From: Jesse Brandeburg <jesse.brandeburg@intel.com>
Date: Thu, 8 Nov 2018 14:55:32 -0800
Subject: ethernet/intel: consolidate NAPI and NAPI exit
Patch-mainline: v5.0-rc1
Git-commit: 0bcd952feec7042d9a5383b639c8edc943402add
References: bsc#1118657 FATE#325278
While reviewing code, I noticed that Eric Dumazet recommends that
drivers check the return code of napi_complete_done, and use that
to decide to enable interrupts or not when exiting poll. One of
the Intel drivers was already fixed (ixgbe).
Upon looking at the Intel drivers as a whole, we are handling our
polling and NAPI exit in a few different ways based on whether we
have multiqueue and whether we have Tx cleanup included. Several
drivers had the bug of exiting NAPI with return 0, which appears
to mess up the accounting in the stack.
Consolidate all the NAPI routines to do best known way of exiting
and to just mostly look like each other.
1) check return code of napi_complete_done to control interrupt enable
2) return the actual amount of work done.
3) return budget immediately if need NAPI poll again
Tested the changes on e1000e with a high interrupt rate set, and
it shows about an 8% reduction in the CPU utilization when busy
polling because we aren't re-enabling interrupts when we're about
to be polled.
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Acked-by: Thomas Bogendoerfer <tbogendoerfer@suse.de>
---
drivers/net/ethernet/intel/e100.c | 10 ++++++----
drivers/net/ethernet/intel/e1000/e1000_main.c | 11 ++++++-----
drivers/net/ethernet/intel/e1000e/netdev.c | 17 +++++++++--------
drivers/net/ethernet/intel/fm10k/fm10k_main.c | 10 +++++-----
drivers/net/ethernet/intel/i40e/i40e_txrx.c | 9 +++++----
drivers/net/ethernet/intel/iavf/iavf_txrx.c | 9 +++++----
drivers/net/ethernet/intel/ice/ice_txrx.c | 10 ++++++----
drivers/net/ethernet/intel/igb/igb_main.c | 10 ++++++----
drivers/net/ethernet/intel/igbvf/netdev.c | 9 ++++++---
drivers/net/ethernet/intel/igc/igc_main.c | 10 ++++++----
drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 22 +++++++++++++---------
11 files changed, 73 insertions(+), 54 deletions(-)
--- a/drivers/net/ethernet/intel/e100.c
+++ b/drivers/net/ethernet/intel/e100.c
@@ -2251,11 +2251,13 @@ static int e100_poll(struct napi_struct
e100_rx_clean(nic, &work_done, budget);
e100_tx_clean(nic);
- /* If budget not fully consumed, exit the polling mode */
- if (work_done < budget) {
- napi_complete_done(napi, work_done);
+ /* If budget fully consumed, continue polling */
+ if (work_done == budget)
+ return budget;
+
+ /* only re-enable interrupt if stack agrees polling is really done */
+ if (likely(napi_complete_done(napi, work_done)))
e100_enable_irq(nic);
- }
return work_done;
}
--- a/drivers/net/ethernet/intel/e1000/e1000_main.c
+++ b/drivers/net/ethernet/intel/e1000/e1000_main.c
@@ -3831,14 +3831,15 @@ static int e1000_clean(struct napi_struc
adapter->clean_rx(adapter, &adapter->rx_ring[0], &work_done, budget);
- if (!tx_clean_complete)
- work_done = budget;
+ if (!tx_clean_complete || work_done == budget)
+ return budget;
- /* If budget not fully consumed, exit the polling mode */
- if (work_done < budget) {
+ /* Exit the polling mode, but don't re-enable interrupts if stack might
+ * poll us due to busy-polling
+ */
+ if (likely(napi_complete_done(napi, work_done))) {
if (likely(adapter->itr_setting & 3))
e1000_set_itr(adapter);
