| Submitter | Xiaofeng Yan |
|---|---|
| Date | April 24, 2012, 5:50 a.m. |
| Message ID | <fde21e1126e23439ccd9dc7ba8e7cc9e96df3799.1335245655.git.xiaofeng.yan@windriver.com> |
| Download | mbox | patch |
| Permalink | /patch/26367/ |
| State | New |
| Headers | show |
Comments
Op 24 apr. 2012, om 07:50 heeft Xiaofeng Yan het volgende geschreven: > From: Xiaofeng Yan <xiaofeng.yan@windriver.com> > > LSB Test Suite complain "no library libgdk_pixbuf_xlib-2.0.so.0" \ > because of having "--without-x11" for gdk-pixbuf_2.24.1.bb. > For passing lsb test I add PACKAGECONFIG support to the recipe and \ > have poky-lsb turn on the poky-lsb distro for this recipe. > > [YOCTO #2284] Shouldn't this be an 'x11' DISTRO_FEATURE instead?
On 2012?04?24? 13:53, Koen Kooi wrote: > Op 24 apr. 2012, om 07:50 heeft Xiaofeng Yan het volgende geschreven: > >> From: Xiaofeng Yan<xiaofeng.yan@windriver.com> >> >> LSB Test Suite complain "no library libgdk_pixbuf_xlib-2.0.so.0" \ >> because of having "--without-x11" for gdk-pixbuf_2.24.1.bb. >> For passing lsb test I add PACKAGECONFIG support to the recipe and \ >> have poky-lsb turn on the poky-lsb distro for this recipe. >> >> [YOCTO #2284] > Shouldn't this be an 'x11' DISTRO_FEATURE instead? because "--without-x11" is disabled when x11 exists in core-image-sato. Using 'x11' DISTRO_FEATURE instead will change the default set, which will cause enable "--with-x11" when x11 is in DISTRO_FEATURE. LSB Test Suite need this configuration but other image may not need it even if having x11. > > _______________________________________________ > Openembedded-core mailing list > Openembedded-core@lists.openembedded.org > http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core >
Op 24 apr. 2012, om 08:12 heeft Xiaofeng Yan het volgende geschreven: > On 2012?04?24? 13:53, Koen Kooi wrote: >> Op 24 apr. 2012, om 07:50 heeft Xiaofeng Yan het volgende geschreven: >> >> >>> From: Xiaofeng Yan <xiaofeng.yan@windriver.com> >>> >>> >>> LSB Test Suite complain "no library libgdk_pixbuf_xlib-2.0.so.0" \ >>> because of having "--without-x11" for gdk-pixbuf_2.24.1.bb. >>> For passing lsb test I add PACKAGECONFIG support to the recipe and \ >>> have poky-lsb turn on the poky-lsb distro for this recipe. >>> >>> [YOCTO #2284] >>> >> Shouldn't this be an 'x11' DISTRO_FEATURE instead? >> > because "--without-x11" is disabled when x11 exists in core-image-sato. Using 'x11' DISTRO_FEATURE instead will change the default set, which will cause enable "--with-x11" when x11 is in DISTRO_FEATURE. LSB Test Suite need this configuration but other image may not need it even if having x11. What use-case is there for having 'x11' in DISTRO_FEATURES but not wanting x11 support in pixbuf? Regardless of that, poky-lsb distro overrides have no place in oe-core, they should go into their own layer as bbappends. regards, Koen
On Tue, 2012-04-24 at 08:29 +0200, Koen Kooi wrote: > Op 24 apr. 2012, om 08:12 heeft Xiaofeng Yan het volgende geschreven: > > > On 2012?04?24? 13:53, Koen Kooi wrote: > >> Op 24 apr. 2012, om 07:50 heeft Xiaofeng Yan het volgende geschreven: > >> > >> > >>> From: Xiaofeng Yan <xiaofeng.yan@windriver.com> > >>> > >>> > >>> LSB Test Suite complain "no library libgdk_pixbuf_xlib-2.0.so.0" \ > >>> because of having "--without-x11" for gdk-pixbuf_2.24.1.bb. > >>> For passing lsb test I add PACKAGECONFIG support to the recipe and \ > >>> have poky-lsb turn on the poky-lsb distro for this recipe. > >>> > >>> [YOCTO #2284] > >>> > >> Shouldn't this be an 'x11' DISTRO_FEATURE instead? > >> > > because "--without-x11" is disabled when x11 exists in core-image-sato. Using 'x11' DISTRO_FEATURE instead will change the default set, which will cause enable "--with-x11" when x11 is in DISTRO_FEATURE. LSB Test Suite need this configuration but other image may not need it even if having x11. > > What use-case is there for having 'x11' in DISTRO_FEATURES but not wanting x11 support in pixbuf? > > Regardless of that, poky-lsb distro overrides have no place in oe-core, they should go into their own layer as bbappends. We should be using the linuxstdbase override and also make it conditional on x11 being in DISTRO_FEATURES. As far as I can tell nothing uses/cares about this library apart from lsb right now. If that changes I'm happy to build more of the time but until we need it, its pointless. We should also split it into a separate package. Cheers, Richard
