To fix these package qa warnings
WARNING: For recipe binutils, the following files were installed but not shipped in any package:
WARNING: /usr/bin/ld.bfd
WARNING: /usr/bin/elfedit
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
ofono: fix packaging and clean up custom do_install
The makefile installs the udev rule in the proper place already. Also fix:
WARNING: For recipe ofono, the following files were installed but not shipped in any package:
WARNING: /lib/udev/rules.d/97-ofono.rules
WARNING: /lib/systemd/system/ofono.service
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yu Ke [Wed, 13 Jul 2011 08:58:27 +0000 (16:58 +0800)]
libzypp: fix for non /usr/lib libdir case
this patch has two fixes:
1. the src tool/CMakeLists.txt has hardcode "lib", so add a patch to fix it
2. the recipe has hardcode "-DLIB=lib", so replace it with libdir
Yu Ke [Wed, 13 Jul 2011 08:57:39 +0000 (16:57 +0800)]
sat-solver: fix for non /usr/lib libdir case
when libdir set to non /usr/lib like /usr/lib64, there is do_configure error:
"
-- Libraries will be installed in /usr/lib
CMake Error at /home/kyu3/sdb/lib64/tmp/sysroots/x86_64-linux/usr/share/cmake-2.8/Modules/FindPackageHandleStandardArgs.cmake:91 (MESSAGE):
Please install 'check' and 'check-devel' packages (missing: CHECK_LIBRARY)
Call Stack (most recent call first):
/home/kyu3/sdb/lib64/tmp/sysroots/x86_64-linux/usr/share/cmake-2.8/Modules/FindPackageHandleStandardArgs.cmake:252 (_FPHSA_FAILURE_MESSAGE)
cmake/modules/FindCheck.cmake:17 (FIND_PACKAGE_HANDLE_STANDARD_ARGS)
CMakeLists.txt:35 (FIND_PACKAGE)
-- Configuring incomplete, errors occurred!
"
The reason is that cmake did not search /usr/lib64 for libcheck, thus report error
add ${libdir} to the CMAKE_SYSTEM_LIBRARY_PATH can fix this issue.
Phil Blundell [Wed, 13 Jul 2011 09:03:42 +0000 (10:03 +0100)]
eglibc: ensure that NSS libs go into the right package when ${base_libdir} and ${libdir} are the same
Previously, libnss_hesiod et al. could get swallowed by the wildcarded glob in ${PN}-dev.
Fix it by moving eglibc-extra-nss earlier in the PACKAGES and adjusting its FILES to ensure
that the devel symlinks don't get caught.
Signed-off-by: Phil Blundell <philb@gnu.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yu Ke [Thu, 7 Jul 2011 14:22:11 +0000 (22:22 +0800)]
sato-sdk: add clutter for sato-sdk image
Yocto ADT has Eclips clutter template, so it is better to have
clutter in SDK image to support ADT. this patch add clutter
and clutter-box2d app to sato-sdk image for this purpose
Fix [YOCTO #891]
Signed-off-by: Yu Ke <ke.yu@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 12 Jul 2011 13:29:28 +0000 (14:29 +0100)]
insane.bbclass: skip rdepends QA checks for kernel / modules
The kernel and module recipes have very few dbg/dev packages, however
they can easily have false positive results from the rdepends QA checks
(e.g. kernel-module-lirc-dev). Thus disable these tests for any recipe
that inherits kernel or module-base.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Darren Hart [Sat, 9 Jul 2011 02:02:12 +0000 (19:02 -0700)]
bb-matrix: initial scripts to record TIME(1) metrics for BB and PM combinations
The bb-matrix.sh script will run a bitbake command, building core-image-minimal
by default, for various combinations of BB_NUMBER_THREADS and PARALLEL_MAKE. It
records all relevant metrics of the TIME(1) command for each combination in a
data file.
The bb-matrix-plot.sh script can be used to visualize each of these metrics via
a 3d surface plot, either solid surface or wireframe with a value-map
projection on the XY plane.
Signed-off-by: Darren Hart <dvhart@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
root@usrp-e1xx:/etc/pulse# pulseaudio --system
[..]
W: protocol-native.c: Failed to get GID of group 'pulse-access'
W: protocol-native.c: Failed to check group membership.
W: protocol-native.c: Denied access to client with invalid authorization data.
* add more required modules
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Andrea Adami [Mon, 11 Jul 2011 22:33:11 +0000 (00:33 +0200)]
kernel.bbclass: handle embedding of initramfs images
* "All 2.6 Linux kernels contain a gzipped "cpio" format archive, which is
* extracted into rootfs when the kernel boots up."