- napi_complete_done(napi, work_done);
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000_irq_enable(adapter);
}
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -2669,9 +2669,9 @@ err:
/**
* e1000e_poll - NAPI Rx polling callback
* @napi: struct associated with this polling callback
- * @weight: number of packets driver is allowed to process this poll
+ * @budget: number of packets driver is allowed to process this poll
**/
-static int e1000e_poll(struct napi_struct *napi, int weight)
+static int e1000e_poll(struct napi_struct *napi, int budget)
{
struct e1000_adapter *adapter = container_of(napi, struct e1000_adapter,
napi);
@@ -2685,16 +2685,17 @@ static int e1000e_poll(struct napi_struc
(adapter->rx_ring->ims_val & adapter->tx_ring->ims_val))
tx_cleaned = e1000_clean_tx_irq(adapter->tx_ring);
- adapter->clean_rx(adapter->rx_ring, &work_done, weight);
+ adapter->clean_rx(adapter->rx_ring, &work_done, budget);
- if (!tx_cleaned)
- work_done = weight;
+ if (!tx_cleaned || work_done == budget)
+ return budget;
- /* If weight not fully consumed, exit the polling mode */
- if (work_done < weight) {
+ /* Exit the polling mode, but don't re-enable interrupts if stack might
+ * poll us due to busy-polling
+ */
+ if (likely(napi_complete_done(napi, work_done))) {
if (adapter->itr_setting & 3)
e1000_set_itr(adapter);
- napi_complete_done(napi, work_done);
if (!test_bit(__E1000_DOWN, &adapter->state)) {
if (adapter->msix_entries)
ew32(IMS, adapter->rx_ring->ims_val);
--- a/drivers/net/ethernet/intel/fm10k/fm10k_main.c
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_main.c
@@ -1482,11 +1482,11 @@ static int fm10k_poll(struct napi_struct
if (!clean_complete)
return budget;
- /* all work done, exit the polling mode */
- napi_complete_done(napi, work_done);
-
- /* re-enable the q_vector */
- fm10k_qv_enable(q_vector);
+ /* Exit the polling mode, but don't re-enable interrupts if stack might
+ * poll us due to busy-polling
+ */
+ if (likely(napi_complete_done(napi, work_done)))
+ fm10k_qv_enable(q_vector);
return min(work_done, budget - 1);
}
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -2671,10 +2671,11 @@ tx_only:
if (vsi->back->flags & I40E_TXR_FLAGS_WB_ON_ITR)
q_vector->arm_wb_state = false;
- /* Work is done so exit the polling mode and re-enable the interrupt */
- napi_complete_done(napi, work_done);
-
- i40e_update_enable_itr(vsi, q_vector);
+ /* Exit the polling mode, but don't re-enable interrupts if stack might
+ * poll us due to busy-polling
+ */
+ if (likely(napi_complete_done(napi, work_done)))
+ i40e_update_enable_itr(vsi, q_vector);
return min(work_done, budget - 1);
}
--- a/drivers/net/ethernet/intel/iavf/iavf_txrx.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_txrx.c
@@ -1761,10 +1761,11 @@ tx_only:
if (vsi->back->flags & IAVF_TXR_FLAGS_WB_ON_ITR)
q_vector->arm_wb_state = false;
- /* Work is done so exit the polling mode and re-enable the interrupt */
- napi_complete_done(napi, work_done);
-
- iavf_update_enable_itr(vsi, q_vector);
+ /* Exit the polling mode, but don't re-enable interrupts if stack might
+ * poll us due to busy-polling
+ */
+ if (likely(napi_complete_done(napi, work_done)))
+ iavf_update_enable_itr(vsi, q_vector);
return min(work_done, budget - 1);
}
--- a/drivers/net/ethernet/intel/ice/ice_txrx.c
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.c
