summaryrefslogtreecommitdiff
path: root/drivers/net/ethernet/intel/idpf/idpf_txrx.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/net/ethernet/intel/idpf/idpf_txrx.c')
-rw-r--r--drivers/net/ethernet/intel/idpf/idpf_txrx.c2120
1 files changed, 1189 insertions, 931 deletions
diff --git a/drivers/net/ethernet/intel/idpf/idpf_txrx.c b/drivers/net/ethernet/intel/idpf/idpf_txrx.c
index af2879f03b8d..1d91c56f7469 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_txrx.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_txrx.c
@@ -1,42 +1,36 @@
// SPDX-License-Identifier: GPL-2.0-only
/* Copyright (C) 2023 Intel Corporation */
-#include <net/libeth/rx.h>
-
#include "idpf.h"
+#include "idpf_ptp.h"
#include "idpf_virtchnl.h"
+#include "xdp.h"
+#include "xsk.h"
-static bool idpf_chk_linearize(struct sk_buff *skb, unsigned int max_bufs,
- unsigned int count);
+#define idpf_tx_buf_next(buf) (*(u32 *)&(buf)->priv)
+LIBETH_SQE_CHECK_PRIV(u32);
/**
- * idpf_buf_lifo_push - push a buffer pointer onto stack
- * @stack: pointer to stack struct
- * @buf: pointer to buf to push
+ * idpf_chk_linearize - Check if skb exceeds max descriptors per packet
+ * @skb: send buffer
+ * @max_bufs: maximum scatter gather buffers for single packet
+ * @count: number of buffers this packet needs
*
- * Returns 0 on success, negative on failure
- **/
-static int idpf_buf_lifo_push(struct idpf_buf_lifo *stack,
- struct idpf_tx_stash *buf)
+ * Make sure we don't exceed maximum scatter gather buffers for a single
+ * packet.
+ * TSO case has been handled earlier from idpf_features_check().
+ */
+static bool idpf_chk_linearize(const struct sk_buff *skb,
+ unsigned int max_bufs,
+ unsigned int count)
{
- if (unlikely(stack->top == stack->size))
- return -ENOSPC;
-
- stack->bufs[stack->top++] = buf;
-
- return 0;
-}
+ if (likely(count <= max_bufs))
+ return false;
-/**
- * idpf_buf_lifo_pop - pop a buffer pointer from stack
- * @stack: pointer to stack struct
- **/
-static struct idpf_tx_stash *idpf_buf_lifo_pop(struct idpf_buf_lifo *stack)
-{
- if (unlikely(!stack->top))
- return NULL;
+ if (skb_is_gso(skb))
+ return false;
- return stack->bufs[--stack->top];
+ return true;
}
/**
@@ -60,32 +54,23 @@ void idpf_tx_timeout(struct net_device *netdev, unsigned int txqueue)
}
}
-/**
- * idpf_tx_buf_rel - Release a Tx buffer
- * @tx_q: the queue that owns the buffer
- * @tx_buf: the buffer to free
- */
-static void idpf_tx_buf_rel(struct idpf_tx_queue *tx_q,
- struct idpf_tx_buf *tx_buf)
+static void idpf_tx_buf_clean(struct idpf_tx_queue *txq)
{
- if (tx_buf->skb) {
- if (dma_unmap_len(tx_buf, len))
- dma_unmap_single(tx_q->dev,
- dma_unmap_addr(tx_buf, dma),
- dma_unmap_len(tx_buf, len),
- DMA_TO_DEVICE);
- dev_kfree_skb_any(tx_buf->skb);
- } else if (dma_unmap_len(tx_buf, len)) {
- dma_unmap_page(tx_q->dev,
- dma_unmap_addr(tx_buf, dma),
- dma_unmap_len(tx_buf, len),
- DMA_TO_DEVICE);
- }
+ struct libeth_sq_napi_stats ss = { };
+ struct xdp_frame_bulk bq;
+ struct libeth_cq_pp cp = {
+ .dev = txq->dev,
+ .bq = &bq,
+ .ss = &ss,
+ };
+
+ xdp_frame_bulk_init(&bq);
- tx_buf->next_to_watch = NULL;
- tx_buf->skb = NULL;
- tx_buf->compl_tag = IDPF_SPLITQ_TX_INVAL_COMPL_TAG;
- dma_unmap_len_set(tx_buf, len, 0);
+ /* Free all the Tx buffer sk_buffs */
+ for (u32 i = 0; i < txq->buf_pool_size; i++)
+ libeth_tx_complete_any(&txq->tx_buf[i], &cp);
+
+ xdp_flush_frame_bulk(&bq);
}
/**
@@ -94,32 +79,17 @@ static void idpf_tx_buf_rel(struct idpf_tx_queue *tx_q,
*/
static void idpf_tx_buf_rel_all(struct idpf_tx_queue *txq)
{
- struct idpf_buf_lifo *buf_stack;
- u16 i;
-
/* Buffers already cleared, nothing to do */
if (!txq->tx_buf)
return;
- /* Free all the Tx buffer sk_buffs */
- for (i = 0; i < txq->desc_count; i++)
- idpf_tx_buf_rel(txq, &txq->tx_buf[i]);
+ if (idpf_queue_has(XSK, txq))
+ idpf_xsksq_clean(txq);
+ else
+ idpf_tx_buf_clean(txq);
kfree(txq->tx_buf);
txq->tx_buf = NULL;
-
- if (!idpf_queue_has(FLOW_SCH_EN, txq))
- return;
-
- buf_stack = &txq->stash->buf_stack;
- if (!buf_stack->bufs)
- return;
-
- for (i = 0; i < buf_stack->size; i++)
- kfree(buf_stack->bufs[i]);
-
- kfree(buf_stack->bufs);
- buf_stack->bufs = NULL;
}
/**
@@ -130,11 +100,24 @@ static void idpf_tx_buf_rel_all(struct idpf_tx_queue *txq)
*/
static void idpf_tx_desc_rel(struct idpf_tx_queue *txq)
{
+ bool xdp = idpf_queue_has(XDP, txq);
+
+ if (xdp)
+ libeth_xdpsq_deinit_timer(txq->timer);
+
idpf_tx_buf_rel_all(txq);
+ if (!xdp)
+ netdev_tx_reset_subqueue(txq->netdev, txq->idx);
+
+ idpf_xsk_clear_queue(txq, VIRTCHNL2_QUEUE_TYPE_TX);
+
if (!txq->desc_ring)
return;
+ if (!xdp && txq->refillq)
+ kfree(txq->refillq->ring);
+
dmam_free_coherent(txq->dev, txq->size, txq->desc_ring, txq->dma);
txq->desc_ring = NULL;
txq->next_to_use = 0;
@@ -149,12 +132,14 @@ static void idpf_tx_desc_rel(struct idpf_tx_queue *txq)
*/
static void idpf_compl_desc_rel(struct idpf_compl_queue *complq)
{
- if (!complq->comp)
+ idpf_xsk_clear_queue(complq, VIRTCHNL2_QUEUE_TYPE_TX_COMPLETION);
+
+ if (!complq->desc_ring)
return;
dma_free_coherent(complq->netdev->dev.parent, complq->size,
- complq->comp, complq->dma);
- complq->comp = NULL;
+ complq->desc_ring, complq->dma);
+ complq->desc_ring = NULL;
complq->next_to_use = 0;
complq->next_to_clean = 0;
}
@@ -191,45 +176,18 @@ static void idpf_tx_desc_rel_all(struct idpf_vport *vport)
*/
static int idpf_tx_buf_alloc_all(struct idpf_tx_queue *tx_q)
{
- struct idpf_buf_lifo *buf_stack;
- int buf_size;
- int i;
-
/* Allocate book keeping buffers only. Buffers to be supplied to HW
* are allocated by kernel network stack and received as part of skb
*/
- buf_size = sizeof(struct idpf_tx_buf) * tx_q->desc_count;
- tx_q->tx_buf = kzalloc(buf_size, GFP_KERNEL);
+ if (idpf_queue_has(FLOW_SCH_EN, tx_q))
+ tx_q->buf_pool_size = U16_MAX;
+ else
+ tx_q->buf_pool_size = tx_q->desc_count;
+ tx_q->tx_buf = kcalloc(tx_q->buf_pool_size, sizeof(*tx_q->tx_buf),
+ GFP_KERNEL);
if (!tx_q->tx_buf)
return -ENOMEM;
- /* Initialize tx_bufs with invalid completion tags */
- for (i = 0; i < tx_q->desc_count; i++)
- tx_q->tx_buf[i].compl_tag = IDPF_SPLITQ_TX_INVAL_COMPL_TAG;
-
- if (!idpf_queue_has(FLOW_SCH_EN, tx_q))
- return 0;
-
- buf_stack = &tx_q->stash->buf_stack;
-
- /* Initialize tx buf stack for out-of-order completions if
- * flow scheduling offload is enabled
- */
- buf_stack->bufs = kcalloc(tx_q->desc_count, sizeof(*buf_stack->bufs),
- GFP_KERNEL);
- if (!buf_stack->bufs)
- return -ENOMEM;
-
- buf_stack->size = tx_q->desc_count;
- buf_stack->top = tx_q->desc_count;
-
- for (i = 0; i < tx_q->desc_count; i++) {
- buf_stack->bufs[i] = kzalloc(sizeof(*buf_stack->bufs[i]),
- GFP_KERNEL);
- if (!buf_stack->bufs[i])
- return -ENOMEM;
- }
-
return 0;
}
@@ -244,6 +202,7 @@ static int idpf_tx_desc_alloc(const struct idpf_vport *vport,
struct idpf_tx_queue *tx_q)
{
struct device *dev = tx_q->dev;
+ struct idpf_sw_queue *refillq;
int err;
err = idpf_tx_buf_alloc_all(tx_q);
@@ -267,6 +226,33 @@ static int idpf_tx_desc_alloc(const struct idpf_vport *vport,
tx_q->next_to_clean = 0;
idpf_queue_set(GEN_CHK, tx_q);
+ idpf_xsk_setup_queue(vport, tx_q, VIRTCHNL2_QUEUE_TYPE_TX);
+
+ if (!idpf_queue_has(FLOW_SCH_EN, tx_q))
+ return 0;
+
+ refillq = tx_q->refillq;
+ refillq->desc_count = tx_q->buf_pool_size;
+ refillq->ring = kcalloc(refillq->desc_count, sizeof(u32),
+ GFP_KERNEL);
+ if (!refillq->ring) {
+ err = -ENOMEM;
+ goto err_alloc;
+ }
+
+ for (unsigned int i = 0; i < refillq->desc_count; i++)
+ refillq->ring[i] =
+ FIELD_PREP(IDPF_RFL_BI_BUFID_M, i) |
+ FIELD_PREP(IDPF_RFL_BI_GEN_M,
+ idpf_queue_has(GEN_CHK, refillq));
+
+ /* Go ahead and flip the GEN bit since this counts as filling
+ * up the ring, i.e. we already ring wrapped.
+ */
+ idpf_queue_change(GEN_CHK, refillq);
+
+ tx_q->last_re = tx_q->desc_count - IDPF_TX_SPLITQ_RE_MIN_GAP;
+
return 0;
err_alloc:
@@ -285,18 +271,25 @@ err_alloc:
static int idpf_compl_desc_alloc(const struct idpf_vport *vport,
struct idpf_compl_queue *complq)
{
- complq->size = array_size(complq->desc_count, sizeof(*complq->comp));
+ u32 desc_size;
- complq->comp = dma_alloc_coherent(complq->netdev->dev.parent,
- complq->size, &complq->dma,
- GFP_KERNEL);
- if (!complq->comp)
+ desc_size = idpf_queue_has(FLOW_SCH_EN, complq) ?
+ sizeof(*complq->comp) : sizeof(*complq->comp_4b);
+ complq->size = array_size(complq->desc_count, desc_size);
+
+ complq->desc_ring = dma_alloc_coherent(complq->netdev->dev.parent,
+ complq->size, &complq->dma,
+ GFP_KERNEL);
+ if (!complq->desc_ring)
return -ENOMEM;
complq->next_to_use = 0;
complq->next_to_clean = 0;
idpf_queue_set(GEN_CHK, complq);
+ idpf_xsk_setup_queue(vport, complq,
+ VIRTCHNL2_QUEUE_TYPE_TX_COMPLETION);
+
return 0;
}
@@ -317,8 +310,6 @@ static int idpf_tx_desc_alloc_all(struct idpf_vport *vport)
for (i = 0; i < vport->num_txq_grp; i++) {
for (j = 0; j < vport->txq_grps[i].num_txq; j++) {
struct idpf_tx_queue *txq = vport->txq_grps[i].txqs[j];
- u8 gen_bits = 0;
- u16 bufidx_mask;
err = idpf_tx_desc_alloc(vport, txq);
if (err) {
@@ -327,34 +318,6 @@ static int idpf_tx_desc_alloc_all(struct idpf_vport *vport)
i);
goto err_out;
}
-
- if (!idpf_is_queue_model_split(vport->txq_model))
- continue;
-
- txq->compl_tag_cur_gen = 0;
-
- /* Determine the number of bits in the bufid
- * mask and add one to get the start of the
- * generation bits
- */
- bufidx_mask = txq->desc_count - 1;
- while (bufidx_mask >> 1) {
- txq->compl_tag_gen_s++;
- bufidx_mask = bufidx_mask >> 1;
- }
- txq->compl_tag_gen_s++;
-
- gen_bits = IDPF_TX_SPLITQ_COMPL_TAG_WIDTH -
- txq->compl_tag_gen_s;
- txq->compl_tag_gen_max = GETMAXVAL(gen_bits);
-
- /* Set bufid mask based on location of first
- * gen bit; it cannot simply be the descriptor
- * ring size-1 since we can have size values
- * where not all of those bits are set.
