lib: Allow pure acknowledgment packets in FRCT

This adds the logic to send a pure acknowledgment packet without any
data to send. This needed the event filter for the fqueue, as these
non-data packets should not trigger application PKT events. The
default timeout is now 10ms, until we have FRCP tuning as part of
fccntl.

Karn's algorithm seems to be very unstable with low (sub-ms) RTT
estimates. Doubling RTO (every RTO) seems still too slow to prevent
rtx storms when the measured rtt suddenly spikes several orders of
magnitude. Just assuming the ACK'd packet is the last one transmitted
seems to be a lot more stable. It can lead to temporary
underestimation, but this is not a throughput-killer in FRCP.

Changes most time units to nanoseconds for faster computation.

Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>

1 files changed, 17 insertions, 33 deletions

diff --git a/src/lib/rxmwheel.c b/src/lib/rxmwheel.c
index 9602c5f9..0572c7b7 100644
--- a/src/lib/rxmwheel.c
+++ b/src/lib/rxmwheel.c
@@ -22,15 +22,15 @@
 
 #include <ouroboros/list.h>
 
-#define RXMQ_S     16                 /* defines #slots           */
-#define RXMQ_M     24                 /* defines max delay  (us)  */
-#define RXMQ_R     (RXMQ_M - RXMQ_S)  /* defines resolution (us)  */
+#define RXMQ_S     14                 /* defines #slots           */
+#define RXMQ_M     34                 /* defines max delay  (ns)  */
+#define RXMQ_R     (RXMQ_M - RXMQ_S)  /* defines resolution (ns)  */
 #define RXMQ_SLOTS (1 << RXMQ_S)
 #define RXMQ_MAX   (1 << RXMQ_M)      /* us                       */
 
-/* Small inacurracy to avoid slow division by MILLION. */
-#define ts_to_us(ts) (ts.tv_sec * MILLION + (ts.tv_nsec >> 10))
-#define ts_to_slot(ts) ((ts_to_us(ts) >> RXMQ_R) & (RXMQ_SLOTS - 1))
+/* Overflow limits range to about 6 hours. */
+#define ts_to_ns(ts) (ts.tv_sec * BILLION + ts.tv_nsec)
+#define ts_to_slot(ts) ((ts_to_ns(ts) >> RXMQ_R) & (RXMQ_SLOTS - 1))
 
 struct rxm {
         struct list_head     next;
@@ -95,22 +95,6 @@ static struct rxmwheel * rxmwheel_create(void)
         return rw;
 }
 
-static void check_probe(struct frcti * frcti,
-                        uint32_t       seqno)
-{
-        /* Disable rtt probe on retransmitted packet! */
-
-        pthread_rwlock_wrlock(&frcti->lock);
-
-        if (frcti->probe && ((frcti->rttseq + 1) == seqno)) {
-                /* Backoff to avoid never updating rtt */
-                frcti->srtt_us += frcti->mdev_us;
-                frcti->probe = false;
-        }
-
-        pthread_rwlock_unlock(&frcti->lock);
-}
-
 static void rxmwheel_move(struct rxmwheel * rw)
 {
         struct timespec    now;
@@ -159,11 +143,15 @@ static void rxmwheel_move(struct rxmwheel * rw)
 
                         shm_du_buff_ack(r->sdb);
 
-                        pthread_rwlock_rdlock(&r->frcti->lock);
+                        pthread_rwlock_wrlock(&r->frcti->lock);
 
                         snd_lwe = snd_cr->lwe;
                         rcv_lwe = rcv_cr->lwe;
                         rto     = r->frcti->rto;
+                        /* Assume last RTX is the one that's ACK'd. */
+                        if (r->frcti->probe
+                            && (r->frcti->rttseq + 1) == r->seqno)
+                                r->frcti->t_probe = now;
 
                         pthread_rwlock_unlock(&r->frcti->lock);
 
@@ -175,13 +163,11 @@ static void rxmwheel_move(struct rxmwheel * rw)
                         }
 
                         /* Check for r-timer expiry. */
-                        if (ts_to_us(now) - r->t0 > r->frcti->r) {
+                        if (ts_to_ns(now) - r->t0 > r->frcti->r) {
                                 ipcp_sdb_release(r->sdb);
                                 free(r);
-                                shm_rbuff_set_acl(ai.flows[fd].rx_rb,
-                                                  ACL_FLOWDOWN);
-                                shm_rbuff_set_acl(ai.flows[fd].tx_rb,
-                                                  ACL_FLOWDOWN);
+                                shm_rbuff_set_acl(f->rx_rb, ACL_FLOWDOWN);
+                                shm_rbuff_set_acl(f->tx_rb, ACL_FLOWDOWN);
                                 continue;
                         }
 
@@ -201,9 +187,7 @@ static void rxmwheel_move(struct rxmwheel * rw)
 
                         ipcp_sdb_release(r->sdb);
 
-                        check_probe(r->frcti, r->seqno);
-
-                        ((struct frct_pci *) head)->ackno = ntoh32(rcv_lwe);
+                        ((struct frct_pci *) head)->ackno = hton32(rcv_lwe);
 
                         /* Retransmit the copy. */
                         if (shm_rbuff_write_b(f->tx_rb, idx, NULL)) {
@@ -224,7 +208,7 @@ static void rxmwheel_move(struct rxmwheel * rw)
                         r->sdb  = sdb;
 
                         /* Schedule at least in the next time slot */
-                        rslot = (slot + MAX(rto >> RXMQ_R, 1))
+                        rslot = (slot + MAX((rto >> RXMQ_R), 1))
                                 & (RXMQ_SLOTS - 1);
 
                         list_add_tail(&r->next, &rw->wheel[rslot]);
@@ -251,7 +235,7 @@ static int rxmwheel_add(struct rxmwheel *    rw,
 
         clock_gettime(PTHREAD_COND_CLOCK, &now);
 
-        r->t0    = ts_to_us(now);
+        r->t0    = ts_to_ns(now);
         r->mul   = 0;
         r->seqno = seqno;
         r->sdb   = sdb;