These can cause build problems. We only want macros in acinclude.m4 which
aren't available for aclocal to get elsewhere.
Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
The networkmanager build runs dbus-binding-tool, which only exists if
dbus-glib-native has been built, and will fail otherwise.
Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Without it, we can hit errors like:
| CMake Error at .../build/tmp/sysroots/qemux86-64/usr/share/cmake/Modules/OpenLMIMacros.cmake:105 (message):
| KonkretCMPI failed: No such file or directory
Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Relying on 'make -e' isn't ideal, as it's less implicit behavior, relies on
variable exports, and hides the variables being used from someone reading the
recipe.
Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
The llvm-config wrapper has an error message that is quite abrupt, expand it so
it helps the user.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
geany-plugins comes around as a loose collection of modules. I thought about
using PACKAGES_DYNAMIC but this way seems better maintainable:
* a removed plugin will at least complain for license file missing
* a new plugin will complain for unpacked files
Signed-off-by: Andreas Müller <schnitzeltony@googlemail.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
The do_compile requires it, otherwise it can't find the native module
correctly when perl-native is mirrored from another build.
Fixed:
| Compilation failed in require.
| BEGIN failed--compilation aborted.
| Can't locate loadable object for module File::Glob in @INC
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Fix opie-image:
ERROR: Nothing PROVIDES 'obexftp'
ERROR: obexftp was skipped: Recipe is blacklisted: bluez5 conflicts with
bluez4 and bluez5 is selected in DISTRO_FEATURES
Somehow on 2015-05-11 obex* had been blacklisted and on the same day Dmitry
sent the patches #93471 and #93473.
Openobex was then fixed (bluez 4 vs. 5 ) with commit id
3d6d36d24b.
Originally
Signed-off-by: Dmitry Eremin-Solenikov <dmitry_eremin@mentor.com>
Signed-off-by: Andrea Adami <andrea.adami@gmail.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
On some targets clang erroniously detects an uninitialized variable.
Backport the fix from upstream.
Signed-off-by: Daniel McGregor <daniel.mcgregor@vecima.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Backported commit e3928d572b12f4e85a2a989482ac2d68c1bc2734
("Force text segment alignment to 0x08000000 for i386 with gold linker")
from libhugetlbfs -next branch to fix the build failure with gold linker.
And, enable x86-32 in COMPATIBLE_HOST.
Signed-off-by: Yang Shi <yang.shi@linaro.org>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
- Use '-fPIC' to replace 'PIC';
- Use 'pie' to replace 'PIE'
- Remove unknown '--add-needed'
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Sync with upstream to have the following fix:
...
commit 53aa9cdf7f0746f91f39521d020b024103041384
Author: Peter Jones <pjones@redhat.com>
Date: Thu Mar 3 09:50:03 2016 -0500
Explicitly pass a mode to efi_set_variable() in all cases.
We'll have to do this for efivar-0.24 .
Signed-off-by: Peter Jones <pjones@redhat.com>
...
It fixed compile failure:
...
efibootmgr.c: In function 'make_boot_var':
efibootmgr.c:305:7: error: too few arguments to function 'efi_set_variable'
rc = efi_set_variable(boot->guid, boot->name, boot->data,
^
...
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Upstream removed the use of the nvme headers completely,
so no more conflicts. Also drop gcc options patch, ubuntu
12.04 is too old to support in meta-oe/master.
Signed-off-by: Koen Kooi <koen.kooi@linaro.org>
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
The revision currently used is quite ancient (beginning of 2014) and we
need bugfixes commited since then. Rather than backport patches it's
cleaner to uprev.
Signed-off-by: Ioan-Adrian Ratiu <adrian.ratiu@ni.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
- New python bindings supported
- New board config options supported
- PACKAGECONFIG options added for other configure flags
Signed-off-by: Jack Mitchell <jack@embed.me.uk>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
The theory behind this bbclass was reasonable, with the primary
goal being to avoid multiple downloads of gnulib, but it neglected
the fact that packages would be shipping a specific version of the
./bootstrap which will support some flags but maybe not all the
latest ones from the latest gnulib/build-aux/bootstrap file.
I attempted to simply update the two pkgs to use the latest copy
of bootstrap from gnulib but this of course triggers the descent
into autoconf hell that we all know and love. Rather than futzing
with the packages configure.ac and deviating from what the pkg
maintainers intended and tested, we can just let the packages have
independent calls to ./bootstrap with whatever flags are needed.