* (http://www.kernel.org/doc/Documentation/filesystems/ramfs-rootfs-initramfs.txt)
*
* In some cases it can be useful to take advantage of this ability and embed your
* customized rootfs ("the move to early userspace is necessary because finding
* and mounting the real root device is complex").
*
* Therefore, some code was added and since 2009 OpenEmbedded metadata
* provides a convenient way to include your initramfs by setting
* INITRAMFS_IMAGE = "your_initramfs_image.bb"
* and specifying the chosen output format in the initramfs image recipe, e.g.
* IMAGE_FSTYPES = "cpio.gz cpio.lzma"
*
*
* Patch was imported from the OpenEmbedded git server
* (git://git.openembedded.org/openembedded)
*
* Add satndalone task "builtin_initramfs".
* commit 72761e468bb3e905459f2b81ce1bc4d80419481f
* Signed-off-by:Paul Sokolovsky <pmiscml@gmail.com>
*
* 'Remove usage of non-standard variables (DEPLOY_TO) and random renaming
* of output.'
* commit 456ba7ffd159821e86ad7ad4b66ec7d5790bd054
* Richard Purdie <rpurdie@rpsys.net>
*
* 'Fis typo in do_builtin_initramfs.'
* commit 37f2fe4b801df832e93553a08eff24fec736c7d4
* Signed-off-by:Paul Sokolovsky <pmiscml@gmail.com>
*
* 'comment initramfs stuff till rootfs locking works properly (initramfs
* should be optional as well)'
* commit 2818d8b7be1a0d9a59ad3528091d47517d59328b
* Signed-off-by: Koen Kooi <koen@openembedded.org>
*
* 'get initramfs stuff working
* commit 1642b3e8fc81c666713951fdd4e7ff9a50d1c5a8
* Signed-off-by: Thomas Kunze <thommycheck@gmx.de>
*
* 'Fix INITRAMFS logic to stop breaking builds for people not using it'
* commit dcf3049eb9eedf0838af59969b3f70a43298d3d7
* Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
*
* '-change initramfs-logic'
* commit 3e3f297457138e96e2b652658675796853eb0293
* Signed-off-by: Thomas Kunze <thommycheck@gmx.de>
*
* 'move initramfs stuff to configure so we can do postprocessing
* on it with do_configure_append'
* commit fc03e2be0b4470748a8b7707cea7293303adc424
* Signed-off-by: Koen Kooi <koen@openembedded.org>
*
* 'copy gz, lzo, lzma initramfs cpio images.'
* commit 572abc3fdd1076ca35d8c15d269cc9d862101805
* Signed-off-by: Andrea Adami <andrea.adami@gmail.com>
*
* 'move the kernel size check to linux-kexecboot.inc.'
* commit 45f82a941c77e9d747814fa1e337ba803475d327
* Signed-off-by: Andrea Adami <andrea.adami@gmail.com>
*
* Finally, two refinements as discussed on openembedded-core
* mailing list:
* - replace "if image != '' and image is not None:" with "if image"
* - add cpio.xz support
Signed-off-by: Andrea Adami <andrea.adami@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Andrea Adami [Mon, 11 Jul 2011 22:33:12 +0000 (00:33 +0200)]
image_types.bbclass: add support for tar.xz, cpio.xz, cpio.lzma
* Once we add support for building kernel containing initramfs images
* we expect compressed cpio archives.
* Thus, add support for lzma and xz compressors for the initramfs images.
* While there, add support for tar.xz as done in org.openembedded.dev.
*
* Patch was imported from the OpenEmbedded git server
* (git://git.openembedded.org/openembedded)
*
* 'bitbake.conf: add cpio and ext4 .xz support'
* commit 7977d3325a09e2edf4a8a1b02f48085f88f7a631
* Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
*
* 'bitbake.conf: Define tar.xz image types.'
* 1fca8dc50a821f4589fa5e339e6e279dd3c0dafe
* Signed-off-by: Tom Rini <tom_rini@mentor.com>
*
* 'bitbake.conf: add XZ_INTEGRITY_CHECK and LZMA_COMPRESSION_LEVEL variables'
* commit 312b42bcf2cc9b11266ba497ade68e8ddabb3007
* Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
*
* 'bitbake.conf: unbreak and improve lzma image compression handling and adapt
* linux-kexecboot.inc'
* commit c382a9ce64bf4291572665652138aa01b2e54306
* Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
* Signed-off-by: Denis 'GNUtoo' Carikli <GNUtoo@no-log.org>
Signed-off-by: Andrea Adami <andrea.adami@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Chris Elston [Mon, 11 Jul 2011 12:14:15 +0000 (13:14 +0100)]
Add support for BAD_RECOMMENDATIONS to rootfs_ipk
As discussed on IRC on 30/06/11, this patch adds support for
BAD_RECOMMENDATIONS to rootfs_ipk, which is a list of packages NOT to
install if suggested or recommended by a recipe. Taken from
http://thread.gmane.org/gmane.comp.handhelds.openembedded/30417.
