From 4e8231f1c22d36bbf1eed20b2a54609f2e4c49ed Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Wed, 26 Jan 2022 17:43:33 +0200
Subject: [PATCH 01/15] net/mlx5e: Prepare non-linear legacy RQ for XDP multi
 buffer support

mlx5e_skb_from_cqe_nonlinear creates an xdp_buff first, putting the
first fragment as the linear part, and the rest of fragments as
fragments to struct skb_shared_info in the tailroom. Then it creates an
SKB in place, based on the xdp_buff. The XDP program is not called in
this commit yet.

This commit contains no functional change, except the SKB is built over
the whole frag_stride of the first fragment, instead of the minimal size
required (headroom, data and skb_shared_info).

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 .../net/ethernet/mellanox/mlx5/core/en_rx.c   | 75 +++++++++++++++----
 1 file changed, 61 insertions(+), 14 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index 4b8699f392003..dd8ff62e16939 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -1567,45 +1567,92 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe,
 			     struct mlx5e_wqe_frag_info *wi, u32 cqe_bcnt)
 {
 	struct mlx5e_rq_frag_info *frag_info = &rq->wqe.info.arr[0];
+	struct mlx5e_wqe_frag_info *head_wi = wi;
 	u16 rx_headroom = rq->buff.headroom;
 	struct mlx5e_dma_info *di = wi->di;
+	struct skb_shared_info *sinfo;
 	u32 frag_consumed_bytes;
-	u32 first_frag_size;
+	struct xdp_buff xdp;
 	struct sk_buff *skb;
+	u32 truesize;
 	void *va;
 
 	va = page_address(di->page) + wi->offset;
 	frag_consumed_bytes = min_t(u32, frag_info->frag_size, cqe_bcnt);
-	first_frag_size = MLX5_SKB_FRAG_SZ(rx_headroom + frag_consumed_bytes);
 
 	dma_sync_single_range_for_cpu(rq->pdev, di->addr, wi->offset,
-				      first_frag_size, DMA_FROM_DEVICE);
+				      rq->buff.frame0_sz, DMA_FROM_DEVICE);
 	net_prefetch(va + rx_headroom);
 
-	/* XDP is not supported in this configuration, as incoming packets
-	 * might spread among multiple pages.
-	 */
-	skb = mlx5e_build_linear_skb(rq, va, first_frag_size, rx_headroom,
-				     frag_consumed_bytes, 0);
-	if (unlikely(!skb))
-		return NULL;
-
-	page_ref_inc(di->page);
+	mlx5e_fill_xdp_buff(rq, va, rx_headroom, frag_consumed_bytes, &xdp);
+	sinfo = xdp_get_shared_info_from_buff(&xdp);
+	truesize = 0;
 
 	cqe_bcnt -= frag_consumed_bytes;
 	frag_info++;
 	wi++;
 
 	while (cqe_bcnt) {
+		skb_frag_t *frag;
+
+		di = wi->di;
+
 		frag_consumed_bytes = min_t(u32, frag_info->frag_size, cqe_bcnt);
 
-		mlx5e_add_skb_frag(rq, skb, wi->di, wi->offset,
-				   frag_consumed_bytes, frag_info->frag_stride);
+		dma_sync_single_for_cpu(rq->pdev, di->addr + wi->offset,
+					frag_consumed_bytes, DMA_FROM_DEVICE);
+
+		if (!xdp_buff_has_frags(&xdp)) {
+			/* Init on the first fragment to avoid cold cache access
+			 * when possible.
+			 */
+			sinfo->nr_frags = 0;
+			sinfo->xdp_frags_size = 0;
+			xdp_buff_set_frags_flag(&xdp);
+		}
+
+		frag = &sinfo->frags[sinfo->nr_frags++];
+		__skb_frag_set_page(frag, di->page);
+		skb_frag_off_set(frag, wi->offset);
+		skb_frag_size_set(frag, frag_consumed_bytes);
+
+		if (page_is_pfmemalloc(di->page))
+			xdp_buff_set_frag_pfmemalloc(&xdp);
+
+		sinfo->xdp_frags_size += frag_consumed_bytes;
+		truesize += frag_info->frag_stride;
+
 		cqe_bcnt -= frag_consumed_bytes;
 		frag_info++;
 		wi++;
 	}
 
+	di = head_wi->di;
+
+	skb = mlx5e_build_linear_skb(rq, xdp.data_hard_start, rq->buff.frame0_sz,
+				     xdp.data - xdp.data_hard_start,
+				     xdp.data_end - xdp.data,
+				     xdp.data - xdp.data_meta);
+	if (unlikely(!skb))
+		return NULL;
+
+	page_ref_inc(di->page);
+
+	if (unlikely(xdp_buff_has_frags(&xdp))) {
+		int i;
+
+		/* sinfo->nr_frags is reset by build_skb, calculate again. */
+		xdp_update_skb_shared_info(skb, wi - head_wi - 1,
+					   sinfo->xdp_frags_size, truesize,
+					   xdp_buff_is_frag_pfmemalloc(&xdp));
+
+		for (i = 0; i < sinfo->nr_frags; i++) {
+			skb_frag_t *frag = &sinfo->frags[i];
+
+			page_ref_inc(skb_frag_page(frag));
+		}
+	}
+
 	return skb;
 }
 

From 9cb9482ef10e7e524f73878cb20de138be7e1a1f Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Thu, 27 Jan 2022 14:14:53 +0200
Subject: [PATCH 02/15] net/mlx5e: Use fragments of the same size in non-linear
 legacy RQ with XDP

XDP multi buffer implementation in the kernel assumes that all fragments
have the same size. bpf_xdp_frags_increase_tail uses this assumption to
get the size of the last fragment, and __xdp_build_skb_from_frame uses
it to calculate truesize as nr_frags * xdpf->frame_sz.

The current implementation of mlx5e uses fragments of different size in
non-linear legacy RQ. Specifically, the last fragment can be larger than
the others. It's an optimization for packets smaller than MTU.

This commit adapts mlx5e to the kernel limitations and makes it use
fragments of the same size, in order to add support for XDP multi
buffer. The change is applied only if XDP is active, otherwise the old
optimization still applies.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 .../ethernet/mellanox/mlx5/core/en/params.c   | 28 +++++++++++++------
 1 file changed, 19 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
index 5c4711be6faed..822fbb9b80e72 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
@@ -398,8 +398,12 @@ void mlx5e_build_create_cq_param(struct mlx5e_create_cq_param *ccp, struct mlx5e
 	};
 }
 
-static int mlx5e_max_nonlinear_mtu(int first_frag_size, int frag_size)
+static int mlx5e_max_nonlinear_mtu(int first_frag_size, int frag_size, bool xdp)
 {
+	if (xdp)
+		/* XDP requires all fragments to be of the same size. */
+		return first_frag_size + (MLX5E_MAX_RX_FRAGS - 1) * frag_size;
+
 	/* Optimization for small packets: the last fragment is bigger than the others. */
 	return first_frag_size + (MLX5E_MAX_RX_FRAGS - 2) * frag_size + PAGE_SIZE;
 }
@@ -438,12 +442,14 @@ static int mlx5e_build_rq_frags_info(struct mlx5_core_dev *mdev,
 	headroom = mlx5e_get_linear_rq_headroom(params, xsk);
 	first_frag_size_max = SKB_WITH_OVERHEAD(frag_size_max - headroom);
 
-	max_mtu = mlx5e_max_nonlinear_mtu(first_frag_size_max, frag_size_max);
+	max_mtu = mlx5e_max_nonlinear_mtu(first_frag_size_max, frag_size_max,
+					  params->xdp_prog);
 	if (byte_count > max_mtu) {
 		frag_size_max = PAGE_SIZE;
 		first_frag_size_max = SKB_WITH_OVERHEAD(frag_size_max - headroom);
 
-		max_mtu = mlx5e_max_nonlinear_mtu(first_frag_size_max, frag_size_max);
+		max_mtu = mlx5e_max_nonlinear_mtu(first_frag_size_max, frag_size_max,
+						  params->xdp_prog);
 		if (byte_count > max_mtu) {
 			mlx5_core_err(mdev, "MTU %u is too big for non-linear legacy RQ (max %d)\n",
 				      params->sw_mtu, max_mtu);
@@ -463,14 +469,18 @@ static int mlx5e_build_rq_frags_info(struct mlx5_core_dev *mdev,
 		info->arr[i].frag_size = frag_size;
 		buf_size += frag_size;
 
-		if (i == 0) {
-			/* Ensure that headroom and tailroom are included. */
-			frag_size += headroom;
-			frag_size += SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
+		if (params->xdp_prog) {
+			/* XDP multi buffer expects fragments of the same size. */
+			info->arr[i].frag_stride = frag_size_max;
+		} else {
+			if (i == 0) {
+				/* Ensure that headroom and tailroom are included. */
+				frag_size += headroom;
+				frag_size += SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
+			}
+			info->arr[i].frag_stride = roundup_pow_of_two(frag_size);
 		}
 
