net/mlx5e: Make XDP xmit functions more generic

Convert the XDP xmit functions to use the generic xdp_frame API
in XDP_TX flow.
Same functions will be used later in this series to transmit
the XDP redirect-out packets as well.
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Eugenia Emantayev <eugenia@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>

drivers/net/ethernet/mellanox/mlx5/core/en.h

@@ -395,6 +395,17 @@ struct mlx5e_txqsq {
 	} recover;
 } ____cacheline_aligned_in_smp;
 
+struct mlx5e_dma_info {
+	struct page     *page;
+	dma_addr_t      addr;
+};
+
+struct mlx5e_xdp_info {
+	struct xdp_frame      *xdpf;
+	dma_addr_t            dma_addr;
+	struct mlx5e_dma_info di;
+};
+
 struct mlx5e_xdpsq {
 	/* data path */
 
@@ -406,7 +417,7 @@ struct mlx5e_xdpsq {
 
 	/* write@xmit, read@completion */
 	struct {
-		struct mlx5e_dma_info     *di;
+		struct mlx5e_xdp_info     *xdpi;
 		bool                       doorbell;
 		bool                       redirect_flush;
 	} db;
@@ -419,6 +430,7 @@ struct mlx5e_xdpsq {
 	__be32                     mkey_be;
 	u8                         min_inline_mode;
 	unsigned long              state;
+	unsigned int               hw_mtu;
 
 	/* control path */
 	struct mlx5_wq_ctrl        wq_ctrl;
@@ -455,11 +467,6 @@ mlx5e_wqc_has_room_for(struct mlx5_wq_cyc *wq, u16 cc, u16 pc, u16 n)
 	return (mlx5_wq_cyc_ctr2ix(wq, cc - pc) >= n) || (cc == pc);
 }
 
-struct mlx5e_dma_info {
-	struct page	*page;
-	dma_addr_t	addr;
-};
-
 struct mlx5e_wqe_frag_info {
 	struct mlx5e_dma_info *di;
 	u32 offset;
@@ -562,7 +569,6 @@ struct mlx5e_rq {
 
 	/* XDP */
 	struct bpf_prog       *xdp_prog;
-	unsigned int           hw_mtu;
 	struct mlx5e_xdpsq     xdpsq;
 	DECLARE_BITMAP(flags, 8);
 	struct page_pool      *page_pool;

drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c

@@ -33,6 +33,23 @@
 #include <linux/bpf_trace.h>
 #include "en/xdp.h"
 
+static inline bool
+mlx5e_xmit_xdp_buff(struct mlx5e_xdpsq *sq, struct mlx5e_dma_info *di,
+		    struct xdp_buff *xdp)
+{
+	struct mlx5e_xdp_info xdpi;
+
+	xdpi.xdpf = convert_to_xdp_frame(xdp);
+	if (unlikely(!xdpi.xdpf))
+		return false;
+	xdpi.dma_addr = di->addr + (xdpi.xdpf->data - (void *)xdpi.xdpf);
+	dma_sync_single_for_device(sq->pdev, xdpi.dma_addr,
+				   xdpi.xdpf->len, PCI_DMA_TODEVICE);
+	xdpi.di = *di;
+
+	return mlx5e_xmit_xdp_frame(sq, &xdpi);
+}
+
 /* returns true if packet was consumed by xdp */
 bool mlx5e_xdp_handle(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
 		      void *va, u16 *rx_headroom, u32 *len)
@@ -58,22 +75,24 @@ bool mlx5e_xdp_handle(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
 		*len = xdp.data_end - xdp.data;
 		return false;
 	case XDP_TX:
-		if (unlikely(!mlx5e_xmit_xdp_frame(rq, di, &xdp)))
-			trace_xdp_exception(rq->netdev, prog, act);
+		if (unlikely(!mlx5e_xmit_xdp_buff(&rq->xdpsq, di, &xdp)))
+			goto xdp_abort;
+		__set_bit(MLX5E_RQ_FLAG_XDP_XMIT, rq->flags); /* non-atomic */
 		return true;
 	case XDP_REDIRECT:
 		/* When XDP enabled then page-refcnt==1 here */
 		err = xdp_do_redirect(rq->netdev, &xdp, prog);
-		if (!err) {
-			__set_bit(MLX5E_RQ_FLAG_XDP_XMIT, rq->flags);
-			rq->xdpsq.db.redirect_flush = true;
-			mlx5e_page_dma_unmap(rq, di);
-		}
+		if (unlikely(err))
+			goto xdp_abort;
+		__set_bit(MLX5E_RQ_FLAG_XDP_XMIT, rq->flags);
+		rq->xdpsq.db.redirect_flush = true;
+		mlx5e_page_dma_unmap(rq, di);
 		rq->stats->xdp_redirect++;
 		return true;
 	default:
 		bpf_warn_invalid_xdp_action(act);
 	case XDP_ABORTED:
+xdp_abort:
 		trace_xdp_exception(rq->netdev, prog, act);
 	case XDP_DROP:
 		rq->stats->xdp_drop++;
@@ -81,27 +100,27 @@ bool mlx5e_xdp_handle(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
 	}
 }
 
