cvs-commit/Week-of-Mon-20070514/155476.html
Fixes bug #537
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since CC is already passed in TARGET_CONFIGURE_OPTS at ./configure
time, there's no point in overriding it with a wrong value at
compilation time. This breaks external toolchain builds since
CC=$(TARGET_CC) is wrong because it excludes the mandatory --sysroot
option.
Fixes bug #525
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Fix the little build systems hacks that no longer applied to the
version of Perl being used. With these fixes in place, microperl
builds and runs properly simple Perl commands on the target.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since the recent update of Kconfig to the Kconfig infrastructure found
in 2.6.30 kernels, make oldconfig was broken because it didn't
generate the .config.cmd file. This is done by a call to the
conf_write_autoconf() function.
Moreover, this process was also broken because the
conf_get_autoconfig_name() function was reading the environment
variable BUILDROOT_AUTOCONF while the main Makefile is passing it as
KCONFIG_AUTOCONFIG. We also remove the fallback on
"$(BR2_DEPENDS_DIR)/config/auto.conf" since we are not able to expand
BR2_DEPENDS_DIR at this point. Therefore, the KCONFIG_AUTOCONFIG
envionment variable *must* be set.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Closes#289
This package is a wireless configuration utility for the linux mac80211
kernel stack.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Tested-By: Will Newton <will.newton@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Closes#437
- Remove a few obsolete patches.
- Add a patch to disable syscall tests unsupported by uClibc.
- Add a patch to fix sched_getaffinity build.
Signed-off-by: Will Newton <will.newton@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Closes#427
- Fix generation of configure script and Makefiles.
- Add dependencies on libsoup and enchant.
- Allow configuration of rendering target (currently X11 or DirectFB).
Signed-off-by: Will Newton <will.newton@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The installation to the target directory was broken for two reasons:
* Due to how the lvm2.mk file was written, for each file listed in
$(LVM2_TARGET_SBINS) and $(LVM2_TARGET_DMSETUP_SBINS), it was
reinstalling all the files listed in those variables. When a target
contains several elements, there's no need to make a loop to handle
each of the elements: the rule will get executed as many times as
they are elements in the target.
* For some odd reason, lvm2 build system installs executable with 555
and libraries with 444. Therefore, once copied to the target
directory, these executables and libraries cannot be
overwritten... for example by a later execution of Buildroot. We
fix this up by changing the permissions of the binaries and
libraries installed by lvm2 to saner values (755 and 644)
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Buildroot Makefile to compile pciutils wasn't passing TARGET_LDFLAGS
to pciutils build system. Therefore, at link time, the pciutils build
system was not taking into account the important --sysroot option,
which was breaking the link with external toolchains.
Fixes bug #523.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Closes#247
wide collection of NTFS utilities from http://www.linux-ntfs.org/
[Peter: misc Config.in cleanups/fixes]
Signed-off-by: Olaf Rempel <razzor@kopf-tisch.de>
Tested-By: Will Newton <will.newton@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Coreutils 6.9 was broken with glibc >= 2.6, due to a coreutils
internal function being named like a glibc function. This has been
fixed in more recent coreutils version, by
http://lists.pld-linux.org/mailman/pipermail/pld-cvs-commit/Week-of-Mon-20070514/155466.html.
Therefore, we upgrade coreutils to its latest version, 7.4, which
raised two problems:
* Recent coreutils releases are not anymore available as .bz2
archives, only .xz archives. Since this archive format is not
supported by Buildroot yet, and the corresponding tools are not
widely available yet, we fallback to the bigger .gz format for the
coreutils package.
* The rename bug detection script m4/rename.m4 was broken, leading
coreutils to try to include windows.h and compile some
Windows-specific code. We introduce a patch to fix this, patch
which has been taken from gnulib. We also make sure that this
workaround is nevery compiled in by passing
gl_cv_func_rename_dest_exists_bug=no to the configure script.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
BR2_UCLIBC_PROGRAM_INVOCATION is a toolchain configuration option,
like BR2_INET_IPV6, BR2_INET_RPC, on which some packages
depend. Therefore, it should be handled like BR2_INET_IPV6 and
BR2_INET_RPC in order to work properly with external toolchains.
