* the section bellow the comment adds only HF variants, VFP is already
mixed in the softfp sections above (unlike armv5, armv6 tune files
where it really was above VFP/DSP section)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Martin Jansa [Fri, 18 Dec 2015 13:14:24 +0000 (14:14 +0100)]
sanity.bbclass: add more information to error message about TUNE_PKGARCH missing in PACKAGE_ARCHS
* sometimes it's hard to see what's wrong here
* error message:
Error, the PACKAGE_ARCHS variable does not contain TUNE_PKGARCH (cortexa7t2hfhf-vfp-vfpv4-neon).
doesn't help much to understand what exactly went wrong and where
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Trevor Woerner [Sat, 28 Nov 2015 14:29:51 +0000 (09:29 -0500)]
mkefidisk.sh: add boot log on console
Hooking up a serial console is a "developer mode", the chances are pretty good
developers are interested in watching the kernel boot log on the console so
they can spot any problems or diagnose any failed boots (e.g. can't find root
fs).
Signed-off-by: Trevor Woerner <twoerner@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Trevor Woerner [Sat, 28 Nov 2015 14:43:02 +0000 (09:43 -0500)]
mkefidisk.sh: add startup script for automated boot
Regardless of which image is built using which layers, try to ensure the image
on the SD device being prepared is the one that is booted automatically when
the board is powered.
Signed-off-by: Trevor Woerner <twoerner@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Maxin B. John [Wed, 2 Dec 2015 08:39:35 +0000 (10:39 +0200)]
libjpeg: Replace libjpeg with libjpeg-turbo
Moving libjpeg-turbo from meta-oe as a replacement for libjpeg
package. libjpeg-turbo has same API/ABI as libjpeg. It is
relatively faster in JPEG compression/decompression than libjpeg.
libjpeg-turbo doesn't support x32 ABI. Work around missing x32 ABI support by
building with "--without-simd" option.
Move to recipes-graphics [RB]
[YOCTO #8628]
Signed-off-by: Maxin B. John <maxin.john@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Robert Yang [Wed, 30 Dec 2015 01:39:11 +0000 (17:39 -0800)]
python3: fix installed-vs-shipped when 64bit + multilib
Fixed:
python3-3.5.0: python3: Files/directories were installed but not shipped in any package:
/usr/lib
/usr/lib/python3.5
/usr/lib/python3.5/config-3.5m
/usr/lib/python3.5/config-3.5m/python.o
/usr/lib/python3.5/config-3.5m/Setup.local
/usr/lib/python3.5/config-3.5m/Setup
/usr/lib/python3.5/config-3.5m/python-config.py
/usr/lib/python3.5/config-3.5m/install-sh
/usr/lib/python3.5/config-3.5m/config.c.in
/usr/lib/python3.5/config-3.5m/makesetup
/usr/lib/python3.5/config-3.5m/Setup.config
/usr/lib/python3.5/config-3.5m/config.c
/usr/lib/python3.5/config-3.5m/Makefile
/usr/lib/python3.5/config-3.5m/libpython3.5m.a
Please set FILES such that these items are packaged. Alternatively if they are unneeded, avoid installing them or delete them within do_install.
python3: 14 installed and not shipped files. [installed-vs-shipped]
Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Martin Jansa [Mon, 4 Jan 2016 14:13:14 +0000 (15:13 +0100)]
sstate-sysroot-cruft.sh: Extend the whitelist
* add more php5 entries
* add builder user (from builder.bb)
* add .bashrc and .profile in HOME directory of created user
* add imgdata directory generated by image.py
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Maxin B. John [Tue, 29 Dec 2015 13:11:59 +0000 (15:11 +0200)]
iptables: upgrade to 1.6.0
1.4.21 -> 1.6.0
xtables_globals structure layout has changed.
* Refreshed below listed patches to work with this release:
1. 0001-configure-Add-option-to-enable-disable-libnfnetlink.patch
2. 0001-fix-build-with-musl.patch
* Added PACKAGECONFIG for libnftnl
Signed-off-by: Maxin B. John <maxin.john@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Daniel Istrate [Tue, 5 Jan 2016 14:53:34 +0000 (16:53 +0200)]
scripts/oe-selftest: Allow to run tests on random/all MACHINEs
Add an option for random MACHINE into oe-selftest:
--machine [random/all]
1. random: will set a random MACHINE for each test
2. all: will run tests for all machines
Custom machine sets only weak default values (??=) for MACHINE in machine.inc.