Note that current support for this in oe.dev may also be broken,
depending on the version of opkg in use.
Signed-off-by: Chris Elston <celston@katalix.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yu Ke [Mon, 11 Jul 2011 07:23:18 +0000 (15:23 +0800)]
python-native: fix for non /usr/lib libdir case
the get_python_lib() use hard code "lib". when other recipe
building use python-native, the STAGING_LIBDIR is non /usr/lib,
so the hard code "lib" will break the build
This patch replace the hardcoded "lib" with value from libdir
to fix this issue.
Signed-off-by: Yu Ke <ke.yu@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yu Ke [Mon, 11 Jul 2011 07:23:14 +0000 (15:23 +0800)]
native.bbclass: generate libdir from libdir_native
currently, the libdir in native.bbclass is inherited from target libdir
defined in bitbake.conf. This is not correct in concept, instead, it
should be inherited from libdir_native. Current code works because both
the target libdir and libdir_native are the same value "/usr/lib".
But if libdir changes (e.g. to /usr/lib64), then build will break.
This patch make libdir generated from libdir_native to fix this issue.
Signed-off-by: Yu Ke <ke.yu@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This disables linking against (lib)hal but enables the udev-hal compat layer so the hal-detect plugin keeps working. The hal-detect plugin is removed from the default dependencies to discourage its usage, but can be installed manually if needed.
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Otavio Salvador [Mon, 11 Jul 2011 17:27:19 +0000 (17:27 +0000)]
cmake: update to 2.8.5 release
Acked-by: Tom Rini <tom_rini@mentor.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 6 Jul 2011 13:05:28 +0000 (14:05 +0100)]
update-rc.d: Switch from a tag to a sepcific revision
Unfortunately the git fetcher cannot resolve a named tag into a revision
without network access. Since its desireable for OE-Core to function
in a networkless environment, we need to specify the revision in the
recipe which corresponds to the 0.7 tag.
[YOCTO #1140]
Acked-by: Joshua Lock <josh@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
XCB support is needed in libx1l, it has been enabled in libx11-trim for
sometime and was not in full version. The usage here is for LSB testing,
which uses full libx11.
* replace shebang of ldd by /bin/sh, so ldd doesn't depend on bash
anymore. Tested with busybox sh.
* added to do_configure_prepend because do_patch is only a python function
rootfs_rpm: Add 50M to IMAGE_ROOTFS_EXTRA_SPACE for zypper /var space
[YOCTO #1171]
This is a temporary fix to address the fact that zypper's db configuation
uses a large amount of /var space. When the correct'ed zypper config can
be applied, this fix can be removed. The zypper bug for yocto is 1174
http://bugzilla.pokylinux.org/show_bug.cgi?id=1174
Yu Ke [Mon, 13 Jun 2011 12:20:53 +0000 (20:20 +0800)]
combo-layer-tool: add tool to manipulate combo layers
This patch adds the script "combo-layer" to manipulate combo layer
repos. A combo layer repo is a repo containing multiple component
repos, e.g. oe-core, bitbake, BSP repos. The combo layer repo needs to
be updated by syncing with the component repo upstream. This script
is written to assist the combo layer handling.
The combo layer tool provides three functionalities:
- init: when the combo layer repo and component repo does not exist,
init will "git init" the combo layer repo, and also "git clone" the
component repos
- update: combo layer tool will pull the latest commit from component
repo upstream, and apply the commits since last update commit to the
combo repo. If the user specifies interactive mode(--interactive),
they can edit the patch list to select which commits to apply.
- splitpatch: split the combo repo commit into separate patches per
component repo, to facilitate upstream submission.
Combo layer tool uses a config file to define the component repo info.
Please check the combo-layer.conf.example for a detailed explanation
of the config file fields.
Signed-off-by: Yu Ke <ke.yu@intel.com> Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Paul Eggleton [Thu, 7 Jul 2011 12:58:55 +0000 (13:58 +0100)]
bitbake.conf: update OLDEST_KERNEL to 2.6.16
Since we no longer support 2.4, update this setting to 2.6.16, to line up
with the most accepted setting from OE. (This affects eglibc's kernel
support, and 2.6.16 is the minimum version for glibc 2.9 onwards.)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Otavio Salvador [Fri, 8 Jul 2011 13:47:42 +0000 (13:47 +0000)]
cmake.bbclass: use CPPFLAGS and CXXFLAGS
Some classes, as for example nativesdk, defines CPPFLAGS and CXXFLAGS
to be passed to compiler. Using those makes more sense and avoid some
hacks on packages using CMake.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
gcc recipes has changed, currently ${PV} is 4.6 and ${BINV} is 4.6.1 version.