- */
- txq->compl_tag_bufid_m =
- GETMAXVAL(txq->compl_tag_gen_s);
}
if (!idpf_is_queue_model_split(vport->txq_model))
@@ -383,12 +346,12 @@ err_out:
*/
static void idpf_rx_page_rel(struct libeth_fqe *rx_buf)
{
- if (unlikely(!rx_buf->page))
+ if (unlikely(!rx_buf->netmem))
return;
- page_pool_put_full_page(rx_buf->page->pp, rx_buf->page, false);
+ libeth_rx_recycle_slow(rx_buf->netmem);
- rx_buf->page = NULL;
+ rx_buf->netmem = 0;
rx_buf->offset = 0;
}
@@ -426,6 +389,11 @@ static void idpf_rx_buf_rel_bufq(struct idpf_buf_queue *bufq)
if (!bufq->buf)
return;
+ if (idpf_queue_has(XSK, bufq)) {
+ idpf_xskfq_rel(bufq);
+ return;
+ }
+
/* Free all the bufs allocated and given to hw on Rx queue */
for (u32 i = 0; i < bufq->desc_count; i++)
idpf_rx_page_rel(&bufq->buf[i]);
@@ -474,14 +442,14 @@ static void idpf_rx_desc_rel(struct idpf_rx_queue *rxq, struct device *dev,
if (!rxq)
return;
- if (rxq->skb) {
- dev_kfree_skb_any(rxq->skb);
- rxq->skb = NULL;
- }
+ if (!idpf_queue_has(XSK, rxq))
+ libeth_xdp_return_stash(&rxq->xdp);
if (!idpf_is_queue_model_split(model))
idpf_rx_buf_rel_all(rxq);
+ idpf_xsk_clear_queue(rxq, VIRTCHNL2_QUEUE_TYPE_RX);
+
rxq->next_to_alloc = 0;
rxq->next_to_clean = 0;
rxq->next_to_use = 0;
@@ -504,6 +472,7 @@ static void idpf_rx_desc_rel_bufq(struct idpf_buf_queue *bufq,
return;
idpf_rx_buf_rel_bufq(bufq);
+ idpf_xsk_clear_queue(bufq, VIRTCHNL2_QUEUE_TYPE_RX_BUFFER);
bufq->next_to_alloc = 0;
bufq->next_to_clean = 0;
@@ -586,6 +555,7 @@ static int idpf_rx_hdr_buf_alloc_all(struct idpf_buf_queue *bufq)
struct libeth_fq fq = {
.count = bufq->desc_count,
.type = LIBETH_FQE_HDR,
+ .xdp = idpf_xdp_enabled(bufq->q_vector->vport),
.nid = idpf_q_vector_to_mem(bufq->q_vector),
};
int ret;
@@ -603,18 +573,18 @@ static int idpf_rx_hdr_buf_alloc_all(struct idpf_buf_queue *bufq)
}
/**
- * idpf_rx_post_buf_refill - Post buffer id to refill queue
+ * idpf_post_buf_refill - Post buffer id to refill queue
* @refillq: refill queue to post to
* @buf_id: buffer id to post
*/
-static void idpf_rx_post_buf_refill(struct idpf_sw_queue *refillq, u16 buf_id)
+static void idpf_post_buf_refill(struct idpf_sw_queue *refillq, u16 buf_id)
{
u32 nta = refillq->next_to_use;
/* store the buffer ID and the SW maintained GEN bit to the refillq */
refillq->ring[nta] =
- FIELD_PREP(IDPF_RX_BI_BUFID_M, buf_id) |
- FIELD_PREP(IDPF_RX_BI_GEN_M,
+ FIELD_PREP(IDPF_RFL_BI_BUFID_M, buf_id) |
+ FIELD_PREP(IDPF_RFL_BI_GEN_M,
idpf_queue_has(GEN_CHK, refillq));
if (unlikely(++nta == refillq->desc_count)) {
@@ -785,10 +755,14 @@ static int idpf_rx_bufs_init(struct idpf_buf_queue *bufq,
.count = bufq->desc_count,
.type = type,
.hsplit = idpf_queue_has(HSPLIT_EN, bufq),
+ .xdp = idpf_xdp_enabled(bufq->q_vector->vport),
.nid = idpf_q_vector_to_mem(bufq->q_vector),
};
int ret;
+ if (idpf_queue_has(XSK, bufq))
+ return idpf_xskfq_init(bufq);
+
ret = libeth_rx_fq_create(&fq, &bufq->q_vector->napi);
if (ret)
return ret;
@@ -812,6 +786,8 @@ int idpf_rx_bufs_init_all(struct idpf_vport *vport)
bool split = idpf_is_queue_model_split(vport->rxq_model);
int i, j, err;
+ idpf_xdp_copy_prog_to_rqs(vport, vport->xdp_prog);
+
for (i = 0; i < vport->num_rxq_grp; i++) {
struct idpf_rxq_group *rx_qgrp = &vport->rxq_grps[i];
u32 truesize = 0;
@@ -882,6 +858,8 @@ static int idpf_rx_desc_alloc(const struct idpf_vport *vport,
rxq->next_to_use = 0;
idpf_queue_set(GEN_CHK, rxq);
+ idpf_xsk_setup_queue(vport, rxq, VIRTCHNL2_QUEUE_TYPE_RX);
+
return 0;
}
@@ -907,9 +885,10 @@ static int idpf_bufq_desc_alloc(const struct idpf_vport *vport,
bufq->next_to_alloc = 0;
bufq->next_to_clean = 0;
bufq->next_to_use = 0;
-
idpf_queue_set(GEN_CHK, bufq);
+ idpf_xsk_setup_queue(vport, bufq, VIRTCHNL2_QUEUE_TYPE_RX_BUFFER);
+
return 0;
}
@@ -943,8 +922,8 @@ static int idpf_rx_desc_alloc_all(struct idpf_vport *vport)
err = idpf_rx_desc_alloc(vport, q);
if (err) {
pci_err(vport->adapter->pdev,
- "Memory allocation for Rx Queue %u failed\n",
- i);
+ "Memory allocation for Rx queue %u from queue group %u failed\n",
+ j, i);
goto err_out;
}
}
@@ -960,8 +939,8 @@ static int idpf_rx_desc_alloc_all(struct idpf_vport *vport)
err = idpf_bufq_desc_alloc(vport, q);
if (err) {
pci_err(vport->adapter->pdev,
- "Memory allocation for Rx Buffer Queue %u failed\n",
- i);
+ "Memory allocation for Rx Buffer Queue %u from queue group %u failed\n",
+ j, i);
goto err_out;
}
}
@@ -975,6 +954,341 @@ err_out:
return err;
}
+static int idpf_init_queue_set(const struct idpf_queue_set *qs)
+{
+ const struct idpf_vport *vport = qs->vport;
+ bool splitq;
+ int err;
+
+ splitq = idpf_is_queue_model_split(vport->rxq_model);
+
+ for (u32 i = 0; i < qs->num; i++) {
+ const struct idpf_queue_ptr *q = &qs->qs[i];
+ struct idpf_buf_queue *bufq;
+
+ switch (q->type) {
+ case VIRTCHNL2_QUEUE_TYPE_RX:
+ err = idpf_rx_desc_alloc(vport, q->rxq);
+ if (err)
+ break;
+
+ err = idpf_xdp_rxq_info_init(q->rxq);
+ if (err)
+ break;
+
+ if (!splitq)
+ err = idpf_rx_bufs_init_singleq(q->rxq);
+
+ break;
+ case VIRTCHNL2_QUEUE_TYPE_RX_BUFFER:
+ bufq = q->bufq;
+
+ err = idpf_bufq_desc_alloc(vport, bufq);
+ if (err)
+ break;
+
+ for (u32 j = 0; j < bufq->q_vector->num_bufq; j++) {
+ struct idpf_buf_queue * const *bufqs;
+ enum libeth_fqe_type type;
+ u32 ts;
+
+ bufqs = bufq->q_vector->bufq;
+ if (bufqs[j] != bufq)
+ continue;
+
+ if (j) {
+ type = LIBETH_FQE_SHORT;
+ ts = bufqs[j - 1]->truesize >> 1;
+ } else {
+ type = LIBETH_FQE_MTU;
+ ts = 0;
+ }
+
+ bufq->truesize = ts;
+
+ err = idpf_rx_bufs_init(bufq, type);
+ break;
+ }
+
+ break;
+ case VIRTCHNL2_QUEUE_TYPE_TX:
+ err = idpf_tx_desc_alloc(vport, q->txq);
+ break;
+ case VIRTCHNL2_QUEUE_TYPE_TX_COMPLETION:
+ err = idpf_compl_desc_alloc(vport, q->complq);
+ break;
+ default:
+ continue;
+ }
+
+ if (err)
+ return err;
+ }
+
+ return 0;
+}
+
+static void idpf_clean_queue_set(const struct idpf_queue_set *qs)
+{
+ const struct idpf_vport *vport = qs->vport;
+ struct device *dev = vport->netdev->dev.parent;
+
+ for (u32 i = 0; i < qs->num; i++) {
+ const struct idpf_queue_ptr *q = &qs->qs[i];
+
+ switch (q->type) {
+ case VIRTCHNL2_QUEUE_TYPE_RX:
+ idpf_xdp_rxq_info_deinit(q->rxq, vport->rxq_model);
+ idpf_rx_desc_rel(q->rxq, dev, vport->rxq_model);
+ break;
+ case VIRTCHNL2_QUEUE_TYPE_RX_BUFFER:
+ idpf_rx_desc_rel_bufq(q->bufq, dev);
+ break;
+ case VIRTCHNL2_QUEUE_TYPE_TX:
+ idpf_tx_desc_rel(q->txq);
+
+ if (idpf_queue_has(XDP, q->txq)) {
+ q->txq->pending = 0;
+ q->txq->xdp_tx = 0;
+ } else {
+ q->txq->txq_grp->num_completions_pending = 0;
+ }
+
+ writel(q->txq->next_to_use, q->txq->tail);
+ break;
+ case VIRTCHNL2_QUEUE_TYPE_TX_COMPLETION:
+ idpf_compl_desc_rel(q->complq);
+ q->complq->num_completions = 0;
+ break;
+ default:
+ break;
+ }
+ }
+}
+
+static void idpf_qvec_ena_irq(struct idpf_q_vector *qv)
+{
+ if (qv->num_txq) {
+ u32 itr;
+
+ if (IDPF_ITR_IS_DYNAMIC(qv->tx_intr_mode))
+ itr = qv->vport->tx_itr_profile[qv->tx_dim.profile_ix];
+ else
+ itr = qv->tx_itr_value;
+
+ idpf_vport_intr_write_itr(qv, itr, true);
+ }
+
+ if (qv->num_rxq) {
+ u32 itr;
+
+ if (IDPF_ITR_IS_DYNAMIC(qv->rx_intr_mode))
+ itr = qv->vport->rx_itr_profile[qv->rx_dim.profile_ix];
+ else
+ itr = qv->rx_itr_value;
+
+ idpf_vport_intr_write_itr(qv, itr, false);
+ }
+
+ if (qv->num_txq || qv->num_rxq)
+ idpf_vport_intr_update_itr_ena_irq(qv);
+}
+
+/**
+ * idpf_vector_to_queue_set - create a queue set associated with the given
+ * queue vector
+ * @qv: queue vector corresponding to the queue pair
+ *
+ * Returns a pointer to a dynamically allocated array of pointers to all
+ * queues associated with a given queue vector (@qv).
+ * Please note that the caller is responsible to free the memory allocated
+ * by this function using kfree().
+ *
+ * Return: &idpf_queue_set on success, %NULL in case of error.
+ */
+static struct idpf_queue_set *
+idpf_vector_to_queue_set(struct idpf_q_vector *qv)
+{
+ bool xdp = qv->vport->xdp_txq_offset && !qv->num_xsksq;
+ struct idpf_vport *vport = qv->vport;
+ struct idpf_queue_set *qs;
+ u32 num;
+
+ num = qv->num_rxq + qv->num_bufq + qv->num_txq + qv->num_complq;
+ num += xdp ? qv->num_rxq * 2 : qv->num_xsksq * 2;
+ if (!num)
+ return NULL;
+
+ qs = idpf_alloc_queue_set(vport, num);
+ if (!qs)
+ return NULL;
+
+ num = 0;
+
+ for (u32 i = 0; i < qv->num_bufq; i++) {
+ qs->qs[num].type = VIRTCHNL2_QUEUE_TYPE_RX_BUFFER;
+ qs->qs[num++].bufq = qv->bufq[i];
+ }
+
+ for (u32 i = 0; i < qv->num_rxq; i++) {
+ qs->qs[num].type = VIRTCHNL2_QUEUE_TYPE_RX;
+ qs->qs[num++].rxq = qv->rx[i];
+ }
+
+ for (u32 i = 0; i < qv->num_txq; i++) {
+ qs->qs[num].type = VIRTCHNL2_QUEUE_TYPE_TX;
+ qs->qs[num++].txq = qv->tx[i];
+ }
+
+ for (u32 i = 0; i < qv->num_complq; i++) {
+ qs->qs[num].type = VIRTCHNL2_QUEUE_TYPE_TX_COMPLETION;
+ qs->qs[num++].complq = qv->complq[i];
+ }
+
+ if (!vport->xdp_txq_offset)
+ goto finalize;
+
+ if (xdp) {
+ for (u32 i = 0; i < qv->num_rxq; i++) {
+ u32 idx = vport->xdp_txq_offset + qv->rx[i]->idx;
+
+ qs->qs[num].type = VIRTCHNL2_QUEUE_TYPE_TX;
+ qs->qs[num++].txq = vport->txqs[idx];
+
+ qs->qs[num].type = VIRTCHNL2_QUEUE_TYPE_TX_COMPLETION;
+ qs->qs[num++].complq = vport->txqs[idx]->complq;
+ }
+ } else {
+ for (u32 i = 0; i < qv->num_xsksq; i++) {
+ qs->qs[num].type = VIRTCHNL2_QUEUE_TYPE_TX;
+ qs->qs[num++].txq = qv->xsksq[i];
+
+ qs->qs[num].type = VIRTCHNL2_QUEUE_TYPE_TX_COMPLETION;
+ qs->qs[num++].complq = qv->xsksq[i]->complq;
+ }
+ }
+
+finalize:
+ if (num != qs->num) {
+ kfree(qs);
+ return NULL;
+ }
+
+ return qs;
+}
+
+static int idpf_qp_enable(const struct idpf_queue_set *qs, u32 qid)
+{
+ struct idpf_vport *vport = qs->vport;
+ struct idpf_q_vector *q_vector;
+ int err;
+
+ q_vector = idpf_find_rxq_vec(vport, qid);
+
+ err = idpf_init_queue_set(qs);
+ if (err) {
+ netdev_err(vport->netdev, "Could not initialize queues in pair %u: %pe\n",
+ qid, ERR_PTR(err));
+ return err;
+ }
+
+ if (!vport->xdp_txq_offset)
+ goto config;
+
+ q_vector->xsksq = kcalloc(DIV_ROUND_UP(vport->num_rxq_grp,
+ vport->num_q_vectors),
+ sizeof(*q_vector->xsksq), GFP_KERNEL);
+ if (!q_vector->xsksq)
+ return -ENOMEM;
+
+ for (u32 i = 0; i < qs->num; i++) {
+ const struct idpf_queue_ptr *q = &qs->qs[i];
+
+ if (q->type != VIRTCHNL2_QUEUE_TYPE_TX)
+ continue;
+
+ if (!idpf_queue_has(XSK, q->txq))
+ continue;
+
+ idpf_xsk_init_wakeup(q_vector);
+
+ q->txq->q_vector = q_vector;
+ q_vector->xsksq[q_vector->num_xsksq++] = q->txq;
+ }
+
+config:
+ err = idpf_send_config_queue_set_msg(qs);
+ if (err) {
+ netdev_err(vport->netdev, "Could not configure queues in pair %u: %pe\n",
+ qid, ERR_PTR(err));
+ return err;
+ }
+
+ err = idpf_send_enable_queue_set_msg(qs);
+ if (err) {
+ netdev_err(vport->netdev, "Could not enable queues in pair %u: %pe\n",
+ qid, ERR_PTR(err));
+ return err;
+ }
+
+ napi_enable(&q_vector->napi);
+ idpf_qvec_ena_irq(q_vector);
+
+ netif_start_subqueue(vport->netdev, qid);
+
+ return 0;
+}
+
+static int idpf_qp_disable(const struct idpf_queue_set *qs, u32 qid)
+{
+ struct idpf_vport *vport = qs->vport;
+ struct idpf_q_vector *q_vector;
+ int err;
+
+ q_vector = idpf_find_rxq_vec(vport, qid);
+ netif_stop_subqueue(vport->netdev, qid);
+
+ writel(0, q_vector->intr_reg.dyn_ctl);
+ napi_disable(&q_vector->napi);
+
+ err = idpf_send_disable_queue_set_msg(qs);
+ if (err) {
+ netdev_err(vport->netdev, "Could not disable queues in pair %u: %pe\n",
+ qid, ERR_PTR(err));
+ return err;
+ }
+
+ idpf_clean_queue_set(qs);
+
+ kfree(q_vector->xsksq);
+ q_vector->num_xsksq = 0;
+
+ return 0;
+}
+
+/**
+ * idpf_qp_switch - enable or disable queues associated with queue pair
+ * @vport: vport to switch the pair for
+ * @qid: index of the queue pair to switch
+ * @en: whether to enable or disable the pair
+ *
+ * Return: 0 on success, -errno on failure.