This let test cases that require a specific MACHINE to be able to
override it, using (?= or =).
e.g.:
oe-selftest --run-tests signing --machine random -->
will run all tests switching MACHINE randomly for each test
oe-selftest --run-tests signing --machine all -->
for each machine will run all tests
oe-selftest --run-all-tests --machine random
Also update oeqa/selftest/base.py to accomodate this feature.
Fix for [YOCTO #5880].
Signed-off-by: Daniel Istrate <daniel.alexandrux.istrate@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Daniel Istrate [Mon, 4 Jan 2016 13:43:56 +0000 (15:43 +0200)]
oe-selftest: Improved --list-classes when determining test names
--list-classes does a weak validation when determining test names:
(if method.startswith("test_") which could report any class attribute
that starts with 'test_' as a valid test case.
This fix checks that the class attribute that starts with 'test_'
is also callable (is a method).
Fix for [YOCTO #8862]
Signed-off-by: Daniel Istrate <daniel.alexandrux.istrate@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Humberto Ibarra [Wed, 23 Dec 2015 23:36:08 +0000 (17:36 -0600)]
scripts/oe-selftest: Remove extra coverage data added to unittests
Coverage data tracking initiates too early, causing coverage data from the
oe-selftest environment setting to be added to each run. Even when no tests are run
oe-selftest reports around 24% of coverage due to this extra data.
Change the custom resultclass used by the TextTestRunner to one generated from the
command arguments. The generated class processes coverage when needed, running
coverage setup just before the first testcase is run and reporting after the last
one finished.
[Yocto #8846]
Signed-off-by: Humberto Ibarra <humberto.ibarra.lopez@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Noor Ahsan [Wed, 23 Dec 2015 13:50:35 +0000 (15:50 +0200)]
wic: rawcopy: Copy source file to build folder
When a file is given using --sourceparams then wic directly use that file
instead of copying them to build folder. At time of assembling it os.rename
is called which renames all the files to name. In that process the original
file is renamed. When image recipe is rebuilt then wic complains about
missing file which was renamed in previous build.
[YOCTO #8854]
Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ross Burton [Mon, 4 Jan 2016 10:20:25 +0000 (10:20 +0000)]
systemd: add more compression and importd PACKAGECONFIGs
Add explicit PACKAGECONFIGs for all of compression formats that systemd can
detect at configure time, and an explicit enable/disable for importd. importd
defaults to disabled as it needs curl and all the compression formats to be
enabled.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Ross Burton [Mon, 4 Jan 2016 13:02:11 +0000 (13:02 +0000)]
gcc-sanitizers: link directly against sysroot libstc++
Instead of building a shadow libstdc++-v3 directory with symlinks to the sysroot
libstdc++-v3.la, fiddle the Makefiles so that it doesn't attempt to link to a
in-tree library at all.
This fixes builds where .la files are not being installed into the sysroot at
all.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Phil Blundell [Tue, 14 May 2013 10:59:37 +0000 (11:59 +0100)]
openjade: Fix build if not installing libtool .la files
The openjade build system is slightly funky and does strange things with library
dependencies. Fix it up so that it depends on the .so rather than the .la
(which may not exist) for libraries from the sysroot.
Signed-off-by: Phil Blundell <philb@brightsign.biz> Signed-off-by: Ross Burton <ross.burton@intel.com>
Andre McCurdy [Tue, 22 Dec 2015 23:57:58 +0000 (15:57 -0800)]
security_flags.inc: disable -fstack-protector-XXX for valgrind
Valgrind (v3.11.0) expects to build with stack protection disabled
and includes -fno-stack-protector in its default CFLAGS. However, the
CFLAGS provided by OE are included on the compiler command line after
the defaults so any -fstack-protector-all / -fstack-protector-strong
option provided by security_flags.inc will cause problems.
| .../build-bcm97425vms/tmp/work/mips32el-rdk-linux/valgrind/3.11.0-r0/valgrind-3.11.0/coregrind/m_mallocfree.c:892: undefined reference to `__stack_chk_guard'
| .../build-bcm97425vms/tmp/work/mips32el-rdk-linux/valgrind/3.11.0-r0/valgrind-3.11.0/coregrind/m_mallocfree.c:947: undefined reference to `__stack_chk_fail'
Signed-off-by: Andre McCurdy <armccurdy@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Qt4 is no longer supportd upstream, but it is still needed for LSB compliancy.
