I see, thanks for the explanation.
On 17 December 2013 10:02, Ben Pfaff <b...@nicira.com> wrote:
> On Tue, Dec 17, 2013 at 08:48:15AM -0800, Joe Stringer wrote:
>> On 15 December 2013 17:57, Ethan Jackson <et...@nicira.com> wrote:
>> > + if (!revalidator->n_udumps) {
>> > + if (revalidator->dump_seq != seq_read(udpif->dump_seq)) {
>> > + revalidator->dump_seq = seq_read(udpif->dump_seq);
>> > + revalidator_sweep(revalidator);
>> > + } else {
>> > + ovs_mutex_cond_wait(&revalidator->wake_cond,
>> > + &revalidator->mutex);
>> > + }
>>
>> I didn't review the whole patch, but I did notice this earlier:-
>> Couldn't udpif->dump_seq change between the two seq_read() calls? The
>> example seq usage in seq.h reads the seq once and uses that variable
>> to compare and update. This should ensure that we handle all changes
>> to dump_seq.
>
> I think it's OK here if the seq value changes between the two reads.
> The goal, as I understand it, is just to make sure that
> revalidator_sweep() runs after the most recent change to the seq, and I
> think that this code will accomplish that.
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
