summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
...
* pango: Use gnomebase class, not gnomeRichard Purdie2012-01-162-2/+2
| | | | | | | | | | | There was a circular dependency introduced by the recent gconf changes to depend on gtk+. The issue is that gtk+ depends on pango and pango depends on gconf. This patch changes to use the gnonebase class since pango has no need of gconf/mime/gtk-icon-cache and hence removes the circular dependency. Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* gconf: enable gtk+ 2.0 support to build gconf-sanity-check-2Koen Kooi2012-01-161-9/+11
| | | | | | | | | This is needed for e.g. gnome-session: gnome-session[424]: WARNING: Failed to run gconf-sanity-check-2: Failed to execute child process "/usr/libexec/gconf-sanity-check-2" (No such file or directory) Signed-off-by: Koen Kooi <koen@dominion.thruhere.net> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* buildstats: tolerate absence of /proc/diskstatsJean-François Dagenais2012-01-161-10/+15
| | | | | | | | | | In OpenVZ containers (and probably lx containers as well), the diskstats entry is not even present. Use the "NoLogicalDrive" introduced by Elizabeth Flanagan in such case. This allows the bitbaking to occure within such containers. Signed-off-by: Jean-François Dagenais <jeff.dagenais@gmail.com>
* dhcp: package /var/lib/dhcp with dhcp-client rather than dhcp-serverSteve Sakoman2012-01-162-3/+3
| | | | | | | | Otherwise dhclient encounters errors of this type: dhclient: can't create /var/lib/dhcp/dhclient-c0c60402-0bc5-4bd7-bc3b-49a27fa37d72-eth1.lease: No such file or directory Signed-off-by: Steve Sakoman <steve@sakoman.com>
* cairo 1.10.2: enable tee backendKoen Kooi2012-01-162-2/+4
| | | | | | | | | | This is needed for firefox: | checking CAIRO_LIBS... -lcairo -lpixman-1 -lfreetype -lfontconfig | checking for cairo-tee >= 1.10... Package cairo-tee was not found in the pkg-config search path. Perhaps you should add the directory containing `cairo-tee.pc' to the PKG_CONFIG_PATH environment variable No package 'cairo-tee' found | configure: error: Library requirements (cairo-tee >= 1.10) not met; consider adjusting the PKG_CONFIG_PATH environment variable if your libraries are in a nonstandard prefix so pkg-config can find them. Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
* pango.inc: add directory "/etc/pango"Xiaofeng Yan2012-01-162-2/+6
| | | | | | | | | | | Command "pango-querymodules > /etc/pango/pango.modules" can't work when \ starting up yocto because of no directory "/etc/pango". It will cause \ messy code when gtk-demo running. [YOCTO #1900] [RP: PR bump] Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
* cairo.inc: Change configuration of directfb to gtk-directfbXiaofeng Yan2012-01-161-1/+1
| | | | | | | | | "direcdtfb" is a wrong configuration in PACKAGECONFIG. It should be "gtk-directfb". PACKAGECONFIG can't select directfb as backend rightly, or else. [YOCTO #1900] Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
* gtk.inc: Change configuration of directfb to gtk-directfbXiaofeng Yan2012-01-161-1/+1
| | | | | | | | | "direcdtfb" is a wrong configuration in PACKAGECONFIG. It should be "gtk-directfb". PACKAGECONFIG can't select directfb as backend rightly, or else. [YOCTO #1900] Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com>
* patch.bbclass: Ensure the DATE and SRCDATE variable exclusions apply to the ↵Richard Purdie2012-01-131-1/+3
| | | | | | | | | | | correct function People have noticed that sstate is now getting invalidated very readily. The issue is that the code using these variables was factored into a new function but the variable exclusion was not. This patch moves the variable exclusion to the correct place allowing the sstate checksums to work correctly. Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* multilib.conf: Clean up file and add missing entries for various dependenciesRichard Purdie2012-01-111-15/+22
| | | | | | | This patch adds in various missing dependencies to ensure the set of recipes listed for multilib support can be cleanly built. Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* binutils-cross-canadian: Clear BBCLASSEXTEND as a native version of this ↵Richard Purdie2012-01-111-0/+2
| | | | | | recipe makes no sense Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* python-native: RPROVIDE python-core-native since pycurl-native depends on itRichard Purdie2012-01-111-1/+1
| | | | Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* libxft: Drop BBCLASSEXTEND nativesdk since dependencies freetype and ↵Richard Purdie2012-01-111-1/+1
| | | | | | fontconfig aren't class extended Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* lzo: Add BBCLASSEXTEND nativesdk so lzop nativesdk works (it depends on this)Richard Purdie2012-01-111-1/+1
| | | | Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* rpm: Fix magic file name magic -> magic.mgcSaul Wold2012-01-111-2/+2
| | | | | Signed-off-by: Saul Wold <sgw@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* package.bbclass: per recipe PRSERV_HOST supportLianhao Lu2012-01-111-0/+8