Qt4 recipes have been moved to a separate meta-qt4 layer to be consistent with
meta-qt3 and meta-qt5.
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Jens Rehsack [Tue, 5 Jan 2016 16:11:42 +0000 (17:11 +0100)]
udev-extraconf: introduce multiple blacklist files for more complex setups
In cases where a target image wants prevent the recovery partition is mounted
automatically, but the recovery partition identifier moves with the device
(internal flash, sd card, usb stick, ...), device/machine dependend extra
blacklists might be desired.
The grep utility prints the file name for each match when there is more
than one file to search. Add -h to suppress the prefixing of file names
on output.
Signed-off-by: Jens Rehsack <sno@netbsd.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Tue, 5 Jan 2016 17:52:06 +0000 (17:52 +0000)]
image/rootfs-postcommands: Separate out post rootfs commands to separate class
Reading image.bbclass is a little difficult as it has many post rootfs
helper functions and its hard to separate those from the core contents
of the rootfs/image code.
Moving it to a separate class would be one way of making it clearer
what these functions are. There are some comment layout improvements
but no code changes.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Alejandro Joya [Tue, 17 Nov 2015 22:13:16 +0000 (16:13 -0600)]
packagegroup-core-boot:replace busybox to variable
It replaces the use of busybox as hardcoded dependency to more dynamic
this wouldn't affect the way that the system is build only it let a more
flexible replacement in the core.
Signed-off-by: Alejandro Joya <alejandro.joya.cruz@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Alejandro Joya [Tue, 17 Nov 2015 22:13:15 +0000 (16:13 -0600)]
initramfs-framework_1.0:replace busybox for variable.
It replaces the use of busybox as hardcoded dependency to more dynamic
this wouldn't affect the way that the initrams is build, just it let a more
flexible replacement in the core.
Signed-off-by: Alejandro Joya <alejandro.joya.cruz@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Alejandro Joya [Tue, 17 Nov 2015 22:13:14 +0000 (16:13 -0600)]
core-image-minimal-initramfs: replace base-utils
It replaces the use of busybox as hardcoded dependency to more dynamic
this wouldn't affect the way that the system is build only it let a more
flexible replacement in the core.
Signed-off-by: Alejandro Joya <alejandro.joya.cruz@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Wed, 23 Dec 2015 13:45:23 +0000 (13:45 +0000)]
oeqa/selftest: support sets in devtool comparisons
The devtool and recipetool tests do literal string comparisons, but for some
fields the ordering could be irrelevant and potentially non-deterministic. For
example, the recipetool_create_simple test started failing with:
AssertionError: 'GPLv2 Unknown' != 'Unknown GPLv2' : values for LICENSE do not match
The ordering of the LICENSE field isn't relevant. So, if the expected value is
a set(), split the string into a set too and compare those.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Ross Burton [Tue, 22 Dec 2015 22:12:30 +0000 (22:12 +0000)]
xorg-lib: allow native building without x11 DISTRO_FEATURES
The Xorg libraries use REQUIRED_DISTRO_FEATURES to stop building on
distributions without the x11 feature but this stops people building native
tooling that uses libX11, such as libsdl-native.
Signed-off-by: Ross Burton <ross.burton@intel.com>
busybox: generalize recipe to work with arbitrary install directories
Allow configuring base_bindir, base_sbindir, bindir and sbindir
arbitrarily.
Also change the temporary symlinks created in the postinst script from
relative to absolute to make the code simpler and more flexible. There was
no reason to use relative links anyways, as directly afterwards
update-alternatives would replace them by absolute links anyways.
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net> Signed-off-by: Ross Burton <ross.burton@intel.com>
Andre McCurdy [Tue, 22 Dec 2015 19:56:56 +0000 (11:56 -0800)]
cairo: update 1.14.4 -> 1.14.6
Release 1.14.6 (2015-12-09 Bryce Harrington <bryce@osg.samsung.com>)
========================================================================
Simple bugfix release to fix one Windows issue.