+ */
+int idpf_qp_switch(struct idpf_vport *vport, u32 qid, bool en)
+{
+ struct idpf_q_vector *q_vector = idpf_find_rxq_vec(vport, qid);
+ struct idpf_queue_set *qs __free(kfree) = NULL;
+
+ if (idpf_find_txq_vec(vport, qid) != q_vector)
+ return -EINVAL;
+
+ qs = idpf_vector_to_queue_set(q_vector);
+ if (!qs)
+ return -ENOMEM;
+
+ return en ? idpf_qp_enable(qs, qid) : idpf_qp_disable(qs, qid);
+}
+
/**
* idpf_txq_group_rel - Release all resources for txq groups
* @vport: vport to release txq groups on
@@ -995,6 +1309,11 @@ static void idpf_txq_group_rel(struct idpf_vport *vport)
struct idpf_txq_group *txq_grp = &vport->txq_grps[i];
for (j = 0; j < txq_grp->num_txq; j++) {
+ if (flow_sch_en) {
+ kfree(txq_grp->txqs[j]->refillq);
+ txq_grp->txqs[j]->refillq = NULL;
+ }
+
kfree(txq_grp->txqs[j]);
txq_grp->txqs[j] = NULL;
}
@@ -1004,9 +1323,6 @@ static void idpf_txq_group_rel(struct idpf_vport *vport)
kfree(txq_grp->complq);
txq_grp->complq = NULL;
-
- if (flow_sch_en)
- kfree(txq_grp->stashes);
}
kfree(vport->txq_grps);
vport->txq_grps = NULL;
@@ -1088,8 +1404,12 @@ static void idpf_vport_queue_grp_rel_all(struct idpf_vport *vport)
*/
void idpf_vport_queues_rel(struct idpf_vport *vport)
{
+ idpf_xdp_copy_prog_to_rqs(vport, NULL);
+
idpf_tx_desc_rel_all(vport);
idpf_rx_desc_rel_all(vport);
+
+ idpf_xdpsqs_put(vport);
idpf_vport_queue_grp_rel_all(vport);
kfree(vport->txqs);
@@ -1108,6 +1428,8 @@ void idpf_vport_queues_rel(struct idpf_vport *vport)
*/
static int idpf_vport_init_fast_path_txqs(struct idpf_vport *vport)
{
+ struct idpf_ptp_vport_tx_tstamp_caps *caps = vport->tx_tstamp_caps;
+ struct work_struct *tstamp_task = &vport->tstamp_task;
int i, j, k = 0;
vport->txqs = kcalloc(vport->num_txq, sizeof(*vport->txqs),
@@ -1122,6 +1444,12 @@ static int idpf_vport_init_fast_path_txqs(struct idpf_vport *vport)
for (j = 0; j < tx_grp->num_txq; j++, k++) {
vport->txqs[k] = tx_grp->txqs[j];
vport->txqs[k]->idx = k;
+
+ if (!caps)
+ continue;
+
+ vport->txqs[k]->cached_tstamp_caps = caps;
+ vport->txqs[k]->tstamp_task = tstamp_task;
}
}
@@ -1155,6 +1483,18 @@ void idpf_vport_init_num_qs(struct idpf_vport *vport,
if (idpf_is_queue_model_split(vport->rxq_model))
vport->num_bufq = le16_to_cpu(vport_msg->num_rx_bufq);
+ vport->xdp_prog = config_data->xdp_prog;
+ if (idpf_xdp_enabled(vport)) {
+ vport->xdp_txq_offset = config_data->num_req_tx_qs;
+ vport->num_xdp_txq = le16_to_cpu(vport_msg->num_tx_q) -
+ vport->xdp_txq_offset;
+ vport->xdpsq_share = libeth_xdpsq_shared(vport->num_xdp_txq);
+ } else {
+ vport->xdp_txq_offset = 0;
+ vport->num_xdp_txq = 0;
+ vport->xdpsq_share = false;
+ }
+
/* Adjust number of buffer queues per Rx queue group. */
if (!idpf_is_queue_model_split(vport->rxq_model)) {
vport->num_bufqs_per_qgrp = 0;
@@ -1226,22 +1566,17 @@ int idpf_vport_calc_total_qs(struct idpf_adapter *adapter, u16 vport_idx,
int dflt_splitq_txq_grps = 0, dflt_singleq_txqs = 0;
int dflt_splitq_rxq_grps = 0, dflt_singleq_rxqs = 0;
u16 num_req_tx_qs = 0, num_req_rx_qs = 0;
+ struct idpf_vport_user_config_data *user;
struct idpf_vport_config *vport_config;
u16 num_txq_grps, num_rxq_grps;
- u32 num_qs;
+ u32 num_qs, num_xdpsq;
vport_config = adapter->vport_config[vport_idx];
if (vport_config) {
num_req_tx_qs = vport_config->user_config.num_req_tx_qs;
num_req_rx_qs = vport_config->user_config.num_req_rx_qs;
} else {
- int num_cpus;
-
- /* Restrict num of queues to cpus online as a default
- * configuration to give best performance. User can always
- * override to a max number of queues via ethtool.
- */
- num_cpus = num_online_cpus();
+ u32 num_cpus = netif_get_num_default_rss_queues();
dflt_splitq_txq_grps = min_t(int, max_q->max_txq, num_cpus);
dflt_singleq_txqs = min_t(int, max_q->max_txq, num_cpus);
@@ -1276,6 +1611,24 @@ int idpf_vport_calc_total_qs(struct idpf_adapter *adapter, u16 vport_idx,
vport_msg->num_rx_bufq = 0;
}
+ if (!vport_config)
+ return 0;
+
+ user = &vport_config->user_config;
+ user->num_req_rx_qs = le16_to_cpu(vport_msg->num_rx_q);
+ user->num_req_tx_qs = le16_to_cpu(vport_msg->num_tx_q);
+
+ if (vport_config->user_config.xdp_prog)
+ num_xdpsq = libeth_xdpsq_num(user->num_req_rx_qs,
+ user->num_req_tx_qs,
+ vport_config->max_q.max_txq);
+ else
+ num_xdpsq = 0;
+
+ vport_msg->num_tx_q = cpu_to_le16(user->num_req_tx_qs + num_xdpsq);
+ if (idpf_is_queue_model_split(le16_to_cpu(vport_msg->txq_model)))
+ vport_msg->num_tx_complq = vport_msg->num_tx_q;
+
return 0;
}
@@ -1325,14 +1678,13 @@ static void idpf_vport_calc_numq_per_grp(struct idpf_vport *vport,
static void idpf_rxq_set_descids(const struct idpf_vport *vport,
struct idpf_rx_queue *q)
{
- if (idpf_is_queue_model_split(vport->rxq_model)) {
- q->rxdids = VIRTCHNL2_RXDID_2_FLEX_SPLITQ_M;
- } else {
- if (vport->base_rxd)
- q->rxdids = VIRTCHNL2_RXDID_1_32B_BASE_M;
- else
- q->rxdids = VIRTCHNL2_RXDID_2_FLEX_SQ_NIC_M;
- }
+ if (idpf_is_queue_model_split(vport->rxq_model))
+ return;
+
+ if (vport->base_rxd)
+ q->rxdids = VIRTCHNL2_RXDID_1_32B_BASE_M;
+ else
+ q->rxdids = VIRTCHNL2_RXDID_2_FLEX_SQ_NIC_M;
}
/**
@@ -1359,7 +1711,6 @@ static int idpf_txq_group_alloc(struct idpf_vport *vport, u16 num_txq)
for (i = 0; i < vport->num_txq_grp; i++) {
struct idpf_txq_group *tx_qgrp = &vport->txq_grps[i];
struct idpf_adapter *adapter = vport->adapter;
- struct idpf_txq_stash *stashes;
int j;
tx_qgrp->vport = vport;
@@ -1372,15 +1723,6 @@ static int idpf_txq_group_alloc(struct idpf_vport *vport, u16 num_txq)
goto err_alloc;
}
- if (split && flow_sch_en) {
- stashes = kcalloc(num_txq, sizeof(*stashes),
- GFP_KERNEL);
- if (!stashes)
- goto err_alloc;
-
- tx_qgrp->stashes = stashes;
- }
-
for (j = 0; j < tx_qgrp->num_txq; j++) {
struct idpf_tx_queue *q = tx_qgrp->txqs[j];
@@ -1390,6 +1732,7 @@ static int idpf_txq_group_alloc(struct idpf_vport *vport, u16 num_txq)
q->tx_min_pkt_len = idpf_get_min_tx_pkt_len(adapter);
q->netdev = vport->netdev;
q->txq_grp = tx_qgrp;
+ q->rel_q_id = j;
if (!split) {
q->clean_budget = vport->compln_clean_budget;
@@ -1400,12 +1743,14 @@ static int idpf_txq_group_alloc(struct idpf_vport *vport, u16 num_txq)
if (!flow_sch_en)
continue;
- if (split) {
- q->stash = &stashes[j];
- hash_init(q->stash->sched_buf_hash);
- }
-
idpf_queue_set(FLOW_SCH_EN, q);
+
+ q->refillq = kzalloc(sizeof(*q->refillq), GFP_KERNEL);
+ if (!q->refillq)
+ goto err_alloc;
+
+ idpf_queue_set(GEN_CHK, q->refillq);
+ idpf_queue_set(RFL_GEN_CHK, q->refillq);
}
if (!split)
@@ -1548,7 +1893,6 @@ skip_splitq_rx_init:
setup_rxq:
q->desc_count = vport->rxq_desc_count;
q->rx_ptype_lkup = vport->rx_ptype_lkup;
- q->netdev = vport->netdev;
q->bufq_sets = rx_qgrp->splitq.bufq_sets;
q->idx = (i * num_rxq) + j;
q->rx_buffer_low_watermark = IDPF_LOW_WATERMARK;
@@ -1609,15 +1953,19 @@ int idpf_vport_queues_alloc(struct idpf_vport *vport)
if (err)
goto err_out;
- err = idpf_tx_desc_alloc_all(vport);
+ err = idpf_vport_init_fast_path_txqs(vport);
if (err)
goto err_out;
- err = idpf_rx_desc_alloc_all(vport);
+ err = idpf_xdpsqs_get(vport);
if (err)
goto err_out;
- err = idpf_vport_init_fast_path_txqs(vport);
+ err = idpf_tx_desc_alloc_all(vport);
+ if (err)
+ goto err_out;
+
+ err = idpf_rx_desc_alloc_all(vport);
if (err)
goto err_out;
@@ -1630,150 +1978,43 @@ err_out:
}
/**
- * idpf_tx_handle_sw_marker - Handle queue marker packet
- * @tx_q: tx queue to handle software marker
- */
-static void idpf_tx_handle_sw_marker(struct idpf_tx_queue *tx_q)
-{
- struct idpf_netdev_priv *priv = netdev_priv(tx_q->netdev);
- struct idpf_vport *vport = priv->vport;
- int i;
-
- idpf_queue_clear(SW_MARKER, tx_q);
- /* Hardware must write marker packets to all queues associated with
- * completion queues. So check if all queues received marker packets
- */
- for (i = 0; i < vport->num_txq; i++)
- /* If we're still waiting on any other TXQ marker completions,
- * just return now since we cannot wake up the marker_wq yet.
- */
- if (idpf_queue_has(SW_MARKER, vport->txqs[i]))
- return;
-
- /* Drain complete */
- set_bit(IDPF_VPORT_SW_MARKER, vport->flags);
- wake_up(&vport->sw_marker_wq);
-}
-
-/**
- * idpf_tx_splitq_clean_hdr - Clean TX buffer resources for header portion of
- * packet
- * @tx_q: tx queue to clean buffer from
- * @tx_buf: buffer to be cleaned
- * @cleaned: pointer to stats struct to track cleaned packets/bytes
- * @napi_budget: Used to determine if we are in netpoll
+ * idpf_tx_read_tstamp - schedule a work to read Tx timestamp value
+ * @txq: queue to read the timestamp from
+ * @skb: socket buffer to provide Tx timestamp value
+ *
+ * Schedule a work to read Tx timestamp value generated once the packet is
+ * transmitted.
*/
-static void idpf_tx_splitq_clean_hdr(struct idpf_tx_queue *tx_q,
- struct idpf_tx_buf *tx_buf,
- struct idpf_cleaned_stats *cleaned,
- int napi_budget)
+static void idpf_tx_read_tstamp(struct idpf_tx_queue *txq, struct sk_buff *skb)
{
- napi_consume_skb(tx_buf->skb, napi_budget);
-
- if (dma_unmap_len(tx_buf, len)) {
- dma_unmap_single(tx_q->dev,
- dma_unmap_addr(tx_buf, dma),
- dma_unmap_len(tx_buf, len),
- DMA_TO_DEVICE);
-
- dma_unmap_len_set(tx_buf, len, 0);
- }
-
- /* clear tx_buf data */
- tx_buf->skb = NULL;
+ struct idpf_ptp_vport_tx_tstamp_caps *tx_tstamp_caps;
+ struct idpf_ptp_tx_tstamp_status *tx_tstamp_status;
- cleaned->bytes += tx_buf->bytecount;
- cleaned->packets += tx_buf->gso_segs;
-}
-
-/**
- * idpf_tx_clean_stashed_bufs - clean bufs that were stored for
- * out of order completions
- * @txq: queue to clean
- * @compl_tag: completion tag of packet to clean (from completion descriptor)
- * @cleaned: pointer to stats struct to track cleaned packets/bytes
- * @budget: Used to determine if we are in netpoll
- */
-static void idpf_tx_clean_stashed_bufs(struct idpf_tx_queue *txq,
- u16 compl_tag,
- struct idpf_cleaned_stats *cleaned,
- int budget)
-{
- struct idpf_tx_stash *stash;
- struct hlist_node *tmp_buf;
+ tx_tstamp_caps = txq->cached_tstamp_caps;
+ spin_lock_bh(&tx_tstamp_caps->status_lock);
- /* Buffer completion */
- hash_for_each_possible_safe(txq->stash->sched_buf_hash, stash, tmp_buf,
- hlist, compl_tag) {
- if (unlikely(stash->buf.compl_tag != (int)compl_tag))
+ for (u32 i = 0; i < tx_tstamp_caps->num_entries; i++) {
+ tx_tstamp_status = &tx_tstamp_caps->tx_tstamp_status[i];
+ if (tx_tstamp_status->state != IDPF_PTP_FREE)
continue;
- if (stash->buf.skb) {
- idpf_tx_splitq_clean_hdr(txq, &stash->buf, cleaned,
- budget);
- } else if (dma_unmap_len(&stash->buf, len)) {
- dma_unmap_page(txq->dev,
- dma_unmap_addr(&stash->buf, dma),
- dma_unmap_len(&stash->buf, len),
- DMA_TO_DEVICE);
- dma_unmap_len_set(&stash->buf, len, 0);
- }
-
- /* Push shadow buf back onto stack */
- idpf_buf_lifo_push(&txq->stash->buf_stack, stash);
+ tx_tstamp_status->skb = skb;
+ tx_tstamp_status->state = IDPF_PTP_REQUEST;
- hash_del(&stash->hlist);
- }
-}
-
-/**
- * idpf_stash_flow_sch_buffers - store buffer parameters info to be freed at a
- * later time (only relevant for flow scheduling mode)
- * @txq: Tx queue to clean
- * @tx_buf: buffer to store
- */
-static int idpf_stash_flow_sch_buffers(struct idpf_tx_queue *txq,
- struct idpf_tx_buf *tx_buf)
-{
- struct idpf_tx_stash *stash;
-
- if (unlikely(!dma_unmap_addr(tx_buf, dma) &&
- !dma_unmap_len(tx_buf, len)))
- return 0;
-
- stash = idpf_buf_lifo_pop(&txq->stash->buf_stack);
- if (unlikely(!stash)) {
- net_err_ratelimited("%s: No out-of-order TX buffers left!\n",
- netdev_name(txq->netdev));
-
- return -ENOMEM;
+ /* Fetch timestamp from completion descriptor through
+ * virtchnl msg to report to stack.
+ */
+ queue_work(system_unbound_wq, txq->tstamp_task);
+ break;
}
- /* Store buffer params in shadow buffer */
- stash->buf.skb = tx_buf->skb;
- stash->buf.bytecount = tx_buf->bytecount;
- stash->buf.gso_segs = tx_buf->gso_segs;
- dma_unmap_addr_set(&stash->buf, dma, dma_unmap_addr(tx_buf, dma));
- dma_unmap_len_set(&stash->buf, len, dma_unmap_len(tx_buf, len));
- stash->buf.compl_tag = tx_buf->compl_tag;
-
- /* Add buffer to buf_hash table to be freed later */
- hash_add(txq->stash->sched_buf_hash, &stash->hlist,
- stash->buf.compl_tag);
-
- memset(tx_buf, 0, sizeof(struct idpf_tx_buf));
-
- /* Reinitialize buf_id portion of tag */
- tx_buf->compl_tag = IDPF_SPLITQ_TX_INVAL_COMPL_TAG;
-
- return 0;
+ spin_unlock_bh(&tx_tstamp_caps->status_lock);
}
#define idpf_tx_splitq_clean_bump_ntc(txq, ntc, desc, buf) \
do { \
- (ntc)++; \
- if (unlikely(!(ntc))) { \
- ntc -= (txq)->desc_count; \
+ if (unlikely(++(ntc) == (txq)->desc_count)) { \
+ ntc = 0; \
buf = (txq)->tx_buf; \
desc = &(txq)->flex_tx[0]; \
} else { \
@@ -1800,156 +2041,98 @@ do { \
*/
static void idpf_tx_splitq_clean(struct idpf_tx_queue *tx_q, u16 end,
int napi_budget,
- struct idpf_cleaned_stats *cleaned,
+ struct libeth_sq_napi_stats *cleaned,
bool descs_only)
{
union idpf_tx_flex_desc *next_pending_desc = NULL;
union idpf_tx_flex_desc *tx_desc;
- s16 ntc = tx_q->next_to_clean;
+ u32 ntc = tx_q->next_to_clean;
+ struct libeth_cq_pp cp = {
+ .dev = tx_q->dev,
+ .ss = cleaned,
+ .napi = napi_budget,
+ };
struct idpf_tx_buf *tx_buf;
+ if (descs_only) {
+ /* Bump ring index to mark as cleaned. */
+ tx_q->next_to_clean = end;
+ return;
+ }
+
tx_desc = &tx_q->flex_tx[ntc];
next_pending_desc = &tx_q->flex_tx[end];
tx_buf = &tx_q->tx_buf[ntc];
- ntc -= tx_q->desc_count;
while (tx_desc != next_pending_desc) {
- union idpf_tx_flex_desc *eop_desc;
+ u32 eop_idx;
/* If this entry in the ring was used as a context descriptor,
- * it's corresponding entry in the buffer ring will have an
- * invalid completion tag since no buffer was used. We can
- * skip this descriptor since there is no buffer to clean.