Since we move it out of toolchain/uClibc/Config.in into
toolchain/Config.in.2, we rename the option to BR2_PROGRAM_INVOCATION
(since BR2_INET_RPC and others don't have UCLIBC in their name).
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The update of kconfig and the addition of xconfig support generates
more files in package/config/. Tell git to ignore them.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Closes#491
- Deletion of reject files no longer required as they have gone from
the upstream release.
Signed-off-by: Will Newton <will.newton@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The AUTO_CONFIGURE_OPTS is not used anywhere in the tree. autoconf
based packages should used the Makefile.autotools.in machinery
instead.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The cross-compile fix included in mainline of libXt between 1.0.5 and
1.0.6 is not sufficient. It allows to pass a different compiler for
the host tools, but still uses the TARGET_CFLAGS/TARGET_LDFLAGS to
compile the host tools, which obviously isn't correct.
The new patch allows to use the CFLAGS_FOR_BUILD/LDFLAGS_FOR_BUILD
variables as the CFLAGS/LDFLAGS to build the host tools.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This version bump is required by the bump of xlib_libXt from 1.0.5 to
1.0.6 (1.0.6 requires at least xutil/macros 1.2).
The linuxdoc patch is removed, since it has been merged upstream, at
http://cgit.freedesktop.org/xorg/util/macros/commit/?id=e20c503d24e5ccf03035aa0f6d833d235d2179df
The other patch exposing STAGING_DIR is kept, as it is Buildroot
specific. It has just been updated to resolve conflicts.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Closes#243.
CTorrent is a BitTorrent client implemented in C++
to be lightweight and quick.
This is the enhanced version from
http://www.rahul.net/dholmes/ctorrent/
[Peter: minor kconfig tweaks]
Signed-off-by: Olaf Rempel <razzor@kopf-tisch.de>
Tested-by: Will Newton <will.newton@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Closes#451
Move from old unmaintained dosfstools-2.11 (packaged as mkdosfs) to
now maintained dosfstools-3.0.3.
Allow each binary to be installed independently.
Signed-off-by: Mike Crowe <mac@mcrowe.com>
Tested-by: Will Newton <will.newton@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
It involved:
* Modifying the top-level Makefile to create the xconfig target, and
the $(CONFIG)/qconf target to compile the qconf utility
* Modifying the package/config/Makefile magic so that C++ files and
binaries gets linked properly, re-using the
package/config/Makefile.kconfig logic inherited from the kernel
* Hacking a little bit package/config/Makefile.kconfig in a way
similar to the modifications done for the mconf (menuconfig
interface)
Signed-off-by: Alper Yildirim <ayildirim@aselsan.com>
Running a simple Qt GUI application on an embedded systems without
QtNetwork installed doesn't work, the dynamic loader complains that
QtNetwork is not present.
While I'm not a Qt build system expert, it looks like this dependency
really exists, since the src/gui/gui.pri file in Qt sources contains
the following line:
embedded: QT += network
Which would mean that the GUI module needs the network module to
operate properly. And a few files in src/gui/embedded/ also include
some QtNetwork stuff. This isn't really surprising since the Qt QWS
mechanism uses communication between processes through a Unix socket.
Therefore, we add a dependency on the network module in the GUI
module. We use a 'select' dependency here, since it's might not be
really obvious to the user that the GUI module requires the network
one.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Set QMAKE_LFLAGS so that at link time --sysroot is passed to the
linker, and Qt finds the external libraries. This fixes Qt build when
zlib or other libraries are external, instead of using Qt's internal
copy.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Closes#287.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Tested-by: Will Newton <will.newton@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Put the Qt options into a submenu to make the configuration interface
easier to use. The comment on the dependency on C++ has to be put
before the menuconfig definition, otherwise kconfig gets confused and
doesn't put the suboptions into a submenu.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
A while ago, Trolltech (now part of Nokia) included Qtopia directly
inside Qt. So let's rename the qtopia4 package to simply qt, which
will be more familiar to our users. Of course, the version we compile
is Qt for Embedded Linux.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The XSERVER variable used to be defined by package/Makefile.in because
the X server package name was different depending on the type of X
server that was choosen. Nowadays, the name of the package is always
xserver_xorg-server, so there's no point in having this XSERVER
intermediate variable.
