On Fri, Feb 13, 2015 at 02:55:58PM -0800, Ben Pfaff wrote: > On Fri, Jan 30, 2015 at 11:41:51AM +0900, Simon Horman wrote: > > This is in preparation for supporting group mod and desc reply > > messages with an NMX selection method group experimenter property. > > > > NMX selection method > > Signed-off-by: Simon Horman <simon.hor...@netronome.com> > > > > --- > > v2 Use list of struct field_array of TLVs rather than OF1.1 match > > for fields field of NMX selection method property > > It might be simpler to use an actual array for struct field_array. I > guess that we can be sure there will be no more than MFF_N_IDS > elements?
As duplicates are rejected I believe that we can make that assumption. I think that should simplify the code somewhat at the expense of some memory cost if the property is present. I'm assuming that you are happy about the latter and I'll see about making it so. > In oxm_pull_field_arary(), DBG seems like a pretty low log level for > decoding errors. I suspect that elsewhere in the same file we use > INFO or even WARN for similar messages, will you look around a bit? Thanks, it looks like it would be appropriate for all but the last message in oxm_pull_field_arary() to be WARN. > I am not sure that nx_pull_match_entry() does what you want because it > validates each value as a value, whereas we want to validate it as a > mask. True. The value is later passed to mf_is_mask_valid so it is validated as a mask. But it seems that I should avoid it also being validated as a value. > I might have more comments later. _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
