When user_frac in the polling subsystem is low it is going to busy the

CPU for too long period than necessary.  Additively, interfaces are kept
polled (in the tick) even if no more packets are available.
In order to avoid such situations a new generic mechanism can be
implemented in proactive way, keeping track of the time spent on any
packet and fragmenting the time for any tick, stopping the processing
as soon as possible.

In order to implement such mechanism, the polling handler needs to
change, returning the number of packets processed.
While the intended logic is not part of this patch, the polling KPI is
broken by this commit, adding an int return value and the new flag
IFCAP_POLLING_NOCOUNT (which will signal that the return value is
meaningless for the installed handler and checking should be skipped).

Bump __FreeBSD_version in order to signal such situation.

Reviewed by:	emaste
Sponsored by:	Sandvine Incorporated

1 files changed, 13 insertions, 7 deletions

diff --git a/sys/dev/sf/if_sf.c b/sys/dev/sf/if_sf.c
index 2abd80c..5105c70 100644
--- a/sys/dev/sf/if_sf.c
+++ b/sys/dev/sf/if_sf.c
@@ -160,7 +160,7 @@ static void sf_stats_update(struct sf_softc *);
 #ifndef __NO_STRICT_ALIGNMENT
 static __inline void sf_fixup_rx(struct mbuf *);
 #endif
-static void sf_rxeof(struct sf_softc *);
+static int sf_rxeof(struct sf_softc *);
 static void sf_txeof(struct sf_softc *);
 static int sf_encap(struct sf_softc *, struct mbuf **);
 static void sf_start(struct ifnet *);
@@ -193,7 +193,7 @@ static int sf_miibus_writereg(device_t, int, int, int);
 static void sf_miibus_statchg(device_t);
 static void sf_link_task(void *, int);
 #ifdef DEVICE_POLLING
-static void sf_poll(struct ifnet *ifp, enum poll_cmd cmd, int count);
+static int sf_poll(struct ifnet *ifp, enum poll_cmd cmd, int count);
 #endif
 
 static uint32_t csr_read_4(struct sf_softc *, int);
@@ -1526,19 +1526,20 @@ sf_fixup_rx(struct mbuf *m)
  * completely unuseable on the Alpha. Our only recourse is to copy received
  * packets into properly aligned buffers before handing them off.
  */
-static void
+static int
 sf_rxeof(struct sf_softc *sc)
 {
 	struct mbuf		*m;
 	struct ifnet		*ifp;
 	struct sf_rxdesc	*rxd;
 	struct sf_rx_rcdesc	*cur_cmp;
-	int			cons, eidx, prog;
+	int			cons, eidx, prog, rx_npkts;
 	uint32_t		status, status2;
 
 	SF_LOCK_ASSERT(sc);
 
 	ifp = sc->sf_ifp;
+	rx_npkts = 0;
 
 	bus_dmamap_sync(sc->sf_cdata.sf_rx_ring_tag,
 	    sc->sf_cdata.sf_rx_ring_map,
@@ -1652,6 +1653,7 @@ sf_rxeof(struct sf_softc *sc)
 		SF_UNLOCK(sc);
 		(*ifp->if_input)(ifp, m);
 		SF_LOCK(sc);
+		rx_npkts++;
 
 		/* Clear completion status. */
 		cur_cmp->sf_rx_status1 = 0;
@@ -1675,6 +1677,7 @@ sf_rxeof(struct sf_softc *sc)
 		    (csr_read_4(sc, SF_RXDQ_PTR_Q1) & ~SF_RXDQ_PRODIDX) |
 		    (eidx & SF_RXDQ_PRODIDX));
 	}
+	return (rx_npkts);
 }
 
 /*
@@ -1789,22 +1792,24 @@ sf_txthresh_adjust(struct sf_softc *sc)
 }
 
 #ifdef DEVICE_POLLING
-static void
+static int
 sf_poll(struct ifnet *ifp, enum poll_cmd cmd, int count)
 {
 	struct sf_softc		*sc;
 	uint32_t		status;
+	int			rx_npkts;
 
 	sc = ifp->if_softc;
+	rx_npkts = 0;
 	SF_LOCK(sc);
 
 	if ((ifp->if_drv_flags & IFF_DRV_RUNNING) == 0) {
 		SF_UNLOCK(sc);
-		return;
+		return (rx_npkts);
 	}
 
 	sc->rxcycles = count;
-	sf_rxeof(sc);
+	rx_npkts = sf_rxeof(sc);
 	sf_txeof(sc);
 	if (!IFQ_DRV_IS_EMPTY(&ifp->if_snd))
 		sf_start_locked(ifp);
@@ -1839,6 +1844,7 @@ sf_poll(struct ifnet *ifp, enum poll_cmd cmd, int count)
 	}
 
 	SF_UNLOCK(sc);
+	return (rx_npkts);
 }
 #endif /* DEVICE_POLLING */