On Fri, Jul 29, 2016 at 01:19:03PM -0700, Jarno Rajahalme wrote: > > > On Jul 29, 2016, at 12:33 PM, Ben Pfaff <b...@ovn.org> wrote: > > > > On Thu, Jul 28, 2016 at 05:55:57PM -0700, Jarno Rajahalme wrote: > >> Adding groups support for bundles is simpler if also groups are > >> modified under ofproto_mutex. > >> > >> Eliminate the search for rules when deleting a group so that we will > >> not keep the mutex for too long. > >> > >> Signed-off-by: Jarno Rajahalme <ja...@ovn.org> > > > > It's confusing that rule_collection_remove() and > > rule_collection_remove_postponed() have unrelated effects. > > > > We want a verb that is the opposite of "add", how about > rule_collection_subtract(), rule_collection_retract() or > rule_collection_withdraw()?
I think that "remove" is OK for rule_collection_remove(). > Of should we be more careful about naming functions that change change > the rule collection vs. operate on the rules in the collection? E.g., > rename rule_collection_add() as rule_collection_add_rule() and > rule_collection_remove() as rule_collection_remove_rule(). Functions > named like "rule_collection_VERB()" would then be reserved for > functions that VERB on each rule? I'm not enthusiastic about having the name of the type twice in the function name. There might be a useful distinction here between the generic object collection functions, which should naturally begin with <object>_collection_, and functions that aren't generic object collection functions but happen to take one as an argument. The latter maybe should not have names that begin with <object>_collection_ since that is kind of reserved. So, can we rename rename rule_collection_remove_postponed() so that it fits this pattern? Maybe something like remove_rules_postponed()? Acked-by: Ben Pfaff <b...@ovn.org> _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