-		info->arr[i].frag_stride = roundup_pow_of_two(frag_size);
-
 		i++;
 	}
 	info->num_frags = i;

From d51f4a4cca6f629927b1871af0c6f5ce7a9022e9 Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Wed, 9 Feb 2022 15:32:35 +0200
Subject: [PATCH 03/15] net/mlx5e: Use page-sized fragments with XDP multi
 buffer

The implementation of XDP in mlx5e assumes that the frame size is equal
to the page size. Force this limitation in the non-linear mode for XDP
multi buffer.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en/params.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
index 822fbb9b80e72..9646867872c14 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
@@ -444,7 +444,7 @@ static int mlx5e_build_rq_frags_info(struct mlx5_core_dev *mdev,
 
 	max_mtu = mlx5e_max_nonlinear_mtu(first_frag_size_max, frag_size_max,
 					  params->xdp_prog);
-	if (byte_count > max_mtu) {
+	if (byte_count > max_mtu || params->xdp_prog) {
 		frag_size_max = PAGE_SIZE;
 		first_frag_size_max = SKB_WITH_OVERHEAD(frag_size_max - headroom);
 

From ea5d49bdae8b4c9dcdac574eef96b1bd47000c2a Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Thu, 27 Jan 2022 17:01:05 +0200
Subject: [PATCH 04/15] net/mlx5e: Add XDP multi buffer support to the
 non-linear legacy RQ

This commit adds XDP multi buffer support to the RX path in the
non-linear legacy RQ mode. mlx5e_xdp_handle is called from
mlx5e_skb_from_cqe_nonlinear.

XDP_TX action for fragmented XDP frames is not yet supported and
blocked.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c |  5 +++++
 drivers/net/ethernet/mellanox/mlx5/core/en_rx.c  | 13 +++++++++++++
 2 files changed, 18 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
index 6aa77f0e094e6..3a837030e96ee 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
@@ -68,6 +68,11 @@ mlx5e_xmit_xdp_buff(struct mlx5e_xdpsq *sq, struct mlx5e_rq *rq,
 	if (unlikely(!xdpf))
 		return false;
 
+	if (unlikely(xdp_frame_has_frags(xdpf))) {
+		xdp_return_frame(xdpf);
+		return false;
+	}
+
 	xdptxd.data = xdpf->data;
 	xdptxd.len  = xdpf->len;
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index dd8ff62e16939..e9ad5b3a30ed5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -1572,6 +1572,7 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe,
 	struct mlx5e_dma_info *di = wi->di;
 	struct skb_shared_info *sinfo;
 	u32 frag_consumed_bytes;
+	struct bpf_prog *prog;
 	struct xdp_buff xdp;
 	struct sk_buff *skb;
 	u32 truesize;
@@ -1582,6 +1583,7 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe,
 
 	dma_sync_single_range_for_cpu(rq->pdev, di->addr, wi->offset,
 				      rq->buff.frame0_sz, DMA_FROM_DEVICE);
+	net_prefetchw(va); /* xdp_frame data area */
 	net_prefetch(va + rx_headroom);
 
 	mlx5e_fill_xdp_buff(rq, va, rx_headroom, frag_consumed_bytes, &xdp);
@@ -1629,6 +1631,17 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe,
 
 	di = head_wi->di;
 
+	prog = rcu_dereference(rq->xdp_prog);
+	if (prog && mlx5e_xdp_handle(rq, di, prog, &xdp)) {
+		if (test_bit(MLX5E_RQ_FLAG_XDP_XMIT, rq->flags)) {
+			int i;
+
+			for (i = wi - head_wi; i < rq->wqe.info.num_frags; i++)
+				mlx5e_put_rx_frag(rq, &head_wi[i], true);
+		}
+		return NULL; /* page/packet was consumed by XDP */
+	}
+
 	skb = mlx5e_build_linear_skb(rq, xdp.data_hard_start, rq->buff.frame0_sz,
 				     xdp.data - xdp.data_hard_start,
 				     xdp.data_end - xdp.data,

From ddc87e7d477552084fd185b315a39d5067b01773 Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Fri, 28 Jan 2022 13:13:11 +0200
Subject: [PATCH 05/15] net/mlx5e: Store DMA address inside struct page

Use page_pool_set_dma_addr() to store the DMA address of a page inside
struct page, in order to avoid passing struct mlx5e_dma_info to XDP
handlers. Previously, struct mlx5e_dma_info was used to pass both the
DMA address and the page, and it worked well for the single-fragment
case.

When XDP multi buffer is in use, and a fragmented xdp_frame has to be
transmitted, the driver needs to know the DMA addresses of fragments,
however, the array of fragments in struct skb_shared_info doesn't
contain them. In order to pass the DMA addresses, the driver puts them
into struct page itself, which is accessible from the array of fragments
in struct skb_shared_info. The existing XDP handlers are modified to
remove the dependency on struct mlx5e_dma_info.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en.h  |  2 +-
 .../net/ethernet/mellanox/mlx5/core/en/txrx.h |  6 +--
 .../net/ethernet/mellanox/mlx5/core/en/xdp.c  | 14 +++----
 .../net/ethernet/mellanox/mlx5/core/en/xdp.h  |  2 +-
 .../net/ethernet/mellanox/mlx5/core/en_main.c |  2 +-
 .../net/ethernet/mellanox/mlx5/core/en_rx.c   | 40 ++++++++++---------
 6 files changed, 33 insertions(+), 33 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 2704c7537481c..f5b2449fa15a1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -515,7 +515,7 @@ struct mlx5e_xdp_info {
 		} frame;
 		struct {
 			struct mlx5e_rq *rq;
-			struct mlx5e_dma_info di;
+			struct page *page;
 		} page;
 	};
 };
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
index 210d23bf3701d..c208ea307bffb 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
@@ -44,10 +44,8 @@ int mlx5e_napi_poll(struct napi_struct *napi, int budget);
 int mlx5e_poll_ico_cq(struct mlx5e_cq *cq);
 
 /* RX */
-void mlx5e_page_dma_unmap(struct mlx5e_rq *rq, struct mlx5e_dma_info *dma_info);
-void mlx5e_page_release_dynamic(struct mlx5e_rq *rq,
-				struct mlx5e_dma_info *dma_info,
-				bool recycle);
+void mlx5e_page_dma_unmap(struct mlx5e_rq *rq, struct page *page);
+void mlx5e_page_release_dynamic(struct mlx5e_rq *rq, struct page *page, bool recycle);
 INDIRECT_CALLABLE_DECLARE(bool mlx5e_post_rx_wqes(struct mlx5e_rq *rq));
 INDIRECT_CALLABLE_DECLARE(bool mlx5e_post_rx_mpwqes(struct mlx5e_rq *rq));
 int mlx5e_poll_rx_cq(struct mlx5e_cq *cq, int budget);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
index 3a837030e96ee..91dd5c59657b7 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
@@ -57,7 +57,7 @@ int mlx5e_xdp_max_mtu(struct mlx5e_params *params, struct mlx5e_xsk_param *xsk)
 
 static inline bool
 mlx5e_xmit_xdp_buff(struct mlx5e_xdpsq *sq, struct mlx5e_rq *rq,
-		    struct mlx5e_dma_info *di, struct xdp_buff *xdp)
+		    struct page *page, struct xdp_buff *xdp)
 {
 	struct mlx5e_xmit_data xdptxd;
 	struct mlx5e_xdp_info xdpi;
@@ -110,13 +110,13 @@ mlx5e_xmit_xdp_buff(struct mlx5e_xdpsq *sq, struct mlx5e_rq *rq,
 
 		xdpi.mode = MLX5E_XDP_XMIT_MODE_PAGE;
 
-		dma_addr = di->addr + (xdpf->data - (void *)xdpf);
+		dma_addr = page_pool_get_dma_addr(page) + (xdpf->data - (void *)xdpf);
 		dma_sync_single_for_device(sq->pdev, dma_addr, xdptxd.len,
 					   DMA_TO_DEVICE);
 
 		xdptxd.dma_addr = dma_addr;
 		xdpi.page.rq    = rq;
-		xdpi.page.di    = *di;
+		xdpi.page.page = page;
 	}
 
 	return INDIRECT_CALL_2(sq->xmit_xdp_frame, mlx5e_xmit_xdp_frame_mpwqe,
@@ -124,7 +124,7 @@ mlx5e_xmit_xdp_buff(struct mlx5e_xdpsq *sq, struct mlx5e_rq *rq,
 }
 