@@ -1103,10 +1103,12 @@ int ice_napi_poll(struct napi_struct *na
if (!clean_complete)
return budget;
- /* Work is done so exit the polling mode and re-enable the interrupt */
- napi_complete_done(napi, work_done);
- if (test_bit(ICE_FLAG_MSIX_ENA, pf->flags))
- ice_irq_dynamic_ena(&vsi->back->hw, vsi, q_vector);
+ /* Exit the polling mode, but don't re-enable interrupts if stack might
+ * poll us due to busy-polling
+ */
+ if (likely(napi_complete_done(napi, work_done)))
+ if (test_bit(ICE_FLAG_MSIX_ENA, pf->flags))
+ ice_irq_dynamic_ena(&vsi->back->hw, vsi, q_vector);
return min(work_done, budget - 1);
}
--- a/drivers/net/ethernet/intel/igb/igb_main.c
+++ b/drivers/net/ethernet/intel/igb/igb_main.c
@@ -7670,11 +7670,13 @@ static int igb_poll(struct napi_struct *
if (!clean_complete)
return budget;
- /* If not enough Rx work done, exit the polling mode */
- napi_complete_done(napi, work_done);
- igb_ring_irq_enable(q_vector);
+ /* Exit the polling mode, but don't re-enable interrupts if stack might
+ * poll us due to busy-polling
+ */
+ if (likely(napi_complete_done(napi, work_done)))
+ igb_ring_irq_enable(q_vector);
- return 0;
+ return min(work_done, budget - 1);
}
/**
--- a/drivers/net/ethernet/intel/igbvf/netdev.c
+++ b/drivers/net/ethernet/intel/igbvf/netdev.c
@@ -1209,10 +1209,13 @@ static int igbvf_poll(struct napi_struct
igbvf_clean_rx_irq(adapter, &work_done, budget);
- /* If not enough Rx work done, exit the polling mode */
- if (work_done < budget) {
- napi_complete_done(napi, work_done);
+ if (work_done == budget)
+ return budget;
+ /* Exit the polling mode, but don't re-enable interrupts if stack might
+ * poll us due to busy-polling
+ */
+ if (likely(napi_complete_done(napi, work_done))) {
if (adapter->requested_itr & 3)
igbvf_set_itr(adapter);
--- a/drivers/net/ethernet/intel/igc/igc_main.c
+++ b/drivers/net/ethernet/intel/igc/igc_main.c
@@ -2852,11 +2852,13 @@ static int igc_poll(struct napi_struct *
if (!clean_complete)
return budget;
- /* If not enough Rx work done, exit the polling mode */
- napi_complete_done(napi, work_done);
- igc_ring_irq_enable(q_vector);
+ /* Exit the polling mode, but don't re-enable interrupts if stack might
+ * poll us due to busy-polling
+ */
+ if (likely(napi_complete_done(napi, work_done)))
+ igc_ring_irq_enable(q_vector);
- return 0;
+ return min(work_done, budget - 1);
}
/**
--- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
+++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
@@ -1316,16 +1316,20 @@ static int ixgbevf_poll(struct napi_stru
/* If all work not completed, return budget and keep polling */
if (!clean_complete)
return budget;
- /* all work done, exit the polling mode */
- napi_complete_done(napi, work_done);
- if (adapter->rx_itr_setting == 1)
- ixgbevf_set_itr(q_vector);
- if (!test_bit(__IXGBEVF_DOWN, &adapter->state) &&
- !test_bit(__IXGBEVF_REMOVING, &adapter->state))
- ixgbevf_irq_enable_queues(adapter,
- BIT(q_vector->v_idx));
- return 0;
+ /* Exit the polling mode, but don't re-enable interrupts if stack might
+ * poll us due to busy-polling
+ */
+ if (likely(napi_complete_done(napi, work_done))) {
+ if (adapter->rx_itr_setting == 1)
+ ixgbevf_set_itr(q_vector);
+ if (!test_bit(__IXGBEVF_DOWN, &adapter->state) &&
+ !test_bit(__IXGBEVF_REMOVING, &adapter->state))
+ ixgbevf_irq_enable_queues(adapter,
+ BIT(q_vector->v_idx));
+ }
+
+ return min(work_done, budget - 1);
}
/**