* do_split_packages and package_do_split_locales dynamically add
packages in order of appearance of os.walk() or os.listdir().
This order varies between multiple build hosts (and probably
also between successive builds).
* Sort the list of dynamically added packages, to get a consistent
and reproducible order.
* This reduces the diffs in buildhistory and improves
comparability between builds.
Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Dexuan Cui [Fri, 9 Mar 2012 07:05:12 +0000 (15:05 +0800)]
image_types: ensure .rootfs.ext3 is created before vmdk is created.
In the case of self-hosted-image.bb, IMAGE_FSTYPES = "vmdk", so the
variables alltypes and subimages don't contain ext3, and .rootfs.ext3
won't be created, and finally the generated .hddimg and .vmdk don't have
an actual rootfs -- the size of the .vmdk file is only about 9MB.
* Both files contain -L../../WebCore/release -L../../JavaScriptCore/release -lwebcore -ljscore
* These paths and libs don't get installed.
* When building a shared lib with libtool, that links
with libQtWebKit, libtool adds these unneeded parameters
to the linker command-line, thus linking fails.
Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Otavio Salvador [Sat, 3 Mar 2012 04:12:18 +0000 (04:12 +0000)]
image.bbclass: fix rootfs generation without package management tools
When building without package management tools the
ROOTFS_BOOTSTRAP_INSTALL packages need to be available to provide the
bare minimal for rootfs unconfigure postinsts to be on first boot.
Those packages where being include in core-image.bbclass' based images
however every image needs those available for proper rootfs generation.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Bruce Ashfield [Fri, 2 Mar 2012 16:36:07 +0000 (11:36 -0500)]
linux-yocto: respect FILESPATH directories
During the work to enhance the ability to specify out of tree kernel
features, an assumption was made about PN being part of a patch
path. This assumption is incorrect, since patches can be anywhere in
the valid FILESPATH.
To make locating the patches in WORKDIR simple, we can just query
patch.bbclass and return both the absolute directory of the patch
and the subdirectory as it was specified on the src_uri.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Richard Purdie [Mon, 12 Mar 2012 22:13:47 +0000 (22:13 +0000)]
wireless-tools: Clarify LICENSE field to avoid warnings
The source is unclear which version of the license applies. Since BSD is listed,
we have plenty of choice so lets pick some versions. Most people would pick
BSD out of the three anyway.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Mon, 12 Mar 2012 21:30:15 +0000 (21:30 +0000)]
librsvg: Fix unpackaged directory warnings
This patch fixes the QA warnings:
WARNING: For recipe librsvg, the following files/directories were installed but not shipped in any package:
WARNING: /usr/lib/gtk-3.0
WARNING: /usr/lib/gtk-3.0/engines
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Mon, 12 Mar 2012 21:24:48 +0000 (21:24 +0000)]
sato-screenshot: Remove unneeded .la file
This resolves the QA warning:
WARNING: For recipe sato-screenshot, the following files/directories were installed but not shipped in any package:
WARNING: /usr/lib/matchbox-panel/libscreenshot.la
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Nitin A Kamble [Mon, 12 Mar 2012 16:28:10 +0000 (09:28 -0700)]
cmake.bbclass: add ${base_libdir} to CMAKE_LIBRARY_PATH
Some libraries like libcrypto.so are installed at base_libdir
instead of libdir. So add the base_libdir to CMAKE_LIBRARY_PATH
so that these libraries can be found correctly.
This resolves an issues with libzypp, which was not finding the
libcrypo library correctly in an x32 build.
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sun, 11 Mar 2012 18:26:15 +0000 (18:26 +0000)]
package.bbclass: Drop python depends from rpmdeps
This python depends is added by rpmdeps every time it finds a python
script. This is not necessary since we handle this in otherways. It
also breaks things like nativesdk since the dependency is not renamed.