This patch makes all packages use xserver_xorg-server directly as a
dependency, and removes the XSERVER variable from package/Makefile.in.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The option BR2_PACKAGE_TINY, BR2_PACKAGE_XORG and BR2_PACKAGE_XGGI
don't exist, so do not use them to find the value of the $(XSERVER)
variable.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The BR2_PACKAGE_XSERVER_none option has gone, so use BR2_PACKAGE_XORG7
instead. It is more logical to test if BR2_PACKAGE_XORG7 is set to
know if an X.org server is available, than testing if
BR2_PACKAGE_XSERVER_none is set to know if an X.org server is *not*
available.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Currently, the X.org server selection always trigger the compilation
and installation of libXt (the X toolkit) and libXaw (Athena Widget
toolkits), which is are legacy toolkits not used by modern toolkits
(Gtk, Qt, etc.). They are only used by old X11 applications that are
typically not used on embedded systems.
Not compiling them save compilation time, and more importantly, space
on the target.
We also remove the compilation on xlib_libxkbui, libxmu and libxpm,
because they all depend on libXt.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reorganize the list of depencies with a sane indentation level, and
one dependency per line. Yes, the list is long, but is easier to read,
and future patches modifying this list will be easier to review.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
A dependency on BR2_PACKAGE_XORG7 is enough, no need to depend on
BR2_PACKAGE_XORG_XSERVER.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Use select independs of depends on for library dependencies. We keep
the depends on for the X.org dependency so that dillo can only be
selected once X.org is enabled (which, contrary to the library
dependencies, is a pretty obvious dependency for the user). We then
get rid of the configuration interface comment.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The dependency of Midori on X.org should be pretty obvious to the
user, and we don't want to unnecessarly clutter the configuration
interface with comments.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
LibMMS is a common library for parsing mms:// and mmsh:// type network
streams.
Signed-off-by: Sven Neumann <s.neumann@raumfeld.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
freetype is only used by examples, but if it is not found
and the host has a freetype-config script, then the host
include dirs are added to the search path causing trouble.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Explictly disable the build of the aasink plug-in. Without this fix
the plug-in would be built if the aalib headers are found on the host.
Signed-off-by: Sven Neumann <s.neumann@raumfeld.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Bump SaWMan to 1.4.1
The current version of sawman (1.2.0-rc1) won't compile with directFB >= 1.4
Signed-off-by: Lionel Landwerlin <lionel.landwerlin@openwide.fr>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
makedepend is a program, not a library - So makedepend for the target
should get installed to target_dir and not to staging_dir.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Since the move of many host binaries to $(HOST_DIR), the TARGET_PATH
for internal toolchain had been changed to include $(HOST_DIR)/bin and
$(HOST_DIR)/usr/bin, but this change was not reflected to the
TARGET_PATH used in external toolchain configuration. This patches
fixes this.
We also remove $(STAGING_DIR)/bin and $(STAGING_DIR)/usr/bin from the
TARGET_PATH in external toolchain configuration, since the STAGING_DIR
is not supposed to contain host binaries.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Many packages used to depend on BR2_PACKAGE_XORG||BR2_PACKAGE_XORG7,
but this is useless since BR2_PACKAGE_XORG is a non-existing
configuration option. So, these depencies gets simplified to
BR2_PACKAGE_XORG7 only.