The goal of this commit is to move the prepend out to the packages
and then delete the class without any real functional change ; i.e.
a purely mechanical change. Then we can adjust each package to
ensure it will still build with a modern host, in an independent
fashion, while keeping the main advantage of not fetching gnulib
two extra times for netcf and fontforge.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Joe MacDonald <joe_macdonald@mentor.com>
WARNING: acpitests-20140828-r0 do_populate_lic: acpitests: No generic license file exists for: Intel in any provider
this was copied from the acpitest web page
Signed-off-by: Armin Kuster <akuster808@gmail.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
* Update do_compile_append where it was updating *.pc from ${S} and
removing staging dir folder. Now we have seperate build folder
where packages get build. We have to search pc in ${B} not in ${S}.
After updating the tmp dir path is removed from nice.pc file.
* Remove PNBLACKLIST for libnice and farsight2 recipes.
Signed-off-by: Noor Ahsan <noor_ahsan@mentor.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
* set NOAUTOPACKAGEDEBUG to disable automatic creation of -dbg FILES split
* it doesn't work well with our do_split_packages calls and in the end breaks
installing llvm3.3-dbg (e.g. when you enable IMAGE_GEN_DEBUGFS)
* it shows a bit ugly error where multiple messages are written over
each other:
WARNING: [log_check] In line: [ConfiCollected errors:
]
WARNING: [log_check] luneos-dev-image: found an error message in the logfile (keyword 'Collected errors'):
[log_check] ConfiCollected errors:
ERROR:
ConfiCollected errors:
* satisfy_dependencies_for: Cannot satisfy the following dependencies for llvm3.3-dbg:
* libllvm3.3-lto-dbg * libllvm3.3-llvm-3.3-dbg *
* opkg_install: Cannot install package llvm3.3-dbg.
guring gstreamer1.0-libav-dbg.
but the main point is that do_split_packages returns some package
names of packages which in the end aren't created, because there
are no files in them after other -dbg packages eat them all due
to automatic -dbg creation
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Change expression used in do_integer_mul and do_uint64_mul
benchmarks so GCC doesn't optimize-away the loops,
Signed-off-by: Roy Li <rongqing.li@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
- Remove EXTRA_OECONF to disable X11 support since direct X11 window rendering has been retired
Signed-off-by: Derek Straka <derek@asterius.io>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
* also fix following QA warning:
WARNING: pidgin-2.10.12-r0 do_package: QA Issue: pidgin: Files/directories were installed but not shipped in any package:
/usr/share/appdata
/usr/share/appdata/pidgin.appdata.xml
Please set FILES such that these items are packaged. Alternatively if they are unneeded, avoid installing them or delete them within do_install.
pidgin: 2 installed and not shipped files. [installed-vs-shipped]
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
* disable gstreamer and vv, because gstreamer 0.10 is in meta-multimedia now
* disable farsight2/farstream, because farsight2 is blacklisted
* don't enable X11 support without x11 in DISTRO_FEATURES
* allow to disable all UI, useful for DISTROs who want just the purple libraries
like LuneOS
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Add additional patches to fix issues seen during testing on qemux86.
Signed-off-by: Paul Barker <paul@paulbarker.me.uk>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
On some hosts, when building rrdtool, we would meet do_configure error like
below.
| *** Module name IN: /usr/lib64/perl/5.22.0/POSIX.pm
| *** p1: /usr/lib64/ p3: perl/5.22.0/ p5: POSIX.pm
| *** Module name OUT: /path/to/builddir/tmp/sysroots/x86_64-linux/usr/lib/perl-native/perl/5.22.1/POSIX.pm
| POSIX object version 1.53_01 does not match $POSIX::VERSION 1.53 at /usr/lib64/perl/5.22.0/XSLoader.pm line 108.
Set PERLHOSTLIB to "" to fix the above error.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
The opencv-java is not always generated so we ought to include it,
only if it is available. The opencv-locales is not available and ought
to be skipped.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Protobuf has two main features, a compiler and a library. Run-time
applications only need the libraries to run.
Signed-off-by: Felipe F. Tonello <eu@felipetonello.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
replaces/obsoletes: commit 0d31f8448412 [Added protobuf-c criu dependency]
in meta-virtualization
This version of the recipe also adds a separated package for the compiler.
Signed-off-by: Felipe F. Tonello <eu@felipetonello.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
If wl_surface is destroyed while work thread is still running,
segmentation fault occurs.
Signed-off-by: Neena Busireddy <neenareddy.busireddy@nxp.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
v7.4.1689 is currently used in Debian Testing and Arch Linux.
The tags file used as an index to vim's documentation is not correctly created
when cross compiling in v7.4.1689. Patch 7.4.1733 is backported to fix this
issue.
Signed-off-by: Paul Barker <paul@paulbarker.me.uk>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Update the license start and end to account for release notes
Signed-off-by: Derek Straka <derek@asterius.io>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
