net/mlx5e: XDP, Exclude headroom and tailroom from memory calculations

In the case of XDP Multi-Buffer with Striding RQ, an extra
page is allocated for the linear part of non-linear SKBs.

Including headroom and tailroom in the calculation may
result in an unnecessary increase in the amount of memory
allocated. This could be critical, particularly for large
MTUs (e.g. 7975B) and large RQ sizes (e.g. 8192).

In this case, the requested page pool size is 64K, but
32K would be sufficient. This causes a failure due to
exceeding the page pool size limit of 32K.

Exclude headroom and tailroom from SKB size calculations
to reduce page pool size.

Signed-off-by: Carolina Jubran <cjubran@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

@@ -257,6 +257,7 @@ static u32 mlx5e_rx_get_linear_stride_sz(struct mlx5_core_dev *mdev,
 					 struct mlx5e_xsk_param *xsk,
 					 bool mpwqe)
 {
+	bool no_head_tail_room;
 	u32 sz;
 
 	/* XSK frames are mapped as individual pages, because frames may come in
@@ -265,7 +266,13 @@ static u32 mlx5e_rx_get_linear_stride_sz(struct mlx5_core_dev *mdev,
 	if (xsk)
 		return mpwqe ? 1 << mlx5e_mpwrq_page_shift(mdev, xsk) : PAGE_SIZE;
 
-	sz = roundup_pow_of_two(mlx5e_rx_get_linear_sz_skb(params, false));
+	no_head_tail_room = params->xdp_prog && mpwqe && !mlx5e_rx_is_linear_skb(mdev, params, xsk);
+
+	/* When no_head_tail_room is set, headroom and tailroom are excluded from skb calculations.
+	 * no_head_tail_room should be set in the case of XDP with Striding RQ
+	 * when SKB is not linear. This is because another page is allocated for the linear part.
+	 */
+	sz = roundup_pow_of_two(mlx5e_rx_get_linear_sz_skb(params, no_head_tail_room));
 
 	/* XDP in mlx5e doesn't support multiple packets per page.
 	 * Do not assume sz <= PAGE_SIZE if params->xdp_prog is set.