Some others were depending on BR2_PACKAGE_TINYX (which doesn't) exist
or BR2_PACKAGE_XSERVER_xorg || BR2_PACKAGE_XSERVER_tiny ||
BR2_PACKAGE_XSERVER_x11r7. Replace all that mess by a simple
dependency on BR2_PACKAGE_XORG7.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The selection of the X.org server type was really strange. The user
had to select between none, tinyx and modular. Now, the menuconfig
interface display a single item for X.org, that can be
selected/unselected. This entry contains a submenu, that allows to
select the type of the X.org server (tinyX or modular) and that allows
(as before) to enable/disable X.org libraries, drivers, etc.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Bump the version of Direct FB to 1.4.1. Fixes bug #467. Thanks to
Lionel Landwerlin <lionel.landwerlin@openwide.fr> for the
notification.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Closes#387.
When tremor is selected, the normal libvorbis shouldn't get built and
tremer should get installed into staging_dir as well.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Ruby 1.9.1 handle cross compilation better, espescially extensions
(I'm working on ARM at91sam9263-ek).
This patch also disable ruby doc generation and ipv6.
[Peter: fixed ipv6 handling]
Signed-off-by: laurent laffont <laurent.laffont@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Enchant is a spell checking library that provides a consistent API across
a number of spell-checking system backends.
Signed-off-by: Will Newton <will.newton@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The fixes made to qmake.conf for the cross-compiling tools were not
sufficient in the case of an external toolchain, since the CFLAGS were
not taken into account. And the --sysroot option contained in the
CFLAGS are very important for external toolchains to work.
Additionally, it makes sure that Qtopia is compiled with the proper
CFLAGS, even if an external toolchain isn't used.
Closes#417,
New package spawn-fcgi, basically a FastCGI process spawner.
This was previously included in lighttpd up to version 1.4.22 and is now
a separate project.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Closes#419.
lighttpd 1.4.23 is out, there are many bugfixes from 1.4.20 currently in
buildroot:
* Fix workaround for incorrect path info/scriptname if fastcgi prefix is
”/” (fixes#729)
* Finally removed spawn-fcgi
* Fix bug with FastCGI request id overflow under high load; just use always
id 1 as we don’t use multiplexing. (thx jgray)
* Workaround broken operating systems: check for trailing ’/’ in filenames
(fixes#1989)
Plus fixes from 1.4.22 and 1.4.21.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Several issues were encountered when trying to build Matchbox. This
patches fixes these issues :
* Selecting BR2_PACKAGE_MATCHBOX should select BR2_PACKAGE_XORG7,
otherwise the x11r7 mk files are not included, and matchbox.mk
refers to non-existing dependencies (such as xlib_libXtst).
* In matchbox.mk, wrong dependencies for Matchbox keyboard. When
pango is available, use it through --enable-pango, otherwise, add
the dependency to xlib_libXft
* Enable AUTORECONF for libXfont and libXft, otherwise the libtool
Buildroot patch has no effect, and these packages fail to build
without the libtool hack.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The patch introduced at c579e32dbe was
incorrect, since __USE_GNU is an internal C library defined, and
should not be used by programs. _GNU_SOURCE should be used instead.
This fixes a build failure on Dropbear when compiling against a
glibc-based toolchain.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Closed#333.
* Refactor the internal representation of SQL expressions so that they
use less memory on embedded platforms.
* Reduce the amount of stack space used
* Fix an 64-bit alignment bug on HP/UX and Sparc
* The sqlite3_create_function() family of interfaces now return
SQLITE_MISUSE instead of SQLITE_ERROR when passed invalid parameter
combinations.
* When new tables are created using CREATE TABLE ... AS SELECT ... the
datatype of the columns is the simplified SQLite datatype (TEXT, INT,
REAL, NUMERIC, or BLOB) instead of a copy of the original datatype
from the source table.
* Resolve race conditions when checking for a hot rollback journal.
* The sqlite3_shutdown() interface frees all mutexes under windows.
* Enhanced robustness against corrupt database files
* Continuing improvements to the test suite and fixes to obscure bugs
and inconsistencies that the test suite improvements are uncovering.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Closes#379
update DirectFB to version 1.4.0
A new stable version of DirectFB has been released upstream,
http://directfb.org/index.php?path=Main%2FNews&entry=2009-06-04-1.niels
Attached patch updates the DirectFB package and introduces a BR2
configuration option for the RGB16 dithering feature that is new in
DirectFB 1.4.0.