+ * it's corresponding entry in the buffer ring is reserved. We
+ * can skip this descriptor since there is no buffer to clean.
*/
- if (unlikely(tx_buf->compl_tag == IDPF_SPLITQ_TX_INVAL_COMPL_TAG))
+ if (tx_buf->type <= LIBETH_SQE_CTX)
goto fetch_next_txq_desc;
- eop_desc = (union idpf_tx_flex_desc *)tx_buf->next_to_watch;
-
- /* clear next_to_watch to prevent false hangs */
- tx_buf->next_to_watch = NULL;
+ if (unlikely(tx_buf->type != LIBETH_SQE_SKB))
+ break;
- if (descs_only) {
- if (idpf_stash_flow_sch_buffers(tx_q, tx_buf))
- goto tx_splitq_clean_out;
+ eop_idx = tx_buf->rs_idx;
+ libeth_tx_complete(tx_buf, &cp);
- while (tx_desc != eop_desc) {
- idpf_tx_splitq_clean_bump_ntc(tx_q, ntc,
- tx_desc, tx_buf);
+ /* unmap remaining buffers */
+ while (ntc != eop_idx) {
+ idpf_tx_splitq_clean_bump_ntc(tx_q, ntc,
+ tx_desc, tx_buf);
- if (dma_unmap_len(tx_buf, len)) {
- if (idpf_stash_flow_sch_buffers(tx_q,
- tx_buf))
- goto tx_splitq_clean_out;
- }
- }
- } else {
- idpf_tx_splitq_clean_hdr(tx_q, tx_buf, cleaned,
- napi_budget);
-
- /* unmap remaining buffers */
- while (tx_desc != eop_desc) {
- idpf_tx_splitq_clean_bump_ntc(tx_q, ntc,
- tx_desc, tx_buf);
-
- /* unmap any remaining paged data */
- if (dma_unmap_len(tx_buf, len)) {
- dma_unmap_page(tx_q->dev,
- dma_unmap_addr(tx_buf, dma),
- dma_unmap_len(tx_buf, len),
- DMA_TO_DEVICE);
- dma_unmap_len_set(tx_buf, len, 0);
- }
- }
+ /* unmap any remaining paged data */
+ libeth_tx_complete(tx_buf, &cp);
}
fetch_next_txq_desc:
idpf_tx_splitq_clean_bump_ntc(tx_q, ntc, tx_desc, tx_buf);
}
-tx_splitq_clean_out:
- ntc += tx_q->desc_count;
tx_q->next_to_clean = ntc;
}
-#define idpf_tx_clean_buf_ring_bump_ntc(txq, ntc, buf) \
-do { \
- (buf)++; \
- (ntc)++; \
- if (unlikely((ntc) == (txq)->desc_count)) { \
- buf = (txq)->tx_buf; \
- ntc = 0; \
- } \
-} while (0)
-
/**
- * idpf_tx_clean_buf_ring - clean flow scheduling TX queue buffers
+ * idpf_tx_clean_bufs - clean flow scheduling TX queue buffers
* @txq: queue to clean
- * @compl_tag: completion tag of packet to clean (from completion descriptor)
+ * @buf_id: packet's starting buffer ID, from completion descriptor
* @cleaned: pointer to stats struct to track cleaned packets/bytes
* @budget: Used to determine if we are in netpoll
*
- * Cleans all buffers associated with the input completion tag either from the
- * TX buffer ring or from the hash table if the buffers were previously
- * stashed. Returns the byte/segment count for the cleaned packet associated
- * this completion tag.
+ * Clean all buffers associated with the packet starting at buf_id. Returns the
+ * byte/segment count for the cleaned packet.
*/
-static bool idpf_tx_clean_buf_ring(struct idpf_tx_queue *txq, u16 compl_tag,
- struct idpf_cleaned_stats *cleaned,
- int budget)
+static void idpf_tx_clean_bufs(struct idpf_tx_queue *txq, u32 buf_id,
+ struct libeth_sq_napi_stats *cleaned,
+ int budget)
{
- u16 idx = compl_tag & txq->compl_tag_bufid_m;
struct idpf_tx_buf *tx_buf = NULL;
- u16 ntc = txq->next_to_clean;
- u16 num_descs_cleaned = 0;
- u16 orig_idx = idx;
-
- tx_buf = &txq->tx_buf[idx];
-
- while (tx_buf->compl_tag == (int)compl_tag) {
- if (tx_buf->skb) {
- idpf_tx_splitq_clean_hdr(txq, tx_buf, cleaned, budget);
- } else if (dma_unmap_len(tx_buf, len)) {
- dma_unmap_page(txq->dev,
- dma_unmap_addr(tx_buf, dma),
- dma_unmap_len(tx_buf, len),
- DMA_TO_DEVICE);
- dma_unmap_len_set(tx_buf, len, 0);
- }
+ struct libeth_cq_pp cp = {
+ .dev = txq->dev,
+ .ss = cleaned,
+ .napi = budget,
+ };
- memset(tx_buf, 0, sizeof(struct idpf_tx_buf));
- tx_buf->compl_tag = IDPF_SPLITQ_TX_INVAL_COMPL_TAG;
+ tx_buf = &txq->tx_buf[buf_id];
+ if (tx_buf->type == LIBETH_SQE_SKB) {
+ if (skb_shinfo(tx_buf->skb)->tx_flags & SKBTX_IN_PROGRESS)
+ idpf_tx_read_tstamp(txq, tx_buf->skb);
- num_descs_cleaned++;
- idpf_tx_clean_buf_ring_bump_ntc(txq, idx, tx_buf);
+ libeth_tx_complete(tx_buf, &cp);
+ idpf_post_buf_refill(txq->refillq, buf_id);
}
- /* If we didn't clean anything on the ring for this completion, there's
- * nothing more to do.
- */
- if (unlikely(!num_descs_cleaned))
- return false;
+ while (idpf_tx_buf_next(tx_buf) != IDPF_TXBUF_NULL) {
+ buf_id = idpf_tx_buf_next(tx_buf);
- /* Otherwise, if we did clean a packet on the ring directly, it's safe
- * to assume that the descriptors starting from the original
- * next_to_clean up until the previously cleaned packet can be reused.
- * Therefore, we will go back in the ring and stash any buffers still
- * in the ring into the hash table to be cleaned later.
- */
- tx_buf = &txq->tx_buf[ntc];
- while (tx_buf != &txq->tx_buf[orig_idx]) {
- idpf_stash_flow_sch_buffers(txq, tx_buf);
- idpf_tx_clean_buf_ring_bump_ntc(txq, ntc, tx_buf);
+ tx_buf = &txq->tx_buf[buf_id];
+ libeth_tx_complete(tx_buf, &cp);
+ idpf_post_buf_refill(txq->refillq, buf_id);
}
-
- /* Finally, update next_to_clean to reflect the work that was just done
- * on the ring, if any. If the packet was only cleaned from the hash
- * table, the ring will not be impacted, therefore we should not touch
- * next_to_clean. The updated idx is used here
- */
- txq->next_to_clean = idx;
-
- return true;
}
/**
@@ -1965,24 +2148,20 @@ static bool idpf_tx_clean_buf_ring(struct idpf_tx_queue *txq, u16 compl_tag,
*/
static void idpf_tx_handle_rs_completion(struct idpf_tx_queue *txq,
struct idpf_splitq_tx_compl_desc *desc,
- struct idpf_cleaned_stats *cleaned,
+ struct libeth_sq_napi_stats *cleaned,
int budget)
{
- u16 compl_tag;
+ /* RS completion contains queue head for queue based scheduling or
+ * completion tag for flow based scheduling.
+ */
+ u16 rs_compl_val = le16_to_cpu(desc->common.q_head_compl_tag.q_head);
if (!idpf_queue_has(FLOW_SCH_EN, txq)) {
- u16 head = le16_to_cpu(desc->q_head_compl_tag.q_head);
-
- return idpf_tx_splitq_clean(txq, head, budget, cleaned, false);
+ idpf_tx_splitq_clean(txq, rs_compl_val, budget, cleaned, false);
+ return;
}
- compl_tag = le16_to_cpu(desc->q_head_compl_tag.compl_tag);
-
- /* If we didn't clean anything on the ring, this packet must be
- * in the hash table. Go clean it there.
- */
- if (!idpf_tx_clean_buf_ring(txq, compl_tag, cleaned, budget))
- idpf_tx_clean_stashed_bufs(txq, compl_tag, cleaned, budget);
+ idpf_tx_clean_bufs(txq, rs_compl_val, cleaned, budget);
}
/**
@@ -2008,21 +2187,21 @@ static bool idpf_tx_clean_complq(struct idpf_compl_queue *complq, int budget,
ntc -= complq->desc_count;
do {
- struct idpf_cleaned_stats cleaned_stats = { };
+ struct libeth_sq_napi_stats cleaned_stats = { };
struct idpf_tx_queue *tx_q;
+ __le16 hw_head;
int rel_tx_qid;
- u16 hw_head;
u8 ctype; /* completion type */
u16 gen;
/* if the descriptor isn't done, no work yet to do */
- gen = le16_get_bits(tx_desc->qid_comptype_gen,
+ gen = le16_get_bits(tx_desc->common.qid_comptype_gen,
IDPF_TXD_COMPLQ_GEN_M);
if (idpf_queue_has(GEN_CHK, complq) != gen)
break;
/* Find necessary info of TX queue to clean buffers */
- rel_tx_qid = le16_get_bits(tx_desc->qid_comptype_gen,
+ rel_tx_qid = le16_get_bits(tx_desc->common.qid_comptype_gen,
IDPF_TXD_COMPLQ_QID_M);
if (rel_tx_qid >= complq->txq_grp->num_txq ||
!complq->txq_grp->txqs[rel_tx_qid]) {
@@ -2032,22 +2211,19 @@ static bool idpf_tx_clean_complq(struct idpf_compl_queue *complq, int budget,
tx_q = complq->txq_grp->txqs[rel_tx_qid];
/* Determine completion type */
- ctype = le16_get_bits(tx_desc->qid_comptype_gen,
+ ctype = le16_get_bits(tx_desc->common.qid_comptype_gen,
IDPF_TXD_COMPLQ_COMPL_TYPE_M);
switch (ctype) {
case IDPF_TXD_COMPLT_RE:
- hw_head = le16_to_cpu(tx_desc->q_head_compl_tag.q_head);
+ hw_head = tx_desc->common.q_head_compl_tag.q_head;
- idpf_tx_splitq_clean(tx_q, hw_head, budget,
- &cleaned_stats, true);
+ idpf_tx_splitq_clean(tx_q, le16_to_cpu(hw_head),
+ budget, &cleaned_stats, true);
break;
case IDPF_TXD_COMPLT_RS:
idpf_tx_handle_rs_completion(tx_q, tx_desc,
&cleaned_stats, budget);
break;
- case IDPF_TXD_COMPLT_SW_MARKER:
- idpf_tx_handle_sw_marker(tx_q);
- break;
default:
netdev_err(tx_q->netdev,
"Unknown TX completion type: %d\n", ctype);
@@ -2099,8 +2275,7 @@ fetch_next_desc:
/* Update BQL */
nq = netdev_get_tx_queue(tx_q->netdev, tx_q->idx);
- dont_wake = !complq_ok || IDPF_TX_BUF_RSV_LOW(tx_q) ||
- np->state != __IDPF_VPORT_UP ||
+ dont_wake = !complq_ok || !test_bit(IDPF_VPORT_UP, np->state) ||
!netif_carrier_ok(tx_q->netdev);
/* Check if the TXQ needs to and can be restarted */
__netif_txq_completed_wake(nq, tx_q->cleaned_pkts, tx_q->cleaned_bytes,
@@ -2121,6 +2296,69 @@ fetch_next_desc:
}
/**
+ * idpf_wait_for_sw_marker_completion - wait for SW marker of disabled Tx queue
+ * @txq: disabled Tx queue
+ *
+ * When Tx queue is requested for disabling, the CP sends a special completion
+ * descriptor called "SW marker", meaning the queue is ready to be destroyed.
+ * If, for some reason, the marker is not received within 500 ms, break the
+ * polling to not hang the driver.
+ */
+void idpf_wait_for_sw_marker_completion(const struct idpf_tx_queue *txq)
+{
+ struct idpf_compl_queue *complq;
+ unsigned long timeout;
+ bool flow, gen_flag;
+ u32 ntc;
+
+ if (!idpf_queue_has(SW_MARKER, txq))
+ return;
+
+ complq = idpf_queue_has(XDP, txq) ? txq->complq : txq->txq_grp->complq;
+ ntc = complq->next_to_clean;
+
+ flow = idpf_queue_has(FLOW_SCH_EN, complq);
+ gen_flag = idpf_queue_has(GEN_CHK, complq);
+
+ timeout = jiffies + msecs_to_jiffies(IDPF_WAIT_FOR_MARKER_TIMEO);
+
+ do {
+ struct idpf_splitq_4b_tx_compl_desc *tx_desc;
+ struct idpf_tx_queue *target;
+ u32 ctype_gen, id;
+
+ tx_desc = flow ? &complq->comp[ntc].common :
+ &complq->comp_4b[ntc];
+ ctype_gen = le16_to_cpu(tx_desc->qid_comptype_gen);
+
+ if (!!(ctype_gen & IDPF_TXD_COMPLQ_GEN_M) != gen_flag) {
+ usleep_range(500, 1000);
+ continue;
+ }
+
+ if (FIELD_GET(IDPF_TXD_COMPLQ_COMPL_TYPE_M, ctype_gen) !=
+ IDPF_TXD_COMPLT_SW_MARKER)
+ goto next;
+
+ id = FIELD_GET(IDPF_TXD_COMPLQ_QID_M, ctype_gen);
+ target = complq->txq_grp->txqs[id];
+
+ idpf_queue_clear(SW_MARKER, target);
+ if (target == txq)
+ break;
+
+next:
+ if (unlikely(++ntc == complq->desc_count)) {
+ ntc = 0;
+ gen_flag = !gen_flag;
+ }
+ } while (time_before(jiffies, timeout));
+
+ idpf_queue_assign(GEN_CHK, complq, gen_flag);
+ complq->next_to_clean = ntc;
+}
+
+/**
* idpf_tx_splitq_build_ctb - populate command tag and size for queue
* based scheduling descriptors
* @desc: descriptor to populate
@@ -2152,68 +2390,55 @@ void idpf_tx_splitq_build_flow_desc(union idpf_tx_flex_desc *desc,
struct idpf_tx_splitq_params *params,
u16 td_cmd, u16 size)
{
- desc->flow.qw1.cmd_dtype = (u16)params->dtype | td_cmd;
+ *(u32 *)&desc->flow.qw1.cmd_dtype = (u8)(params->dtype | td_cmd);
desc->flow.qw1.rxr_bufsize = cpu_to_le16((u16)size);
desc->flow.qw1.compl_tag = cpu_to_le16(params->compl_tag);
}
/**
- * idpf_tx_maybe_stop_common - 1st level check for common Tx stop conditions
+ * idpf_tx_splitq_has_room - check if enough Tx splitq resources are available
* @tx_q: the queue to be checked
- * @size: number of descriptors we want to assure is available
+ * @descs_needed: number of descriptors required for this packet
+ * @bufs_needed: number of Tx buffers required for this packet
*
- * Returns 0 if stop is not needed
+ * Return: 0 if no room available, 1 otherwise
*/
-int idpf_tx_maybe_stop_common(struct idpf_tx_queue *tx_q, unsigned int size)
+static int idpf_txq_has_room(struct idpf_tx_queue *tx_q, u32 descs_needed,
+ u32 bufs_needed)
{
- struct netdev_queue *nq;
-
- if (likely(IDPF_DESC_UNUSED(tx_q) >= size))
+ if (IDPF_DESC_UNUSED(tx_q) < descs_needed ||
+ IDPF_TX_COMPLQ_PENDING(tx_q->txq_grp) >
+ IDPF_TX_COMPLQ_OVERFLOW_THRESH(tx_q->txq_grp->complq) ||
+ idpf_tx_splitq_get_free_bufs(tx_q->refillq) < bufs_needed)
return 0;
-
- u64_stats_update_begin(&tx_q->stats_sync);
- u64_stats_inc(&tx_q->q_stats.q_busy);
- u64_stats_update_end(&tx_q->stats_sync);
-
- nq = netdev_get_tx_queue(tx_q->netdev, tx_q->idx);
-
- return netif_txq_maybe_stop(nq, IDPF_DESC_UNUSED(tx_q), size, size);
+ return 1;
}
/**
* idpf_tx_maybe_stop_splitq - 1st level check for Tx splitq stop conditions
* @tx_q: the queue to be checked
* @descs_needed: number of descriptors required for this packet
+ * @bufs_needed: number of buffers needed for this packet
*
- * Returns 0 if stop is not needed
+ * Return: 0 if stop is not needed
*/
static int idpf_tx_maybe_stop_splitq(struct idpf_tx_queue *tx_q,
- unsigned int descs_needed)
+ u32 descs_needed,
+ u32 bufs_needed)
{
- if (idpf_tx_maybe_stop_common(tx_q, descs_needed))
- goto splitq_stop;
-
- /* If there are too many outstanding completions expected on the
- * completion queue, stop the TX queue to give the device some time to
- * catch up
+ /* Since we have multiple resources to check for splitq, our
+ * start,stop_thrs becomes a boolean check instead of a count
+ * threshold.