 /* returns true if packet was consumed by xdp */
-bool mlx5e_xdp_handle(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
+bool mlx5e_xdp_handle(struct mlx5e_rq *rq, struct page *page,
 		      struct bpf_prog *prog, struct xdp_buff *xdp)
 {
 	u32 act;
@@ -135,7 +135,7 @@ bool mlx5e_xdp_handle(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
 	case XDP_PASS:
 		return false;
 	case XDP_TX:
-		if (unlikely(!mlx5e_xmit_xdp_buff(rq->xdpsq, rq, di, xdp)))
+		if (unlikely(!mlx5e_xmit_xdp_buff(rq->xdpsq, rq, page, xdp)))
 			goto xdp_abort;
 		__set_bit(MLX5E_RQ_FLAG_XDP_XMIT, rq->flags); /* non-atomic */
 		return true;
@@ -147,7 +147,7 @@ bool mlx5e_xdp_handle(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
 		__set_bit(MLX5E_RQ_FLAG_XDP_XMIT, rq->flags);
 		__set_bit(MLX5E_RQ_FLAG_XDP_REDIRECT, rq->flags);
 		if (xdp->rxq->mem.type != MEM_TYPE_XSK_BUFF_POOL)
-			mlx5e_page_dma_unmap(rq, di);
+			mlx5e_page_dma_unmap(rq, page);
 		rq->stats->xdp_redirect++;
 		return true;
 	default:
@@ -384,7 +384,7 @@ static void mlx5e_free_xdpsq_desc(struct mlx5e_xdpsq *sq,
 			break;
 		case MLX5E_XDP_XMIT_MODE_PAGE:
 			/* XDP_TX from the regular RQ */
-			mlx5e_page_release_dynamic(xdpi.page.rq, &xdpi.page.di, recycle);
+			mlx5e_page_release_dynamic(xdpi.page.rq, xdpi.page.page, recycle);
 			break;
 		case MLX5E_XDP_XMIT_MODE_XSK:
 			/* AF_XDP send */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
index 20d8af66c0724..8a92cf0079912 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
@@ -47,7 +47,7 @@
 
 struct mlx5e_xsk_param;
 int mlx5e_xdp_max_mtu(struct mlx5e_params *params, struct mlx5e_xsk_param *xsk);
-bool mlx5e_xdp_handle(struct mlx5e_rq *rq, struct mlx5e_dma_info *di,
+bool mlx5e_xdp_handle(struct mlx5e_rq *rq, struct page *page,
 		      struct bpf_prog *prog, struct xdp_buff *xdp);
 void mlx5e_xdp_mpwqe_complete(struct mlx5e_xdpsq *sq);
 bool mlx5e_poll_xdpsq_cq(struct mlx5e_cq *cq);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 91b90bbb2b286..f21cae712ce5d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -780,7 +780,7 @@ static void mlx5e_free_rq(struct mlx5e_rq *rq)
 		 * entered, and it's safe to call mlx5e_page_release_dynamic
 		 * directly.
 		 */
-		mlx5e_page_release_dynamic(rq, dma_info, false);
+		mlx5e_page_release_dynamic(rq, dma_info->page, false);
 	}
 
 	xdp_rxq_info_unreg(&rq->xdp_rxq);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index e9ad5b3a30ed5..56bb58704bf98 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -222,8 +222,7 @@ static inline u32 mlx5e_decompress_cqes_start(struct mlx5e_rq *rq,
 	return mlx5e_decompress_cqes_cont(rq, wq, 1, budget_rem) - 1;
 }
 
-static inline bool mlx5e_rx_cache_put(struct mlx5e_rq *rq,
-				      struct mlx5e_dma_info *dma_info)
+static inline bool mlx5e_rx_cache_put(struct mlx5e_rq *rq, struct page *page)
 {
 	struct mlx5e_page_cache *cache = &rq->page_cache;
 	u32 tail_next = (cache->tail + 1) & (MLX5E_CACHE_SIZE - 1);
@@ -234,12 +233,13 @@ static inline bool mlx5e_rx_cache_put(struct mlx5e_rq *rq,
 		return false;
 	}
 
-	if (!dev_page_is_reusable(dma_info->page)) {
+	if (!dev_page_is_reusable(page)) {
 		stats->cache_waive++;
 		return false;
 	}
 
-	cache->page_cache[cache->tail] = *dma_info;
+	cache->page_cache[cache->tail].page = page;
+	cache->page_cache[cache->tail].addr = page_pool_get_dma_addr(page);
 	cache->tail = tail_next;
 	return true;
 }
@@ -287,6 +287,7 @@ static inline int mlx5e_page_alloc_pool(struct mlx5e_rq *rq,
 		dma_info->page = NULL;
 		return -ENOMEM;
 	}
+	page_pool_set_dma_addr(dma_info->page, dma_info->addr);
 
 	return 0;
 }
@@ -300,26 +301,27 @@ static inline int mlx5e_page_alloc(struct mlx5e_rq *rq,
 		return mlx5e_page_alloc_pool(rq, dma_info);
 }
 
-void mlx5e_page_dma_unmap(struct mlx5e_rq *rq, struct mlx5e_dma_info *dma_info)
+void mlx5e_page_dma_unmap(struct mlx5e_rq *rq, struct page *page)
 {
-	dma_unmap_page_attrs(rq->pdev, dma_info->addr, PAGE_SIZE, rq->buff.map_dir,
+	dma_addr_t dma_addr = page_pool_get_dma_addr(page);
+
+	dma_unmap_page_attrs(rq->pdev, dma_addr, PAGE_SIZE, rq->buff.map_dir,
 			     DMA_ATTR_SKIP_CPU_SYNC);
+	page_pool_set_dma_addr(page, 0);
 }
 
-void mlx5e_page_release_dynamic(struct mlx5e_rq *rq,
-				struct mlx5e_dma_info *dma_info,
-				bool recycle)
+void mlx5e_page_release_dynamic(struct mlx5e_rq *rq, struct page *page, bool recycle)
 {
 	if (likely(recycle)) {
-		if (mlx5e_rx_cache_put(rq, dma_info))
+		if (mlx5e_rx_cache_put(rq, page))
 			return;
 
-		mlx5e_page_dma_unmap(rq, dma_info);
-		page_pool_recycle_direct(rq->page_pool, dma_info->page);
+		mlx5e_page_dma_unmap(rq, page);
+		page_pool_recycle_direct(rq->page_pool, page);
 	} else {
-		mlx5e_page_dma_unmap(rq, dma_info);
-		page_pool_release_page(rq->page_pool, dma_info->page);
-		put_page(dma_info->page);
+		mlx5e_page_dma_unmap(rq, page);
+		page_pool_release_page(rq->page_pool, page);
+		put_page(page);
 	}
 }
 
@@ -334,7 +336,7 @@ static inline void mlx5e_page_release(struct mlx5e_rq *rq,
 		 */
 		xsk_buff_free(dma_info->xsk);
 	else
-		mlx5e_page_release_dynamic(rq, dma_info, recycle);
+		mlx5e_page_release_dynamic(rq, dma_info->page, recycle);
 }
 
 static inline int mlx5e_get_rx_frag(struct mlx5e_rq *rq,
@@ -1544,7 +1546,7 @@ mlx5e_skb_from_cqe_linear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe,
 
 		net_prefetchw(va); /* xdp_frame data area */
 		mlx5e_fill_xdp_buff(rq, va, rx_headroom, cqe_bcnt, &xdp);
-		if (mlx5e_xdp_handle(rq, di, prog, &xdp))
+		if (mlx5e_xdp_handle(rq, di->page, prog, &xdp))
 			return NULL; /* page/packet was consumed by XDP */
 
 		rx_headroom = xdp.data - xdp.data_hard_start;
@@ -1632,7 +1634,7 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe,
 	di = head_wi->di;
 
 	prog = rcu_dereference(rq->xdp_prog);
-	if (prog && mlx5e_xdp_handle(rq, di, prog, &xdp)) {
+	if (prog && mlx5e_xdp_handle(rq, di->page, prog, &xdp)) {
 		if (test_bit(MLX5E_RQ_FLAG_XDP_XMIT, rq->flags)) {
 			int i;
 
@@ -1934,7 +1936,7 @@ mlx5e_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi,
 
 		net_prefetchw(va); /* xdp_frame data area */
 		mlx5e_fill_xdp_buff(rq, va, rx_headroom, cqe_bcnt, &xdp);
-		if (mlx5e_xdp_handle(rq, di, prog, &xdp)) {
+		if (mlx5e_xdp_handle(rq, di->page, prog, &xdp)) {
 			if (__test_and_clear_bit(MLX5E_RQ_FLAG_XDP_XMIT, rq->flags))
 				__set_bit(page_idx, wi->xdp_xmit_bitmap); /* non-atomic */
 			return NULL; /* page/packet was consumed by XDP */

From 49529a1726850bddd7a991de6e8ef52c65c49fbe Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Fri, 28 Jan 2022 16:43:14 +0200
Subject: [PATCH 06/15] net/mlx5e: Move mlx5e_xdpi_fifo_push out of
 xmit_xdp_frame

The implementations of xmit_xdp_frame get the xdpi parameter of type
struct mlx5e_xdp_info for the sole purpose of calling
mlx5e_xdpi_fifo_push() on success.