-bool mlx5e_xmit_xdp_frame(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
-			  const struct xdp_buff *xdp)
+bool mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xdp_info *xdpi)
 {
-	struct mlx5e_xdpsq       *sq   = &rq->xdpsq;
 	struct mlx5_wq_cyc       *wq   = &sq->wq;
 	u16                       pi   = mlx5_wq_cyc_ctr2ix(wq, sq->pc);
 	struct mlx5e_tx_wqe      *wqe  = mlx5_wq_cyc_get_wqe(wq, pi);
 
+	struct mlx5e_rq *rq = container_of(sq, struct mlx5e_rq, xdpsq);
+
 	struct mlx5_wqe_ctrl_seg *cseg = &wqe->ctrl;
 	struct mlx5_wqe_eth_seg  *eseg = &wqe->eth;
-	struct mlx5_wqe_data_seg *dseg;
+	struct mlx5_wqe_data_seg *dseg = wqe->data;
 
-	ptrdiff_t data_offset = xdp->data - xdp->data_hard_start;
-	dma_addr_t dma_addr  = di->addr + data_offset;
-	unsigned int dma_len = xdp->data_end - xdp->data;
+	struct xdp_frame *xdpf = xdpi->xdpf;
+	dma_addr_t dma_addr  = xdpi->dma_addr;
+	unsigned int dma_len = xdpf->len;
 
 	struct mlx5e_rq_stats *stats = rq->stats;
 
 	prefetchw(wqe);
 
-	if (unlikely(dma_len < MLX5E_XDP_MIN_INLINE || rq->hw_mtu < dma_len)) {
+	if (unlikely(dma_len < MLX5E_XDP_MIN_INLINE || sq->hw_mtu < dma_len)) {
 		stats->xdp_drop++;
 		return false;
 	}
@@ -116,15 +135,11 @@ bool mlx5e_xmit_xdp_frame(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
 		return false;
 	}
 
-	dma_sync_single_for_device(sq->pdev, dma_addr, dma_len, PCI_DMA_TODEVICE);
-
 	cseg->fm_ce_se = 0;
 
-	dseg = (struct mlx5_wqe_data_seg *)eseg + 1;
-
 	/* copy the inline part if required */
 	if (sq->min_inline_mode != MLX5_INLINE_MODE_NONE) {
-		memcpy(eseg->inline_hdr.start, xdp->data, MLX5E_XDP_MIN_INLINE);
+		memcpy(eseg->inline_hdr.start, xdpf->data, MLX5E_XDP_MIN_INLINE);
 		eseg->inline_hdr.sz = cpu_to_be16(MLX5E_XDP_MIN_INLINE);
 		dma_len  -= MLX5E_XDP_MIN_INLINE;
 		dma_addr += MLX5E_XDP_MIN_INLINE;
@@ -140,8 +155,7 @@ bool mlx5e_xmit_xdp_frame(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
 	/* move page to reference to sq responsibility,
 	 * and mark so it's not put back in page-cache.
 	 */
-	__set_bit(MLX5E_RQ_FLAG_XDP_XMIT, rq->flags); /* non-atomic */
-	sq->db.di[pi] = *di;
+	sq->db.xdpi[pi] = *xdpi;
 	sq->pc++;
 
 	sq->db.doorbell = true;
@@ -184,17 +198,17 @@ bool mlx5e_poll_xdpsq_cq(struct mlx5e_cq *cq)
 		wqe_counter = be16_to_cpu(cqe->wqe_counter);
 