The easiest solution is just to ignore this dependency.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 9 Mar 2012 06:14:04 +0000 (14:14 +0800)]
rpm 5.4.0: Add rpmdeps-oecore to replace rpmdeps for package.bbclass
This is a performance enhancement by adding a binary allowing batch processing
of individual file dependencies. The second patch in this series uses the binary
this patch creates.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* Otherwise, recipes that only append to SRC_URI instead of
overwriting it will get the full path of the recipe written
into the Source: field of the package.
Signed-off-by: Andreas Oberritter <obi@opendreambox.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Darren Hart [Wed, 7 Mar 2012 08:06:46 +0000 (00:06 -0800)]
kernel.bbclass: Remove warnings for modutils and modprobe.d
Fixes [Yocto #2036]
The source and build directories are unused, remove them.
The modutils and modprobe.d directories may be used if modules are built that
are either autoloaded or have modprobe.d entries. This isn't known at install
time, so check after the package split if these directories are empty and
remove them if they are.
Signed-off-by: Darren Hart <dvhart@linux.intel.com> CC: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
external-csl-toolchain: override PKGV, not PKGV_${PN}
This is needed to ensure that the RDEPENDS_${PN}-dev,
RDEPENDS_${PN}-staticdev, and RRECOMMENDS_${PN}-dbg are accurate with respect
to the PKGV of the ${PN} package, and incidentally fixes an rpm packaging
issue, as rpm dislikes the recipe PV as PKGV.
Signed-off-by: Christopher Larson <chris_larson@mentor.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Zhai Edwin [Tue, 6 Mar 2012 17:03:17 +0000 (09:03 -0800)]
shadow-sysroot: Fix for multilib
Fix following error in multilib build:
"ERROR: Task do_package_setscene depends upon nonexistant task
poky/meta/recipes-extended/shadow/shadow-sysroot_4.1.4.3.bb:do_populate_sysroot_setscene"
>From richard.purdie@linuxfoundation.org
Signed-off-by: Zhai Edwin <edwin.zhai@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Robert Yang [Tue, 6 Mar 2012 17:03:11 +0000 (09:03 -0800)]
logrotate 3.7.9: Allow rotate log across filesystems
The logrotate can't save the log across the different filesystems since
it used the "rename(const char *oldpath, const char *newpath)" to save
the file, fix it to act as the "mv" command(first rename, if failed,
read and write) to allow save the log across the different filesystems.
* config.c: Remove the check for different filesystems
* logrotate.c: Act as the "mv" command when rotate log
* logrotate.8: Update the mannual
* logrotate.8: Fix a bug in the mannual(\f should be \fR)
[YOCTO #718]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Bruce Ashfield [Tue, 6 Mar 2012 17:03:09 +0000 (09:03 -0800)]
linux-yocto: use bareclone option
Instead of specifying nocheckout=1, it provides a better workflow if
the linux-yocto recipes use the new bareclone=1 option to get a mirrored
copy of the source repository. This allows all local and remote branches
to be available during the kernel build and development cycle.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ken Werner [Tue, 6 Mar 2012 17:03:06 +0000 (09:03 -0800)]
image_types: add IMAGE_ROOTFS_ALIGNMENT
Introduce a new variable called IMAGE_ROOTFS_ALIGNMENT that allows to control
the aligment of the size of the rootfs. Its default value is set to 1KiB so
that the existing behaviour is not changed. In case the SD card emulation of
a QEMU system emulator gets used you may set the alignment to 2MiB.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Chase Maupin [Wed, 7 Mar 2012 20:48:52 +0000 (14:48 -0600)]
bluez4: add dependency on libcheck to enable test modules
* The later versions of bluez such as 4.98 require libcheck
in order to build the test packages as enabled by the
--enable-test option passed in EXTRA_OECONF.
* Failure to have the libcheck package available results in the
configure script ignoring the --enable-test option and not
building the test modules.
* Adding libcheck, which is greater than the minimum required
version of the check library of 0.9.6 allows building the
test modules.
* Build tested for am335x-evm MACHINE type and run time tested
by back porting this fix to the oe-classic and arago setup
and booting on am335x-evm.