This commit moves this call outside of xmit_xdp_frame, shifting this
responsibility to the caller. It will allow more fine-grained handling
of XDP info for cases when an xdp_frame is fragmented.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en.h    |  1 -
 .../net/ethernet/mellanox/mlx5/core/en/xdp.c    | 17 ++++++++++-------
 .../net/ethernet/mellanox/mlx5/core/en/xdp.h    |  2 --
 .../net/ethernet/mellanox/mlx5/core/en/xsk/tx.c |  4 +++-
 4 files changed, 13 insertions(+), 11 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index f5b2449fa15a1..d084f930bb370 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -537,7 +537,6 @@ struct mlx5e_xdpsq;
 typedef int (*mlx5e_fp_xmit_xdp_frame_check)(struct mlx5e_xdpsq *);
 typedef bool (*mlx5e_fp_xmit_xdp_frame)(struct mlx5e_xdpsq *,
 					struct mlx5e_xmit_data *,
-					struct mlx5e_xdp_info *,
 					int);
 
 struct mlx5e_xdpsq {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
index 91dd5c59657b7..b1114f8540572 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
@@ -119,8 +119,12 @@ mlx5e_xmit_xdp_buff(struct mlx5e_xdpsq *sq, struct mlx5e_rq *rq,
 		xdpi.page.page = page;
 	}
 
-	return INDIRECT_CALL_2(sq->xmit_xdp_frame, mlx5e_xmit_xdp_frame_mpwqe,
-			       mlx5e_xmit_xdp_frame, sq, &xdptxd, &xdpi, 0);
+	if (unlikely(!INDIRECT_CALL_2(sq->xmit_xdp_frame, mlx5e_xmit_xdp_frame_mpwqe,
+				      mlx5e_xmit_xdp_frame, sq, &xdptxd, 0)))
+		return false;
+
+	mlx5e_xdpi_fifo_push(&sq->db.xdpi_fifo, &xdpi);
+	return true;
 }
 
 /* returns true if packet was consumed by xdp */
@@ -261,7 +265,7 @@ INDIRECT_CALLABLE_SCOPE int mlx5e_xmit_xdp_frame_check_mpwqe(struct mlx5e_xdpsq
 
 INDIRECT_CALLABLE_SCOPE bool
 mlx5e_xmit_xdp_frame_mpwqe(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
-			   struct mlx5e_xdp_info *xdpi, int check_result)
+			   int check_result)
 {
 	struct mlx5e_tx_mpwqe *session = &sq->mpwqe;
 	struct mlx5e_xdpsq_stats *stats = sq->stats;
@@ -289,7 +293,6 @@ mlx5e_xmit_xdp_frame_mpwqe(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptx
 	if (unlikely(mlx5e_xdp_mpqwe_is_full(session, sq->max_sq_mpw_wqebbs)))
 		mlx5e_xdp_mpwqe_complete(sq);
 
-	mlx5e_xdpi_fifo_push(&sq->db.xdpi_fifo, xdpi);
 	stats->xmit++;
 	return true;
 }
@@ -308,7 +311,7 @@ INDIRECT_CALLABLE_SCOPE int mlx5e_xmit_xdp_frame_check(struct mlx5e_xdpsq *sq)
 
 INDIRECT_CALLABLE_SCOPE bool
 mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
-		     struct mlx5e_xdp_info *xdpi, int check_result)
+		     int check_result)
 {
 	struct mlx5_wq_cyc       *wq   = &sq->wq;
 	u16                       pi   = mlx5_wq_cyc_ctr2ix(wq, sq->pc);
@@ -358,7 +361,6 @@ mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
 
 	sq->doorbell_cseg = cseg;
 
-	mlx5e_xdpi_fifo_push(&sq->db.xdpi_fifo, xdpi);
 	stats->xmit++;
 	return true;
 }
@@ -537,12 +539,13 @@ int mlx5e_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **frames,
 		xdpi.frame.dma_addr = xdptxd.dma_addr;
 
 		ret = INDIRECT_CALL_2(sq->xmit_xdp_frame, mlx5e_xmit_xdp_frame_mpwqe,
-				      mlx5e_xmit_xdp_frame, sq, &xdptxd, &xdpi, 0);
+				      mlx5e_xmit_xdp_frame, sq, &xdptxd, 0);
 		if (unlikely(!ret)) {
 			dma_unmap_single(sq->pdev, xdptxd.dma_addr,
 					 xdptxd.len, DMA_TO_DEVICE);
 			break;
 		}
+		mlx5e_xdpi_fifo_push(&sq->db.xdpi_fifo, &xdpi);
 		nxmit++;
 	}
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
index 8a92cf0079912..ce31828b6c19d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
@@ -59,11 +59,9 @@ int mlx5e_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **frames,
 
 INDIRECT_CALLABLE_DECLARE(bool mlx5e_xmit_xdp_frame_mpwqe(struct mlx5e_xdpsq *sq,
 							  struct mlx5e_xmit_data *xdptxd,
-							  struct mlx5e_xdp_info *xdpi,
 							  int check_result));
 INDIRECT_CALLABLE_DECLARE(bool mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq,
 						    struct mlx5e_xmit_data *xdptxd,
-						    struct mlx5e_xdp_info *xdpi,
 						    int check_result));
 INDIRECT_CALLABLE_DECLARE(int mlx5e_xmit_xdp_frame_check_mpwqe(struct mlx5e_xdpsq *sq));
 INDIRECT_CALLABLE_DECLARE(int mlx5e_xmit_xdp_frame_check(struct mlx5e_xdpsq *sq));
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c
index 8e96260fce1df..5a889835039c3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c
@@ -103,12 +103,14 @@ bool mlx5e_xsk_tx(struct mlx5e_xdpsq *sq, unsigned int budget)
 		xsk_buff_raw_dma_sync_for_device(pool, xdptxd.dma_addr, xdptxd.len);
 
 		ret = INDIRECT_CALL_2(sq->xmit_xdp_frame, mlx5e_xmit_xdp_frame_mpwqe,
-				      mlx5e_xmit_xdp_frame, sq, &xdptxd, &xdpi, check_result);
+				      mlx5e_xmit_xdp_frame, sq, &xdptxd, check_result);
 		if (unlikely(!ret)) {
 			if (sq->mpwqe.wqe)
 				mlx5e_xdp_mpwqe_complete(sq);
 
 			mlx5e_xsk_tx_post_err(sq, &xdpi);
+		} else {
+			mlx5e_xdpi_fifo_push(&sq->db.xdpi_fifo, &xdpi);
 		}
 
 		flush = true;

From c090451633f89c292316e928f970d082d7e8295f Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Mon, 31 Jan 2022 17:28:40 +0200
Subject: [PATCH 07/15] net/mlx5e: Remove assignment of inline_hdr.sz on XDP TX

When MPWQE is disabled, mlx5e_open_xdpsq prefills the common fields of
WQEs in the XDP SQ to save time when sending packets. One of such fields
is eseg->inline_hdr.sz, which can be either 0 or MLX5E_XDP_MIN_INLINE,
depending on the inline mode of the SQ.

The inline mode can't change during the lifetime of the SQ, so setting
this field again in mlx5e_xmit_xdp_frame is redundant. Moreover, the
xmit function only sets it to MLX5E_XDP_MIN_INLINE, but not to 0 in the
other case.

This commit removes the redundant assignment in mlx5e_xmit_xdp_frame.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
index b1114f8540572..9478df6c87bc5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
@@ -343,7 +343,6 @@ mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
 	/* copy the inline part if required */
 	if (sq->min_inline_mode != MLX5_INLINE_MODE_NONE) {
 		memcpy(eseg->inline_hdr.start, xdptxd->data, sizeof(eseg->inline_hdr.start));
-		eseg->inline_hdr.sz = cpu_to_be16(MLX5E_XDP_MIN_INLINE);
 		memcpy(dseg, xdptxd->data + sizeof(eseg->inline_hdr.start),
 		       MLX5E_XDP_MIN_INLINE - sizeof(eseg->inline_hdr.start));
 		dma_len  -= MLX5E_XDP_MIN_INLINE;

From 9ded70fa1d8186135090f09638db7d8126dca0db Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Mon, 31 Jan 2022 17:41:31 +0200
Subject: [PATCH 08/15] net/mlx5e: Don't prefill WQEs in XDP SQ in the multi
 buffer mode

When MPWQE is disabled, mlx5e_open_xdpsq() prefills the common fields of
WQEs in the XDP SQ to save time when sending packets.
mlx5e_xmit_xdp_frame() runs on the prefilled fields, however, sending
multi buffer XDP frames would require changing some of these fields on a
per-packet basis. Besides that, mlx5e_xmit_xdp_frame() will be used as a
fallback to send multi buffer XDP frames when MPWQE is enabled (MPWQE
can only handle linear packets).

