when X11 is not in distro features it tries to link with
libGL which does not exist and compile fails
Add a patch to add foreign option to AUTOMAKE_OPTIONS
which lets it configure and relax automake rules to not
check for some GNU specific files
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
When a connection upgrade is requested and when the
request sent an error reply, it happened most often
that a segmentation fault occured.
The patch applied here is a backport from the
upstream (see https://gnunet.org/git/libmicrohttpd.git/
commit b4216c60fdb5b48f6cfec416301fc63a1167e6cd).
Change-Id: I6847550ba2c4fc24d5caf8912ac6d5ac89ae01fb
Signed-off-by: José Bollo <jose.bollo@iot.bzh>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
With opencv-3.2, the configuration detects if the target supports
half-precision floating-point format. This fails to compile for some
Intel targets such as skylake with an error such as:
error: '_mm_cvtph_ps' was not declared in this scope
The configuration used to work in opencv-3.1 so revert two commits to
drop the FP16 detection even though it may make opencv slower.
The only change in the configure log is:
- FP16: Compiler support is available
+ FP16: Compiler support is not available
The patch should be dropped when a newer version of opencv that includes commit:
e5d9b608c4
is available. Backporting that fix was complicated.
Signed-off-by: Randy MacLeod <Randy.MacLeod@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
With opencv-3.2, the configuration detects if the target supports
half-precision floating-point format. This fails to compile for some
Intel targets such as skylake with an error such as:
error: '_mm_cvtph_ps' was not declared in this scope
The configuration worked in opencv-3.1 so revert two commits to
drop the FP16 detection even though it may make opencv slower.
The only change in the configure log is:
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Utility for IPMI control which is for interfacing with
devices that support the Intelligent Platform Management
Interface specification.
Signed-off-by: Jackie Huang <jackie.huang@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
OpenIPMI is an effort to create a full-function IPMI
system to allow full access to all IPMI information
on a server and to abstract it to a level that will
make it easy to use.
Signed-off-by: Jackie Huang <jackie.huang@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Now we've separted libdevmapper out of lvm2 recipe, we need to change
multipath-tools to depend on libdevmapper instead of lvm2 to avoid
build error like below.
fatal error: libdevmapper.h: No such file or directory
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
When 'cryptsetup' PACKAGECONFIG is enabled for systemd, there would
be circular dependency issue among 'systemd', 'cryptsetup' and 'lvm2'.
To fix this problem, we've separated libdevmapper into a new recipe.
Let cryptsetup depend on libdevmapper instead of lvm2 to fix the issue.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
If we enable 'cryptsetup' PACKAGECONFIG for systemd, there would be
circular dependency issue among 'systemd', 'cryptsetup' and 'lvm2'.
cryptsetup only needs libdevmapper from lvm2. So we separte libdevmapper
into a new recipe to solve this circular dependency issue.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Don't blindly assume base_sbindir is /sbin. Pass its real value
instead to the build process to make sure binaries end up in the
right place even when base_sbin happens to point somewhere else.
Signed-off-by: Krisztian Litkey <krisztian.litkey@intel.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Update mcelog to 1.49 to support Denverton SoC, and delete the additional
patches which are already involved in v1.49.
Signed-off-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
* Use cmake as build system. Projects also using cmake and depending on fltk
would not find all required components. Note that FindFLTK.cmake is part of
cmake.
* Add a native recipe and split out common - we need fluid (Fast Light User
Interface Designer) as working binary
Signed-off-by: Andreas Müller <schnitzeltony@googlemail.com>
Subpackage python3-${PN} is not right for multilib and it will be
extended to such as lib32-python3-lib32-opencv.
Replace PN with BPN to fix it.
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
The current setting of S is not right for multilib. Remove the setting
and use the default value.
And library install directory is not right for multilib either. It fails
with [installed-vs-shipped] QA error:
| ERROR: gflags-2.2.0-r0 do_package: QA Issue: gflags: Files/directories
| were installed but not shipped in any package:
| /usr/lib/libgflags.so
| /usr/lib/libgflags_nothreads.so.2.2
Set cmake variable LIB_INSTALL_DIR with ${baselib} to fix the issue.
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Additional patches are upstream to fix build with gcc6 and clang
Add systemd unit file
Create missing symlink for dev shared object
Signed-off-by: Khem Raj <raj.khem@gmail.com>
* based on discussion in pndeprecated thread:
https://patchwork.openembedded.org/patch/137573/
update the messages to warn possible users that the
recipe will be removed before the end of the next development
cycle (before Yocto 2.4 is released).
* updated with:
sed -i 's/^\(PNBLACKLIST.*".*\)"/\1 - the recipe will be removed on 2017-09-01 unless the issue is fixed"/g' `git grep PNBLACKLIST | sed 's/:.*//g' | sort -u | xargs`
* then noticed couple recipes being blacklisted only based on
DISTRO_FEATURES, so removed those:
meta-networking/recipes-support/lksctp-tools/lksctp-tools_1.0.17.bb
meta-oe/recipes-connectivity/bluez/bluez-hcidump_2.5.bb
meta-oe/recipes-connectivity/bluez/bluez4_4.101.bb
meta-oe/recipes-connectivity/bluez/gst-plugin-bluetooth_4.101.bb
meta-oe/recipes-navigation/foxtrotgps/foxtrotgps_1.1.1.bb
meta-oe/recipes-navigation/gypsy/gypsy.inc
meta-oe/recipes-navigation/navit/navit.inc
meta-oe/recipes-support/opensync/libsyncml_0.5.4.bb
* if it isn't fixed by this date, it's fair game to be removed
whenever someone gets around to i
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
* fixes:
No GNU_HASH in the elf binary:
'/tmp/work/armv5te-oe-linux-gnueabi/pngcheck/2.3.0-r0/packages-split/pngcheck/usr/bin/pngcheck'
[ldflags]
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
The recipe opencv_3.2.bb inherits to either distutils-base or
distutils3-base classes or both which do add appropriate
build dependency on python or python3 respectively.
There's no need to have python in DEPENDS explicitly.
It can even make things needlessly complicated in case a user
wants to build python3-only images.
The redundant build dependency on python is removed from DEPENDS.
Signed-off-by: Dmitry Rozhkov <dmitry.rozhkov@linux.intel.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
it has a tclsh need for executing a script during build
Fixes
| ./gen-rules.tcl
| /usr/bin/env: ‘tclsh’: No such file or directory
| make: *** [Makefile:26: 40-usb_modeswitch.rules] Error 127
| WARNING: /mnt/b/build/tmp-musl/work/all-oe-linux/usb-modeswitch-data/20170205-r0/temp/run.do_compile.30695:1 exit 1 from 'exit 1'
Signed-off-by: Khem Raj <raj.khem@gmail.com>