*/
- if (unlikely(IDPF_TX_COMPLQ_PENDING(tx_q->txq_grp) >
- IDPF_TX_COMPLQ_OVERFLOW_THRESH(tx_q->txq_grp->complq)))
- goto splitq_stop;
-
- /* Also check for available book keeping buffers; if we are low, stop
- * the queue to wait for more completions
- */
- if (unlikely(IDPF_TX_BUF_RSV_LOW(tx_q)))
- goto splitq_stop;
-
- return 0;
+ if (netif_subqueue_maybe_stop(tx_q->netdev, tx_q->idx,
+ idpf_txq_has_room(tx_q, descs_needed,
+ bufs_needed),
+ 1, 1))
+ return 0;
-splitq_stop:
u64_stats_update_begin(&tx_q->stats_sync);
u64_stats_inc(&tx_q->q_stats.q_busy);
u64_stats_update_end(&tx_q->stats_sync);
- netif_stop_subqueue(tx_q->netdev, tx_q->idx);
return -EBUSY;
}
@@ -2236,8 +2461,6 @@ void idpf_tx_buf_hw_update(struct idpf_tx_queue *tx_q, u32 val,
nq = netdev_get_tx_queue(tx_q->netdev, tx_q->idx);
tx_q->next_to_use = val;
- idpf_tx_maybe_stop_common(tx_q, IDPF_TX_DESC_NEEDED);
-
/* Force memory writes to complete before letting h/w
* know there are new descriptors to fetch. (Only
* applicable for weak-ordered memory model archs,
@@ -2251,14 +2474,16 @@ void idpf_tx_buf_hw_update(struct idpf_tx_queue *tx_q, u32 val,
}
/**
- * idpf_tx_desc_count_required - calculate number of Tx descriptors needed
+ * idpf_tx_res_count_required - get number of Tx resources needed for this pkt
* @txq: queue to send buffer on
* @skb: send buffer
+ * @bufs_needed: (output) number of buffers needed for this skb.
*
- * Returns number of data descriptors needed for this skb.
+ * Return: number of data descriptors and buffers needed for this skb.
*/
-unsigned int idpf_tx_desc_count_required(struct idpf_tx_queue *txq,
- struct sk_buff *skb)
+unsigned int idpf_tx_res_count_required(struct idpf_tx_queue *txq,
+ struct sk_buff *skb,
+ u32 *bufs_needed)
{
const struct skb_shared_info *shinfo;
unsigned int count = 0, i;
@@ -2269,6 +2494,7 @@ unsigned int idpf_tx_desc_count_required(struct idpf_tx_queue *txq,
return count;
shinfo = skb_shinfo(skb);
+ *bufs_needed += shinfo->nr_frags;
for (i = 0; i < shinfo->nr_frags; i++) {
unsigned int size;
@@ -2298,65 +2524,89 @@ unsigned int idpf_tx_desc_count_required(struct idpf_tx_queue *txq,
}
/**
- * idpf_tx_dma_map_error - handle TX DMA map errors
- * @txq: queue to send buffer on
- * @skb: send buffer
- * @first: original first buffer info buffer for packet
- * @idx: starting point on ring to unwind
+ * idpf_tx_splitq_bump_ntu - adjust NTU and generation
+ * @txq: the tx ring to wrap
+ * @ntu: ring index to bump
*/
-void idpf_tx_dma_map_error(struct idpf_tx_queue *txq, struct sk_buff *skb,
- struct idpf_tx_buf *first, u16 idx)
+static unsigned int idpf_tx_splitq_bump_ntu(struct idpf_tx_queue *txq, u16 ntu)
{
- u64_stats_update_begin(&txq->stats_sync);
- u64_stats_inc(&txq->q_stats.dma_map_errs);
- u64_stats_update_end(&txq->stats_sync);
+ ntu++;
- /* clear dma mappings for failed tx_buf map */
- for (;;) {
- struct idpf_tx_buf *tx_buf;
+ if (ntu == txq->desc_count)
+ ntu = 0;
- tx_buf = &txq->tx_buf[idx];
- idpf_tx_buf_rel(txq, tx_buf);
- if (tx_buf == first)
- break;
- if (idx == 0)
- idx = txq->desc_count;
- idx--;
- }
+ return ntu;
+}
- if (skb_is_gso(skb)) {
- union idpf_tx_flex_desc *tx_desc;
+/**
+ * idpf_tx_get_free_buf_id - get a free buffer ID from the refill queue
+ * @refillq: refill queue to get buffer ID from
+ * @buf_id: return buffer ID
+ *
+ * Return: true if a buffer ID was found, false if not
+ */
+static bool idpf_tx_get_free_buf_id(struct idpf_sw_queue *refillq,
+ u32 *buf_id)
+{
+ u32 ntc = refillq->next_to_clean;
+ u32 refill_desc;
- /* If we failed a DMA mapping for a TSO packet, we will have
- * used one additional descriptor for a context
- * descriptor. Reset that here.
- */
- tx_desc = &txq->flex_tx[idx];
- memset(tx_desc, 0, sizeof(struct idpf_flex_tx_ctx_desc));
- if (idx == 0)
- idx = txq->desc_count;
- idx--;
+ refill_desc = refillq->ring[ntc];
+
+ if (unlikely(idpf_queue_has(RFL_GEN_CHK, refillq) !=
+ !!(refill_desc & IDPF_RFL_BI_GEN_M)))
+ return false;
+
+ *buf_id = FIELD_GET(IDPF_RFL_BI_BUFID_M, refill_desc);
+
+ if (unlikely(++ntc == refillq->desc_count)) {
+ idpf_queue_change(RFL_GEN_CHK, refillq);
+ ntc = 0;
}
- /* Update tail in case netdev_xmit_more was previously true */
- idpf_tx_buf_hw_update(txq, idx, false);
+ refillq->next_to_clean = ntc;
+
+ return true;
}
/**
- * idpf_tx_splitq_bump_ntu - adjust NTU and generation
- * @txq: the tx ring to wrap
- * @ntu: ring index to bump
+ * idpf_tx_splitq_pkt_err_unmap - Unmap buffers and bump tail in case of error
+ * @txq: Tx queue to unwind
+ * @params: pointer to splitq params struct
+ * @first: starting buffer for packet to unmap
*/
-static unsigned int idpf_tx_splitq_bump_ntu(struct idpf_tx_queue *txq, u16 ntu)
+static void idpf_tx_splitq_pkt_err_unmap(struct idpf_tx_queue *txq,
+ struct idpf_tx_splitq_params *params,
+ struct idpf_tx_buf *first)
{
- ntu++;
+ struct idpf_sw_queue *refillq = txq->refillq;
+ struct libeth_sq_napi_stats ss = { };
+ struct idpf_tx_buf *tx_buf = first;
+ struct libeth_cq_pp cp = {
+ .dev = txq->dev,
+ .ss = &ss,
+ };
- if (ntu == txq->desc_count) {
- ntu = 0;
- txq->compl_tag_cur_gen = IDPF_TX_ADJ_COMPL_TAG_GEN(txq);
+ u64_stats_update_begin(&txq->stats_sync);
+ u64_stats_inc(&txq->q_stats.dma_map_errs);
+ u64_stats_update_end(&txq->stats_sync);
+
+ libeth_tx_complete(tx_buf, &cp);
+ while (idpf_tx_buf_next(tx_buf) != IDPF_TXBUF_NULL) {
+ tx_buf = &txq->tx_buf[idpf_tx_buf_next(tx_buf)];
+ libeth_tx_complete(tx_buf, &cp);
}
- return ntu;
+ /* Update tail in case netdev_xmit_more was previously true. */
+ idpf_tx_buf_hw_update(txq, params->prev_ntu, false);
+
+ if (!refillq)
+ return;
+
+ /* Restore refillq state to avoid leaking tags. */
+ if (params->prev_refill_gen != idpf_queue_has(RFL_GEN_CHK, refillq))
+ idpf_queue_change(RFL_GEN_CHK, refillq);
+ refillq->next_to_clean = params->prev_refill_ntc;
}
/**
@@ -2380,6 +2630,7 @@ static void idpf_tx_splitq_map(struct idpf_tx_queue *tx_q,
struct netdev_queue *nq;
struct sk_buff *skb;
skb_frag_t *frag;
+ u32 next_buf_id;
u16 td_cmd = 0;
dma_addr_t dma;
@@ -2395,17 +2646,19 @@ static void idpf_tx_splitq_map(struct idpf_tx_queue *tx_q,
dma = dma_map_single(tx_q->dev, skb->data, size, DMA_TO_DEVICE);
tx_buf = first;
-
- params->compl_tag =
- (tx_q->compl_tag_cur_gen << tx_q->compl_tag_gen_s) | i;
+ first->nr_frags = 0;
for (frag = &skb_shinfo(skb)->frags[0];; frag++) {
unsigned int max_data = IDPF_TX_MAX_DESC_DATA_ALIGNED;
- if (dma_mapping_error(tx_q->dev, dma))
- return idpf_tx_dma_map_error(tx_q, skb, first, i);
+ if (unlikely(dma_mapping_error(tx_q->dev, dma))) {
+ idpf_tx_buf_next(tx_buf) = IDPF_TXBUF_NULL;
+ return idpf_tx_splitq_pkt_err_unmap(tx_q, params,
+ first);
+ }
- tx_buf->compl_tag = params->compl_tag;
+ first->nr_frags++;
+ tx_buf->type = LIBETH_SQE_FRAG;
/* record length, and DMA address */
dma_unmap_len_set(tx_buf, len, size);
@@ -2459,29 +2712,13 @@ static void idpf_tx_splitq_map(struct idpf_tx_queue *tx_q,
idpf_tx_splitq_build_desc(tx_desc, params, td_cmd,
max_data);
- tx_desc++;
- i++;
-
- if (i == tx_q->desc_count) {
+ if (unlikely(++i == tx_q->desc_count)) {
tx_desc = &tx_q->flex_tx[0];
i = 0;
- tx_q->compl_tag_cur_gen =
- IDPF_TX_ADJ_COMPL_TAG_GEN(tx_q);
+ } else {
+ tx_desc++;
}
- /* Since this packet has a buffer that is going to span
- * multiple descriptors, it's going to leave holes in
- * to the TX buffer ring. To ensure these holes do not
- * cause issues in the cleaning routines, we will clear
- * them of any stale data and assign them the same
- * completion tag as the current packet. Then when the
- * packet is being cleaned, the cleaning routines will
- * simply pass over these holes and finish cleaning the
- * rest of the packet.
- */
- memset(&tx_q->tx_buf[i], 0, sizeof(struct idpf_tx_buf));
- tx_q->tx_buf[i].compl_tag = params->compl_tag;
-
/* Adjust the DMA offset and the remaining size of the
* fragment. On the first iteration of this loop,
* max_data will be >= 12K and <= 16K-1. On any
@@ -2504,40 +2741,51 @@ static void idpf_tx_splitq_map(struct idpf_tx_queue *tx_q,
break;
idpf_tx_splitq_build_desc(tx_desc, params, td_cmd, size);
- tx_desc++;
- i++;
- if (i == tx_q->desc_count) {
+ if (unlikely(++i == tx_q->desc_count)) {
tx_desc = &tx_q->flex_tx[0];
i = 0;
- tx_q->compl_tag_cur_gen = IDPF_TX_ADJ_COMPL_TAG_GEN(tx_q);
+ } else {
+ tx_desc++;
+ }
+
+ if (idpf_queue_has(FLOW_SCH_EN, tx_q)) {
+ if (unlikely(!idpf_tx_get_free_buf_id(tx_q->refillq,
+ &next_buf_id))) {
+ idpf_tx_buf_next(tx_buf) = IDPF_TXBUF_NULL;
+ return idpf_tx_splitq_pkt_err_unmap(tx_q, params,
+ first);
+ }
+ } else {
+ next_buf_id = i;
}
+ idpf_tx_buf_next(tx_buf) = next_buf_id;
+ tx_buf = &tx_q->tx_buf[next_buf_id];
size = skb_frag_size(frag);
data_len -= size;
dma = skb_frag_dma_map(tx_q->dev, frag, 0, size,
DMA_TO_DEVICE);
-
- tx_buf = &tx_q->tx_buf[i];
}
/* record SW timestamp if HW timestamp is not available */
skb_tx_timestamp(skb);
+ first->type = LIBETH_SQE_SKB;
+
/* write last descriptor with RS and EOP bits */
+ first->rs_idx = i;
+ idpf_tx_buf_next(tx_buf) = IDPF_TXBUF_NULL;
td_cmd |= params->eop_cmd;
idpf_tx_splitq_build_desc(tx_desc, params, td_cmd, size);
i = idpf_tx_splitq_bump_ntu(tx_q, i);
- /* set next_to_watch value indicating a packet is present */
- first->next_to_watch = tx_desc;
-
tx_q->txq_grp->num_completions_pending++;
/* record bytecount for BQL */
nq = netdev_get_tx_queue(tx_q->netdev, tx_q->idx);
- netdev_tx_sent_queue(nq, first->bytecount);
+ netdev_tx_sent_queue(nq, first->bytes);
idpf_tx_buf_hw_update(tx_q, i, netdev_xmit_more());
}
@@ -2618,111 +2866,6 @@ int idpf_tso(struct sk_buff *skb, struct idpf_tx_offload_params *off)
return 1;
}
-/**
- * __idpf_chk_linearize - Check skb is not using too many buffers
- * @skb: send buffer
- * @max_bufs: maximum number of buffers
- *
- * For TSO we need to count the TSO header and segment payload separately. As
- * such we need to check cases where we have max_bufs-1 fragments or more as we
- * can potentially require max_bufs+1 DMA transactions, 1 for the TSO header, 1
- * for the segment payload in the first descriptor, and another max_buf-1 for
- * the fragments.
- */
-static bool __idpf_chk_linearize(struct sk_buff *skb, unsigned int max_bufs)
-{
- const struct skb_shared_info *shinfo = skb_shinfo(skb);
- const skb_frag_t *frag, *stale;
- int nr_frags, sum;
-
- /* no need to check if number of frags is less than max_bufs - 1 */
- nr_frags = shinfo->nr_frags;
- if (nr_frags < (max_bufs - 1))
- return false;
-
- /* We need to walk through the list and validate that each group
- * of max_bufs-2 fragments totals at least gso_size.
- */
- nr_frags -= max_bufs - 2;
- frag = &shinfo->frags[0];
-
- /* Initialize size to the negative value of gso_size minus 1. We use
- * this as the worst case scenario in which the frag ahead of us only
- * provides one byte which is why we are limited to max_bufs-2
- * descriptors for a single transmit as the header and previous
- * fragment are already consuming 2 descriptors.
- */
- sum = 1 - shinfo->gso_size;
-
- /* Add size of frags 0 through 4 to create our initial sum */
- sum += skb_frag_size(frag++);
- sum += skb_frag_size(frag++);
- sum += skb_frag_size(frag++);
- sum += skb_frag_size(frag++);
- sum += skb_frag_size(frag++);
-
- /* Walk through fragments adding latest fragment, testing it, and
- * then removing stale fragments from the sum.
- */
- for (stale = &shinfo->frags[0];; stale++) {
- int stale_size = skb_frag_size(stale);
-
- sum += skb_frag_size(frag++);
-
- /* The stale fragment may present us with a smaller
- * descriptor than the actual fragment size. To account
- * for that we need to remove all the data on the front and
- * figure out what the remainder would be in the last
- * descriptor associated with the fragment.
- */
- if (stale_size > IDPF_TX_MAX_DESC_DATA) {
- int align_pad = -(skb_frag_off(stale)) &
- (IDPF_TX_MAX_READ_REQ_SIZE - 1);
-
- sum -= align_pad;
- stale_size -= align_pad;
-
- do {
- sum -= IDPF_TX_MAX_DESC_DATA_ALIGNED;
- stale_size -= IDPF_TX_MAX_DESC_DATA_ALIGNED;
- } while (stale_size > IDPF_TX_MAX_DESC_DATA);
- }
-
- /* if sum is negative we failed to make sufficient progress */
- if (sum < 0)
- return true;
-
- if (!nr_frags--)
- break;
-
- sum -= stale_size;
- }
-
- return false;
-}
-
-/**
- * idpf_chk_linearize - Check if skb exceeds max descriptors per packet
- * @skb: send buffer
- * @max_bufs: maximum scatter gather buffers for single packet
- * @count: number of buffers this packet needs
- *
- * Make sure we don't exceed maximum scatter gather buffers for a single
- * packet. We have to do some special checking around the boundary (max_bufs-1)
- * if TSO is on since we need count the TSO header and payload separately.