In order to prepare for XDP multi buffer support, this commit introduces
a mode for mlx5e_xmit_xdp_frame() that fills all the fields itself.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en.h  |  1 +
 .../ethernet/mellanox/mlx5/core/en/params.c   |  4 ++-
 .../ethernet/mellanox/mlx5/core/en/params.h   |  2 ++
 .../net/ethernet/mellanox/mlx5/core/en/xdp.c  | 32 +++++++++++++++++--
 .../mellanox/mlx5/core/en/xsk/setup.c         |  2 +-
 .../net/ethernet/mellanox/mlx5/core/en_main.c | 10 +++++-
 6 files changed, 46 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index d084f930bb370..cf935a7bf3879 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -405,6 +405,7 @@ enum {
 	MLX5E_SQ_STATE_VLAN_NEED_L2_INLINE,
 	MLX5E_SQ_STATE_PENDING_XSK_TX,
 	MLX5E_SQ_STATE_PENDING_TLS_RX_RESYNC,
+	MLX5E_SQ_STATE_XDP_MULTIBUF,
 };
 
 struct mlx5e_tx_mpwqe {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
index 9646867872c14..08fd1370a8b0a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
@@ -843,6 +843,7 @@ static void mlx5e_build_async_icosq_param(struct mlx5_core_dev *mdev,
 
 void mlx5e_build_xdpsq_param(struct mlx5_core_dev *mdev,
 			     struct mlx5e_params *params,
+			     struct mlx5e_xsk_param *xsk,
 			     struct mlx5e_sq_param *param)
 {
 	void *sqc = param->sqc;
@@ -851,6 +852,7 @@ void mlx5e_build_xdpsq_param(struct mlx5_core_dev *mdev,
 	mlx5e_build_sq_param_common(mdev, param);
 	MLX5_SET(wq, wq, log_wq_sz, params->log_sq_size);
 	param->is_mpw = MLX5E_GET_PFLAG(params, MLX5E_PFLAG_XDP_TX_MPWQE);
+	param->is_xdp_mb = !mlx5e_rx_is_linear_skb(params, xsk);
 	mlx5e_build_tx_cq_param(mdev, params, &param->cqp);
 }
 
@@ -870,7 +872,7 @@ int mlx5e_build_channel_param(struct mlx5_core_dev *mdev,
 	async_icosq_log_wq_sz = mlx5e_build_async_icosq_log_wq_sz(mdev);
 
 	mlx5e_build_sq_param(mdev, params, &cparam->txq_sq);
-	mlx5e_build_xdpsq_param(mdev, params, &cparam->xdp_sq);
+	mlx5e_build_xdpsq_param(mdev, params, NULL, &cparam->xdp_sq);
 	mlx5e_build_icosq_param(mdev, icosq_log_wq_sz, &cparam->icosq);
 	mlx5e_build_async_icosq_param(mdev, async_icosq_log_wq_sz, &cparam->async_icosq);
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/params.h b/drivers/net/ethernet/mellanox/mlx5/core/en/params.h
index 47a368112e313..f5c46e78eebc5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/params.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/params.h
@@ -31,6 +31,7 @@ struct mlx5e_sq_param {
 	struct mlx5_wq_param       wq;
 	bool                       is_mpw;
 	bool                       is_tls;
+	bool                       is_xdp_mb;
 	u16                        stop_room;
 };
 
@@ -155,6 +156,7 @@ void mlx5e_build_tx_cq_param(struct mlx5_core_dev *mdev,
 			     struct mlx5e_cq_param *param);
 void mlx5e_build_xdpsq_param(struct mlx5_core_dev *mdev,
 			     struct mlx5e_params *params,
+			     struct mlx5e_xsk_param *xsk,
 			     struct mlx5e_sq_param *param);
 int mlx5e_build_channel_param(struct mlx5_core_dev *mdev,
 			      struct mlx5e_params *params,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
index 9478df6c87bc5..b220e613d102c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
@@ -323,6 +323,7 @@ mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
 
 	dma_addr_t dma_addr = xdptxd->dma_addr;
 	u32 dma_len = xdptxd->len;
+	u16 ds_cnt, inline_hdr_sz;
 
 	struct mlx5e_xdpsq_stats *stats = sq->stats;
 
@@ -338,7 +339,8 @@ mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
 	if (unlikely(check_result < 0))
 		return false;
 
-	cseg->fm_ce_se = 0;
+	ds_cnt = MLX5E_TX_WQE_EMPTY_DS_COUNT + 1;
+	inline_hdr_sz = 0;
 
 	/* copy the inline part if required */
 	if (sq->min_inline_mode != MLX5_INLINE_MODE_NONE) {
@@ -347,7 +349,9 @@ mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
 		       MLX5E_XDP_MIN_INLINE - sizeof(eseg->inline_hdr.start));
 		dma_len  -= MLX5E_XDP_MIN_INLINE;
 		dma_addr += MLX5E_XDP_MIN_INLINE;
+		inline_hdr_sz = MLX5E_XDP_MIN_INLINE;
 		dseg++;
+		ds_cnt++;
 	}
 
 	/* write the dma part */
@@ -356,7 +360,31 @@ mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
 
 	cseg->opmod_idx_opcode = cpu_to_be32((sq->pc << 8) | MLX5_OPCODE_SEND);
 
-	sq->pc++;
+	if (unlikely(test_bit(MLX5E_SQ_STATE_XDP_MULTIBUF, &sq->state))) {
+		u8 num_pkts = 1;
+		u8 num_wqebbs;
+
+		memset(&cseg->signature, 0, sizeof(*cseg) -
+		       sizeof(cseg->opmod_idx_opcode) - sizeof(cseg->qpn_ds));
+		memset(eseg, 0, sizeof(*eseg) - sizeof(eseg->trailer));
+
+		eseg->inline_hdr.sz = cpu_to_be16(inline_hdr_sz);
+		dseg->lkey = sq->mkey_be;
+
+		cseg->qpn_ds = cpu_to_be32((sq->sqn << 8) | ds_cnt);
+
+		num_wqebbs = DIV_ROUND_UP(ds_cnt, MLX5_SEND_WQEBB_NUM_DS);
+		sq->db.wqe_info[pi] = (struct mlx5e_xdp_wqe_info) {
+			.num_wqebbs = num_wqebbs,
+			.num_pkts = num_pkts,
+		};
+
+		sq->pc += num_wqebbs;
+	} else {
+		cseg->fm_ce_se = 0;
+
+		sq->pc++;
+	}
 
 	sq->doorbell_cseg = cseg;
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c
index 25eac9e203423..3ad7f1301fa8d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c
@@ -43,7 +43,7 @@ static void mlx5e_build_xsk_cparam(struct mlx5_core_dev *mdev,
 				   struct mlx5e_channel_param *cparam)
 {
 	mlx5e_build_rq_param(mdev, params, xsk, q_counter, &cparam->rq);
-	mlx5e_build_xdpsq_param(mdev, params, &cparam->xdp_sq);
+	mlx5e_build_xdpsq_param(mdev, params, xsk, &cparam->xdp_sq);
 }
 
 static int mlx5e_init_xsk_rq(struct mlx5e_channel *c,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index f21cae712ce5d..95cec28486855 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -1666,13 +1666,21 @@ int mlx5e_open_xdpsq(struct mlx5e_channel *c, struct mlx5e_params *params,
 	csp.wq_ctrl         = &sq->wq_ctrl;
 	csp.min_inline_mode = sq->min_inline_mode;
 	set_bit(MLX5E_SQ_STATE_ENABLED, &sq->state);
+
+	/* Don't enable multi buffer on XDP_REDIRECT SQ, as it's not yet
+	 * supported by upstream, and there is no defined trigger to allow
+	 * transmitting redirected multi-buffer frames.
+	 */
+	if (param->is_xdp_mb && !is_redirect)
+		set_bit(MLX5E_SQ_STATE_XDP_MULTIBUF, &sq->state);
+
 	err = mlx5e_create_sq_rdy(c->mdev, param, &csp, 0, &sq->sqn);
 	if (err)
 		goto err_free_xdpsq;
 
 	mlx5e_set_xmit_fp(sq, param->is_mpw);
 
-	if (!param->is_mpw) {
+	if (!param->is_mpw && !test_bit(MLX5E_SQ_STATE_XDP_MULTIBUF, &sq->state)) {
 		unsigned int ds_cnt = MLX5E_XDP_TX_DS_COUNT;
 		unsigned int inline_hdr_sz = 0;
 		int i;