On 2012?04?24? 15:29, Richard Purdie wrote: > On Tue, 2012-04-24 at 08:29 +0200, Koen Kooi wrote: >> Op 24 apr. 2012, om 08:12 heeft Xiaofeng Yan het volgende geschreven: >> >>> On 2012?04?24? 13:53, Koen Kooi wrote: >>>> Op 24 apr. 2012, om 07:50 heeft Xiaofeng Yan het volgende geschreven: >>>> >>>> >>>>> From: Xiaofeng Yan<xiaofeng.yan@windriver.com> >>>>> >>>>> >>>>> LSB Test Suite complain "no library libgdk_pixbuf_xlib-2.0.so.0" \ >>>>> because of having "--without-x11" for gdk-pixbuf_2.24.1.bb. >>>>> For passing lsb test I add PACKAGECONFIG support to the recipe and \ >>>>> have poky-lsb turn on the poky-lsb distro for this recipe. >>>>> >>>>> [YOCTO #2284] >>>>> >>>> Shouldn't this be an 'x11' DISTRO_FEATURE instead? >>>> >>> because "--without-x11" is disabled when x11 exists in core-image-sato. Using 'x11' DISTRO_FEATURE instead will change the default set, which will cause enable "--with-x11" when x11 is in DISTRO_FEATURE. LSB Test Suite need this configuration but other image may not need it even if having x11. >> What use-case is there for having 'x11' in DISTRO_FEATURES but not wanting x11 support in pixbuf? >> >> Regardless of that, poky-lsb distro overrides have no place in oe-core, they should go into their own layer as bbappends. > We should be using the linuxstdbase override and also make it > conditional on x11 being in DISTRO_FEATURES. You mean I should use the following methods in gdk-pixbuf_2.24.1.bb EXTRA_OECONF_append_linuxstdbase = " ${@base_contains('DISTRO_FEATURES', 'x11', '--with-x11', '--without-x11', d)}" The above variable should be putted into gdk-pixbuf_2.24.1.bb because poky-lsb.conf is parsed before this bb file. right ?? > As far as I can tell nothing uses/cares about this library apart from > lsb right now. If that changes I'm happy to build more of the time but > until we need it, its pointless. We should also split it into a separate > package. > > Cheers, > > Richard > > > > _______________________________________________ > Openembedded-core mailing list > Openembedded-core@lists.openembedded.org > http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
On Tue, 2012-04-24 at 23:02 +0200, Tasslehoff Kjappfot wrote: > boost.inc: > > # To enable python, uncomment the following: > #BOOST_LIBS += "python" > #DEPENDS += "python" > #PYTHON_ROOT = "${STAGING_DIR_HOST}/${prefix}" > #PYTHON_VERSION = "2.5" > > If this is uncommented, stuff fails with messages like: > > | ...failed gcc.compile.c++ > bin.v2/libs/python/build/69ffc88faf507005827aa061bd65b9bd/wrapper.o... > | gcc.compile.c++ > bin.v2/libs/python/build/69ffc88faf507005827aa061bd65b9bd/import.o > | In file included from > /src/openembedded/build/tmp-angstrom_v2012_05-eglibc/sysroots/x86_64-linux/usr/include/python2.7/Python.h:58:0, > | from ./boost/python/detail/wrap_python.hpp:142, > | from ./boost/python/detail/prefix.hpp:13, > | from ./boost/python/ssize_t.hpp:9, > | from ./boost/python/object.hpp:8, > | from ./boost/python/import.hpp:8, > | from libs/python/src/import.cpp:6: > | > /src/openembedded/build/tmp-angstrom_v2012_05-eglibc/sysroots/x86_64-linux/usr/include/python2.7/pyport.h:849:2: > error: #error "LONG_BIT definition appears wrong for platform (bad > gcc/glibc config?)." > > I have problems figuring out how to solve this (other than by > reinstalling Ubuntu 32-bit). You have to wonder why its looking in the native sysroot and not the target one. It would probably find better values in the target one... Cheers, Richard
boost.inc:
# To enable python, uncomment the following:
#BOOST_LIBS += "python"
#DEPENDS += "python"
#PYTHON_ROOT = "${STAGING_DIR_HOST}/${prefix}"
#PYTHON_VERSION = "2.5"
If this is uncommented, stuff fails with messages like:
| ...failed gcc.compile.c++
bin.v2/libs/python/build/69ffc88faf507005827aa061bd65b9bd/wrapper.o...