- * E.g.: a packet with 7 fragments can require 9 DMA transactions; 1 for TSO
- * header, 1 for segment payload, and then 7 for the fragments.
- */
-static bool idpf_chk_linearize(struct sk_buff *skb, unsigned int max_bufs,
- unsigned int count)
-{
- if (likely(count < max_bufs))
- return false;
- if (skb_is_gso(skb))
- return __idpf_chk_linearize(skb, max_bufs);
-
- return count > max_bufs;
-}
/**
* idpf_tx_splitq_get_ctx_desc - grab next desc and update buffer ring
@@ -2731,15 +2874,12 @@ static bool idpf_chk_linearize(struct sk_buff *skb, unsigned int max_bufs,
* Since the TX buffer rings mimics the descriptor ring, update the tx buffer
* ring entry to reflect that this index is a context descriptor
*/
-static struct idpf_flex_tx_ctx_desc *
+static union idpf_flex_tx_ctx_desc *
idpf_tx_splitq_get_ctx_desc(struct idpf_tx_queue *txq)
{
- struct idpf_flex_tx_ctx_desc *desc;
+ union idpf_flex_tx_ctx_desc *desc;
int i = txq->next_to_use;
- memset(&txq->tx_buf[i], 0, sizeof(struct idpf_tx_buf));
- txq->tx_buf[i].compl_tag = IDPF_SPLITQ_TX_INVAL_COMPL_TAG;
-
/* grab the next descriptor */
desc = &txq->flex_ctx[i];
txq->next_to_use = idpf_tx_splitq_bump_ntu(txq, i);
@@ -2765,6 +2905,88 @@ netdev_tx_t idpf_tx_drop_skb(struct idpf_tx_queue *tx_q, struct sk_buff *skb)
return NETDEV_TX_OK;
}
+#if (IS_ENABLED(CONFIG_PTP_1588_CLOCK))
+/**
+ * idpf_tx_tstamp - set up context descriptor for hardware timestamp
+ * @tx_q: queue to send buffer on
+ * @skb: pointer to the SKB we're sending
+ * @off: pointer to the offload struct
+ *
+ * Return: Positive index number on success, negative otherwise.
+ */
+static int idpf_tx_tstamp(struct idpf_tx_queue *tx_q, struct sk_buff *skb,
+ struct idpf_tx_offload_params *off)
+{
+ int err, idx;
+
+ /* only timestamp the outbound packet if the user has requested it */
+ if (likely(!(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)))
+ return -1;
+
+ if (!idpf_ptp_get_txq_tstamp_capability(tx_q))
+ return -1;
+
+ /* Tx timestamps cannot be sampled when doing TSO */
+ if (off->tx_flags & IDPF_TX_FLAGS_TSO)
+ return -1;
+
+ /* Grab an open timestamp slot */
+ err = idpf_ptp_request_ts(tx_q, skb, &idx);
+ if (err) {
+ u64_stats_update_begin(&tx_q->stats_sync);
+ u64_stats_inc(&tx_q->q_stats.tstamp_skipped);
+ u64_stats_update_end(&tx_q->stats_sync);
+
+ return -1;
+ }
+
+ off->tx_flags |= IDPF_TX_FLAGS_TSYN;
+
+ return idx;
+}
+
+/**
+ * idpf_tx_set_tstamp_desc - Set the Tx descriptor fields needed to generate
+ * PHY Tx timestamp
+ * @ctx_desc: Context descriptor
+ * @idx: Index of the Tx timestamp latch
+ */
+static void idpf_tx_set_tstamp_desc(union idpf_flex_tx_ctx_desc *ctx_desc,
+ u32 idx)
+{
+ ctx_desc->tsyn.qw1 = le64_encode_bits(IDPF_TX_DESC_DTYPE_CTX,
+ IDPF_TX_CTX_DTYPE_M) |
+ le64_encode_bits(IDPF_TX_CTX_DESC_TSYN,
+ IDPF_TX_CTX_CMD_M) |
+ le64_encode_bits(idx, IDPF_TX_CTX_TSYN_REG_M);
+}
+#else /* CONFIG_PTP_1588_CLOCK */
+static int idpf_tx_tstamp(struct idpf_tx_queue *tx_q, struct sk_buff *skb,
+ struct idpf_tx_offload_params *off)
+{
+ return -1;
+}
+
+static void idpf_tx_set_tstamp_desc(union idpf_flex_tx_ctx_desc *ctx_desc,
+ u32 idx)
+{ }
+#endif /* CONFIG_PTP_1588_CLOCK */
+
+/**
+ * idpf_tx_splitq_need_re - check whether RE bit needs to be set
+ * @tx_q: pointer to Tx queue
+ *
+ * Return: true if RE bit needs to be set, false otherwise
+ */
+static bool idpf_tx_splitq_need_re(struct idpf_tx_queue *tx_q)
+{
+ int gap = tx_q->next_to_use - tx_q->last_re;
+
+ gap += (gap < 0) ? tx_q->desc_count : 0;
+
+ return gap >= IDPF_TX_SPLITQ_RE_MIN_GAP;
+}
+
/**
* idpf_tx_splitq_frame - Sends buffer on Tx ring using flex descriptors
* @skb: send buffer
@@ -2775,12 +2997,16 @@ netdev_tx_t idpf_tx_drop_skb(struct idpf_tx_queue *tx_q, struct sk_buff *skb)
static netdev_tx_t idpf_tx_splitq_frame(struct sk_buff *skb,
struct idpf_tx_queue *tx_q)
{
- struct idpf_tx_splitq_params tx_params = { };
+ struct idpf_tx_splitq_params tx_params = {
+ .prev_ntu = tx_q->next_to_use,
+ };
+ union idpf_flex_tx_ctx_desc *ctx_desc;
struct idpf_tx_buf *first;
- unsigned int count;
- int tso;
+ u32 count, buf_count = 1;
+ int tso, idx;
+ u32 buf_id;
- count = idpf_tx_desc_count_required(tx_q, skb);
+ count = idpf_tx_res_count_required(tx_q, skb, &buf_count);
if (unlikely(!count))
return idpf_tx_drop_skb(tx_q, skb);
@@ -2790,7 +3016,7 @@ static netdev_tx_t idpf_tx_splitq_frame(struct sk_buff *skb,
/* Check for splitq specific TX resources */
count += (IDPF_TX_DESCS_PER_CACHE_LINE + tso);
- if (idpf_tx_maybe_stop_splitq(tx_q, count)) {
+ if (idpf_tx_maybe_stop_splitq(tx_q, count, buf_count)) {
idpf_tx_buf_hw_update(tx_q, tx_q->next_to_use, false);
return NETDEV_TX_BUSY;
@@ -2798,8 +3024,7 @@ static netdev_tx_t idpf_tx_splitq_frame(struct sk_buff *skb,
if (tso) {
/* If tso is needed, set up context desc */
- struct idpf_flex_tx_ctx_desc *ctx_desc =
- idpf_tx_splitq_get_ctx_desc(tx_q);
+ ctx_desc = idpf_tx_splitq_get_ctx_desc(tx_q);
ctx_desc->tso.qw1.cmd_dtype =
cpu_to_le16(IDPF_TX_DESC_DTYPE_FLEX_TSO_CTX |
@@ -2817,36 +3042,53 @@ static netdev_tx_t idpf_tx_splitq_frame(struct sk_buff *skb,
u64_stats_update_end(&tx_q->stats_sync);
}
- /* record the location of the first descriptor for this packet */
- first = &tx_q->tx_buf[tx_q->next_to_use];
- first->skb = skb;
-
- if (tso) {
- first->gso_segs = tx_params.offload.tso_segs;
- first->bytecount = skb->len +
- ((first->gso_segs - 1) * tx_params.offload.tso_hdr_len);
- } else {
- first->gso_segs = 1;
- first->bytecount = max_t(unsigned int, skb->len, ETH_ZLEN);
+ idx = idpf_tx_tstamp(tx_q, skb, &tx_params.offload);
+ if (idx != -1) {
+ ctx_desc = idpf_tx_splitq_get_ctx_desc(tx_q);
+ idpf_tx_set_tstamp_desc(ctx_desc, idx);
}
if (idpf_queue_has(FLOW_SCH_EN, tx_q)) {
+ struct idpf_sw_queue *refillq = tx_q->refillq;
+
+ /* Save refillq state in case of a packet rollback. Otherwise,
+ * the tags will be leaked since they will be popped from the
+ * refillq but never reposted during cleaning.
+ */
+ tx_params.prev_refill_gen =
+ idpf_queue_has(RFL_GEN_CHK, refillq);
+ tx_params.prev_refill_ntc = refillq->next_to_clean;
+
+ if (unlikely(!idpf_tx_get_free_buf_id(tx_q->refillq,
+ &buf_id))) {
+ if (tx_params.prev_refill_gen !=
+ idpf_queue_has(RFL_GEN_CHK, refillq))
+ idpf_queue_change(RFL_GEN_CHK, refillq);
+ refillq->next_to_clean = tx_params.prev_refill_ntc;
+
+ tx_q->next_to_use = tx_params.prev_ntu;
+ return idpf_tx_drop_skb(tx_q, skb);
+ }
+ tx_params.compl_tag = buf_id;
+
tx_params.dtype = IDPF_TX_DESC_DTYPE_FLEX_FLOW_SCHE;
tx_params.eop_cmd = IDPF_TXD_FLEX_FLOW_CMD_EOP;
- /* Set the RE bit to catch any packets that may have not been
- * stashed during RS completion cleaning. MIN_GAP is set to
- * MIN_RING size to ensure it will be set at least once each
- * time around the ring.
+ /* Set the RE bit to periodically "clean" the descriptor ring.
+ * MIN_GAP is set to MIN_RING size to ensure it will be set at
+ * least once each time around the ring.
*/
- if (!(tx_q->next_to_use % IDPF_TX_SPLITQ_RE_MIN_GAP)) {
+ if (idpf_tx_splitq_need_re(tx_q)) {
tx_params.eop_cmd |= IDPF_TXD_FLEX_FLOW_CMD_RE;
tx_q->txq_grp->num_completions_pending++;
+ tx_q->last_re = tx_q->next_to_use;
}
if (skb->ip_summed == CHECKSUM_PARTIAL)
tx_params.offload.td_cmd |= IDPF_TXD_FLEX_FLOW_CMD_CS_EN;
} else {
+ buf_id = tx_q->next_to_use;
+
tx_params.dtype = IDPF_TX_DESC_DTYPE_FLEX_L2TAG1_L2TAG2;
tx_params.eop_cmd = IDPF_TXD_LAST_DESC_CMD;
@@ -2854,6 +3096,18 @@ static netdev_tx_t idpf_tx_splitq_frame(struct sk_buff *skb,
tx_params.offload.td_cmd |= IDPF_TX_FLEX_DESC_CMD_CS_EN;
}
+ first = &tx_q->tx_buf[buf_id];
+ first->skb = skb;
+
+ if (tso) {
+ first->packets = tx_params.offload.tso_segs;
+ first->bytes = skb->len +
+ ((first->packets - 1) * tx_params.offload.tso_hdr_len);
+ } else {
+ first->packets = 1;
+ first->bytes = max_t(unsigned int, skb->len, ETH_ZLEN);
+ }
+
idpf_tx_splitq_map(tx_q, &tx_params, first);
return NETDEV_TX_OK;
@@ -2868,10 +3122,11 @@ static netdev_tx_t idpf_tx_splitq_frame(struct sk_buff *skb,
*/
netdev_tx_t idpf_tx_start(struct sk_buff *skb, struct net_device *netdev)
{
- struct idpf_vport *vport = idpf_netdev_to_vport(netdev);
+ const struct idpf_vport *vport = idpf_netdev_to_vport(netdev);
struct idpf_tx_queue *tx_q;
- if (unlikely(skb_get_queue_mapping(skb) >= vport->num_txq)) {
+ if (unlikely(skb_get_queue_mapping(skb) >=
+ vport->num_txq - vport->num_xdp_txq)) {
dev_kfree_skb_any(skb);
return NETDEV_TX_OK;
@@ -2908,7 +3163,7 @@ idpf_rx_hash(const struct idpf_rx_queue *rxq, struct sk_buff *skb,
{
u32 hash;
- if (!libeth_rx_pt_has_hash(rxq->netdev, decoded))
+ if (!libeth_rx_pt_has_hash(rxq->xdp_rxq.dev, decoded))
return;
hash = le16_to_cpu(rx_desc->hash1) |
@@ -2928,13 +3183,13 @@ idpf_rx_hash(const struct idpf_rx_queue *rxq, struct sk_buff *skb,
* skb->protocol must be set before this function is called
*/
static void idpf_rx_csum(struct idpf_rx_queue *rxq, struct sk_buff *skb,
- struct idpf_rx_csum_decoded csum_bits,
+ struct libeth_rx_csum csum_bits,
struct libeth_rx_pt decoded)
{
bool ipv4, ipv6;
/* check if Rx checksum is enabled */
- if (!libeth_rx_pt_has_checksum(rxq->netdev, decoded))
+ if (!libeth_rx_pt_has_checksum(rxq->xdp_rxq.dev, decoded))
return;
/* check if HW has decoded the packet and checksum */
@@ -2956,7 +3211,7 @@ static void idpf_rx_csum(struct idpf_rx_queue *rxq, struct sk_buff *skb,
if (unlikely(csum_bits.l4e))
goto checksum_fail;
- if (csum_bits.raw_csum_inv ||
+ if (!csum_bits.raw_csum_valid ||
decoded.inner_prot == LIBETH_RX_PT_INNER_SCTP) {
skb->ip_summed = CHECKSUM_UNNECESSARY;
return;
@@ -2979,10 +3234,10 @@ checksum_fail:
*
* Return: parsed checksum status.
**/
-static struct idpf_rx_csum_decoded
+static struct libeth_rx_csum
idpf_rx_splitq_extract_csum_bits(const struct virtchnl2_rx_flex_desc_adv_nic_3 *rx_desc)
{
- struct idpf_rx_csum_decoded csum = { };
+ struct libeth_rx_csum csum = { };
u8 qword0, qword1;
qword0 = rx_desc->status_err0_qw0;
@@ -2998,9 +3253,9 @@ idpf_rx_splitq_extract_csum_bits(const struct virtchnl2_rx_flex_desc_adv_nic_3 *
qword1);
csum.ipv6exadd = FIELD_GET(VIRTCHNL2_RX_FLEX_DESC_ADV_STATUS0_IPV6EXADD_M,
qword0);
- csum.raw_csum_inv =
- le16_get_bits(rx_desc->ptype_err_fflags0,
- VIRTCHNL2_RX_FLEX_DESC_ADV_RAW_CSUM_INV_M);
+ csum.raw_csum_valid =
+ !le16_get_bits(rx_desc->ptype_err_fflags0,
+ VIRTCHNL2_RX_FLEX_DESC_ADV_RAW_CSUM_INV_M);
csum.raw_csum = le16_to_cpu(rx_desc->misc.raw_cs);
return csum;
@@ -3041,14 +3296,11 @@ static int idpf_rx_rsc(struct idpf_rx_queue *rxq, struct sk_buff *skb,
return -EINVAL;
rsc_segments = DIV_ROUND_UP(skb->data_len, rsc_seg_len);
- if (unlikely(rsc_segments == 1))
- return 0;
NAPI_GRO_CB(skb)->count = rsc_segments;
skb_shinfo(skb)->gso_size = rsc_seg_len;
skb_reset_network_header(skb);
- len = skb->len - skb_transport_offset(skb);
if (ipv4) {
struct iphdr *ipv4h = ip_hdr(skb);
@@ -3057,6 +3309,7 @@ static int idpf_rx_rsc(struct idpf_rx_queue *rxq, struct sk_buff *skb,
/* Reset and set transport header offset in skb */
skb_set_transport_header(skb, sizeof(struct iphdr));
+ len = skb->len - skb_transport_offset(skb);
/* Compute the TCP pseudo header checksum*/
tcp_hdr(skb)->check =
@@ -3066,6 +3319,7 @@ static int idpf_rx_rsc(struct idpf_rx_queue *rxq, struct sk_buff *skb,
skb_shinfo(skb)->gso_type = SKB_GSO_TCPV6;
skb_set_transport_header(skb, sizeof(struct ipv6hdr));
+ len = skb->len - skb_transport_offset(skb);
tcp_hdr(skb)->check =
~tcp_v6_check(len, &ipv6h->saddr, &ipv6h->daddr, 0);
}
@@ -3080,7 +3334,34 @@ static int idpf_rx_rsc(struct idpf_rx_queue *rxq, struct sk_buff *skb,
}
/**
- * idpf_rx_process_skb_fields - Populate skb header fields from Rx descriptor
+ * idpf_rx_hwtstamp - check for an RX timestamp and pass up the stack
+ * @rxq: pointer to the rx queue that receives the timestamp
+ * @rx_desc: pointer to rx descriptor containing timestamp
+ * @skb: skb to put timestamp in
+ */
+static void
+idpf_rx_hwtstamp(const struct idpf_rx_queue *rxq,
+ const struct virtchnl2_rx_flex_desc_adv_nic_3 *rx_desc,
+ struct sk_buff *skb)
+{
+ u64 cached_time, ts_ns;
+ u32 ts_high;
+
+ if (!(rx_desc->ts_low & VIRTCHNL2_RX_FLEX_TSTAMP_VALID))
+ return;
+
+ cached_time = READ_ONCE(rxq->cached_phc_time);
+
+ ts_high = le32_to_cpu(rx_desc->ts_high);
+ ts_ns = idpf_ptp_tstamp_extend_32b_to_64b(cached_time, ts_high);
+
+ *skb_hwtstamps(skb) = (struct skb_shared_hwtstamps) {
+ .hwtstamp = ns_to_ktime(ts_ns),
+ };
+}
+
+/**
+ * __idpf_rx_process_skb_fields - Populate skb header fields from Rx descriptor
* @rxq: Rx descriptor ring packet is being transacted on
* @skb: pointer to current skb being populated
* @rx_desc: Receive descriptor
@@ -3090,10 +3371,10 @@ static int idpf_rx_rsc(struct idpf_rx_queue *rxq, struct sk_buff *skb,
* other fields within the skb.
