> The new failure case in ofproto_delete_flow() looks like a problem for > fail-open, which never retries because previously it could never fail. > Is there a reason we can't block waiting for the write lock here? The > uses of ofproto_delete_flow() (in-band, fail-open) are not performance > sensitive fast paths. > > Ditto for ofproto_flush__().
This is totally fine. I've changed it to just take the lock. I also noticed that I had messed up ofproto_flush__() and only destroyed it if rule->pending instead of the other way around. > I think that some of the changes here are because Clang can't annotate > a return value as acquired. If so, would you mind mentioning that in > the commit message? Sure, I'll send a new version of the patch with an updated commit message and the other changes. X-CudaMail-Whitelist-To: dev@openvswitch.org _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