| | | | | | | [YOCTO #1126] Added per recipe PRSERV_HOST PRSERV_PORT support. Signed-off-by: Lianhao Lu <lianhao.lu@intel.com>
* meta/PRService: Added export/import fuctions.Lianhao Lu2012-01-119-27/+260
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | [YOCTO #1556] - Modified meta/class/package.bbclass and prserv.bbclass according to the change in PR service by adding PACKAGE_ARCH into the query tuple. - Added prexport.bbclass, primport.bbclass to export/import AUTOPR values from/to PRService. - Move PR service related common code to lib/oe/prservice.py. - Supported reading the AUTOPR values from the exported .inc file instead of reading it from remote PR service. - Created a new script bitbake-prserv-tool to export/import the AUTOPR values from/to the PR service. Typical usage scenario of the export/import is: 1. bitbake-prserv-tool export <file> to export the AUTOPR values from the current PR service into an exported .inc file. 2. Others may use that exported .inc file(to be included in the local.conf) to lockdown and reproduce the same AUTOPR when generating package feeds. 3. Others may "bitbake-prserv-tool import <file>" to import the AUTOPR values into their own PR service and the AUTOPR values will be incremented from there. Signed-off-by: Lianhao Lu <lianhao.lu@intel.com>
* Incremental rpm image generation(Add config sample)Robert Yang2012-01-111-0/+7
| | | | | | | | | Add the config sample for incremental image generation to meta-yocto/conf/local.conf.sample.extended [YOCTO #1651] Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
* Incremental rpm image generationRobert Yang2012-01-113-13/+77
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Incremental rpm image generation, the rootfs would be totally removed and re-created in the second generation by default, but with INC_RPM_IMAGE_GEN = "1", the rpm based rootfs would be kept, and will do update(remove/add some pkgs) on it. NOTE: This is not suggested when you want to create a productive rootfs For example: 1) Add the follow config option to a conf file: INC_RPM_IMAGE_GEN = "1" 2) bitbake core-image-sato modify a package bitbake core-image-sato The rootfs would not be totally removed and re-created in the second generation, it would be simply updated based on the "package". Implatation: 1) Figure out the pkg which need to be removed or re-installed, then use 'rpm -e to remove the old one. Use the rpm's BUILDTIME to determine which pkg has been rebuilt. 2) Figure out the pkg which is newly added, and use 'rpm -U' to install it. This only for the rpm based rootfs, the deb and ipk based rootfs would be done later. [YOCTO #1651] Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
* Revert "site/common-uclibc: Cache gl_cv_func_wcwidth_works"Richard Purdie2012-01-101-1/+0
| | | | This reverts commit a471d342424502de48f28c4f8297d490f2cd1443.
* distro_tracking: update manual check infoDongxiao Xu2012-01-101-0/+1
| | | | | | | updated kexec-tools manual check information Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* license.bbclass base.bbclass: support for 'or' operand in LICENSE and for ↵Andrei Gherzan2012-01-102-2/+46
| | | | | | | | | | | | | | | | SPDX license names A new function was defined in license.bbclass in order to correctly exclude packages where OE-Style licence naming is used. In this way licenses as GPL-3, GPLv3, GPLv3.0 etc will be excluded from a non-GPLv3 build. This function takes into consideration if 'or' operand is used. The function defined in license.bbclass is called in base.bbclass where packages are excluded based on INCOMPATIBLE_LICENSE variable. [YOCTO #1884] [YOCTO #1844] Signed-off-by: Andrei Gherzan <andrei at gherzan.ro> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* distro_tracking_fields.inc: upgrade tcf-agentDexuan Cui2012-01-101-2/+2
| | | | | | | Upgraded the field RECIPE_MANUAL_CHECK_DATE. Also changed the MAINTAINER to Lianhao who volunteered to take the recipe. Signed-off-by: Dexuan Cui <dexuan.cui@intel.com>
* distro_tracking: Update Manual ChecksSaul Wold2012-01-101-25/+25
| | | | Signed-off-by: Saul Wold <sgw@linux.intel.com>
* task-core-tools: adapt to kexec-tools package splitAndrea Adami2012-01-101-2/+2
| | | | | | | | * The recipe packages now kexec and kdump separately. * For standard purposes only kexec is needed. * Bump PR. Signed-off-by: Andrea Adami <andrea.adami@gmail.com>
* kexec-tools: split packaging of kexec and kdumpAndrea Adami2012-01-101-1/+6
| | | | | | | | | | | | | * from meta-oe (originally from org.openembedded.dev) * There are no reasons to install kdump when only kexec is needed. * * In oe-core/meta there are no references to kexec-tools so * the recipes inn external layers rdepending on kexec-tools should be * fixed and rdepend on kexec and/or kdump. * * Bump PR. Signed-off-by: Andrea Adami <andrea.adami@gmail.com>
* util-linux: .pc files should go in the -dev packages for subpackagesColin Walters2012-01-101-2/+2
| | | | | The .pc files were ending up in util-linux-dev, not the correct subpackage like libuuid1-dev.