Signed-off-by: Sven Neumann <s.neumann@raumfeld.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Closes#349.
Attached patch updates libsoup to version 2.26.2.
This is relevant for bug 303 as it allows to compile gvfs with much
less patching.
Signed-off-by: Sven Neumann <s.neumann@raumfeld.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Add support for configuring security context support to allow building for
non-SELinux targets.
Remove some obsolete patches.
Signed-off-by: Will Newton <will.newton@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Closes#409.
php-5.2.10 is out, mostly misc bugfixes, one small security fix:
Security Enhancements and Fixes in PHP 5.2.10:
* Fixed bug #48378 (exif_read_data() segfaults on certain corrupted .jpeg
files). (Pierre)
Key enhancements in PHP 5.2.10 include:
* Added "ignore_errors" option to http fopen wrapper. (David Zulke, Sara)
* Fixed memory corruptions while reading properties of zip files. (Ilia)
* Fixed memory leak in ob_get_clean/ob_get_flush. (Christian)
* Fixed segfault on invalid session.save_path. (Hannes)
* Fixed leaks in imap when a mail_criteria is used. (Pierre)
* Changed default value of array_unique()'s optional sorting type parameter
back to SORT_STRING to fix backwards compatibility breakage introduced in PHP
5.2.9. (Moriyoshi)
* Fixed bug #47940 (memory leaks in imap_body). (Pierre, Jake Levitt)
* Fixed bug #47903 ("@" operator does not work with string offsets).
(Felipe)
* Fixed bug #47644 (Valid integers are truncated with json_decode()).
(Scott)
* Fixed bug #47564 (unpacking unsigned long 32bit big endian returns wrong
result). (Ilia)
* Fixed bug #47365 (ip2long() may allow some invalid values on certain
64bit systems).
* Over 100 bug fixes.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Override nfs mount support based on BR2_INET_RPC setting, so BR will DTRT.
This way you will automatically get NFS support in busybox if available in
the toolchain, and the build won't break if the defconfig enables it but
the user has it disabled in the toolchain.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
As device-mapper has moved to lvm2, dmraid must now require lvm2.
Signed-off-by: Nigel Kukard <nkukard@lbsd.net>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Also add a config option for the app plugin that was moved from
to gst-plugins-base from gst-plugins-bad in version 0.10.22.
Closes#393.
Signed-off-by: Will Newton <will.newton@gmail.com>
gstreamer uses libxml2 to support the plugin registry and pipeline
load/save features. This change allows the user to build gstreamer without
libxml2 which saves a large amount of space on the target.
Closes#373.
Signed-off-by: Will Newton <will.newton@gmail.com>
* Upgrade to version 0.98
* Use --disable-Werror, otherwise it doesn't build since warnings are
treated as errors
* Don't depend on tiff, which isn't mandatory, and doesn't
build. Once tiff is fixed, we can re-add it as a dependency.
dmraid is hard coded with -L$(DESTDIR)$(libdir) which tries to link in
the host systems' libs
Signed-off-by: Nigel Kukard <nkukard@lbsd.net>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
neon should depend on host-pkgconfig rather than pkgconfig.
Closes#385.
Signed-off-by: Will Newton <will.newton@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
0.15.0 seems to never have been in Debian (it is atleast not on
snapshot.debian.net), so use the real upstream (freedesktop) instead.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The avahi udhcpc script used to start up avahi-autoipd at defconfig
(init) time, and then kill it if a dhcp lease was acquired, leading
to extra network traffic.
Change it to only start avahi-autoipd on timeout and dhcp nak.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Look at BR2_STRIP_none to decide whether to strip diffutils binaries.