From 39a1665d16a2adba6c0b05019068749af3cd05ee Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Mon, 31 Jan 2022 19:27:15 +0200
Subject: [PATCH 09/15] net/mlx5e: Implement sending multi buffer XDP frames

xmit_xdp_frame is extended to support sending fragmented XDP frames. The
next commit will start using this functionality.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en.h  |  1 +
 .../net/ethernet/mellanox/mlx5/core/en/xdp.c  | 96 +++++++++++++++----
 .../net/ethernet/mellanox/mlx5/core/en/xdp.h  |  2 +
 .../ethernet/mellanox/mlx5/core/en/xsk/tx.c   |  3 +-
 4 files changed, 80 insertions(+), 22 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index cf935a7bf3879..8653ac0fd865c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -538,6 +538,7 @@ struct mlx5e_xdpsq;
 typedef int (*mlx5e_fp_xmit_xdp_frame_check)(struct mlx5e_xdpsq *);
 typedef bool (*mlx5e_fp_xmit_xdp_frame)(struct mlx5e_xdpsq *,
 					struct mlx5e_xmit_data *,
+					struct skb_shared_info *,
 					int);
 
 struct mlx5e_xdpsq {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
index b220e613d102c..52e0f0028c35a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
@@ -120,7 +120,7 @@ mlx5e_xmit_xdp_buff(struct mlx5e_xdpsq *sq, struct mlx5e_rq *rq,
 	}
 
 	if (unlikely(!INDIRECT_CALL_2(sq->xmit_xdp_frame, mlx5e_xmit_xdp_frame_mpwqe,
-				      mlx5e_xmit_xdp_frame, sq, &xdptxd, 0)))
+				      mlx5e_xmit_xdp_frame, sq, &xdptxd, NULL, 0)))
 		return false;
 
 	mlx5e_xdpi_fifo_push(&sq->db.xdpi_fifo, &xdpi);
@@ -263,13 +263,27 @@ INDIRECT_CALLABLE_SCOPE int mlx5e_xmit_xdp_frame_check_mpwqe(struct mlx5e_xdpsq
 	return MLX5E_XDP_CHECK_OK;
 }
 
+INDIRECT_CALLABLE_SCOPE bool
+mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
+		     struct skb_shared_info *sinfo, int check_result);
+
 INDIRECT_CALLABLE_SCOPE bool
 mlx5e_xmit_xdp_frame_mpwqe(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
-			   int check_result)
+			   struct skb_shared_info *sinfo, int check_result)
 {
 	struct mlx5e_tx_mpwqe *session = &sq->mpwqe;
 	struct mlx5e_xdpsq_stats *stats = sq->stats;
 
+	if (unlikely(sinfo)) {
+		/* MPWQE is enabled, but a multi-buffer packet is queued for
+		 * transmission. MPWQE can't send fragmented packets, so close
+		 * the current session and fall back to a regular WQE.
+		 */
+		if (unlikely(sq->mpwqe.wqe))
+			mlx5e_xdp_mpwqe_complete(sq);
+		return mlx5e_xmit_xdp_frame(sq, xdptxd, sinfo, 0);
+	}
+
 	if (unlikely(xdptxd->len > sq->hw_mtu)) {
 		stats->err++;
 		return false;
@@ -297,9 +311,9 @@ mlx5e_xmit_xdp_frame_mpwqe(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptx
 	return true;
 }
 
-INDIRECT_CALLABLE_SCOPE int mlx5e_xmit_xdp_frame_check(struct mlx5e_xdpsq *sq)
+static int mlx5e_xmit_xdp_frame_check_stop_room(struct mlx5e_xdpsq *sq, int stop_room)
 {
-	if (unlikely(!mlx5e_wqc_has_room_for(&sq->wq, sq->cc, sq->pc, 1))) {
+	if (unlikely(!mlx5e_wqc_has_room_for(&sq->wq, sq->cc, sq->pc, stop_room))) {
 		/* SQ is full, ring doorbell */
 		mlx5e_xmit_xdp_doorbell(sq);
 		sq->stats->full++;
@@ -309,37 +323,66 @@ INDIRECT_CALLABLE_SCOPE int mlx5e_xmit_xdp_frame_check(struct mlx5e_xdpsq *sq)
 	return MLX5E_XDP_CHECK_OK;
 }
 
+INDIRECT_CALLABLE_SCOPE int mlx5e_xmit_xdp_frame_check(struct mlx5e_xdpsq *sq)
+{
+	return mlx5e_xmit_xdp_frame_check_stop_room(sq, 1);
+}
+
 INDIRECT_CALLABLE_SCOPE bool
 mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
-		     int check_result)
+		     struct skb_shared_info *sinfo, int check_result)
 {
 	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 mlx5_wqe_ctrl_seg *cseg = &wqe->ctrl;
-	struct mlx5_wqe_eth_seg  *eseg = &wqe->eth;
-	struct mlx5_wqe_data_seg *dseg = wqe->data;
+	struct mlx5_wqe_ctrl_seg *cseg;
+	struct mlx5_wqe_data_seg *dseg;
+	struct mlx5_wqe_eth_seg *eseg;
+	struct mlx5e_tx_wqe *wqe;
 
 	dma_addr_t dma_addr = xdptxd->dma_addr;
 	u32 dma_len = xdptxd->len;
 	u16 ds_cnt, inline_hdr_sz;
+	u8 num_wqebbs = 1;
+	int num_frags = 0;
+	u16 pi;
 
 	struct mlx5e_xdpsq_stats *stats = sq->stats;
 
-	net_prefetchw(wqe);
-
 	if (unlikely(dma_len < MLX5E_XDP_MIN_INLINE || sq->hw_mtu < dma_len)) {
 		stats->err++;
 		return false;
 	}
 
-	if (!check_result)
-		check_result = mlx5e_xmit_xdp_frame_check(sq);
+	ds_cnt = MLX5E_TX_WQE_EMPTY_DS_COUNT + 1;
+	if (sq->min_inline_mode != MLX5_INLINE_MODE_NONE)
+		ds_cnt++;
+
+	/* check_result must be 0 if sinfo is passed. */
+	if (!check_result) {
+		int stop_room = 1;
+
+		if (unlikely(sinfo)) {
+			ds_cnt += sinfo->nr_frags;
+			num_frags = sinfo->nr_frags;
+			num_wqebbs = DIV_ROUND_UP(ds_cnt, MLX5_SEND_WQEBB_NUM_DS);
+			/* Assuming MLX5_CAP_GEN(mdev, max_wqe_sz_sq) is big
+			 * enough to hold all fragments.
+			 */
+			stop_room = MLX5E_STOP_ROOM(num_wqebbs);
+		}
+
+		check_result = mlx5e_xmit_xdp_frame_check_stop_room(sq, stop_room);
+	}
 	if (unlikely(check_result < 0))
 		return false;
 
-	ds_cnt = MLX5E_TX_WQE_EMPTY_DS_COUNT + 1;
+	pi = mlx5e_xdpsq_get_next_pi(sq, num_wqebbs);
+	wqe = mlx5_wq_cyc_get_wqe(wq, pi);
+	net_prefetchw(wqe);
+
+	cseg = &wqe->ctrl;
+	eseg = &wqe->eth;
+	dseg = wqe->data;
+
 	inline_hdr_sz = 0;
 
 	/* copy the inline part if required */
@@ -351,7 +394,6 @@ mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
 		dma_addr += MLX5E_XDP_MIN_INLINE;
 		inline_hdr_sz = MLX5E_XDP_MIN_INLINE;
 		dseg++;
-		ds_cnt++;
 	}
 
 	/* write the dma part */
@@ -361,8 +403,8 @@ mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
 	cseg->opmod_idx_opcode = cpu_to_be32((sq->pc << 8) | MLX5_OPCODE_SEND);
 
 	if (unlikely(test_bit(MLX5E_SQ_STATE_XDP_MULTIBUF, &sq->state))) {
-		u8 num_pkts = 1;
-		u8 num_wqebbs;
+		u8 num_pkts = 1 + num_frags;
+		int i;
 
 		memset(&cseg->signature, 0, sizeof(*cseg) -
 		       sizeof(cseg->opmod_idx_opcode) - sizeof(cseg->qpn_ds));
@@ -371,9 +413,21 @@ mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptxd,
 		eseg->inline_hdr.sz = cpu_to_be16(inline_hdr_sz);
 		dseg->lkey = sq->mkey_be;
 
+		for (i = 0; i < num_frags; i++) {
+			skb_frag_t *frag = &sinfo->frags[i];
+			dma_addr_t addr;
+
+			addr = page_pool_get_dma_addr(skb_frag_page(frag)) +
+				skb_frag_off(frag);
+
+			dseg++;
+			dseg->addr = cpu_to_be64(addr);
+			dseg->byte_count = cpu_to_be32(skb_frag_size(frag));
+			dseg->lkey = sq->mkey_be;
+		}
+
 		cseg->qpn_ds = cpu_to_be32((sq->sqn << 8) | ds_cnt);
 
-		num_wqebbs = DIV_ROUND_UP(ds_cnt, MLX5_SEND_WQEBB_NUM_DS);
 		sq->db.wqe_info[pi] = (struct mlx5e_xdp_wqe_info) {
 			.num_wqebbs = num_wqebbs,
 			.num_pkts = num_pkts,
@@ -566,7 +620,7 @@ int mlx5e_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **frames,
 		xdpi.frame.dma_addr = xdptxd.dma_addr;
 
 		ret = INDIRECT_CALL_2(sq->xmit_xdp_frame, mlx5e_xmit_xdp_frame_mpwqe,
-				      mlx5e_xmit_xdp_frame, sq, &xdptxd, 0);
+				      mlx5e_xmit_xdp_frame, sq, &xdptxd, NULL, 0);
 		if (unlikely(!ret)) {
 			dma_unmap_single(sq->pdev, xdptxd.dma_addr,
 					 xdptxd.len, DMA_TO_DEVICE);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
index ce31828b6c19d..4868fa1b82d16 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
@@ -59,9 +59,11 @@ int mlx5e_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **frames,
 
