https://bugzilla.redhat.com/show_bug.cgi?id=1279527
Jaroslav Škarvada jskarvad@redhat.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(jskarvad@redhat.c | |om) |
--- Comment #3 from Jaroslav Škarvada jskarvad@redhat.com --- (In reply to Jan Synacek from comment #2) Thanks for the review.
New version: Spec URL: https://jskarvad.fedorapeople.org/libosmocore/libosmocore.spec SRPM URL: https://jskarvad.fedorapeople.org/libosmocore/libosmocore-0.9.0-2.20151109gi...
[ ]: License field in the package spec file matches the actual license. Note: Checking patched sources after %prep for licenses. Licenses found: "GPL (v2 or later)", "GPL (v3 or later)", "Unknown or generated", "AGPL (v3 or later)", "LGPL (v3 or later)", "BSD GPL (v2)", "GPL (v2 or later) (with incorrect FSF address)", "*No copyright* GPL (v2 or later)", "GPL (v2)". 172 files have unknown license.
Some tests are under different license than the rest of the library:
libosmocore-0.9.0/tests/fr/fr_test.c (AGPLv3+) libosmocore-0.9.0/tests/strrb/strrb_test.c (GPLv3+)
I'm not sure if that's a reason to list them in the specfile, though.
Good catch, it could be resolved (probably) by upgrading the package to GPLv3+ or by removing the tests under question. They are all FSF free, so keeping them there and extending the resulting license.
[ ]: Package must own all directories that it creates. Note: Directories without known owners: /usr/share/doc/libosmocore
Fixed.
[ ]: Package does not own files or directories owned by other packages. Note: Dirs in package are owned also by: /usr/include/osmocom(libosmo- dsp-devel)
Are these owned by a required package? If not, they should be owned by libosmocore.
A bit tricky - it seems osmocom puts header files for all their projects under osmocom and libosmo-dsp was packed first, thus providing this directory. This could be handled better way, but for now I required the libosmo-dsp-devel package to just provide this directory.
[ ]: Fully versioned dependency in subpackages if applicable. Note: No Requires: %{name}%{?_isa} = %{version}-%{release} in libosmocore-doc , libosmocore-debuginfo
I believe this is missing.
It is generated doc, there could be (or could appear anytime later) some arch specific details, thus added the flag.
[ ]: %check is present and all tests pass.
There are tests in the package. Are they not run on purpose?
Added and run.
Rpmlint (installed packages)
libosmocore.x86_64: W: private-shared-object-provides /usr/lib64/libosmocodec.so.0.0.0 libosmocodec.so.0()(64bit) libosmocore.x86_64: W: private-shared-object-provides /usr/lib64/libosmoctrl.so.0.0.0 libosmoctrl.so.0()(64bit) libosmocore.x86_64: W: private-shared-object-provides /usr/lib64/libosmovty.so.3.0.0 libosmovty.so.3()(64bit) libosmocore.x86_64: W: private-shared-object-provides /usr/lib64/libosmogsm.so.5.1.0 libosmogsm.so.5(LIBOSMOGSM_1.0)(64bit) libosmocore.x86_64: W: private-shared-object-provides /usr/lib64/libosmogsm.so.5.1.0 libosmogsm.so.5()(64bit) libosmocore.x86_64: W: private-shared-object-provides /usr/lib64/libosmocore.so.6.0.0 libosmocore.so.6()(64bit) libosmocore.x86_64: W: private-shared-object-provides /usr/lib64/libosmosim.so.0.0.0 libosmosim.so.0()(64bit) libosmocore.x86_64: W: private-shared-object-provides /usr/lib64/libosmogb.so.4.0.0 libosmogb.so.4(LIBOSMOGB_1.0)(64bit) libosmocore.x86_64: W: private-shared-object-provides /usr/lib64/libosmogb.so.4.0.0 libosmogb.so.4()(64bit)
I have no idea if these are valid complaints or fedora-review just had a brain-fart...
I can't reproduce these, no idea why they were flagged as private shared objects. The shared objects are public and they are in the right location. Could you provide more details regarding this?
libosmocore.x86_64: W: unused-direct-shlib-dependency /usr/lib64/libosmosim.so.0.0.0 /lib64/libpthread.so.0
This is "fixed" in the spec, but rpmlint still complains for some reason.
Filled gcc bug 1287805
libosmocore.x86_64: W: undefined-non-weak-symbol /usr/lib64/libosmogb.so.4.0.0 bssgp_prim_cb
Is the library actually usable without this?
Yes, it is. The bssgp_prim_cb is callback function that must be defined by the consumer, see e.g. gprs_ns_test.c.