提交 cd0b6fa6 编写于 作者: A Alexander Duyck 提交者: Jeff Kirsher

i40e: Replace ring container array with linked list

This replaces the ring container array with a linked list.  The idea is
to make the logic much easier to deal with since this will allow us to
call a simple helper function from the q_vectors to go through the
entire list.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NJesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: NKavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
上级 493fb300
......@@ -2516,7 +2516,7 @@ static irqreturn_t i40e_msix_clean_rings(int irq, void *data)
{
struct i40e_q_vector *q_vector = data;
if (!q_vector->tx.ring[0] && !q_vector->rx.ring[0])
if (!q_vector->tx.ring && !q_vector->rx.ring)
return IRQ_HANDLED;
napi_schedule(&q_vector->napi);
......@@ -2533,7 +2533,7 @@ static irqreturn_t i40e_fdir_clean_rings(int irq, void *data)
{
struct i40e_q_vector *q_vector = data;
if (!q_vector->tx.ring[0] && !q_vector->rx.ring[0])
if (!q_vector->tx.ring && !q_vector->rx.ring)
return IRQ_HANDLED;
pr_info("fdir ring cleaning needed\n");
......@@ -2560,14 +2560,14 @@ static int i40e_vsi_request_irq_msix(struct i40e_vsi *vsi, char *basename)
for (vector = 0; vector < q_vectors; vector++) {
struct i40e_q_vector *q_vector = vsi->q_vectors[vector];
if (q_vector->tx.ring[0] && q_vector->rx.ring[0]) {
if (q_vector->tx.ring && q_vector->rx.ring) {
snprintf(q_vector->name, sizeof(q_vector->name) - 1,
"%s-%s-%d", basename, "TxRx", rx_int_idx++);
tx_int_idx++;
} else if (q_vector->rx.ring[0]) {
} else if (q_vector->rx.ring) {
snprintf(q_vector->name, sizeof(q_vector->name) - 1,
"%s-%s-%d", basename, "rx", rx_int_idx++);
} else if (q_vector->tx.ring[0]) {
} else if (q_vector->tx.ring) {
snprintf(q_vector->name, sizeof(q_vector->name) - 1,
"%s-%s-%d", basename, "tx", tx_int_idx++);
} else {
......@@ -2778,40 +2778,26 @@ static irqreturn_t i40e_intr(int irq, void *data)
}
/**
* i40e_map_vector_to_rxq - Assigns the Rx queue to the vector
* i40e_map_vector_to_qp - Assigns the queue pair to the vector
* @vsi: the VSI being configured
* @v_idx: vector index
* @r_idx: rx queue index
* @qp_idx: queue pair index
**/
static void map_vector_to_rxq(struct i40e_vsi *vsi, int v_idx, int r_idx)
static void map_vector_to_qp(struct i40e_vsi *vsi, int v_idx, int qp_idx)
{
struct i40e_q_vector *q_vector = vsi->q_vectors[v_idx];
struct i40e_ring *rx_ring = &(vsi->rx_rings[r_idx]);
rx_ring->q_vector = q_vector;
q_vector->rx.ring[q_vector->rx.count] = rx_ring;
q_vector->rx.count++;
q_vector->rx.latency_range = I40E_LOW_LATENCY;
q_vector->vsi = vsi;
}
/**
* i40e_map_vector_to_txq - Assigns the Tx queue to the vector
* @vsi: the VSI being configured
* @v_idx: vector index
* @t_idx: tx queue index
**/
static void map_vector_to_txq(struct i40e_vsi *vsi, int v_idx, int t_idx)
{
struct i40e_q_vector *q_vector = vsi->q_vectors[v_idx];
struct i40e_ring *tx_ring = &(vsi->tx_rings[t_idx]);
struct i40e_ring *tx_ring = &(vsi->tx_rings[qp_idx]);
struct i40e_ring *rx_ring = &(vsi->rx_rings[qp_idx]);
tx_ring->q_vector = q_vector;
q_vector->tx.ring[q_vector->tx.count] = tx_ring;
tx_ring->next = q_vector->tx.ring;
q_vector->tx.ring = tx_ring;
q_vector->tx.count++;
q_vector->tx.latency_range = I40E_LOW_LATENCY;
q_vector->num_ringpairs++;
q_vector->vsi = vsi;
rx_ring->q_vector = q_vector;
rx_ring->next = q_vector->rx.ring;
q_vector->rx.ring = rx_ring;
q_vector->rx.count++;
}
/**
......@@ -2827,7 +2813,7 @@ static void i40e_vsi_map_rings_to_vectors(struct i40e_vsi *vsi)
{
int qp_remaining = vsi->num_queue_pairs;
int q_vectors = vsi->num_q_vectors;
int qp_per_vector;
int num_ringpairs;
int v_start = 0;
int qp_idx = 0;
......@@ -2835,11 +2821,21 @@ static void i40e_vsi_map_rings_to_vectors(struct i40e_vsi *vsi)
* group them so there are multiple queues per vector.