 INDIRECT_CALLABLE_DECLARE(bool mlx5e_xmit_xdp_frame_mpwqe(struct mlx5e_xdpsq *sq,
 							  struct mlx5e_xmit_data *xdptxd,
+							  struct skb_shared_info *sinfo,
 							  int check_result));
 INDIRECT_CALLABLE_DECLARE(bool mlx5e_xmit_xdp_frame(struct mlx5e_xdpsq *sq,
 						    struct mlx5e_xmit_data *xdptxd,
+						    struct skb_shared_info *sinfo,
 						    int check_result));
 INDIRECT_CALLABLE_DECLARE(int mlx5e_xmit_xdp_frame_check_mpwqe(struct mlx5e_xdpsq *sq));
 INDIRECT_CALLABLE_DECLARE(int mlx5e_xmit_xdp_frame_check(struct mlx5e_xdpsq *sq));
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c
index 5a889835039c3..3ec0c17db0104 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c
@@ -103,7 +103,8 @@ bool mlx5e_xsk_tx(struct mlx5e_xdpsq *sq, unsigned int budget)
 		xsk_buff_raw_dma_sync_for_device(pool, xdptxd.dma_addr, xdptxd.len);
 
 		ret = INDIRECT_CALL_2(sq->xmit_xdp_frame, mlx5e_xmit_xdp_frame_mpwqe,
-				      mlx5e_xmit_xdp_frame, sq, &xdptxd, check_result);
+				      mlx5e_xmit_xdp_frame, sq, &xdptxd, NULL,
+				      check_result);
 		if (unlikely(!ret)) {
 			if (sq->mpwqe.wqe)
 				mlx5e_xdp_mpwqe_complete(sq);

From fbeed25bcc45ef6df4b8fc35ec82edd895fc6587 Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Thu, 10 Mar 2022 18:16:17 +0200
Subject: [PATCH 10/15] net/mlx5e: Unindent the else-block in
 mlx5e_xmit_xdp_buff

The next commit will add more indentation levels to mlx5e_xmit_xdp_buff.
To keep indentation minimal, unindent the else-block of the if-statement
by doing an early return.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 .../net/ethernet/mellanox/mlx5/core/en/xdp.c  | 34 +++++++++++--------
 1 file changed, 20 insertions(+), 14 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
index 52e0f0028c35a..368e549496149 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
@@ -101,24 +101,30 @@ mlx5e_xmit_xdp_buff(struct mlx5e_xdpsq *sq, struct mlx5e_rq *rq,
 		xdptxd.dma_addr     = dma_addr;
 		xdpi.frame.xdpf     = xdpf;
 		xdpi.frame.dma_addr = dma_addr;
-	} else {
-		/* Driver assumes that xdp_convert_buff_to_frame returns
-		 * an xdp_frame that points to the same memory region as
-		 * the original xdp_buff. It allows to map the memory only
-		 * once and to use the DMA_BIDIRECTIONAL mode.
-		 */
-
-		xdpi.mode = MLX5E_XDP_XMIT_MODE_PAGE;
 
-		dma_addr = page_pool_get_dma_addr(page) + (xdpf->data - (void *)xdpf);
-		dma_sync_single_for_device(sq->pdev, dma_addr, xdptxd.len,
-					   DMA_TO_DEVICE);
+		if (unlikely(!INDIRECT_CALL_2(sq->xmit_xdp_frame, mlx5e_xmit_xdp_frame_mpwqe,
+					      mlx5e_xmit_xdp_frame, sq, &xdptxd, NULL, 0)))
+			return false;
 
-		xdptxd.dma_addr = dma_addr;
-		xdpi.page.rq    = rq;
-		xdpi.page.page = page;
+		mlx5e_xdpi_fifo_push(&sq->db.xdpi_fifo, &xdpi);
+		return true;
 	}
 
+	/* Driver assumes that xdp_convert_buff_to_frame returns an xdp_frame
+	 * that points to the same memory region as the original xdp_buff. It
+	 * allows to map the memory only once and to use the DMA_BIDIRECTIONAL
+	 * mode.
+	 */
+
+	xdpi.mode = MLX5E_XDP_XMIT_MODE_PAGE;
+
+	dma_addr = page_pool_get_dma_addr(page) + (xdpf->data - (void *)xdpf);
+	dma_sync_single_for_device(sq->pdev, dma_addr, xdptxd.len, DMA_TO_DEVICE);
+
+	xdptxd.dma_addr = dma_addr;
+	xdpi.page.rq = rq;
+	xdpi.page.page = page;
+
 	if (unlikely(!INDIRECT_CALL_2(sq->xmit_xdp_frame, mlx5e_xmit_xdp_frame_mpwqe,
 				      mlx5e_xmit_xdp_frame, sq, &xdptxd, NULL, 0)))
 		return false;

From a48ad58cec18702249a228267dec29d105bbe5a5 Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Mon, 31 Jan 2022 19:43:53 +0200
Subject: [PATCH 11/15] net/mlx5e: Support multi buffer XDP_TX

This commit enables passing multi buffer XDP frames to the TX handlers
on XDP_TX. Fragments are DMA synchronized to the device and queued to
the xdpi_fifo for a subsequent unmapping.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 .../net/ethernet/mellanox/mlx5/core/en/xdp.c  | 39 +++++++++++++++----
 1 file changed, 31 insertions(+), 8 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
index 368e549496149..f35b62ce4c074 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
@@ -59,20 +59,17 @@ static inline bool
 mlx5e_xmit_xdp_buff(struct mlx5e_xdpsq *sq, struct mlx5e_rq *rq,
 		    struct page *page, struct xdp_buff *xdp)
 {
+	struct skb_shared_info *sinfo = NULL;
 	struct mlx5e_xmit_data xdptxd;
 	struct mlx5e_xdp_info xdpi;
 	struct xdp_frame *xdpf;
 	dma_addr_t dma_addr;
+	int i;
 
 	xdpf = xdp_convert_buff_to_frame(xdp);
 	if (unlikely(!xdpf))
 		return false;
 
-	if (unlikely(xdp_frame_has_frags(xdpf))) {
-		xdp_return_frame(xdpf);
-		return false;
-	}
-
 	xdptxd.data = xdpf->data;
 	xdptxd.len  = xdpf->len;
 
@@ -117,19 +114,45 @@ mlx5e_xmit_xdp_buff(struct mlx5e_xdpsq *sq, struct mlx5e_rq *rq,
 	 */
 
 	xdpi.mode = MLX5E_XDP_XMIT_MODE_PAGE;
+	xdpi.page.rq = rq;
 
 	dma_addr = page_pool_get_dma_addr(page) + (xdpf->data - (void *)xdpf);
 	dma_sync_single_for_device(sq->pdev, dma_addr, xdptxd.len, DMA_TO_DEVICE);
 
+	if (unlikely(xdp_frame_has_frags(xdpf))) {
+		sinfo = xdp_get_shared_info_from_frame(xdpf);
+
+		for (i = 0; i < sinfo->nr_frags; i++) {
+			skb_frag_t *frag = &sinfo->frags[i];
+			dma_addr_t addr;
+			u32 len;
+
+			addr = page_pool_get_dma_addr(skb_frag_page(frag)) +
+				skb_frag_off(frag);
+			len = skb_frag_size(frag);
+			dma_sync_single_for_device(sq->pdev, addr, len,
+						   DMA_TO_DEVICE);
+		}
+	}
+
 	xdptxd.dma_addr = dma_addr;
-	xdpi.page.rq = rq;
-	xdpi.page.page = page;
 
 	if (unlikely(!INDIRECT_CALL_2(sq->xmit_xdp_frame, mlx5e_xmit_xdp_frame_mpwqe,
-				      mlx5e_xmit_xdp_frame, sq, &xdptxd, NULL, 0)))
+				      mlx5e_xmit_xdp_frame, sq, &xdptxd, sinfo, 0)))
 		return false;
 
+	xdpi.page.page = page;
 	mlx5e_xdpi_fifo_push(&sq->db.xdpi_fifo, &xdpi);
+
+	if (unlikely(xdp_frame_has_frags(xdpf))) {
+		for (i = 0; i < sinfo->nr_frags; i++) {
+			skb_frag_t *frag = &sinfo->frags[i];
+
+			xdpi.page.page = skb_frag_page(frag);
+			mlx5e_xdpi_fifo_push(&sq->db.xdpi_fifo, &xdpi);
+		}
+	}
+
 	return true;
 }
 

From 1b8a10bbfe770c53ef77eb84177ea13183cc0711 Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Tue, 1 Feb 2022 14:21:26 +0200
Subject: [PATCH 12/15] net/mlx5e: Permit XDP with non-linear legacy RQ

Now that legacy RQ implements XDP in the non-linear mode, stop blocking
this configuration. Allow non-linear mode only for programs aware of
multi buffer.

