Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 32b3e08f authored by Jesse Brandeburg's avatar Jesse Brandeburg Committed by Jeff Kirsher
Browse files

drivers/net/intel: use napi_complete_done() 



As per Eric Dumazet's previous patches:
(see commit (24d2e4a5) - tg3: use napi_complete_done())

Quoting verbatim:
Using napi_complete_done() instead of napi_complete() allows
us to use /sys/class/net/ethX/gro_flush_timeout

GRO layer can aggregate more packets if the flush is delayed a bit,
without having to set too big coalescing parameters that impact
latencies.
</end quote>

Tested
configuration: low latency via ethtool -C ethx adaptive-rx off
				rx-usecs 10 adaptive-tx off tx-usecs 15
workload: streaming rx using netperf TCP_MAERTS

igb:
MIGRATED TCP MAERTS TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.0.0.1 () port 0 AF_INET : demo
...
Interim result:  941.48 10^6bits/s over 1.000 seconds ending at 1440193171.589

Alignment      Offset         Bytes    Bytes       Recvs   Bytes    Sends
Local  Remote  Local  Remote  Xfered   Per                 Per
Recv   Send    Recv   Send             Recv (avg)          Send (avg)
    8       8      0       0 1176930056  1475.36    797726   16384.00  71905

MIGRATED TCP MAERTS TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.0.0.1 () port 0 AF_INET : demo
...
Interim result:  941.49 10^6bits/s over 0.997 seconds ending at 1440193142.763

Alignment      Offset         Bytes    Bytes       Recvs   Bytes    Sends
Local  Remote  Local  Remote  Xfered   Per                 Per
Recv   Send    Recv   Send             Recv (avg)          Send (avg)
    8       8      0       0 1175182320  50476.00     23282   16384.00  71816

i40e:
Hard to test because the traffic is incoming so fast (24Gb/s) that GRO
always receives 87kB, even at the highest interrupt rate.

Other drivers were only compile tested.

Signed-off-by: default avatarJesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: default avatarAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: default avatarJeff Kirsher <jeffrey.t.kirsher@intel.com>
parent 7709b4c1

drivers/net/ethernet/intel/e1000/e1000_main.c

+1 −1
Original line number Diff line number Diff line
@@ -3820,7 +3820,7 @@ static int e1000_clean(struct napi_struct *napi, int budget) 
	if (work_done < budget) {

		if (likely(adapter->itr_setting & 3))

			e1000_set_itr(adapter);

		napi_complete(napi);

		napi_complete_done(napi, work_done);

		if (!test_bit(__E1000_DOWN, &adapter->flags))

			e1000_irq_enable(adapter);

	}

drivers/net/ethernet/intel/e1000e/netdev.c

+1 −1
Original line number Diff line number Diff line
@@ -2693,7 +2693,7 @@ static int e1000e_poll(struct napi_struct *napi, int weight) 
	if (work_done < weight) {

		if (adapter->itr_setting & 3)

			e1000_set_itr(adapter);

		napi_complete(napi);

		napi_complete_done(napi, work_done);

		if (!test_bit(__E1000_DOWN, &adapter->state)) {

			if (adapter->msix_entries)

				ew32(IMS, adapter->rx_ring->ims_val);

drivers/net/ethernet/intel/fm10k/fm10k_main.c

+12 −9
Original line number Diff line number Diff line
@@ -593,7 +593,7 @@ static void fm10k_receive_skb(struct fm10k_q_vector *q_vector, 
	napi_gro_receive(&q_vector->napi, skb);

}



static bool fm10k_clean_rx_irq(struct fm10k_q_vector *q_vector,

static int fm10k_clean_rx_irq(struct fm10k_q_vector *q_vector,

			      struct fm10k_ring *rx_ring,

			      int budget)

{
@@ -662,7 +662,7 @@ static bool fm10k_clean_rx_irq(struct fm10k_q_vector *q_vector, 
	q_vector->rx.total_packets += total_packets;

	q_vector->rx.total_bytes += total_bytes;



	return total_packets < budget;

	return total_packets;

