On 3 July 2014 10:30, Ben Pfaff <b...@nicira.com> wrote:
> On Wed, Jul 02, 2014 at 08:14:15PM +1200, Joe Stringer wrote:
> > If the datapath doesn't dump a flow for some reason, and the current
> > dump is expected to revalidate all flows in the datapath, then perform
> > revalidation for those flows by fetching them during the sweep phase.
> >
> > Signed-off-by: Joe Stringer <joestrin...@nicira.com>
> > ---
> > This relies on the mask dumping from the previous patch, but doesn't
> > make use of batched flow_get.
>
> I thought that the logic in revalidator_sweep__() was confusing
> before, but this patch left me more confused.
>
> Is this a correct transformation? From:
> HMAP_FOR_EACH_SAFE (ukey, next, hmap_node, revalidator->ukeys) {
> bool missed_reval, delete_flow;
>
> missed_reval = ukey->last_seen && need_revalidate;
> ukey->last_seen++;
> delete_flow = purge || ukey->last_seen == MAX_UKEY_AGE;
> if (!ukey->flow_exists) {
> ukey_delete(revalidator, ukey);
> continue;
> } else if (!delete_flow && missed_reval) {
> delete_flow = handle_missed_revalidation(revalidator, ukey);
> }
>
> if (delete_flow) {
> struct dump_op *op = &ops[n_ops++];
>
> dump_op_init(op, ukey->key, ukey->key_len, ukey);
> if (n_ops == REVALIDATE_MAX_BATCH) {
> push_dump_ops(revalidator, ops, n_ops);
> n_ops = 0;
> }
> }
> }
> to:
> HMAP_FOR_EACH_SAFE (ukey, next, hmap_node, revalidator->ukeys) {
> if (ukey->flow_exists) {
> ukey->last_seen++;
> if (purge
> || ukey->last_seen == MAX_UKEY_AGE
> || (ukey->last_seen > 1
> && need_revalidate
> && handle_missed_revalidation(revalidator, ukey))) {
> struct dump_op *op = &ops[n_ops++];
>
> dump_op_init(op, ukey->key, ukey->key_len, ukey);
> if (n_ops == REVALIDATE_MAX_BATCH) {
> push_dump_ops(revalidator, ops, n_ops);
> n_ops = 0;
> }
> }
> } else {
> ukey_delete(revalidator, ukey);
> }
> }
> I find the latter easier to understand, if it is equivalent.
> (I guess that it will change, though, when you revise the series.)
>
Thanks for this! I got deep enough into it that I know what I'm trying to
achieve, but not the most elegant way to present that. I made some minor
changes in addition to this for v2.
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