Signed-off-by: Chase Maupin <Chase.Maupin@ti.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Mon, 5 Mar 2012 21:01:21 +0000 (13:01 -0800)]
base.bbclass: Fix PACKAGECONFIG handling when no flags are set
When the main PACKAGECONFIG variable was empty with no flags set,
the options were not being added to explicitly disable features.
This patch corrects that problem and ensures the disable fields
are correctly parsed and added to variables.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
soc-family.inc: to be included in machine.conf to add SOC_FAMILY to MACHINEOVERRIDE
Add a soc-family.inc file that can be included in a machine.conf to enable
the use of SOC_FAMILY in MACHINEOVERRIDE, which could be useful to group
multiple machines with the same common base. Some examples can be seen in
meta-ti BSP layer.
Signed-off-by: Denys Dmytriyenko <denys@ti.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
LSB 4.1 complain a host contamination error from libx11 because of absent
patch "makekeys_crosscompile.patch" from libx11-trim-1.4.4.
The error from LSB Test Suite is as follow:
VSW5TESTSUITE PURPOSE 7
Assertion XStringToKeysym-7.(A)
When the string argument is the name of a KeySym in the
table with the prefix XK_ removed, then a call to
XStringToKeysym returns that KeySym.
METH: For each KeySym name in table with code G:
METH: Call XStringToKeysym to obtain the KeySym defined for that string.
METH: Verify that XStringToKeysym did not return NoSymbol.
METH: Verify that the returned string is correct.
CHECK: XStringToKeysym-7 1, line 130
CHECK: XStringToKeysym-7 2, line 140
CHECK: XStringToKeysym-7 3, line 150
CHECK: XStringToKeysym-7 4, line 160
CHECK: XStringToKeysym-7 5, line 170
CHECK: XStringToKeysym-7 6, line 180
CHECK: XStringToKeysym-7 7, line 190
CHECK: XStringToKeysym-7 8, line 200
CHECK: XStringToKeysym-7 9, line 210
CHECK: XStringToKeysym-7 10, line 220
CHECK: XStringToKeysym-7 11, line 230
CHECK: XStringToKeysym-7 12, line 240
CHECK: XStringToKeysym-7 13, line 250
CHECK: XStringToKeysym-7 14, line 260
CHECK: XStringToKeysym-7 15, line 270
CHECK: XStringToKeysym-7 16, line 280
CHECK: XStringToKeysym-7 17, line 290
CHECK: XStringToKeysym-7 18, line 300
CHECK: XStringToKeysym-7 19, line 310
CHECK: XStringToKeysym-7 20, line 320
[YOCTO #1970]
Signed-off-by: Xiaofeng Yan <xiaofeng.yan@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Fri, 2 Mar 2012 11:53:26 +0000 (11:53 +0000)]
base.bbclass: If unpacking again, wipe out ${S}/patches
If we unpack again, its assumed the data in any patches directory is invalid
since do_patch will run again. This ensures old patch data doesn't get
reused in a confused way.
Ideally we should probably wipe out ${S} here but that is probably a change
for another time.
[YOCTO #2043 partially]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Thu, 1 Mar 2012 23:38:00 +0000 (23:38 +0000)]
lib/oe/patch.py: Fix and improve PatchTree() resolver logic
Currently, if PATCHRESOLVE is user and and PatchTree() is being used, you can
get backtraces if patch application fails. This is because even in the failure
case, self._current is incremented, meaning second time around, there are array
range issues.
This patch changes the code so _current is only incremented upon successful
patch application, thereby resolving this failure.
Secondly, if you bitbake -c patch -f a recipe using PatchTree(), the
clean method was unimplemented leading to patch failures.
The other part of this patch changes the logic so a series file and
set of applied patches are maintained in a quilt like fashion. This
means a the Clean method can be implemented correctly and rerunning
the patch task of an existing patches source now works reliably.
