On Mon, Mar 06, 2017 at 12:00:59AM +0000, Jan Scheurich wrote: > > > Signed-off-by: Lorand Jakab <loja...@cisco.com> > > > Signed-off-by: Simon Horman <simon.hor...@netronome.com> > > > Signed-off-by: Jiri Benc <jb...@redhat.com> > > > Signed-off-by: Yi Yang <yi.y.y...@intel.com> > > > Signed-off-by: Jean Tourrilhes <j...@labs.hpe.com> > > > Signed-off-by: Jan Scheurich <jan.scheur...@ericsson.com> > > > Co-authored-by: Zoltan Balogh <zoltan.bal...@ericsson.com> > > > > This sign-off chain is baffling. Please explain it. > > Support for L3 tunneling has been dragging along for several years. We > are just the last ones in a long chain of authors who have submitted > earlier versions of patches for this. We thought it was appropriate to > keep the earlier Signed-off-by tags to reflect that. If there is a > better way to express this, please let us know.
OK, it really was not clear to me that the code had really passed through that many hands. If it really has, then it's fine. > > Clang reports: > > > > ../lib/packets.c:260:22: error: cast from 'char *' to 'ovs_be16 *' (aka > > 'unsigned short *') increases required alignment from 1 to 2 [- > > Werror,-Wcast-align] > > ../lib/packets.c:263:22: error: cast from 'char *' to 'ovs_be16 *' (aka > > 'unsigned short *') increases required alignment from 1 to 2 [- > > Werror,-Wcast-align] > > I think these warnings are artifacts caused by the fact that char* does not > express the actual alignment of the l2.5 or l3 headers. The packet data is > word-aligned and the Ethernet header occupies 14 bytes. Any number of > potentially present VLAN tags cannot change that. Any suggestion how to fix > this? When these warnings come about and the code is actually correct, then the ALIGNED_CAST macro is the right way to suppress them. _______________________________________________ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev
