netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Saeed Mahameed <saeed@kernel.org>
To: "David S. Miller" <davem@davemloft.net>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Eric Dumazet <edumazet@google.com>
Cc: Saeed Mahameed <saeedm@nvidia.com>,
	netdev@vger.kernel.org, Tariq Toukan <tariqt@nvidia.com>,
	Gal Pressman <gal@nvidia.com>,
	Leon Romanovsky <leonro@nvidia.com>,
	Carolina Jubran <cjubran@nvidia.com>
Subject: [net-next 14/15] net/mlx5e: XSK, Exclude tailroom from non-linear SKBs memory calculations
Date: Fri, 26 Jan 2024 14:36:15 -0800	[thread overview]
Message-ID: <20240126223616.98696-15-saeed@kernel.org> (raw)
In-Reply-To: <20240126223616.98696-1-saeed@kernel.org>

From: Carolina Jubran <cjubran@nvidia.com>

Packet data buffers lack reserved headroom or tailroom,
and SKBs are allocated on a side memory when needed.

Exclude the tailroom from the SKB size calculations.

Signed-off-by: Carolina Jubran <cjubran@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 .../net/ethernet/mellanox/mlx5/core/en/params.c   | 15 +++++++++++----
 1 file changed, 11 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
index 5d213a9886f1..b9d39ef8053c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
@@ -240,11 +240,14 @@ static u32 mlx5e_rx_get_linear_sz_xsk(struct mlx5e_params *params,
 	return xsk->headroom + hw_mtu;
 }
 
-static u32 mlx5e_rx_get_linear_sz_skb(struct mlx5e_params *params, bool xsk)
+static u32 mlx5e_rx_get_linear_sz_skb(struct mlx5e_params *params, bool no_head_tail_room)
 {
-	/* SKBs built on XDP_PASS on XSK RQs don't have headroom. */
-	u16 headroom = xsk ? 0 : mlx5e_get_linear_rq_headroom(params, NULL);
 	u32 hw_mtu = MLX5E_SW2HW_MTU(params, params->sw_mtu);
+	u16 headroom;
+
+	if (no_head_tail_room)
+		return SKB_DATA_ALIGN(hw_mtu);
+	headroom = mlx5e_get_linear_rq_headroom(params, NULL);
 
 	return MLX5_SKB_FRAG_SZ(headroom + hw_mtu);
 }
@@ -289,7 +292,11 @@ bool mlx5e_rx_is_linear_skb(struct mlx5_core_dev *mdev,
 	if (params->packet_merge.type != MLX5E_PACKET_MERGE_NONE)
 		return false;
 
-	/* Both XSK and non-XSK cases allocate an SKB on XDP_PASS. Packet data
+	/* Call mlx5e_rx_get_linear_sz_skb with the no_head_tail_room parameter set
+	 * to exclude headroom and tailroom from calculations.
+	 * no_head_tail_room is true when SKB is built on XDP_PASS on XSK RQs
+	 * since packet data buffers don't have headroom and tailroom resreved for the SKB.
+	 * Both XSK and non-XSK cases allocate an SKB on XDP_PASS. Packet data
 	 * must fit into a CPU page.
 	 */
 	if (mlx5e_rx_get_linear_sz_skb(params, xsk) > PAGE_SIZE)
-- 
2.43.0


  parent reply	other threads:[~2024-01-26 22:37 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-01-26 22:36 [pull request][net-next 00/15] mlx5 updates 2024-01-26 Saeed Mahameed
2024-01-26 22:36 ` [net-next 01/15] xfrm: generalize xdo_dev_state_update_curlft to allow statistics update Saeed Mahameed
2024-01-26 22:36 ` [net-next 02/15] xfrm: get global statistics from the offloaded device Saeed Mahameed
2024-01-26 22:36 ` [net-next 03/15] net/mlx5e: Connect mlx5 IPsec statistics with XFRM core Saeed Mahameed
2024-01-26 22:36 ` [net-next 04/15] net/mlx5e: Delete obsolete IPsec code Saeed Mahameed
2024-01-26 22:36 ` [net-next 05/15] Documentation: Fix counter name of mlx5 vnic reporter Saeed Mahameed
2024-01-26 22:36 ` [net-next 06/15] net/mlx5: Rename mlx5_sf_dev_remove Saeed Mahameed
2024-01-26 22:36 ` [net-next 07/15] net/mlx5: remove fw_fatal reporter dump option for non PF Saeed Mahameed
2024-01-26 22:36 ` [net-next 08/15] net/mlx5: remove fw " Saeed Mahameed
2024-01-26 22:36 ` [net-next 09/15] net/mlx5: SF, Stop waiting for FW as teardown was called Saeed Mahameed
2024-01-26 22:36 ` [net-next 10/15] net/mlx5: Return specific error code for timeout on wait_fw_init Saeed Mahameed
2024-01-26 22:36 ` [net-next 11/15] net/mlx5: Remove initial segmentation duplicate definitions Saeed Mahameed
2024-01-26 22:36 ` [net-next 12/15] net/mlx5: Change missing SyncE capability print to debug Saeed Mahameed
2024-01-26 22:36 ` [net-next 13/15] net/mlx5: DR, Change SWS usage to debug fs seq_file interface Saeed Mahameed
2024-01-29 20:55   ` Simon Horman
2024-01-31  1:42     ` Yevgeny Kliteynik
2024-01-26 22:36 ` Saeed Mahameed [this message]
2024-01-26 22:36 ` [net-next 15/15] net/mlx5e: XDP, Exclude headroom and tailroom from memory calculations Saeed Mahameed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20240126223616.98696-15-saeed@kernel.org \
    --to=saeed@kernel.org \
    --cc=cjubran@nvidia.com \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=gal@nvidia.com \
    --cc=kuba@kernel.org \
    --cc=leonro@nvidia.com \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=saeedm@nvidia.com \
    --cc=tariqt@nvidia.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).