aboutsummaryrefslogtreecommitdiff
path: root/drivers/net/ethernet/mellanox/mlx4/en_tx.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/net/ethernet/mellanox/mlx4/en_tx.c')
-rw-r--r--drivers/net/ethernet/mellanox/mlx4/en_tx.c78
1 files changed, 36 insertions, 42 deletions
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_tx.c b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
index 13457032d15..5045bab5963 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
@@ -44,16 +44,6 @@
#include "mlx4_en.h"
-enum {
- MAX_INLINE = 104, /* 128 - 16 - 4 - 4 */
- MAX_BF = 256,
-};
-
-static int inline_thold __read_mostly = MAX_INLINE;
-
-module_param_named(inline_thold, inline_thold, int, 0444);
-MODULE_PARM_DESC(inline_thold, "threshold for using inline data");
-
int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv,
struct mlx4_en_tx_ring **pring, int qpn, u32 size,
u16 stride, int node, int queue_index)
@@ -75,8 +65,7 @@ int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv,
ring->size = size;
ring->size_mask = size - 1;
ring->stride = stride;
-
- inline_thold = min(inline_thold, MAX_INLINE);
+ ring->inline_thold = priv->prof->inline_thold;
tmp = size * sizeof(struct mlx4_en_tx_info);
ring->tx_info = vmalloc_node(tmp, node);
@@ -119,12 +108,12 @@ int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv,
ring->buf = ring->wqres.buf.direct.buf;
- en_dbg(DRV, priv, "Allocated TX ring (addr:%p) - buf:%p size:%d "
- "buf_size:%d dma:%llx\n", ring, ring->buf, ring->size,
- ring->buf_size, (unsigned long long) ring->wqres.buf.direct.map);
+ en_dbg(DRV, priv, "Allocated TX ring (addr:%p) - buf:%p size:%d buf_size:%d dma:%llx\n",
+ ring, ring->buf, ring->size, ring->buf_size,
+ (unsigned long long) ring->wqres.buf.direct.map);
ring->qpn = qpn;
- err = mlx4_qp_alloc(mdev->dev, ring->qpn, &ring->qp);
+ err = mlx4_qp_alloc(mdev->dev, ring->qpn, &ring->qp, GFP_KERNEL);
if (err) {
en_err(priv, "Failed allocating qp %d\n", ring->qpn);
goto err_map;
@@ -133,7 +122,7 @@ int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv,
err = mlx4_bf_alloc(mdev->dev, &ring->bf, node);
if (err) {
- en_dbg(DRV, priv, "working without blueflame (%d)", err);
+ en_dbg(DRV, priv, "working without blueflame (%d)\n", err);
ring->bf.uar = &mdev->priv_uar;
ring->bf.uar->map = mdev->uar_map;
ring->bf_enabled = false;
@@ -325,7 +314,7 @@ static u32 mlx4_en_free_tx_desc(struct mlx4_en_priv *priv,
}
}
}
- dev_kfree_skb(skb);
+ dev_kfree_skb_any(skb);
return tx_info->nr_txbb;
}
@@ -362,9 +351,8 @@ int mlx4_en_free_tx_buf(struct net_device *dev, struct mlx4_en_tx_ring *ring)
return cnt;
}
-static int mlx4_en_process_tx_cq(struct net_device *dev,
- struct mlx4_en_cq *cq,
- int budget)
+static bool mlx4_en_process_tx_cq(struct net_device *dev,
+ struct mlx4_en_cq *cq)
{
struct mlx4_en_priv *priv = netdev_priv(dev);
struct mlx4_cq *mcq = &cq->mcq;
@@ -383,9 +371,10 @@ static int mlx4_en_process_tx_cq(struct net_device *dev,
int factor = priv->cqe_factor;
u64 timestamp = 0;
int done = 0;
+ int budget = priv->tx_work_limit;
if (!priv->port_up)
- return 0;
+ return true;
index = cons_index & size_mask;
cqe = &buf[(index << factor) + factor];
@@ -456,9 +445,9 @@ static int mlx4_en_process_tx_cq(struct net_device *dev,
*/
if (netif_tx_queue_stopped(ring->tx_queue) && txbbs_skipped > 0) {
netif_tx_wake_queue(ring->tx_queue);
- priv->port_stats.wake_queue++;
+ ring->wake_queue++;
}
- return done;
+ return done < budget;
}
void mlx4_en_tx_irq(struct mlx4_cq *mcq)
@@ -478,18 +467,16 @@ int mlx4_en_poll_tx_cq(struct napi_struct *napi, int budget)