 		do {
-			struct mlx5e_dma_info *di;
+			struct mlx5e_xdp_info *xdpi;
 			u16 ci;
 
 			last_wqe = (sqcc == wqe_counter);
 
 			ci = mlx5_wq_cyc_ctr2ix(&sq->wq, sqcc);
-			di = &sq->db.di[ci];
+			xdpi = &sq->db.xdpi[ci];
 
 			sqcc++;
 			/* Recycle RX page */
-			mlx5e_page_release(rq, di, true);
+			mlx5e_page_release(rq, &xdpi->di, true);
 		} while (!last_wqe);
 	} while ((++i < MLX5E_TX_CQ_POLL_BUDGET) && (cqe = mlx5_cqwq_get_cqe(&cq->wq)));
 
@@ -212,15 +226,15 @@ bool mlx5e_poll_xdpsq_cq(struct mlx5e_cq *cq)
 void mlx5e_free_xdpsq_descs(struct mlx5e_xdpsq *sq)
 {
 	struct mlx5e_rq *rq = container_of(sq, struct mlx5e_rq, xdpsq);
-	struct mlx5e_dma_info *di;
+	struct mlx5e_xdp_info *xdpi;
 	u16 ci;
 
 	while (sq->cc != sq->pc) {
 		ci = mlx5_wq_cyc_ctr2ix(&sq->wq, sq->cc);
-		di = &sq->db.di[ci];
+		xdpi = &sq->db.xdpi[ci];
 		sq->cc++;
 
-		mlx5e_page_release(rq, di, false);
+		mlx5e_page_release(rq, &xdpi->di, false);
 	}
 }
 

drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h

@@ -45,8 +45,7 @@ bool mlx5e_xdp_handle(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
 bool mlx5e_poll_xdpsq_cq(struct mlx5e_cq *cq);
 void mlx5e_free_xdpsq_descs(struct mlx5e_xdpsq *sq);
 
-bool mlx5e_xmit_xdp_frame(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
-			  const struct xdp_buff *xdp);
+bool mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xdp_info *xdpi);
 
 static inline void mlx5e_xmit_xdp_doorbell(struct mlx5e_xdpsq *sq)
 {

drivers/net/ethernet/mellanox/mlx5/core/en_main.c

@@ -491,7 +491,6 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c,
 	rq->channel = c;
 	rq->ix      = c->ix;
 	rq->mdev    = mdev;
-	rq->hw_mtu  = MLX5E_SW2HW_MTU(params, params->sw_mtu);
 	rq->stats   = &c->priv->channel_stats[c->ix].rq;
 
 	rq->xdp_prog = params->xdp_prog ? bpf_prog_inc(params->xdp_prog) : NULL;
@@ -969,16 +968,16 @@ static void mlx5e_close_rq(struct mlx5e_rq *rq)
 
 static void mlx5e_free_xdpsq_db(struct mlx5e_xdpsq *sq)
 {
-	kvfree(sq->db.di);
+	kvfree(sq->db.xdpi);
 }
 
 static int mlx5e_alloc_xdpsq_db(struct mlx5e_xdpsq *sq, int numa)
 {
 	int wq_sz = mlx5_wq_cyc_get_size(&sq->wq);
 
-	sq->db.di = kvzalloc_node(array_size(wq_sz, sizeof(*sq->db.di)),
-				  GFP_KERNEL, numa);
-	if (!sq->db.di) {
+	sq->db.xdpi = kvzalloc_node(array_size(wq_sz, sizeof(*sq->db.xdpi)),
+				    GFP_KERNEL, numa);
+	if (!sq->db.xdpi) {
 		mlx5e_free_xdpsq_db(sq);
 		return -ENOMEM;
 	}
@@ -1001,6 +1000,7 @@ static int mlx5e_alloc_xdpsq(struct mlx5e_channel *c,
 	sq->channel   = c;
 	sq->uar_map   = mdev->mlx5e_res.bfreg.map;
 	sq->min_inline_mode = params->tx_min_inline_mode;
+	sq->hw_mtu    = MLX5E_SW2HW_MTU(params, params->sw_mtu);
 
 	param->wq.db_numa_node = cpu_to_node(c->cpu);
 	err = mlx5_wq_cyc_create(mdev, &param->wq, sqc_wq, wq, &sq->wq_ctrl);