XDP performance with linear mode RQ hasn't changed.

Baseline (MTU 1500, TX MPWQE, legacy RQ, single core):
 60-byte packets, XDP_DROP: 11.25 Mpps
 60-byte packets, XDP_TX: 9.0 Mpps
 60-byte packets, XDP_PASS: 668 kpps

Multi buffer (MTU 9000, TX MPWQE, legacy RQ, single core):
 60-byte packets, XDP_DROP: 10.1 Mpps
 60-byte packets, XDP_TX: 6.6 Mpps
 60-byte packets, XDP_PASS: 658 kpps
 8900-byte packets, XDP_DROP: 769 kpps (100% of sent packets)
 8900-byte packets, XDP_TX: 674 kpps (100% of sent packets)
 8900-byte packets, XDP_PASS: 637 kpps

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 .../net/ethernet/mellanox/mlx5/core/en_main.c | 39 +++++++++++++------
 1 file changed, 27 insertions(+), 12 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 95cec28486855..3256d2c375c39 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -3953,6 +3953,31 @@ static bool mlx5e_xsk_validate_mtu(struct net_device *netdev,
 	return true;
 }
 
+static bool mlx5e_params_validate_xdp(struct net_device *netdev, struct mlx5e_params *params)
+{
+	bool is_linear;
+
+	/* No XSK params: AF_XDP can't be enabled yet at the point of setting
+	 * the XDP program.
+	 */
+	is_linear = mlx5e_rx_is_linear_skb(params, NULL);
+
+	if (!is_linear && params->rq_wq_type != MLX5_WQ_TYPE_CYCLIC) {
+		netdev_warn(netdev, "XDP is not allowed with striding RQ and MTU(%d) > %d\n",
+			    params->sw_mtu,
+			    mlx5e_xdp_max_mtu(params, NULL));
+		return false;
+	}
+	if (!is_linear && !params->xdp_prog->aux->xdp_has_frags) {
+		netdev_warn(netdev, "MTU(%d) > %d, too big for an XDP program not aware of multi buffer\n",
+			    params->sw_mtu,
+			    mlx5e_xdp_max_mtu(params, NULL));
+		return false;
+	}
+
+	return true;
+}
+
 int mlx5e_change_mtu(struct net_device *netdev, int new_mtu,
 		     mlx5e_fp_preactivate preactivate)
 {
@@ -3972,10 +3997,7 @@ int mlx5e_change_mtu(struct net_device *netdev, int new_mtu,
 	if (err)
 		goto out;
 
-	if (params->xdp_prog &&
-	    !mlx5e_rx_is_linear_skb(&new_params, NULL)) {
-		netdev_err(netdev, "MTU(%d) > %d is not allowed while XDP enabled\n",
-			   new_mtu, mlx5e_xdp_max_mtu(params, NULL));
+	if (new_params.xdp_prog && !mlx5e_params_validate_xdp(netdev, &new_params)) {
 		err = -EINVAL;
 		goto out;
 	}
@@ -4458,15 +4480,8 @@ static int mlx5e_xdp_allowed(struct mlx5e_priv *priv, struct bpf_prog *prog)
 	new_params = priv->channels.params;
 	new_params.xdp_prog = prog;
 
-	/* No XSK params: AF_XDP can't be enabled yet at the point of setting
-	 * the XDP program.
-	 */
-	if (!mlx5e_rx_is_linear_skb(&new_params, NULL)) {
-		netdev_warn(netdev, "XDP is not allowed with MTU(%d) > %d\n",
-			    new_params.sw_mtu,
-			    mlx5e_xdp_max_mtu(&new_params, NULL));
+	if (!mlx5e_params_validate_xdp(netdev, &new_params))
 		return -EINVAL;
-	}
 
 	return 0;
 }

From 08c34e95422bdecbc1bd67d35bc3ec307e2bafc2 Mon Sep 17 00:00:00 2001
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Date: Tue, 15 Feb 2022 21:01:22 +0200
Subject: [PATCH 13/15] net/mlx5e: Remove MLX5E_XDP_TX_DS_COUNT

After introducing multi-buffer XDP_TX, the MLX5E_XDP_TX_DS_COUNT define
became misleading. It's no longer the DS count of an XDP_TX WQE, this
WQE can be longer because of fragments.

As this define is only used at one place in mlx5e_open_xdpsq(), it's
also not very useful anymore. This commit removes the define and puts
the calculation of ds_count for prefilled single-fragment WQEs inline.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h  | 1 -
 drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 2 +-
 2 files changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
index 4868fa1b82d16..287e17911251b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h
@@ -38,7 +38,6 @@
 #include "en/txrx.h"
 
 #define MLX5E_XDP_MIN_INLINE (ETH_HLEN + VLAN_HLEN)
-#define MLX5E_XDP_TX_DS_COUNT (MLX5E_TX_WQE_EMPTY_DS_COUNT + 1 /* SG DS */)
 
 #define MLX5E_XDP_INLINE_WQE_MAX_DS_CNT 16
 #define MLX5E_XDP_INLINE_WQE_SZ_THRSD \
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 3256d2c375c39..2f1dedc721d1e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -1681,7 +1681,7 @@ int mlx5e_open_xdpsq(struct mlx5e_channel *c, struct mlx5e_params *params,
 	mlx5e_set_xmit_fp(sq, param->is_mpw);
 
 	if (!param->is_mpw && !test_bit(MLX5E_SQ_STATE_XDP_MULTIBUF, &sq->state)) {
-		unsigned int ds_cnt = MLX5E_XDP_TX_DS_COUNT;
+		unsigned int ds_cnt = MLX5E_TX_WQE_EMPTY_DS_COUNT + 1;
 		unsigned int inline_hdr_sz = 0;
 		int i;
 

From 60796198b44ff8c2e3e513f794f96e89b513a64d Mon Sep 17 00:00:00 2001
From: Tariq Toukan <tariqt@nvidia.com>
Date: Wed, 9 Mar 2022 10:44:24 +0200
Subject: [PATCH 14/15] net/mlx5e: Statify function
 mlx5_cmd_trigger_completions

Starting from commit
4cab346bcf74 ("net/mlx5: No command allowed when command interface is not ready"),
no calls to mlx5_cmd_trigger_completions() are external to cmd.c anymore.
Make it a static function.

Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Gal Pressman <gal@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/cmd.c       | 2 +-
 drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h | 1 -
 2 files changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index c2462d37f1b3c..f1329f83f9889 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -1720,7 +1720,7 @@ static void mlx5_cmd_comp_handler(struct mlx5_core_dev *dev, u64 vec, bool force
 	}
 }
 
-void mlx5_cmd_trigger_completions(struct mlx5_core_dev *dev)
+static void mlx5_cmd_trigger_completions(struct mlx5_core_dev *dev)
 {
 	struct mlx5_cmd *cmd = &dev->cmd;
 	unsigned long bitmask;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
index 6f8baa0f2a735..3b231e90111d2 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
@@ -176,7 +176,6 @@ int mlx5_destroy_scheduling_element_cmd(struct mlx5_core_dev *dev, u8 hierarchy,
 					u32 element_id);
 int mlx5_wait_for_pages(struct mlx5_core_dev *dev, int *pages);
 
-void mlx5_cmd_trigger_completions(struct mlx5_core_dev *dev);
 void mlx5_cmd_flush(struct mlx5_core_dev *dev);
 void mlx5_cq_debugfs_init(struct mlx5_core_dev *dev);
 void mlx5_cq_debugfs_cleanup(struct mlx5_core_dev *dev);

From 5dc2b581cd2cf518d28d0c703478a0c1fd54436c Mon Sep 17 00:00:00 2001
From: Saeed Mahameed <saeedm@nvidia.com>
Date: Tue, 9 Feb 2021 14:41:05 -0800
Subject: [PATCH 15/15] net/mlx5e: HTB, remove unused function declaration

There is no function mlx5e_get_sq(), remove the declaration.

Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Moshe Tal <moshet@nvidia.com>
Reviewed-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en/qos.h | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/qos.h b/drivers/net/ethernet/mellanox/mlx5/core/en/qos.h
index b7558907ba208..5d9bd91d86c22 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/qos.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/qos.h
@@ -18,7 +18,6 @@ int mlx5e_qos_cur_leaf_nodes(struct mlx5e_priv *priv);
 
 /* TX datapath API */
 int mlx5e_get_txq_by_classid(struct mlx5e_priv *priv, u16 classid);
-struct mlx5e_txqsq *mlx5e_get_sq(struct mlx5e_priv *priv, int qid);
 
 /* SQ lifecycle */
 int mlx5e_qos_open_queues(struct mlx5e_priv *priv, struct mlx5e_channels *chs);