Jussi Kukkonen [Mon, 31 Oct 2016 13:40:18 +0000 (15:40 +0200)]
distcc: Don't remove users/groups in postrm
There's no way to ensure that files owned by the users aren't left
on the system at postrm time: Removing the user would mean those
files are now owned by a non-existing user, and later may be owned
by a completely unrelated new user.
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jussi Kukkonen [Mon, 31 Oct 2016 13:40:17 +0000 (15:40 +0200)]
avahi: Don't remove users/groups in postrm
There's no way to ensure that files owned by the users aren't left
on the system at postrm time: Removing the user would mean those
files are now owned by a non-existing user, and later may be owned
by a completely unrelated new user.
[YOCTO #10442]
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jussi Kukkonen [Mon, 31 Oct 2016 13:40:16 +0000 (15:40 +0200)]
e2fsprogs: Depend on attr
e2fsprogs builds fine with or without attr but it's possible to
end up with this sequence:
* e2fsprogs configure finds attr/xattr.h (coming from sstate)
* attr starts rebuild, attr/xattr.h is removed from sysroot
* e2fsprogs compile fails: "attr/xattr.h: No such file or directory"
Depend on attr to ensure reproducible build.
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
oeqa/selftest/kernel.py: Add new file destined for kernel related tests
[YP#7202]: Test for linux-dummy
The new kernel.py file is intended for kernel related test cases.
The test for linux-dummy will ensure it is in good shape and can
be used as a kernel replacement at build time. To do this, the
test will first clean sstate for linux-dummy target, ensuring no
file is present in the stamps directory. After, core-image-minimal
is built, ensuring linux-dummy can be used as a kernel substitute.
Signed-off-by: Costin Constantin <costin.c.constantin@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Jussi Kukkonen [Wed, 26 Oct 2016 08:37:39 +0000 (11:37 +0300)]
xserver-xf86-config: Remove legacy drivers from xorg.conf
Trying to configure evdev and mouse leads to errors in X startup
because xf86-input-libinput is now the default. No configuration
should be needed: xf86-input-libinput should just work out-of-the-box.
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Jussi Kukkonen [Wed, 26 Oct 2016 08:37:38 +0000 (11:37 +0300)]
conf: Use xf86-input-libinput by default
Don't install legacy X input drivers for any machines by default,
RRECOMMEND xf86-input-libinput instead.
This is the setup suggested by upstream: install only libinput by
default, but let niche legacy drivers sort higher in configuration
so they get chosen if installed. So the order is:
evdev < libinput < (synaptics|vmmouse|...)
This also removes vmmouse X driver from the qemu config. If a VMware
virtual mouse device really needs to be supported, we should enable
CONFIG_MOUSE_PS2_VMMOUSE in kernel instead: that is directly supported
by the libinput X driver.
Fixes [YOCTO #10195].
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Andreas Müller [Thu, 27 Oct 2016 20:07:19 +0000 (22:07 +0200)]
libsdl2: update to 2.0.5
* 0001-src-video-make-it-compatible-with-wayland-1.10.patch is removed because
the problem has been resolved upstream
* Add 0001-prepend-our-sysroot-path-so-that-make-finds-our-wayl.patch
so that WAYLAND_PROTOCOLS_SYSROOT_DIR is substituted in a way similar to gtk3
Signed-off-by: Andreas Müller <schnitzeltony@googlemail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Olaf Mandel [Fri, 21 Oct 2016 14:06:08 +0000 (14:06 +0000)]
Remove LIC_FILES_CHKSUM from recipes without SRC_URI
LICENSE and LIC_FILES_CHKSUM apply to the sources specified by SRC_URI,
not to the recipe itself. As such a license declaration for a source-less
recipe makes little sense. The LICENSE declaration is mandatory, but
LIC_FILES_CHKSUM can be removed in such cases.
Remove the LIC_FILES_CHKSUM declarations from all recipes that do not
need it.
CC: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Olaf Mandel <o.mandel@menlosystems.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The change to get rid of FuncFailed exceptions changed the behavior
of how missing uid/gid error are be handled. Instead of catching
the exception and handling that via bb.parse.SkipPackage(), a fatal
error was called.
This won't work with recipes that are unused and therefore do not have
UID/GIDs defined. The problem triggers when parsing all recipes (e.g.,
oe-selftest runs bitbake -p).
The right way to handle this is to raise bb.parse.SkipPackage(). This
will error correctly once the recipe is needed.
Signed-off-by: Mikko Ylinen <mikko.ylinen@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Yuanjie Huang [Fri, 21 Oct 2016 02:47:52 +0000 (19:47 -0700)]
binutils: enable warn system direcotories in cross-canadian ld
Warn system directories requires a configuration option to be enabled.
This patch enables the warning for cross-canadian ld, to align with the
cross version.
Signed-off-by: Yuanjie Huang <yuanjie.huang@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
This class enables 'manpages' feature in packages if 'api-documentation'
is in distro features. This ensures that manpages are always built
and installed when API documentation feature is enabled.