For a complete log of changes since 1.14.4, please see:
Mark Hatle [Thu, 19 Nov 2015 03:46:02 +0000 (19:46 -0800)]
rpm: Generate per distribution and multilib macro files
The distribution and multilib macro files are required for rpmbuild, on the
target, to work properly. These override the default settings from the
upstream macro files with the proper values for the OE configuration.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com> Signed-off-by: Robert Yang <liezhi.yang@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Patrick Ohly [Wed, 9 Dec 2015 17:48:28 +0000 (18:48 +0100)]
package_manager.py: add debugging support for rpm scriptlet execution
By default, smart is invoked with --log-level=warning, which hides all
output from pre/post install scriptlets. That makes it hard to debug
scriptlet failure or why they get postponed to first-boot via
/etc/rpm-postinst.
The new ROOTFS_RPM_DEBUG variabled is expected to be set to an integer in
local.conf an incrementally adds more output:
0 = default, only warnings
1 = --log-level=info (includes information about executing scriptlets and their output)
2 = --log-level=debug
3 = --log-level=debug plus dumps of scriplet content and command invocation
The default behavior is not changed yet, but it seems that level 1 would
be a better default.
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Jonathan David [Mon, 30 Nov 2015 21:59:50 +0000 (15:59 -0600)]
xinput-calibrator: get screen geometry when calibrating
Calibration acquires a default resolution of 1280x1024 when running
xinput_calibrator_pointercal.sh, obtain correct screen resolution by
using xrandr to grab user-specified geometry and pass to script
Signed-off-by: Jonathan David <jonathan.david@ni.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Historical releases may disappear from the main mirrors, so
then we need to fall back to the archive (which, in turn, may
not carry the current release, and so can't be the primary mirror).
Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Kai Kang [Tue, 22 Dec 2015 01:04:54 +0000 (09:04 +0800)]
bind: 9.10.2-P4 -> 9.10.3-P2
Upgrade bind from 9.10.2-P4 to 9.10.3-P2.
* update context of 0001-build-use-pkg-config-to-find-libxml2.patch
* add PACKAGECONFIGs readline and libedit. They provide same library, so
should not be set at same time.
Signed-off-by: Kai Kang <kai.kang@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Tanu Kaskinen [Tue, 22 Dec 2015 09:49:36 +0000 (11:49 +0200)]
libsndfile1: enable FLAC/Ogg/Vorbis support
This makes it possible to handle FLAC and Vorbis files in programs
that use libsndfile.
It seems that the external dependencies were disabled just as a
routine measure to make builds deterministic, since the dependencies
were missing from DEPENDS[1]. Therefore it seems likely that nobody
particularly wants to have libsndfile without support for these
formats.
Paul Eggleton [Sun, 20 Dec 2015 20:55:39 +0000 (09:55 +1300)]
buildhistory: improve support for extensible SDK
* Ensure extensible SDK and standard SDK go into their own directories
* Record extra extensible SDK variables
* Write sstate-package-sizes.txt and sstate-task-sizes.txt files so you
can analyse the size of the contents
* Add BUILDHISTORY_SDK_FILES (similar to BUILDHISTORY_IMAGE_FILES) and
default it to pick up config files installed in the extensible SDK
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Paul Eggleton [Sun, 20 Dec 2015 20:04:59 +0000 (09:04 +1300)]
buildhistory: fix not recording SDK information
After OE-Core revision baa4e43a29e45df17eaa3456acc179b08d571db6 we lost
recording SDK the contents in buildhistory. This was due to the
SDK_POSTPROCESS_COMMAND variable being set with = in
populate_sdk_base.bbclass which overwrote any value set with += in
buildhistory.bbclass; to fix it, use _append in buildhistory.bbclass
instead.
Fixes [YOCTO #8839].
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
scripts/oe-selftest: Add support for selftest log with timestamp
Each time oe-selftest runs, the oe-selftest.log file is overwritten.
This patch solves it by adding time stamp to each selftest log file
and doing a symlink named as oe-selftest.log to the last one created.
Signed-off-by: Costin Constantin <costin.c.constantin@intel.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ming Liu [Sun, 20 Dec 2015 16:40:43 +0000 (17:40 +0100)]
ncurses: fixes wrong paths in BINCONFIG
* Removes ncurses-config from BINCONFIG, that file does not exist at all.