*/
static int
-idpf_rx_process_skb_fields(struct idpf_rx_queue *rxq, struct sk_buff *skb,
- const struct virtchnl2_rx_flex_desc_adv_nic_3 *rx_desc)
+__idpf_rx_process_skb_fields(struct idpf_rx_queue *rxq, struct sk_buff *skb,
+ const struct virtchnl2_rx_flex_desc_adv_nic_3 *rx_desc)
{
- struct idpf_rx_csum_decoded csum_bits;
+ struct libeth_rx_csum csum_bits;
struct libeth_rx_pt decoded;
u16 rx_ptype;
@@ -3104,7 +3385,8 @@ idpf_rx_process_skb_fields(struct idpf_rx_queue *rxq, struct sk_buff *skb,
/* process RSS/hash */
idpf_rx_hash(rxq, skb, rx_desc, decoded);
- skb->protocol = eth_type_trans(skb, rxq->netdev);
+ if (idpf_queue_has(PTP, rxq))
+ idpf_rx_hwtstamp(rxq, rx_desc, skb);
if (le16_get_bits(rx_desc->hdrlen_flags,
VIRTCHNL2_RX_FLEX_DESC_ADV_RSC_M))
@@ -3113,30 +3395,27 @@ idpf_rx_process_skb_fields(struct idpf_rx_queue *rxq, struct sk_buff *skb,
csum_bits = idpf_rx_splitq_extract_csum_bits(rx_desc);
idpf_rx_csum(rxq, skb, csum_bits, decoded);
- skb_record_rx_queue(skb, rxq->idx);
-
return 0;
}
-/**
- * idpf_rx_add_frag - Add contents of Rx buffer to sk_buff as a frag
- * @rx_buf: buffer containing page to add
- * @skb: sk_buff to place the data into
- * @size: packet length from rx_desc
- *
- * This function will add the data contained in rx_buf->page to the skb.
- * It will just attach the page as a frag to the skb.
- * The function will then update the page offset.
- */
-void idpf_rx_add_frag(struct idpf_rx_buf *rx_buf, struct sk_buff *skb,
- unsigned int size)
+bool idpf_rx_process_skb_fields(struct sk_buff *skb,
+ const struct libeth_xdp_buff *xdp,
+ struct libeth_rq_napi_stats *rs)
{
- u32 hr = rx_buf->page->pp->p.offset;
+ struct idpf_rx_queue *rxq;
- skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, rx_buf->page,
- rx_buf->offset + hr, size, rx_buf->truesize);
+ rxq = libeth_xdp_buff_to_rq(xdp, typeof(*rxq), xdp_rxq);
+
+ return !__idpf_rx_process_skb_fields(rxq, skb, xdp->desc);
}
+LIBETH_XDP_DEFINE_START();
+LIBETH_XDP_DEFINE_RUN(static idpf_xdp_run_pass, idpf_xdp_run_prog,
+ idpf_xdp_tx_flush_bulk, idpf_rx_process_skb_fields);
+LIBETH_XDP_DEFINE_FINALIZE(static idpf_xdp_finalize_rx, idpf_xdp_tx_flush_bulk,
+ idpf_xdp_tx_finalize);
+LIBETH_XDP_DEFINE_END();
+
/**
* idpf_rx_hsplit_wa - handle header buffer overflows and split errors
* @hdr: Rx buffer for the headers
@@ -3157,51 +3436,28 @@ static u32 idpf_rx_hsplit_wa(const struct libeth_fqe *hdr,
struct libeth_fqe *buf, u32 data_len)
{
u32 copy = data_len <= L1_CACHE_BYTES ? data_len : ETH_HLEN;
+ struct page *hdr_page, *buf_page;
const void *src;
void *dst;
- if (!libeth_rx_sync_for_cpu(buf, copy))
+ if (unlikely(netmem_is_net_iov(buf->netmem)) ||
+ !libeth_rx_sync_for_cpu(buf, copy))
return 0;
- dst = page_address(hdr->page) + hdr->offset + hdr->page->pp->p.offset;
- src = page_address(buf->page) + buf->offset + buf->page->pp->p.offset;
- memcpy(dst, src, LARGEST_ALIGN(copy));
+ hdr_page = __netmem_to_page(hdr->netmem);
+ buf_page = __netmem_to_page(buf->netmem);
+ dst = page_address(hdr_page) + hdr->offset +
+ pp_page_to_nmdesc(hdr_page)->pp->p.offset;
+ src = page_address(buf_page) + buf->offset +
+ pp_page_to_nmdesc(buf_page)->pp->p.offset;
+ memcpy(dst, src, LARGEST_ALIGN(copy));
buf->offset += copy;
return copy;
}
/**
- * idpf_rx_build_skb - Allocate skb and populate it from header buffer
- * @buf: Rx buffer to pull data from
- * @size: the length of the packet
- *
- * This function allocates an skb. It then populates it with the page data from
- * the current receive descriptor, taking care to set up the skb correctly.
- */
-struct sk_buff *idpf_rx_build_skb(const struct libeth_fqe *buf, u32 size)
-{
- u32 hr = buf->page->pp->p.offset;
- struct sk_buff *skb;
- void *va;
-
- va = page_address(buf->page) + buf->offset;
- prefetch(va + hr);
-
- skb = napi_build_skb(va, buf->truesize);
- if (unlikely(!skb))
- return NULL;
-
- skb_mark_for_recycle(skb);
-
- skb_reserve(skb, hr);
- __skb_put(skb, size);
-
- return skb;
-}
-
-/**
* idpf_rx_splitq_test_staterr - tests bits in Rx descriptor
* status and error fields
* @stat_err_field: field from descriptor to test bits in
@@ -3242,13 +3498,18 @@ static bool idpf_rx_splitq_is_eop(struct virtchnl2_rx_flex_desc_adv_nic_3 *rx_de
*/
static int idpf_rx_splitq_clean(struct idpf_rx_queue *rxq, int budget)
{
- int total_rx_bytes = 0, total_rx_pkts = 0;
struct idpf_buf_queue *rx_bufq = NULL;
- struct sk_buff *skb = rxq->skb;
+ struct libeth_rq_napi_stats rs = { };
u16 ntc = rxq->next_to_clean;
+ LIBETH_XDP_ONSTACK_BUFF(xdp);
+ LIBETH_XDP_ONSTACK_BULK(bq);
+
+ libeth_xdp_tx_init_bulk(&bq, rxq->xdp_prog, rxq->xdp_rxq.dev,
+ rxq->xdpsqs, rxq->num_xdp_txq);
+ libeth_xdp_init_buff(xdp, &rxq->xdp, &rxq->xdp_rxq);
/* Process Rx packets bounded by budget */
- while (likely(total_rx_pkts < budget)) {
+ while (likely(rs.packets < budget)) {
struct virtchnl2_rx_flex_desc_adv_nic_3 *rx_desc;
struct libeth_fqe *hdr, *rx_buf = NULL;
struct idpf_sw_queue *refillq = NULL;
@@ -3262,18 +3523,14 @@ static int idpf_rx_splitq_clean(struct idpf_rx_queue *rxq, int budget)
/* get the Rx desc from Rx queue based on 'next_to_clean' */
rx_desc = &rxq->rx[ntc].flex_adv_nic_3_wb;
- /* This memory barrier is needed to keep us from reading
- * any other fields out of the rx_desc
- */
- dma_rmb();
-
/* if the descriptor isn't done, no work yet to do */
gen_id = le16_get_bits(rx_desc->pktlen_gen_bufq_id,
VIRTCHNL2_RX_FLEX_DESC_ADV_GEN_M);
-
if (idpf_queue_has(GEN_CHK, rxq) != gen_id)
break;
+ dma_rmb();
+
rxdid = FIELD_GET(VIRTCHNL2_RX_FLEX_DESC_ADV_RXDID_M,
rx_desc->rxdid_ucast);
if (rxdid != VIRTCHNL2_RXDID_2_FLEX_SPLITQ) {
@@ -3318,84 +3575,47 @@ static int idpf_rx_splitq_clean(struct idpf_rx_queue *rxq, int budget)
hdr = &rx_bufq->hdr_buf[buf_id];
- if (unlikely(!hdr_len && !skb)) {
+ if (unlikely(!hdr_len && !xdp->data)) {
hdr_len = idpf_rx_hsplit_wa(hdr, rx_buf, pkt_len);
- pkt_len -= hdr_len;
+ /* If failed, drop both buffers by setting len to 0 */
+ pkt_len -= hdr_len ? : pkt_len;
u64_stats_update_begin(&rxq->stats_sync);
u64_stats_inc(&rxq->q_stats.hsplit_buf_ovf);
u64_stats_update_end(&rxq->stats_sync);
}
- if (libeth_rx_sync_for_cpu(hdr, hdr_len)) {
- skb = idpf_rx_build_skb(hdr, hdr_len);
- if (!skb)
- break;
-
- u64_stats_update_begin(&rxq->stats_sync);
- u64_stats_inc(&rxq->q_stats.hsplit_pkts);
- u64_stats_update_end(&rxq->stats_sync);
- }
+ if (libeth_xdp_process_buff(xdp, hdr, hdr_len))
+ rs.hsplit++;
- hdr->page = NULL;
+ hdr->netmem = 0;
payload:
- if (!libeth_rx_sync_for_cpu(rx_buf, pkt_len))
- goto skip_data;
-
- if (skb)
- idpf_rx_add_frag(rx_buf, skb, pkt_len);
- else
- skb = idpf_rx_build_skb(rx_buf, pkt_len);
+ libeth_xdp_process_buff(xdp, rx_buf, pkt_len);
+ rx_buf->netmem = 0;
- /* exit if we failed to retrieve a buffer */
- if (!skb)
- break;
-
-skip_data:
- rx_buf->page = NULL;
-
- idpf_rx_post_buf_refill(refillq, buf_id);
+ idpf_post_buf_refill(refillq, buf_id);
IDPF_RX_BUMP_NTC(rxq, ntc);
/* skip if it is non EOP desc */
- if (!idpf_rx_splitq_is_eop(rx_desc) || unlikely(!skb))
+ if (!idpf_rx_splitq_is_eop(rx_desc) || unlikely(!xdp->data))
continue;
- /* pad skb if needed (to make valid ethernet frame) */
- if (eth_skb_pad(skb)) {
- skb = NULL;
- continue;
- }
-
- /* probably a little skewed due to removing CRC */
- total_rx_bytes += skb->len;
-
- /* protocol */
- if (unlikely(idpf_rx_process_skb_fields(rxq, skb, rx_desc))) {
- dev_kfree_skb_any(skb);
- skb = NULL;
- continue;
- }
-
- /* send completed skb up the stack */
- napi_gro_receive(rxq->napi, skb);
- skb = NULL;
-
- /* update budget accounting */
- total_rx_pkts++;
+ idpf_xdp_run_pass(xdp, &bq, rxq->napi, &rs, rx_desc);
}
+ idpf_xdp_finalize_rx(&bq);
+
rxq->next_to_clean = ntc;
+ libeth_xdp_save_buff(&rxq->xdp, xdp);
- rxq->skb = skb;
u64_stats_update_begin(&rxq->stats_sync);
- u64_stats_add(&rxq->q_stats.packets, total_rx_pkts);
- u64_stats_add(&rxq->q_stats.bytes, total_rx_bytes);
+ u64_stats_add(&rxq->q_stats.packets, rs.packets);
+ u64_stats_add(&rxq->q_stats.bytes, rs.bytes);
+ u64_stats_add(&rxq->q_stats.hsplit_pkts, rs.hsplit);
u64_stats_update_end(&rxq->stats_sync);
- /* guarantee a trip back through this routine if there was a failure */
- return total_rx_pkts;
+ return rs.packets;
}
/**
@@ -3463,10 +3683,10 @@ static void idpf_rx_clean_refillq(struct idpf_buf_queue *bufq,
bool failure;
if (idpf_queue_has(RFL_GEN_CHK, refillq) !=
- !!(refill_desc & IDPF_RX_BI_GEN_M))
+ !!(refill_desc & IDPF_RFL_BI_GEN_M))
break;
- buf_id = FIELD_GET(IDPF_RX_BI_BUFID_M, refill_desc);
+ buf_id = FIELD_GET(IDPF_RFL_BI_BUFID_M, refill_desc);
failure = idpf_rx_update_bufq_desc(bufq, buf_id, buf_desc);
if (failure)
break;
@@ -3538,7 +3758,7 @@ static irqreturn_t idpf_vport_intr_clean_queues(int __always_unused irq,
struct idpf_q_vector *q_vector = (struct idpf_q_vector *)data;
q_vector->total_events++;
- napi_schedule(&q_vector->napi);
+ napi_schedule_irqoff(&q_vector->napi);
return IRQ_HANDLED;
}
@@ -3576,11 +3796,11 @@ static void idpf_vport_intr_napi_dis_all(struct idpf_vport *vport)
*/
void idpf_vport_intr_rel(struct idpf_vport *vport)
{
- int i, j, v_idx;
-
- for (v_idx = 0; v_idx < vport->num_q_vectors; v_idx++) {
+ for (u32 v_idx = 0; v_idx < vport->num_q_vectors; v_idx++) {
struct idpf_q_vector *q_vector = &vport->q_vectors[v_idx];
+ kfree(q_vector->xsksq);
+ q_vector->xsksq = NULL;
kfree(q_vector->complq);
q_vector->complq = NULL;
kfree(q_vector->bufq);
@@ -3589,32 +3809,24 @@ void idpf_vport_intr_rel(struct idpf_vport *vport)
q_vector->tx = NULL;
kfree(q_vector->rx);
q_vector->rx = NULL;
-
- free_cpumask_var(q_vector->affinity_mask);
}
- /* Clean up the mapping of queues to vectors */
- for (i = 0; i < vport->num_rxq_grp; i++) {
- struct idpf_rxq_group *rx_qgrp = &vport->rxq_grps[i];
+ kfree(vport->q_vectors);
+ vport->q_vectors = NULL;
+}
- if (idpf_is_queue_model_split(vport->rxq_model))
- for (j = 0; j < rx_qgrp->splitq.num_rxq_sets; j++)
- rx_qgrp->splitq.rxq_sets[j]->rxq.q_vector = NULL;
- else
- for (j = 0; j < rx_qgrp->singleq.num_rxq; j++)
- rx_qgrp->singleq.rxqs[j]->q_vector = NULL;
- }
+static void idpf_q_vector_set_napi(struct idpf_q_vector *q_vector, bool link)
+{
+ struct napi_struct *napi = link ? &q_vector->napi : NULL;
+ struct net_device *dev = q_vector->vport->netdev;
- if (idpf_is_queue_model_split(vport->txq_model))
- for (i = 0; i < vport->num_txq_grp; i++)
- vport->txq_grps[i].complq->q_vector = NULL;
- else
- for (i = 0; i < vport->num_txq_grp; i++)
- for (j = 0; j < vport->txq_grps[i].num_txq; j++)
- vport->txq_grps[i].txqs[j]->q_vector = NULL;
+ for (u32 i = 0; i < q_vector->num_rxq; i++)
+ netif_queue_set_napi(dev, q_vector->rx[i]->idx,
+ NETDEV_QUEUE_TYPE_RX, napi);
- kfree(vport->q_vectors);
- vport->q_vectors = NULL;
+ for (u32 i = 0; i < q_vector->num_txq; i++)
+ netif_queue_set_napi(dev, q_vector->tx[i]->idx,
+ NETDEV_QUEUE_TYPE_TX, napi);
}
/**
@@ -3637,8 +3849,7 @@ static void idpf_vport_intr_rel_irq(struct idpf_vport *vport)
vidx = vport->q_vector_idxs[vector];
irq_num = adapter->msix_entries[vidx].vector;
- /* clear the affinity_mask in the IRQ descriptor */
- irq_set_affinity_hint(irq_num, NULL);
+ idpf_q_vector_set_napi(q_vector, false);
kfree(free_irq(irq_num, q_vector));
}
}
@@ -3652,6 +3863,8 @@ static void idpf_vport_intr_dis_irq_all(struct idpf_vport *vport)
struct idpf_q_vector *q_vector = vport->q_vectors;
int q_idx;
+ writel(0, vport->noirq_dyn_ctl);
+
for (q_idx = 0; q_idx < vport->num_q_vectors; q_idx++)
writel(0, q_vector[q_idx].intr_reg.dyn_ctl);
}
@@ -3659,21 +3872,31 @@ static void idpf_vport_intr_dis_irq_all(struct idpf_vport *vport)
/**
* idpf_vport_intr_buildreg_itr - Enable default interrupt generation settings
* @q_vector: pointer to q_vector
- * @type: itr index
- * @itr: itr value
*/
-static u32 idpf_vport_intr_buildreg_itr(struct idpf_q_vector *q_vector,
- const int type, u16 itr)
+static u32 idpf_vport_intr_buildreg_itr(struct idpf_q_vector *q_vector)
{
- u32 itr_val;
+ u32 itr_val = q_vector->intr_reg.dyn_ctl_intena_m;
+ int type = IDPF_NO_ITR_UPDATE_IDX;
+ u16 itr = 0;
+
+ if (q_vector->wb_on_itr) {
+ /*
+ * Trigger a software interrupt when exiting wb_on_itr, to make
+ * sure we catch any pending write backs that might have been
+ * missed due to interrupt state transition.