This avoids bad quoting in diffutils Makefile.in (generated by automake 1.6)
when STRIP is set to: true -Not_stripping
Signed-off-by: Marc Gauthier <marc@alumni.uwaterloo.ca>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The -D option to install only works portably when the destination is
a file. Specifying the filename avoids the following build error (at
least on some hosts):
"/usr/bin/install: cannot overwrite directory `/tmp/buildroot/
project_build_foobar/uclibc/root/etc/init.d' with non-directory"
Signed-off-by: Marc Gauthier <marc@tensilica.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
mesa wants to compile and run gen_matypes at build time, so make sure it
gets compiled for the host and not for the target.
Closes#323.
Signed-off-by: Jim Persson <spambox@nurd.se>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Olaf Rempel <razzor@kopf-tisch.de>
Acked-by: Thiago A. Correa <thiago.correa@gmail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The pdftops support in cups needs C++ support, whereas the rest doesn't.
Add a BR2_PACKAGE_CUPS_PDFTOPS (depending on C++ support) to enable it
if needed.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
- bump version to 2009.03.8
- internal fuse support, no longer depends on libfuse/pkgconfig
- fix WCHAR dependency
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Some configure scripts seems to ignore CXX settings if it is set to
the empty string, and goes back to the default (<arch>-linux-g++),
so use false instead, as that will loudly break the build if the
C++ compiler is ever used.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The directory argument was introduced to support packages in
subdirectories of package/ several months ago.
Signed-off-by: Markus Heidelberg <markus.heidelberg@web.de>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
The shared-mime-info package contains the core database of common
types and the update-mime-database command used to extend it.
The only thing that is actually needed in the target is the file
mime.cache. In order to generate that file, shared-mime-info is
installed into the staging-dir. The idea is that other packages might
add more mime-info files and these should be collected in
staging-dir. From there the mime.cache file is created using tools
installed in host-dir and copied to target.
From Sven Neumann <s.neumann@raumfeld.com>
Modifications by Thomas Petazzoni :
* Add a patch to the main Makefile.am so that the $(XMLLINT) variable
is used instead of hardcoding xmllint
* Apply the patch in the host build of shared-mime-info
* In the target build, pass an XMLLINT environment variable so that
the xmllint binary isn't searched in $(STAGING_DIR) but in
$(HOST_DIR)
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Some other packages, such as shared-mime-info, might need libxml2 and
related tools (xmllint, etc.) to be available on the host. This patch
modifies tthe libxml2 Makefile to compile libxml2 for the host, in
$(HOST_DIR).
Patch from Sven Neumann <s.neumann@raumfeld.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
And remove the unnedded c_rehash binary while we're at it.
Patch by Gustavo Zacarias <gustavo@zacarias.com.br>, closes#307.
Saves ~250k on PPC with default config (E.G. -Os)
Patch by Olaf Rempel <razzor@kopf-tisch.de>, closes#245.
lzop is a file compressor which is very similar to gzip.
lzop uses the LZO data compression library for compression services.
http://www.lzop.org/
Subentries of a menu automatically gets hidden if the dependencies of the
menu element aren't available, so get rid of the individual depends on
BR2_PACKAGE_NCFTP.
qmake.conf needs to be tweaked to make QT use the correct toolchain for
building, but the current approach relies on the fact that the selected
platform qmake.conf has QMAKE_{CC,CXX,LINK,LINK_SHLIB,AR,OBJCOPY,RANLIB,STRIP}
lines that are then adapted to the correct settings.
This is true for some configs (namely arm and avr32), but not for others -
Causing the build to fail (or build for the host).
Instead add a QTOPIA4_QMAKE_SET macro which removes any existing lines
and add the correct lines to the qmake.conf, so it works no matter if the
platform provides them or not.
Revert r25827 and instead copy over the host aclocal files of the stuff
needed for auto* (libtool + pkgconfig). These logically belong with the
target stuff, as they are used for the target auto* toolchain.
Long term we should probably consider using
aclocal --acdir=$(STAGING_DIR)/usr/share/aclocal, so the host .m4 files
cannot interfere with the target build.