* Adds in ncurses5-config and ncursesw5-config to BINCONFIG.
This fixed following warnings during SDK install:
warning: Removing ncurses-dev-5.9-r15.1@lib32_x86 due to file /usr/bin/ncurses5-config conflicting with ncurses-dev-5.9-r15.1@core2_64
warning: Removing ncurses-dev-5.9-r15.1@lib32_x86 due to file /usr/bin/ncursesw5-config conflicting with ncurses-dev-5.9-r15.1@core2_64
[YOCTO #8801]
Signed-off-by: Ming Liu <peter.x.liu@external.atlascopco.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ross Burton [Fri, 11 Dec 2015 15:08:57 +0000 (15:08 +0000)]
xcb: don't build-depend on python-native
xcb-proto had a dependency on python-native as it builds and installs a Python
module, but xcb-proto is very low in the build to force a dependency on
python-native.
As it turns out libxcb simply asks pkg-config for the full path to the module
xcb-proto installed and uses it directly so there is no need to build
python-native. Instead replace AM_PATH_PYTHON with two explicit variable
definitions (need to set PYTHON for automake's install logic).
Also remove the dependency on xcb-proto-native as libxcb uses the data files and
tooling from the target sysroot, and patching it to use the native sysroot adds
complication for no good reason.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Khem Raj [Tue, 15 Dec 2015 21:20:46 +0000 (21:20 +0000)]
tcmode-default: Use glibc for nativesdk version even on uclibc and musl
We do not have musl or uclibc based systems for building OE itself. Most
of build servers run glibc, there will be other issues to build OE on a
uclibc based build system
Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Bruce Ashfield [Thu, 26 Nov 2015 14:05:07 +0000 (09:05 -0500)]
linux-yocto: remove 3.14 and 3.19 recipes
In preparation for the introduction of new (supported) kernel versions,
we remove the 3.14 and 3.19 variants.
The kernel trees will still be available, but the recipes are removed
from master. These versions can still be built on older releases, or
with private copies of the recipes.
3.14 LTSI is replaced by the 4.1 LTSI kernel, and 3.19 will be replaced
by 4.4+.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
The existing code doesn't tell regular (with .git) and bare cases and
just move the unpacked repo to the place of kernel source. But later
steps will fail on a bare-cloned repo because we can not checkout
directly in a bare cloned repo.
This change performs another clone to fix the issue.
Note: This change doesn't cover the case that S and WORKDIR are same
and the repo is bare cloned.
Signed-off-by: Jianxun Zhang <jianxun.zhang@linux.intel.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Bruce Ashfield [Fri, 11 Dec 2015 18:30:10 +0000 (13:30 -0500)]
yaffs2: fix checkpoint functionality
Updating the linux-yocto SRCREVs to pull in the following change:
Author: He Zhe <zhe.he@windriver.com>
Date: Wed Dec 2 01:31:31 2015 -0500
fs/yaffs2: fix missing checkpoint on yaffs
For yaffs file system, the mode of reading or writing is restricted
at four pointer where are mnt->mnt_flags,mnt->mnt_sb->s_flags,mtd->flags and
dev->read_only,the first three is used handle file and file
system(eg,remount) operation, and last one(dev->read_only) almost is
used handle checkpoint of yaffs2. However, in current code, the
dev->read_only only can be changed at first time when the yaffs2
file system is mounted, later it can't be changed again(eg,mount -o
remount), the result is that the checkpoint's saving operation
always can't succeed if you set readonly mode for yaffs2 file system
when it is mounted at the first time.
To fix this issue, we implement yaffs_remount_fs() which allows the
rootfs to be remounted as r/w.
Signed-off-by: Wenlin Kang <wenlin.kang@windriver.com> Signed-off-by: He Zhe <zhe.he@windriver.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com> Signed-off-by: Ross Burton <ross.burton@intel.com>
Ross Burton [Fri, 18 Dec 2015 16:56:18 +0000 (16:56 +0000)]
conf/local.conf.sample: comment out ASSUME_PROVIDED=libsdl-native
Ubuntu 15.10 and Debian testing can't build qemu-native against the host libsdl.
Now that libsdl-native is buildable, comment out the ASSUME_PROVIDED which meant
it wouldn't be used.
[ YOCTO #8553 ]
Signed-off-by: Ross Burton <ross.burton@intel.com>
