On Mon, 2018-11-12 at 15:07 +0100, Fabiano Fidêncio wrote: > This patch set introduces resources' inheritance for OSes that > derive-from/clone a device that already has its resources set. > > While this patch doesn't cause any harm per si, from the moment we > start > removing the duplications we'll need a new libosinfo release with > this > code included. > > https://gitlab.com/libosinfo/osinfo-db/issues/15 > > Fabiano Fidêncio (3): > os: Add _get_.*_resources_without_inheritance() > os: Add support for Resources inheritance > test-os: Add test_resources_inheritance() > > osinfo/Makefile.am | 1 + > osinfo/osinfo_os.c | 85 ++++++++++++++++- > osinfo/osinfo_os_private.h | 39 ++++++++ > .../test-os-resources-inheritance-1.xml | 25 +++++ > .../test-os-resources-inheritance-2.xml | 10 ++ > tests/test-os.c | 91 > +++++++++++++++++++ > 6 files changed, 247 insertions(+), 4 deletions(-) > create mode 100644 osinfo/osinfo_os_private.h > create mode 100644 tests/dbdata/os/libosinfo.org/test-os-resources- > inheritance-1.xml > create mode 100644 tests/dbdata/os/libosinfo.org/test-os-resources- > inheritance-2.xml > Apart from Pino's comments, I've also realized that we're not filling values of a resource that was previous set (but it's not in this version). Example: In os1, we have: <resources arch="all"> <minimum> <cpu>1000000000</cpu> <ram>1073741824</ram> </minimum> </resources> Is os2, which inherits the values from os1, we have: <resources arch="all"> <minimum> <n-cpus>2</cpu> </minimum> </resources> What I'd expect is that when querying for minimum CPU for os2 I'd get: 1000000000. I'll cover this and Pino's comments and submit a v2. Best Regards, -- Fabiano Fidêncio _______________________________________________ Libosinfo mailing list Libosinfo@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libosinfo