| gcc.compile.c++
bin.v2/libs/python/build/69ffc88faf507005827aa061bd65b9bd/import.o
| In file included from
/src/openembedded/build/tmp-angstrom_v2012_05-eglibc/sysroots/x86_64-linux/usr/include/python2.7/Python.h:58:0,
| from ./boost/python/detail/wrap_python.hpp:142,
| from ./boost/python/detail/prefix.hpp:13,
| from ./boost/python/ssize_t.hpp:9,
| from ./boost/python/object.hpp:8,
| from ./boost/python/import.hpp:8,
| from libs/python/src/import.cpp:6:
|
/src/openembedded/build/tmp-angstrom_v2012_05-eglibc/sysroots/x86_64-linux/usr/include/python2.7/pyport.h:849:2:
error: #error "LONG_BIT definition appears wrong for platform (bad
gcc/glibc config?)."
I have problems figuring out how to solve this (other than by
reinstalling Ubuntu 32-bit).
- Tasslehoff
Ouch. Didn't mean to make this a part of another thread. Sorry about that.
On 04/24/2012 11:44 AM, Richard Purdie wrote: > On Tue, 2012-04-24 at 23:02 +0200, Tasslehoff Kjappfot wrote: >> boost.inc: >> >> # To enable python, uncomment the following: >> #BOOST_LIBS += "python" >> #DEPENDS += "python" >> #PYTHON_ROOT = "${STAGING_DIR_HOST}/${prefix}" >> #PYTHON_VERSION = "2.5" >> >> If this is uncommented, stuff fails with messages like: >> >> | ...failed gcc.compile.c++ >> bin.v2/libs/python/build/69ffc88faf507005827aa061bd65b9bd/wrapper.o... >> | gcc.compile.c++ >> bin.v2/libs/python/build/69ffc88faf507005827aa061bd65b9bd/import.o >> | In file included from >> /src/openembedded/build/tmp-angstrom_v2012_05-eglibc/sysroots/x86_64-linux/usr/include/python2.7/Python.h:58:0, >> | from ./boost/python/detail/wrap_python.hpp:142, >> | from ./boost/python/detail/prefix.hpp:13, >> | from ./boost/python/ssize_t.hpp:9, >> | from ./boost/python/object.hpp:8, >> | from ./boost/python/import.hpp:8, >> | from libs/python/src/import.cpp:6: >> | >> /src/openembedded/build/tmp-angstrom_v2012_05-eglibc/sysroots/x86_64-linux/usr/include/python2.7/pyport.h:849:2: >> error: #error "LONG_BIT definition appears wrong for platform (bad >> gcc/glibc config?)." >> >> I have problems figuring out how to solve this (other than by >> reinstalling Ubuntu 32-bit). > You have to wonder why its looking in the native sysroot and not the > target one. It would probably find better values in the target one... > > Cheers, > > Richard > d-core I thought so too, but I wasn't able to convince it to do so. Will have to try harder :)
Patch
diff --git a/meta/recipes-gnome/gdk-pixbuf/gdk-pixbuf_2.24.1.bb b/meta/recipes-gnome/gdk-pixbuf/gdk-pixbuf_2.24.1.bb index e31f717..33627a3 100644 --- a/meta/recipes-gnome/gdk-pixbuf/gdk-pixbuf_2.24.1.bb +++ b/meta/recipes-gnome/gdk-pixbuf/gdk-pixbuf_2.24.1.bb @@ -28,9 +28,10 @@ LIBV = "2.10.0" EXTRA_OECONF = "\ --without-libtiff \ --with-libpng \ - --without-x11 \ --disable-introspection \ " +PACKAGECONFIG ??= "${@base_contains('DISTRO', 'poky-lsb', 'poky-lsb', '', d)}" +PACKAGECONFIG[poky-lsb] = "--with-x11,--without-x11,," FILES_${PN} = "${bindir}/gdk-pixbuf-query-loaders \ ${libdir}/lib*.so.*"