Change the definition of TARGET_LDFLAGS to use --sysroot
$(STAGING_DIR) instead of -L$(STAGING_DIR)/lib
-L$(STAGING_DIR)/usr/lib. It fixes the following failure while trying
to build mtd-utils :
/usr/local/xtools/arm-unknown-linux-uclibcgnueabi/bin/arm-unknown-linux-uclibcgnueabi-gcc -L/home/thomas/local/buildroot-output/build_arm/staging_dir/lib -L/home/thomas/local/buildroot-output/build_arm/staging_dir/usr/lib -o /home/thomas/local/buildroot-output/build_arm/mtd_orig/flash_eraseall /home/thomas/local/buildroot-output/build_arm/mtd_orig/crc32.o /home/thomas/local/buildroot-output/build_arm/mtd_orig/flash_eraseall.o
/usr/local/xtools/arm-unknown-linux-uclibcgnueabi/lib/gcc/arm-unknown-linux-uclibcgnueabi/4.3.2/../../../../arm-unknown-linux-uclibcgnueabi/bin/ld: cannot find /lib/libc.so.0
collect2: ld returned 1 exit status
make[1]: *** [/home/thomas/local/buildroot-output/build_arm/mtd_orig/flash_eraseall] Error 1
At the same time, simplify the definition of TARGET_CFLAGS, because
the -I$(STAGING_DIR)/include -I$(STAGING_DIR)/usr/include
-I$(TOOLCHAIN_EXTERNAL_PATH)/$(TOOLCHAIN_EXTERNAL_PREFIX)/include are
no longer necessary since we sysroot the toolchain in $(SYSROOT_DIR).
This patch has no effect on non-external toolchain builds.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
LIVE555 was installed on the build system and was recognized by configure:
Checking for LIVE555 Streaming Media libraries ... yes (using distribution version)
Later this led to a linker error:
ld: cannot find -lliveMedia
Disable LIVE555, since it's not available in Buildroot.
Signed-off-by: Markus Heidelberg <markus.heidelberg@web.de>
pango likes to rerun it's configure script at make time, and it then
fails to find the correct pkg-config files (it finds the ones ins
STAGING_DIR) causing problems if they are not compatible.
Fix it by setting PKG_CONFIG_PATH in it's make environment.
Move stamp (dependency) files outside the (version specific) source
directories, so other packages can hardcode dependencies on them instead
of having to use <PACKAGE>_VERSION variables.
This is important as the variables in the make rules are evaluated when
the rules is seen, which might be before the dependent makefile is parsed
(and hence <PACKAGE>_VERSION variable is known, screwing up stuff.
The downside of this is that the package isn't automatically rebuilt
when the version changes (E.G. by a svn update) and you now also have to
remove the stamp files next to $(BUILD_DIR)/<PACKAGE>-* to force a rebuild.
This matches upstream tarball, doesn't screw up existing .config's with
BR2_PACKAGE_PKGCONFIG and makes sure the patch gets applied for target
compilation.
Generate pango.modules using the host tools instead of setting up an init
script to do it at run time. Also apply patches to host build to get
identical pango.modules file.
The host versions shouldn't be visible in Kconfig, so remove the
reference to BR2_PACKAGE_PKGCONFIG everywhere and prefix the host targets
with host-.
At the same time add pkgconfig for the target (E.G. for development) and
let BR2_PACKAGE_PKGCONFIG control that package.
Notice: all defconfigs in the tree have been updated, but make sure to
disable the pkgconfig package (unless you want it) if you use an external
config, otherwise you'll end up with pkgconfig and glib2 in the target.
Choosing between GPL or LGPL doesn't change the build process, so it's
sufficient to have a selection for the commercial license.
Also change the URLs to the licenses to the Qt web page, where the
exception for the LGPL is included.
Signed-off-by: Markus Heidelberg <markus.heidelberg@web.de>
Not particular useful until we add dbus-python (which depends on a fixed up
dbus-glib, which in turn depends on us starting to compile glib/dbus/dbus-glib
for the host as well) - But that's for another day.
