On Thu, Aug 22, 2013 at 1:09 PM, Andy Zhou <az...@nicira.com> wrote: > Thanks a lot for the review comments. > > On Thu, Aug 22, 2013 at 12:10 PM, Jesse Gross <je...@nicira.com> wrote: >> >> On Fri, Aug 16, 2013 at 4:01 PM, Andy Zhou <az...@nicira.com> wrote: >> > diff --git a/datapath/flow.c b/datapath/flow.c >> > index ab803ef..25eb8c1 100644 >> > --- a/datapath/flow.c >> > +++ b/datapath/flow.c >> > void ovs_flow_key_mask(struct sw_flow_key *dst, const struct >> > sw_flow_key *src, >> > const struct sw_flow_mask *mask) >> [...] >> > + for (i = 0; i < mask->range.start; i += sizeof(long)) >> > + *d++ = 0; >> > + >> > + for (i = 0; i < range_n_bytes(&mask->range); i += sizeof(long)) >> > + *d++ = *s++ & *m++; >> > + >> > + for (i = mask->range.end; i < sizeof(*dst); i += sizeof(long)) >> > + *d++ = 0; >> >> Do we actually need to zero out the areas before and after the key? >> Now that we've made the masking, comparison, and hashing functions all >> use the same lengths I don't think anyone will look at those regions. >> > There was a memset(0) in current code. I did not verify it was safe to > remove. Will remove them as suggested if it is.
I think it should be safe but it would be good if you double check. _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