*/
for (; v_start < q_vectors && qp_remaining; v_start++) {
qp_per_vector = DIV_ROUND_UP(qp_remaining, q_vectors - v_start);
for (; qp_per_vector;
qp_per_vector--, qp_idx++, qp_remaining--) {
map_vector_to_rxq(vsi, v_start, qp_idx);
map_vector_to_txq(vsi, v_start, qp_idx);
struct i40e_q_vector *q_vector = vsi->q_vectors[v_start];
num_ringpairs = DIV_ROUND_UP(qp_remaining, q_vectors - v_start);
q_vector->num_ringpairs = num_ringpairs;
q_vector->rx.count = 0;
q_vector->tx.count = 0;
q_vector->rx.ring = NULL;
q_vector->tx.ring = NULL;
while (num_ringpairs--) {
map_vector_to_qp(vsi, v_start, qp_idx);
qp_idx++;
qp_remaining--;
}
}
}
......@@ -3179,16 +3175,17 @@ static void i40e_vsi_free_irq(struct i40e_vsi *vsi)
static void i40e_free_q_vector(struct i40e_vsi *vsi, int v_idx)
{
struct i40e_q_vector *q_vector = vsi->q_vectors[v_idx];
int r_idx;
struct i40e_ring *ring;
if (!q_vector)
return;
/* disassociate q_vector from rings */
for (r_idx = 0; r_idx < q_vector->tx.count; r_idx++)
q_vector->tx.ring[r_idx]->q_vector = NULL;
for (r_idx = 0; r_idx < q_vector->rx.count; r_idx++)
q_vector->rx.ring[r_idx]->q_vector = NULL;
i40e_for_each_ring(ring, q_vector->tx)
ring->q_vector = NULL;
i40e_for_each_ring(ring, q_vector->rx)
ring->q_vector = NULL;
/* only VSI w/ an associated netdev is set up w/ NAPI */
if (vsi->netdev)
......@@ -5312,6 +5309,9 @@ static int i40e_alloc_q_vector(struct i40e_vsi *vsi, int v_idx)
netif_napi_add(vsi->netdev, &q_vector->napi,
i40e_napi_poll, vsi->work_limit);
q_vector->rx.latency_range = I40E_LOW_LATENCY;
q_vector->tx.latency_range = I40E_LOW_LATENCY;
/* tie q_vector and vsi together */
vsi->q_vectors[v_idx] = q_vector;
......
......@@ -1100,27 +1100,28 @@ int i40e_napi_poll(struct napi_struct *napi, int budget)
struct i40e_q_vector *q_vector =
container_of(napi, struct i40e_q_vector, napi);
struct i40e_vsi *vsi = q_vector->vsi;
struct i40e_ring *ring;
bool clean_complete = true;
int budget_per_ring;
int i;
if (test_bit(__I40E_DOWN, &vsi->state)) {
napi_complete(napi);
return 0;
}
/* Since the actual Tx work is minimal, we can give the Tx a larger
* budget and be more aggressive about cleaning up the Tx descriptors.
*/
i40e_for_each_ring(ring, q_vector->tx)
clean_complete &= i40e_clean_tx_irq(ring, vsi->work_limit);
/* We attempt to distribute budget to each Rx queue fairly, but don't
* allow the budget to go below 1 because that would exit polling early.
* Since the actual Tx work is minimal, we can give the Tx a larger
* budget and be more aggressive about cleaning up the Tx descriptors.
*/
budget_per_ring = max(budget/q_vector->num_ringpairs, 1);
for (i = 0; i < q_vector->num_ringpairs; i++) {
clean_complete &= i40e_clean_tx_irq(q_vector->tx.ring[i],
vsi->work_limit);
clean_complete &= i40e_clean_rx_irq(q_vector->rx.ring[i],
budget_per_ring);
}
i40e_for_each_ring(ring, q_vector->rx)
clean_complete &= i40e_clean_rx_irq(ring, budget_per_ring);
/* If work not completed, return budget and polling will return */
if (!clean_complete)
......
......@@ -180,6 +180,7 @@ enum i40e_ring_state_t {
/* struct that defines a descriptor ring, associated with a VSI */
struct i40e_ring {
struct i40e_ring *next; /* pointer to next ring in q_vector */
void *desc; /* Descriptor ring memory */
struct device *dev; /* Used for DMA mapping */
struct net_device *netdev; /* netdev ring maps to */
......@@ -236,9 +237,8 @@ enum i40e_latency_range {
};
struct i40e_ring_container {
#define I40E_MAX_RINGPAIR_PER_VECTOR 8
/* array of pointers to rings */
struct i40e_ring *ring[I40E_MAX_RINGPAIR_PER_VECTOR];
struct i40e_ring *ring;
unsigned int total_bytes; /* total bytes processed this int */
unsigned int total_packets; /* total packets processed this int */
u16 count;
......@@ -246,6 +246,10 @@ struct i40e_ring_container {
u16 itr;
};
/* iterator for handling rings in ring container */
#define i40e_for_each_ring(pos, head) \
for (pos = (head).ring; pos != NULL; pos = pos->next)
void i40e_alloc_rx_buffers(struct i40e_ring *rxr, u16 cleaned_count);
netdev_tx_t i40e_lan_xmit_frame(struct sk_buff *skb, struct net_device *netdev);
void i40e_clean_tx_ring(struct i40e_ring *tx_ring);
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册