struct mlx4_en_cq *cq = container_of(napi, struct mlx4_en_cq, napi);
struct net_device *dev = cq->dev;
struct mlx4_en_priv *priv = netdev_priv(dev);
- int done;
+ int clean_complete;
- done = mlx4_en_process_tx_cq(dev, cq, budget);
+ clean_complete = mlx4_en_process_tx_cq(dev, cq);
+ if (!clean_complete)
+ return budget;
- /* If we used up all the quota - we're probably not done yet... */
- if (done < budget) {
- /* Done for now */
- napi_complete(napi);
- mlx4_en_arm_cq(priv, cq);
- return done;
- }
- return budget;
+ napi_complete(napi);
+ mlx4_en_arm_cq(priv, cq);
+
+ return 0;
}
static struct mlx4_en_tx_desc *mlx4_en_bounce_to_desc(struct mlx4_en_priv *priv,
@@ -520,7 +507,7 @@ static struct mlx4_en_tx_desc *mlx4_en_bounce_to_desc(struct mlx4_en_priv *priv,
return ring->buf + index * TXBB_SIZE;
}
-static int is_inline(struct sk_buff *skb, void **pfrag)
+static int is_inline(int inline_thold, struct sk_buff *skb, void **pfrag)
{
void *ptr;
@@ -580,7 +567,7 @@ static int get_real_size(struct sk_buff *skb, struct net_device *dev,
}
} else {
*lso_header_size = 0;
- if (!is_inline(skb, NULL))
+ if (!is_inline(priv->prof->inline_thold, skb, NULL))
real_size = CTRL_SIZE + (skb_shinfo(skb)->nr_frags + 1) * DS_SIZE;
else
real_size = inline_size(skb);
@@ -596,7 +583,13 @@ static void build_inline_wqe(struct mlx4_en_tx_desc *tx_desc, struct sk_buff *sk
int spc = MLX4_INLINE_ALIGN - CTRL_SIZE - sizeof *inl;
if (skb->len <= spc) {
- inl->byte_count = cpu_to_be32(1 << 31 | skb->len);
+ if (likely(skb->len >= MIN_PKT_LEN)) {
+ inl->byte_count = cpu_to_be32(1 << 31 | skb->len);
+ } else {
+ inl->byte_count = cpu_to_be32(1 << 31 | MIN_PKT_LEN);
+ memset(((void *)(inl + 1)) + skb->len, 0,
+ MIN_PKT_LEN - skb->len);
+ }
skb_copy_from_linear_data(skb, inl + 1, skb_headlen(skb));
if (skb_shinfo(skb)->nr_frags)
memcpy(((void *)(inl + 1)) + skb_headlen(skb), fragptr,
@@ -696,7 +689,7 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
ring->size - HEADROOM - MAX_DESC_TXBBS)) {
/* every full Tx ring stops queue */
netif_tx_stop_queue(ring->tx_queue);
- priv->port_stats.queue_stopped++;
+ ring->queue_stopped++;
/* If queue was emptied after the if, and before the
* stop_queue - need to wake the queue, or else it will remain
@@ -709,7 +702,7 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
if (unlikely(((int)(ring->prod - ring->cons)) <=
ring->size - HEADROOM - MAX_DESC_TXBBS)) {
netif_tx_wake_queue(ring->tx_queue);
- priv->port_stats.wake_queue++;
+ ring->wake_queue++;
} else {
return NETDEV_TX_BUSY;
}
@@ -747,11 +740,11 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
tx_info->data_offset = (void *)data - (void *)tx_desc;
tx_info->linear = (lso_header_size < skb_headlen(skb) &&
- !is_inline(skb, NULL)) ? 1 : 0;
+ !is_inline(ring->inline_thold, skb, NULL)) ? 1 : 0;
data += skb_shinfo(skb)->nr_frags + tx_info->linear - 1;
- if (is_inline(skb, &fragptr)) {
+ if (is_inline(ring->inline_thold, skb, &fragptr)) {
tx_info->inl = 1;
} else {
/* Map fragments */
@@ -881,7 +874,8 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
skb_tx_timestamp(skb);
if (ring->bf_enabled && desc_size <= MAX_BF && !bounce && !vlan_tx_tag_present(skb)) {
- *(__be32 *) (&tx_desc->ctrl.vlan_tag) |= cpu_to_be32(ring->doorbell_qpn);
+ tx_desc->ctrl.bf_qpn |= cpu_to_be32(ring->doorbell_qpn);
+
op_own |= htonl((bf_index & 0xffff) << 8);
/* Ensure new descirptor hits memory
* before setting ownership of this descriptor to HW */
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-12-20 13:07:06 +0000