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Martin Jansa [Wed, 19 Oct 2016 15:31:56 +0000 (17:31 +0200)]
test-dependencies.sh: Strip also '\.bb: .*' before adding failed recipe to list of failed
* format of bitbake tasks changed in: 2c88afb taskdata/runqueue: Rewrite without use of ID indirection
-ERROR: Task 4 (/OE/build/oe-core/openembedded-core/meta/recipes-devtools/rpm/sftp.bb, do_fetch) failed with exit code '1'
+ERROR: Task /OE/build/oe-core/openembedded-core/meta/recipes-devtools/rpm/sftp.bb:do_fetch (/OE/build/oe-core/openembedded-core/meta/recipes-devtools/rpm/sftp.bb:do_fetch) failed with exit code '1'
so strip not only '\.bb, .*' used before, but also '\.bb:.*' to drop
the task name to get recipe name.
* for more details see:
http://lists.openembedded.org/pipermail/openembedded-core/2016-June/123132.html
* without this change you can see test-dependencies.sh trying to rebuild packages
like:
Building recipe: fbprogress (6/21)
Building recipe: fbprogress.bb:do (7/21)
where the later of course doesn't exist as a recipe
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Marek Vasut [Tue, 18 Oct 2016 21:55:09 +0000 (23:55 +0200)]
u-boot: Factor out common bits
The u-boot recipes share a couple of common variables, which makes
updating of the recipes error prone and a toil. Factor those common
bits into u-boot-common_${PV}.inc so that they are in one place.
The u-boot_${PV}.bb now explicitly require u-boot-common_${PV}.inc
before require u-boot.inc , as doing require u-boot-common_${PV}.inc
from u-boot.inc would fail due to ${PV} not being set. This is a
functional change.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Richard Purdie <richard.purdie@linuxfoundation.org> Cc: Denys Dmytriyenko <denis@denix.org> Signed-off-by: Ross Burton <ross.burton@intel.com>
Remove backported 0001-Use-correct-include-dir-with-minitasn.patch and
CVE-2016-7444.patch (which still applied silently and incorrectly:
https://bugzilla.yoctoproject.org/show_bug.cgi?id=10450).
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
buildhistory: fix latest_srcrev in the common case
buildhistory was writing srcrevs.values() as SRCREV when only one
srcrev/branch exists. This returns a view of the dictionary values in python
3, and used to return a list in python 2, neither of which is an appropriate
value for SRCREV. It was resulting in latest_srcrev files like this:
Olaf Mandel [Mon, 17 Oct 2016 08:16:28 +0000 (08:16 +0000)]
combo-layer: handle ambiguous git arguments
If a branch/src-repository has the same name as a file/directory, git
since 1.4.0(?) gives an error like the one below:
ambiguous argument 'bitbake': both revision and filename
Use '--' to separate paths from revisions, like this:
'git <command> [<revision>...] -- [<file>...]'
Add two dashes to make the intent clear.
Signed-off-by: Olaf Mandel <o.mandel@menlosystems.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Olaf Mandel [Mon, 17 Oct 2016 08:16:27 +0000 (08:16 +0000)]
combo-layer: fix crashes on wrong tempfile usage
When calling tempfile.NamedTemporaryFile().write(str()), at least on
Python 3.4.2 this fails with this error:
TypeError: 'str' does not support the buffer interface
Change the file-mode for all such files from binary to text mode.
Signed-off-by: Olaf Mandel <o.mandel@menlosystems.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Olaf Mandel [Fri, 21 Oct 2016 09:24:28 +0000 (09:24 +0000)]
Remove $COREBASE/LICENSE from LIC_FILES_CHKSUM
Several recipes reference the LICENSE file in their LIC_FILES_CHKSUM
variable as ${COREBASE}/LICENSE. This forces distribution providers to
keep this file verbatim or to overload the affected recipes. The section
"Moving to the Yocto Project 1.6 Release" in the Yocto manual suggests
removing the LICENSE file where possible.
Remove LICENSE in cases where COPYING.MIT is also given and replace
LICENSE with COPYING.MIT if the former was the only entry. All modified
recipes specify LICENSE = "MIT" and none of the in-tree files specify a
different license either.
As the packages do not change (the license files are not contained in
them), do not increase PR.
Signed-off-by: Olaf Mandel <o.mandel@menlosystems.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Paul Eggleton [Tue, 18 Oct 2016 02:51:16 +0000 (15:51 +1300)]
devtool: runqemu: work around runqemu script path assumption
The new runqemu script assumes that if OECORE_NATIVE_SYSROOT is set then
it shouldn't try to run bitbake to find out the values of various
variables such as DEPLOY_DIR_IMAGE; this assumption is incorrect for the
extensible SDK. To work around this, clear OECORE_NATIVE_SYSROOT in the
environment when running runqemu.
Fixes [YOCTO #10447].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
