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, 17 insertions, 10 deletions

diff --git a/sys/pci/if_rl.c b/sys/pci/if_rl.c
index dc3e3b5..16e3ccc 100644
--- a/sys/pci/if_rl.c
+++ b/sys/pci/if_rl.c
@@ -205,14 +205,14 @@ static int rl_miibus_readreg(device_t, int, int);
 static void rl_miibus_statchg(device_t);
 static int rl_miibus_writereg(device_t, int, int, int);
 #ifdef DEVICE_POLLING
-static void rl_poll(struct ifnet *ifp, enum poll_cmd cmd, int count);
-static void rl_poll_locked(struct ifnet *ifp, enum poll_cmd cmd, int count);
+static int rl_poll(struct ifnet *ifp, enum poll_cmd cmd, int count);
+static int rl_poll_locked(struct ifnet *ifp, enum poll_cmd cmd, int count);
 #endif
 static int rl_probe(device_t);
 static void rl_read_eeprom(struct rl_softc *, uint8_t *, int, int, int);
 static void rl_reset(struct rl_softc *);
 static int rl_resume(device_t);
-static void rl_rxeof(struct rl_softc *);
+static int rl_rxeof(struct rl_softc *);
 static void rl_setmulti(struct rl_softc *);
 static int rl_shutdown(device_t);
 static void rl_start(struct ifnet *);
@@ -1221,7 +1221,7 @@ rl_list_rx_init(struct rl_softc *sc)
  * on a 32-bit boundary. To achieve this, we pass RL_ETHER_ALIGN (2 bytes)
  * as the offset argument to m_devget().
  */
-static void
+static int
 rl_rxeof(struct rl_softc *sc)
 {
 	struct mbuf		*m;
@@ -1229,6 +1229,7 @@ rl_rxeof(struct rl_softc *sc)
 	uint8_t			*rxbufpos;
 	int			total_len = 0;
 	int			wrap = 0;
+	int			rx_npkts = 0;
 	uint32_t		rxstat;
 	uint16_t		cur_rx;
 	uint16_t		limit;
@@ -1277,7 +1278,7 @@ rl_rxeof(struct rl_softc *sc)
 		    total_len > ETHER_MAX_LEN + ETHER_VLAN_ENCAP_LEN) {
 			ifp->if_ierrors++;
 			rl_init_locked(sc);
-			return;
+			return (rx_npkts);
 		}
 
 		/* No errors; receive the packet. */
@@ -1331,9 +1332,11 @@ rl_rxeof(struct rl_softc *sc)
 		RL_UNLOCK(sc);
 		(*ifp->if_input)(ifp, m);
 		RL_LOCK(sc);
+		rx_npkts++;
 	}
 
 	/* No need to sync Rx memory block as we didn't modify it. */
+	return (rx_npkts);
 }
 
 /*
@@ -1538,26 +1541,29 @@ rl_tick(void *xsc)
 }
 
 #ifdef DEVICE_POLLING
-static void
+static int
 rl_poll(struct ifnet *ifp, enum poll_cmd cmd, int count)
 {
 	struct rl_softc *sc = ifp->if_softc;
+	int rx_npkts = 0;
 
 	RL_LOCK(sc);
 	if (ifp->if_drv_flags & IFF_DRV_RUNNING)
-		rl_poll_locked(ifp, cmd, count);
+		rx_npkts = rl_poll_locked(ifp, cmd, count);
 	RL_UNLOCK(sc);
+	return (rx_npkts);
 }
 
-static void
+static int
 rl_poll_locked(struct ifnet *ifp, enum poll_cmd cmd, int count)
 {
 	struct rl_softc *sc = ifp->if_softc;
+	int rx_npkts;
 
 	RL_LOCK_ASSERT(sc);
 
 	sc->rxcycles = count;
-	rl_rxeof(sc);
+	rx_npkts = rl_rxeof(sc);
 	rl_txeof(sc);
 
 	if (!IFQ_DRV_IS_EMPTY(&ifp->if_snd))
@@ -1569,7 +1575,7 @@ rl_poll_locked(struct ifnet *ifp, enum poll_cmd cmd, int count)
 		/* We should also check the status register. */
 		status = CSR_READ_2(sc, RL_ISR);
 		if (status == 0xffff)
-			return;
+			return (rx_npkts);
 		if (status != 0)
 			CSR_WRITE_2(sc, RL_ISR, status);
 
@@ -1578,6 +1584,7 @@ rl_poll_locked(struct ifnet *ifp, enum poll_cmd cmd, int count)
 		if (status & RL_ISR_SYSTEM_ERR)
 			rl_init_locked(sc);
 	}
+	return (rx_npkts);
 }
 #endif /* DEVICE_POLLING */