Fix these QA warnings.
WARNING: For recipe gcc-runtime, the following files were installed but not shipped in any package:
WARNING: /usr/share/gcc-4.6.1/python/libstdcxx/__init__.py
WARNING: /usr/share/gcc-4.6.1/python/libstdcxx/v6/__init__.py
WARNING: /usr/share/gcc-4.6.1/python/libstdcxx/v6/printers.py
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Richard Purdie [Thu, 7 Jul 2011 16:40:12 +0000 (09:40 -0700)]
image.bbclass: Add LINGUAS_INSTALL to dependency list
Currently if there are locales to install in the image the appropriate
packages might not automatically get built since there is no dependency
being exposed to bitbake.
This patch exposes the missing dependencies to bitbake through the RDEPENDS
field hence ensuring images build successfully when locales are added but
not depended upon anywhere else in the build.
[YOCTO #1215]
Signed-off-by: Saul Wold <sgw@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Dexuan Cui [Thu, 7 Jul 2011 07:45:57 +0000 (15:45 +0800)]
powertop: inherit update-alternatives and use a higher priority than busybox
busybox-1.18.4 installs /bin/powertop and the powertop recipe installs
/usr/bin/powertop. So, in PATH, if /bin appears before /usr/bin, we would
run the version offered by busybox, which has a very limited function (e.g.,
no parameter is accepted) and this causes trouble to eclipse plugin.
We can use update-alternatives for powertop with higher priority to resolve
the issue.
Paul Eggleton [Thu, 7 Jul 2011 10:58:42 +0000 (11:58 +0100)]
default-distrovars.inc: add ipv4 to DISTRO_FEATURES
Since the change in busybox configuration to match OE (OE core rev b5564c4a9cadf306b447180c433b25ec071f8ce1) we now need ipv4 in
DISTRO_FEATURES to get standard IPv4 functionality in busybox.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Joshua Lock [Wed, 29 Jun 2011 21:55:03 +0000 (14:55 -0700)]
sanity: implement network connectivity test
Sanity test to verify files can be fetched from the network using git, http
and https fetchers point users at a page to help get set up in the case of a
failure.
Requires a variable CONNECTIVITY_CHECK_URIS to be set, using the same pattern
as SRC_URI, of URI's to test against.
The variable CONNECTIVITY_CHECK_MSG can be set to provide a custom error
message, such as a pointer to some help, when this check fails.
Addresses [YOCTO #933]
Signed-off-by: Joshua Lock <josh@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Anders Darander [Tue, 5 Jul 2011 14:55:04 +0000 (16:55 +0200)]
module-base: remove KERNEL-SOURCE
kernel-source do not exist in STAGING_KERNEL_DIR any longer.
Remove the exported KERNEL_SOURCE as the STAGING_KERNEL_DIR can
be used in module recipes directly.
Signed-off-by: Anders Darander <anders@chargestorm.se> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Anders Darander [Wed, 6 Jul 2011 06:31:45 +0000 (08:31 +0200)]
task-base: remove modutils reference.
Also remove the other kernel24 references.
Make everything dependent on kernel26 default.
Move the RDEPENDS- and RRECOMMENDS_task-base-kernel26 to _task-base.
Signed-off-by: Anders Darander <anders@chargestorm.se> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Addes the 2011.06 version, while keeping the older 2011.03 version since
other layers might be using it still. This will need to be fixed in the
future as oe-core should only be keeping the current versions and layer
management addresses this somehow.
Richard Purdie [Tue, 5 Jul 2011 13:33:02 +0000 (14:33 +0100)]
openssh/dropbear: No need for each to PROVIDE ssh/sshd
Nothing in the system actually uses the PROVIDES field for these
recipes, its usually the runtime packages that are used. We can
therefore remove the PROVIDES and hence quieten the associated
warnings from bitbake.
If these recipes do really need the PROVIDES, they would be better
as virtuals and adding that to MULTI_PROVIDER_WHITELIST.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Tue, 5 Jul 2011 15:35:03 +0000 (16:35 +0100)]
kernel-yocto: Fix case where recipe is always reparsed
If you include a bitbake variable as a comment in a shell function
then it gets expanded by the bitbake signature handling code.
This could be classed as a bug or a feature depending on your viewpoint
(e.g. a multiline variable included in a comment could actually contain
executable code).
Since we don't always want kernel-yocto to reparse this changes the
syntax of the comment so it doesn't trigger the problem.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