[YOCTO #2043 partially]
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Paul Eggleton [Tue, 28 Feb 2012 17:36:22 +0000 (17:36 +0000)]
qt4: make pulseaudio support conditional upon DISTRO_FEATURES
Add a pulseaudio DISTRO_FEATURES item and disable pulseaudio support in
phonon within Qt4 if it is not present; otherwise it is a race condition
as to whether it is enabled or disabled or breaks the build.
Note that this adds pulseaudio to DISTRO_FEATURES_BACKFILL thus leaving
existing distro configurations unaffected.
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
James Limbouris [Wed, 29 Feb 2012 01:13:46 +0000 (09:13 +0800)]
initscripts: Create the volatiles cache atomically.
If a machine loses power while building the volatiles cache, it will
continue to operate with an incomplete set of volatiles. Fix this by
updating atomically.
Signed-off-by: James Limbouris <james@digitalmatter.com.au> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Bruce Ashfield [Tue, 28 Feb 2012 21:09:58 +0000 (16:09 -0500)]
linux-yocto: make kernel_checkout reentrant
The steps in do_kernel_checkout modify the source tree in WORKDIR.
If it is called multiple times, or interrupted, the tree is left
in an inconsistent state.
This change adds protections around branch names, and around the
manipulations of directories to ensure that it is safe to call
at any point.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Scott Garman [Tue, 28 Feb 2012 21:45:28 +0000 (13:45 -0800)]
sudo: upgrade to 1.8.4
During the upgrade to 1.8.4, the UCB-licensed fnmatch.c was replaced
with a non-recursive BSD-licensed version, hence the removal of UCB
and addition of BSD in the LICENSE field. This led to checksum changes
in the doc/LICENSE file, and we now additionally track the comment
headers in redblack.c.
These changes were confirmed on the sudo mailing list:
Robert Yang [Fri, 2 Mar 2012 10:34:19 +0000 (18:34 +0800)]
V2 Fix libpam's chmod
The libpam's has an error when generating the rootfs:
chmod: cannot access `/usr/sbin/unix_chkpwd': No such file or directory
This is because the following code in libpam_1.1.5.bb:
pkg_postinst_pam-plugin-unix () {
# below is necessary to allow unix_chkpwd get user info from shadow file
# on lsb images
chmod 4755 ${sbindir}/unix_chkpwd
}
This is to set the setuid permission for unix_chkpwd (the lsb test
requires this), but it lacks a "${D}", and we can do this in the install
stage.
[YOCTO #2049]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Gary Thomas [Thu, 1 Mar 2012 17:41:38 +0000 (10:41 -0700)]
busybox: Restore 'date' compatability
Restore CONFIG_FEATURE_DATE_COMPAT so that all versions of 'date',
whether from busybox or coreutils, agree on the format when
setting the date from the command line.
Signed-off-by: Gary Thomas <gary@mlbassoc.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
site.conf.sample: Fix broken SOCKS proxy setup and configuration
SOCKS proxy specification with git was using conflicting methods and
thus was failing when mixed SOCKS needs were in place (requiring no
proxy for some hosts and proxy for the rest)
- GIT_PROXY_COMMAND is an environment variable GIT uses to OVERRIDE
all proxy configuration in ~/.gitconfig or any other gitconfig. By
using it to configure, it was breaking havoc on site git
configuration or the one generated by bitbake in tmp/.
Renamed to OE_GIT_PROXY_COMMAND in meta/conf/site.conf.sample
(with a doc tidbit on the name chosen), meta/classes/base.bbclass.
- The gitconfig generated by bitbake was wrong. There was a typo error
(gitproxy vs gitProxy), thus all lines were being ignored. Fixed in
meta/classes/base.bbclass.
- The gitconfig generated was being placed in
${STAGING_DIR_NATIVE}/usr/etc/gitconfig; git was looking for it in
${STAGING_DIR_NATIVE}/etc/gitconfig. Fixed that in
meta/classes/base.bbclass, at the same time creating a
GIT_CONFIG_PATH variable, since it is also referenced in
generate_git_config() and have all instances refer to that.
Signed-off-by: Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