+ */
+ itr_val |= q_vector->intr_reg.dyn_ctl_swint_trig_m |
+ q_vector->intr_reg.dyn_ctl_sw_itridx_ena_m;
+ type = IDPF_SW_ITR_UPDATE_IDX;
+ itr = IDPF_ITR_20K;
+ }
itr &= IDPF_ITR_MASK;
/* Don't clear PBA because that can cause lost interrupts that
* came in while we were cleaning/polling
*/
- itr_val = q_vector->intr_reg.dyn_ctl_intena_m |
- (type << q_vector->intr_reg.dyn_ctl_itridx_s) |
- (itr << (q_vector->intr_reg.dyn_ctl_intrvl_s - 1));
+ itr_val |= (type << q_vector->intr_reg.dyn_ctl_itridx_s) |
+ (itr << (q_vector->intr_reg.dyn_ctl_intrvl_s - 1));
return itr_val;
}
@@ -3735,7 +3958,7 @@ static void idpf_net_dim(struct idpf_q_vector *q_vector)
idpf_update_dim_sample(q_vector, &dim_sample, &q_vector->tx_dim,
packets, bytes);
- net_dim(&q_vector->tx_dim, dim_sample);
+ net_dim(&q_vector->tx_dim, &dim_sample);
check_rx_itr:
if (!IDPF_ITR_IS_DYNAMIC(q_vector->rx_intr_mode))
@@ -3754,7 +3977,7 @@ check_rx_itr:
idpf_update_dim_sample(q_vector, &dim_sample, &q_vector->rx_dim,
packets, bytes);
- net_dim(&q_vector->rx_dim, dim_sample);
+ net_dim(&q_vector->rx_dim, &dim_sample);
}
/**
@@ -3771,8 +3994,8 @@ void idpf_vport_intr_update_itr_ena_irq(struct idpf_q_vector *q_vector)
/* net_dim() updates ITR out-of-band using a work item */
idpf_net_dim(q_vector);
- intval = idpf_vport_intr_buildreg_itr(q_vector,
- IDPF_NO_ITR_UPDATE_IDX, 0);
+ intval = idpf_vport_intr_buildreg_itr(q_vector);
+ q_vector->wb_on_itr = false;
writel(intval, q_vector->intr_reg.dyn_ctl);
}
@@ -3780,13 +4003,15 @@ void idpf_vport_intr_update_itr_ena_irq(struct idpf_q_vector *q_vector)
/**
* idpf_vport_intr_req_irq - get MSI-X vectors from the OS for the vport
* @vport: main vport structure
- * @basename: name for the vector
*/
-static int idpf_vport_intr_req_irq(struct idpf_vport *vport, char *basename)
+static int idpf_vport_intr_req_irq(struct idpf_vport *vport)
{
struct idpf_adapter *adapter = vport->adapter;
+ const char *drv_name, *if_name, *vec_name;
int vector, err, irq_num, vidx;
- const char *vec_name;
+
+ drv_name = dev_driver_string(&adapter->pdev->dev);
+ if_name = netdev_name(vport->netdev);
for (vector = 0; vector < vport->num_q_vectors; vector++) {
struct idpf_q_vector *q_vector = &vport->q_vectors[vector];
@@ -3804,8 +4029,8 @@ static int idpf_vport_intr_req_irq(struct idpf_vport *vport, char *basename)
else
continue;
- name = kasprintf(GFP_KERNEL, "%s-%s-%d", basename, vec_name,
- vidx);
+ name = kasprintf(GFP_KERNEL, "%s-%s-%s-%d", drv_name, if_name,
+ vec_name, vidx);
err = request_irq(irq_num, idpf_vport_intr_clean_queues, 0,
name, q_vector);
@@ -3814,8 +4039,8 @@ static int idpf_vport_intr_req_irq(struct idpf_vport *vport, char *basename)
"Request_irq failed, error: %d\n", err);
goto free_q_irqs;
}
- /* assign the mask for this irq */
- irq_set_affinity_hint(irq_num, q_vector->affinity_mask);
+
+ idpf_q_vector_set_napi(q_vector, true);
}
return 0;
@@ -3883,6 +4108,8 @@ static void idpf_vport_intr_ena_irq_all(struct idpf_vport *vport)
if (qv->num_txq || qv->num_rxq)
idpf_vport_intr_update_itr_ena_irq(qv);
}
+
+ writel(vport->noirq_dyn_ctl_ena, vport->noirq_dyn_ctl);
}
/**
@@ -4032,7 +4259,9 @@ static bool idpf_rx_splitq_clean_all(struct idpf_q_vector *q_vec, int budget,
struct idpf_rx_queue *rxq = q_vec->rx[i];
int pkts_cleaned_per_q;
- pkts_cleaned_per_q = idpf_rx_splitq_clean(rxq, budget_per_q);
+ pkts_cleaned_per_q = idpf_queue_has(XSK, rxq) ?
+ idpf_xskrq_poll(rxq, budget_per_q) :
+ idpf_rx_splitq_clean(rxq, budget_per_q);
/* if we clean as many as budgeted, we must not be done */
if (pkts_cleaned_per_q >= budget_per_q)
clean_complete = false;
@@ -4042,8 +4271,10 @@ static bool idpf_rx_splitq_clean_all(struct idpf_q_vector *q_vec, int budget,
nid = numa_mem_id();
- for (i = 0; i < q_vec->num_bufq; i++)
- idpf_rx_clean_refillq_all(q_vec->bufq[i], nid);
+ for (i = 0; i < q_vec->num_bufq; i++) {
+ if (!idpf_queue_has(XSK, q_vec->bufq[i]))
+ idpf_rx_clean_refillq_all(q_vec->bufq[i], nid);
+ }
return clean_complete;
}
@@ -4057,7 +4288,7 @@ static int idpf_vport_splitq_napi_poll(struct napi_struct *napi, int budget)
{
struct idpf_q_vector *q_vector =
container_of(napi, struct idpf_q_vector, napi);
- bool clean_complete;
+ bool clean_complete = true;
int work_done = 0;
/* Handle case where we are called by netpoll with a budget of 0 */
@@ -4067,30 +4298,31 @@ static int idpf_vport_splitq_napi_poll(struct napi_struct *napi, int budget)
return 0;
}
- clean_complete = idpf_rx_splitq_clean_all(q_vector, budget, &work_done);
- clean_complete &= idpf_tx_splitq_clean_all(q_vector, budget, &work_done);
+ for (u32 i = 0; i < q_vector->num_xsksq; i++)
+ clean_complete &= idpf_xsk_xmit(q_vector->xsksq[i]);
+
+ clean_complete &= idpf_tx_splitq_clean_all(q_vector, budget,
+ &work_done);
+ clean_complete &= idpf_rx_splitq_clean_all(q_vector, budget,
+ &work_done);
/* If work not completed, return budget and polling will return */
- if (!clean_complete)
+ if (!clean_complete) {
+ idpf_vport_intr_set_wb_on_itr(q_vector);
return budget;
+ }
work_done = min_t(int, work_done, budget - 1);
/* 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 (napi_complete_done(napi, work_done))
idpf_vport_intr_update_itr_ena_irq(q_vector);
-
- /* Switch to poll mode in the tear-down path after sending disable
- * queues virtchnl message, as the interrupts will be disabled after
- * that
- */
- if (unlikely(q_vector->num_txq && idpf_queue_has(POLL_MODE,
- q_vector->tx[0])))
- return budget;
else
- return work_done;
+ idpf_vport_intr_set_wb_on_itr(q_vector);
+
+ return work_done;
}
/**
@@ -4101,8 +4333,8 @@ static int idpf_vport_splitq_napi_poll(struct napi_struct *napi, int budget)
*/
static void idpf_vport_intr_map_vector_to_qs(struct idpf_vport *vport)
{
+ u16 num_txq_grp = vport->num_txq_grp - vport->num_xdp_txq;
bool split = idpf_is_queue_model_split(vport->rxq_model);
- u16 num_txq_grp = vport->num_txq_grp;
struct idpf_rxq_group *rx_qgrp;
struct idpf_txq_group *tx_qgrp;
u32 i, qv_idx, q_index;
@@ -4178,6 +4410,21 @@ static void idpf_vport_intr_map_vector_to_qs(struct idpf_vport *vport)
qv_idx++;
}
+
+ for (i = 0; i < vport->num_xdp_txq; i++) {
+ struct idpf_tx_queue *xdpsq;
+ struct idpf_q_vector *qv;
+
+ xdpsq = vport->txqs[vport->xdp_txq_offset + i];
+ if (!idpf_queue_has(XSK, xdpsq))
+ continue;
+
+ qv = idpf_find_rxq_vec(vport, i);
+ idpf_xsk_init_wakeup(qv);
+
+ xdpsq->q_vector = qv;
+ qv->xsksq[qv->num_xsksq++] = xdpsq;
+ }
}
/**
@@ -4198,6 +4445,8 @@ static int idpf_vport_intr_init_vec_idx(struct idpf_vport *vport)
for (i = 0; i < vport->num_q_vectors; i++)
vport->q_vectors[i].v_idx = vport->q_vector_idxs[i];
+ vport->noirq_v_idx = vport->q_vector_idxs[i];
+
return 0;
}
@@ -4211,6 +4460,8 @@ static int idpf_vport_intr_init_vec_idx(struct idpf_vport *vport)
for (i = 0; i < vport->num_q_vectors; i++)
vport->q_vectors[i].v_idx = vecids[vport->q_vector_idxs[i]];
+ vport->noirq_v_idx = vecids[vport->q_vector_idxs[i]];
+
kfree(vecids);
return 0;
@@ -4223,7 +4474,8 @@ static int idpf_vport_intr_init_vec_idx(struct idpf_vport *vport)
static void idpf_vport_intr_napi_add_all(struct idpf_vport *vport)
{
int (*napi_poll)(struct napi_struct *napi, int budget);
- u16 v_idx;
+ u16 v_idx, qv_idx;
+ int irq_num;
if (idpf_is_queue_model_split(vport->txq_model))
napi_poll = idpf_vport_splitq_napi_poll;
@@ -4232,12 +4484,12 @@ static void idpf_vport_intr_napi_add_all(struct idpf_vport *vport)
for (v_idx = 0; v_idx < vport->num_q_vectors; v_idx++) {
struct idpf_q_vector *q_vector = &vport->q_vectors[v_idx];
+ qv_idx = vport->q_vector_idxs[v_idx];
+ irq_num = vport->adapter->msix_entries[qv_idx].vector;
- netif_napi_add(vport->netdev, &q_vector->napi, napi_poll);
-
- /* only set affinity_mask if the CPU is online */
- if (cpu_online(v_idx))
- cpumask_set_cpu(v_idx, q_vector->affinity_mask);
+ netif_napi_add_config(vport->netdev, &q_vector->napi,
+ napi_poll, v_idx);
+ netif_napi_set_irq(&q_vector->napi, irq_num);
}
}
@@ -4251,9 +4503,13 @@ static void idpf_vport_intr_napi_add_all(struct idpf_vport *vport)
int idpf_vport_intr_alloc(struct idpf_vport *vport)
{
u16 txqs_per_vector, rxqs_per_vector, bufqs_per_vector;
+ struct idpf_vport_user_config_data *user_config;
struct idpf_q_vector *q_vector;
+ struct idpf_q_coalesce *q_coal;
u32 complqs_per_vector, v_idx;
+ u16 idx = vport->idx;
+ user_config = &vport->adapter->vport_config[idx]->user_config;
vport->q_vectors = kcalloc(vport->num_q_vectors,
sizeof(struct idpf_q_vector), GFP_KERNEL);
if (!vport->q_vectors)
@@ -4271,19 +4527,17 @@ int idpf_vport_intr_alloc(struct idpf_vport *vport)
for (v_idx = 0; v_idx < vport->num_q_vectors; v_idx++) {
q_vector = &vport->q_vectors[v_idx];
+ q_coal = &user_config->q_coalesce[v_idx];
q_vector->vport = vport;
- q_vector->tx_itr_value = IDPF_ITR_TX_DEF;
- q_vector->tx_intr_mode = IDPF_ITR_DYNAMIC;
+ q_vector->tx_itr_value = q_coal->tx_coalesce_usecs;
+ q_vector->tx_intr_mode = q_coal->tx_intr_mode;
q_vector->tx_itr_idx = VIRTCHNL2_ITR_IDX_1;
- q_vector->rx_itr_value = IDPF_ITR_RX_DEF;
- q_vector->rx_intr_mode = IDPF_ITR_DYNAMIC;
+ q_vector->rx_itr_value = q_coal->rx_coalesce_usecs;
+ q_vector->rx_intr_mode = q_coal->rx_intr_mode;
q_vector->rx_itr_idx = VIRTCHNL2_ITR_IDX_0;
- if (!zalloc_cpumask_var(&q_vector->affinity_mask, GFP_KERNEL))
- goto error;
-
q_vector->tx = kcalloc(txqs_per_vector, sizeof(*q_vector->tx),
GFP_KERNEL);
if (!q_vector->tx)
@@ -4308,6 +4562,15 @@ int idpf_vport_intr_alloc(struct idpf_vport *vport)
GFP_KERNEL);
if (!q_vector->complq)
goto error;
+
+ if (!vport->xdp_txq_offset)
+ continue;
+
+ q_vector->xsksq = kcalloc(rxqs_per_vector,
+ sizeof(*q_vector->xsksq),
+ GFP_KERNEL);
+ if (!q_vector->xsksq)
+ goto error;
}
return 0;
@@ -4326,7 +4589,6 @@ error:
*/
int idpf_vport_intr_init(struct idpf_vport *vport)
{
- char *int_name;
int err;
err = idpf_vport_intr_init_vec_idx(vport);
@@ -4340,11 +4602,7 @@ int idpf_vport_intr_init(struct idpf_vport *vport)
if (err)
goto unroll_vectors_alloc;
- int_name = kasprintf(GFP_KERNEL, "%s-%s",
- dev_driver_string(&vport->adapter->pdev->dev),
- vport->netdev->name);
-
- err = idpf_vport_intr_req_irq(vport, int_name);
+ err = idpf_vport_intr_req_irq(vport);
if (err)
goto unroll_vectors_alloc;