* util-linux: Enable fallocate and use CACHED_CONFIGUREVARSKhem Raj2012-01-101-4/+2
| | | | | | | | | fallocate is implemented in eglibc 2.11 and all eglibc we use are greater than 2.11 so drop disabling it. Use CACHED_CONFIGUREVARS to specify scanf_cv_type_modifier=as Signed-off-by: Khem Raj <raj.khem@gmail.com>
* findutils: Use CACHED_CONFIGUREVARS to specify cached configure varsKhem Raj2012-01-101-2/+2
| | | | | | | Since this feature is depending on wchar support we only cache is when we have libc-posix-clang-wchar enabled Signed-off-by: Khem Raj <raj.khem@gmail.com>
* uclibc: Configure UCLIBC_HAS_WCHAR based on libc-posix-clang-wchar distro ↵Khem Raj2012-01-101-0/+1
| | | | | | feature Signed-off-by: Khem Raj <raj.khem@gmail.com>
* autotools.bbclass: Introduce CACHED_CONFIGUREVARSKhem Raj2012-01-091-1/+1
| | | | | | | | | | | | | | | This variable is for holding the cached configure variables to be specified in recipes. e.g. CACHED_CONFIGUREVARS += "ac_cv_foo=yes ac_cv_bar=no" This will make sure that the variables are not detected by configure. This is useful in cross builds where some features can not be detected correctly by configure and having it as a variables gives us capability to override it Signed-off-by: Khem Raj <raj.khem@gmail.com>
* libxml2: add shared library version info to libxml shared librariesMatthew McClintock2012-01-092-1/+26
| | | | | | | | | | | | | | This fixes an issue with RPM where it checks version imformation for binaries linked against libxml and fails because it's missing info | error: Failed dependencies: | libxml2.so.2(LIBXML2_2.6.0) is needed by fmc-0.9.7+2-r2.1.ppce500mc | libxml2.so.2(LIBXML2_2.4.30) is needed by fmc-0.9.7+2-r2.1.ppce500mc | ERROR: Function 'do_rootfs' failed (see Note: fmc is just an example recipe/name Signed-off-by: Matthew McClintock <msm@freescale.com>
* eglibc: Upgrade recipes 2.14 -> 2.15Khem Raj2012-01-0919-15/+172
| | | | | | Add patch to provide __finite_sqrt functions for ppc Signed-off-by: Khem Raj <raj.khem@gmail.com>
* update-modules: don't run modutils.sh if systemd is installedSteve Sakoman2012-01-092-2/+4
| | | | | | systemd has its own mechanism for loading kernel modules at startup Signed-off-by: Steve Sakoman <steve@sakoman.com>
* gtk-icon-cache.bbclass: add -f option to gtk-update-icon-cache in postinstSteve Sakoman2012-01-091-1/+1
| | | | | | | | | | The gtk-update-icon-cache utility does not create a cache file if one does not already exist. Since some packages (notably gnome-icon-theme) do not ship a cache file by default, the results of the utility are not saved. Adding the force option writes out the cache even if there wasn't a default cache in the package. Signed-off-by: Steve Sakoman <steve@sakoman.com>
* dbus: change permissions of dbus-daemon-launch-helper to 4755Steve Sakoman2012-01-091-1/+3
| | | | | | | | | | | With current permissions of 4754 on systemd systems various services will fail to activate with "helpful" error messages of the type: 2000-01-01T00:00:25+00:00 omap3-multi dbus[178]: [system] Activated service 'org.freedesktop.nm_dispatcher' failed: Failed to execute program /usr/libexec/dbus-daemon-launch-helper: Success Signed-off-by: Steve Sakoman <steve@sakoman.com>
* Remove last remnants of kernel26 MACHINE_FEATURESSteve Sakoman2012-01-093-4/+3
| | | | | | There is no reason to continue to carry this feature Signed-off-by: Steve Sakoman <steve@sakoman.com>
* libusb-compat: keep *.la in base_libdirAndreas Müller2012-01-091-2/+1
| | | | | | build-tested from scratch in angstrom environment. Signed-off-by: Andreas Müller <schnitzeltony@gmx.de>
* libusb1: keep *.la in base_libdirAndreas Müller2012-01-091-2/+1
| | | | | | | | | | | | | This fixes the following error found during build from scratch for meta-oe / obexd: | arm-angstrom-linux-gnueabi-libtool: link: cannot find the library | `/home/Superandy/tmp/oe-core-eglibc/sysroots/overo/lib/libusb-1.0.la' or | unhandled argument `=/lib/libusb-1.0.la' build-tested from scratch in angstrom environment. Signed-off-by: Andreas Müller <schnitzeltony@gmx.de>