}



#define VXLAN_HLEN (sizeof(struct udphdr) + 8)
@@ -1422,7 +1422,7 @@ static int fm10k_poll(struct napi_struct *napi, int budget) 
	struct fm10k_q_vector *q_vector =

			       container_of(napi, struct fm10k_q_vector, napi);

	struct fm10k_ring *ring;

	int per_ring_budget;

	int per_ring_budget, work_done = 0;

	bool clean_complete = true;



	fm10k_for_each_ring(ring, q_vector->tx)
@@ -1436,16 +1436,19 @@ static int fm10k_poll(struct napi_struct *napi, int budget) 
	else

		per_ring_budget = budget;



	fm10k_for_each_ring(ring, q_vector->rx)

		clean_complete &= fm10k_clean_rx_irq(q_vector, ring,

						     per_ring_budget);

	fm10k_for_each_ring(ring, q_vector->rx) {

		int work = fm10k_clean_rx_irq(q_vector, ring, per_ring_budget);



		work_done += work;

		clean_complete &= !!(work < per_ring_budget);

	}



	/* If all work not completed, return budget and keep polling */

	if (!clean_complete)

		return budget;



	/* all work done, exit the polling mode */

	napi_complete(napi);

	napi_complete_done(napi, work_done);



	/* re-enable the q_vector */

	fm10k_qv_enable(q_vector);

drivers/net/ethernet/intel/i40e/i40e_txrx.c

+6 −2
Original line number Diff line number Diff line
@@ -1825,7 +1825,7 @@ int i40e_napi_poll(struct napi_struct *napi, int budget) 
	bool clean_complete = true;

	bool arm_wb = false;

	int budget_per_ring;

	int cleaned;

	int work_done = 0;



	if (test_bit(__I40E_DOWN, &vsi->state)) {

		napi_complete(napi);
@@ -1851,10 +1851,14 @@ int i40e_napi_poll(struct napi_struct *napi, int budget) 
	budget_per_ring = max(budget/q_vector->num_ringpairs, 1);



	i40e_for_each_ring(ring, q_vector->rx) {

		int cleaned;



		if (ring_is_ps_enabled(ring))

			cleaned = i40e_clean_rx_irq_ps(ring, budget_per_ring);

		else

			cleaned = i40e_clean_rx_irq_1buf(ring, budget_per_ring);



		work_done += cleaned;

		/* if we didn't clean as many as budgeted, we must be done */

		clean_complete &= (budget_per_ring != cleaned);

	}
@@ -1871,7 +1875,7 @@ int i40e_napi_poll(struct napi_struct *napi, int budget) 
		q_vector->arm_wb_state = false;



	/* Work is done so exit the polling mode and re-enable the interrupt */

	napi_complete(napi);

	napi_complete_done(napi, work_done);

	if (vsi->back->flags & I40E_FLAG_MSIX_ENABLED) {

		i40e_update_enable_itr(vsi, q_vector);

	} else { /* Legacy mode */

drivers/net/ethernet/intel/i40evf/i40e_txrx.c

+6 −2
Original line number Diff line number Diff line
@@ -1266,7 +1266,7 @@ int i40evf_napi_poll(struct napi_struct *napi, int budget) 
	bool clean_complete = true;

	bool arm_wb = false;

	int budget_per_ring;

	int cleaned;

	int work_done = 0;



	if (test_bit(__I40E_DOWN, &vsi->state)) {

		napi_complete(napi);
@@ -1292,10 +1292,14 @@ int i40evf_napi_poll(struct napi_struct *napi, int budget) 
	budget_per_ring = max(budget/q_vector->num_ringpairs, 1);



	i40e_for_each_ring(ring, q_vector->rx) {

		int cleaned;



		if (ring_is_ps_enabled(ring))

			cleaned = i40e_clean_rx_irq_ps(ring, budget_per_ring);

		else

			cleaned = i40e_clean_rx_irq_1buf(ring, budget_per_ring);



		work_done += cleaned;

		/* if we didn't clean as many as budgeted, we must be done */

		clean_complete &= (budget_per_ring != cleaned);

	}
@@ -1312,7 +1316,7 @@ int i40evf_napi_poll(struct napi_struct *napi, int budget) 
		q_vector->arm_wb_state = false;



	/* Work is done so exit the polling mode and re-enable the interrupt */

	napi_complete(napi);

	napi_complete_done(napi, work_done);

	i40e_update_enable_itr(vsi, q_vector);

	return 0;

}