<p dir="ltr"><br>
On 5 Feb 2014 15:43, "Laurentiu Palcu" <<a href="mailto:laurentiu.palcu@intel.com">laurentiu.palcu@intel.com</a>> wrote:<br>
><br>
> On Wed, Feb 05, 2014 at 01:09:05PM +0100, David Nyström wrote:<br>
> > ><br>
> > >| NOTE: Installing complementary packages ...<br>
> > >| NOTE: Installing the following packages:<br>
> > >| /media/sdb5/bld/tmp/sysroots/x86_64-linux/usr/bin/opkg-cl: the<br>
> > >``install'' command requires at least one argument<br>
> > >| NOTE: Unable to install packages. Command<br>
> > >/media/sdb5/bld/tmp/sysroots/x86_64-linux/usr/bin/opkg-cl -f<br>
> > >/media/sdb5/bld/tmp/work/qemuppc-poky-linux/core-image-minimal/1.0-r0/opkg.conf<br>
> > >-o<br>
> > >/media/sdb5/bld/tmp/work/qemuppc-poky-linux/core-image-minimal/1.0-r0/rootfs<br>
> > >--force_postinstall --prefer-arch-to-version install returned 1<br>
> I'll send a quick fix for this. This, however, is not the reason of the<br>
> build failure. It's complementary packages that are installed and an<br>
> empty list is passed to install(). Since these are attempt only<br>
> packages, the build does not fail because of this. Anyway, I added a<br>
> simple check and I will send a v3.<br>
><br>
> > >| NOTE: Running intercept scripts:<br>
> > >| NOTE: Executing write_package_manifest ...<br>
> > >| DEBUG: Executing python function write_package_manifest<br>
> > >| NOTE: Building from feeds activated!<br>
> > >| NOTE: Add all feed with URL<br>
> > >file:///media/sdb5/oel/build/tmp/deploy/ipk/all<br>
> > >| NOTE: Add qemuppc feed with URL<br>
> > >file:///media/sdb5/oel/build/tmp/deploy/ipk/qemuppc<br>
> > >| NOTE: Add ppc7400 feed with URL<br>
> > >file:///media/sdb5/oel/build/tmp/deploy/ipk/ppc7400<br>
> > >| DEBUG: Python function write_package_manifest finished<br>
> > >| NOTE: Executing license_create_manifest ...<br>
> > >| DEBUG: Executing shell function license_create_manifest<br>
> > >| ls: cannot access<br>
> > >/media/sdb5/bld/tmp/sysroots/qemuppc/pkgdata/runtime-reverse/base-files: No<br>
> > >such file or directory<br>
> > >| readlink: missing operand<br>
> > >| Try 'readlink --help' for more information.<br>
> > >| basename: missing operand<br>
> > >| Try 'basename --help' for more information.<br>
> > >| WARNING: exit code 1 from a shell command.<br>
> > >| DEBUG: Python function do_rootfs finished<br>
> > >| ERROR: Function failed: license_create_manifest (log file is located at<br>
> > >/media/sdb5/bld/tmp/work/qemuppc-poky-linux/core-image-minimal/1.0-r0/temp/log.do_rootfs.29007)<br>
> Now, this is the interesting part which, btw, happens on master too. That's<br>
> because you're installing from feeds and you're installing an older<br>
> version that has different packages than the latest (these are read from<br>
> package.manifest). The YP 1.5 core-image-minimal package.manifest<br>
> contains update-alternatives-cworth. The latest core-image-minimal does<br>
> not contain this package anymore. It contains update-alternatives-opkg.</p>
<p dir="ltr">Thanks Laurentiu,<br>
The 1.5 repo reference was just to give you an easy testcase. Sorry for creating confusion.<br>
My debug run above was done with a repo created by poky on the master branch.</p>
<p dir="ltr">><br>
> The easiest way would be to add a check in license_create_manifest() to<br>
> just ignore packages that cannot be found. However, since we're dealing<br>
> with licenses here, I honestly don't know if this is the right way of<br>
> doing it...</p>
<p dir="ltr">Good point, I think it might be better to avoid creating a manifest, rather than creating a potentially bogus manifest when using the FEEDS functionality.</p>
<p dir="ltr">The per-package licenses should be already existing in the repo when using a feed.</p>
<p dir="ltr">//DD</p>
<p dir="ltr">Br,<br>
David<br>
><br>
> laurentiu<br>
> _______________________________________________<br>
> Openembedded-core mailing list<br>
> <a href="mailto:Openembedded-core@lists.openembedded.org">Openembedded-core@lists.openembedded.org</a><br>
> <a href="http://lists.openembedded.org/mailman/listinfo/openembedded-core">http://lists.openembedded.org/mailman/listinfo/openembedded-core</a><br>
</p>