* libiconv_1.14.bb: Fix build failureKhem Raj2012-01-092-1/+5015
| | | | | | | | | | Currently libiconv is failing after upgrade to 1.14 this patch fixes the problem. It uses relocatable gnulib module which should be added. Delete unpackaged files. Signed-off-by: Khem Raj <raj.khem@gmail.com>
* site/common-uclibc: Cache gl_cv_func_wcwidth_worksKhem Raj2012-01-091-0/+1
| | | | | | | | | | | | | | Its not computed properly during configure and we configure uclibc with widechar support anyway so its fine to cache it. It fixes the errors like | In file included from ./stdio.h:31:0, | from fpurge.h:20, | from fpurge.c:20: | /home/kraj/work/angstrom/sources/openembedded-core/build/tmp-uclibc/sysroots/qemux86/usr/include/stdio.h:662:40: error: expected declaration specifiers or '...' before '(' token | make[4]: *** [fpurge.o] Error 1 Signed-off-by: Khem Raj <raj.khem@gmail.com>
* add sysroot support for meta-ide-support which is the toolchain within build ↵Jessica Zhang2012-01-091-4/+5
| | | | | | env case Signed-off-by: Jessica Zhang <jessica.zhang@intel.com>
* libxxf86dga: fix compilation with x32 toolchainNitin A Kamble2012-01-092-1/+34
| | | | | | | | | | | | | | | | | Fix type conversion for x32. For x32 the off_t is 64bit and pointers are 32bit. so the conversion of pointer to off_t was resulting into this error: | XF86DGA2.c:931:24: error: cast from pointer to integer of different size [-Werror=pointer-to-int-cast] | cc1: some warnings being treated as errors | | make[2]: *** [XF86DGA2.lo] Error 1 Fixed it by typecasting pointer into unsigned long 1st and then again typecasting unsigned long to off_t. Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
* base-files: filesystems: fix mount orderGary Thomas2012-01-062-5/+6
| | | | | | | | | | | | | | | | | | This changes the order that file system [types] are tried when using 'mount' in busybox when the file system type is not explictly specified. The pervious ordering is incorrect in that less capable file system types are tried first, e.g. ext2 before ext3, which will cause an ext3 file system to be mounted as ext2, disabling some of the ext3 features such as journaling. The change also moves infrequently used file system types to the end as checking for them is just a waste of effort in most cases. The list now also includes ext4. [RP: Add PR bump, unbreak patch application] Signed-off-by: Gary Thomas <gary@mlbassoc.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* valgrind: add powerpc to the compatible host and add ${TCLIBC}-dbg in ↵Zhenhua Luo2012-01-061-2/+4
| | | | | | | | | | | RRECOMMENDS-powerpc 1. add powerpc/powerpc64 into the compatible host list 2. valgrind requires the non-stripped libraries of ${TCLIBC}, so add ${TCLIBC}-dbg into RRECOMMENDS for powerpc Signed-off-by: Zhenhua Luo <b19537@freescale.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* bblayers.conf.sample: Use ?= to set BBLAYERSDongxiao Xu2012-01-061-1/+1
| | | | | | | | Hob may dynamically sets BBLAYERS to bitbake server, thus we need a flexible way to load BBLAYERS value. Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* image.bbclass: Use ?= to set LINGUAS_INSTALLDongxiao Xu2012-01-061-1/+1
| | | | | | | | Hob may dynamically sets BBLAYERS to bitbake server, thus we need a flexible way to load LINGUAS_INSTALL value. Signed-off-by: Dongxiao Xu <dongxiao.xu@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* x11-common: remove extraneous BSP customizationsAndrea Adami2012-01-061-24/+0
| | | | Signed-off-by: Andrea Adami <andrea.adami@gmail.com>
* xorg-driver: add xf86-video-fbdev from meta-openembeddedAndrea Adami2012-01-061-0/+8
| | | | Signed-off-by: Andrea Adami <andrea.adami@gmail.com>
* systemtap: remove not needed fix for automake 1.11.2Nitin A Kamble2012-01-062-50/+2
| | | | | | | | The pkglibexec_SCRIPTS pair is valid. the 1.11.2 introduced an bug to make it invalid. Now the automake 1.11.2 recipe is fixed, so no need for this fix for systemtap recipe. Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>