From: Tariq Toukan <tariqt@nvidia.com>
To: "David S. Miller" <davem@davemloft.net>,
Jakub Kicinski <kuba@kernel.org>
Cc: Eric Dumazet <edumazet@google.com>,
Paolo Abeni <pabeni@redhat.com>,
Jesper Dangaard Brouer <brouer@redhat.com>,
Toke Hoiland-Jorgensen <toke@redhat.com>,
<netdev@vger.kernel.org>, Saeed Mahameed <saeedm@nvidia.com>,
Lorenzo Bianconi <lorenzo@kernel.org>,
Gal Pressman <gal@nvidia.com>,
Henning Fehrmann <henning.fehrmann@aei.mpg.de>,
"Oliver Behnke" <oliver.behnke@aei.mpg.de>,
Tariq Toukan <tariqt@nvidia.com>
Subject: [PATCH net-next 09/15] net/mlx5e: XDP, Consider large muti-buffer packets in Striding RQ params calculations
Date: Mon, 17 Apr 2023 15:18:57 +0300 [thread overview]
Message-ID: <20230417121903.46218-10-tariqt@nvidia.com> (raw)
In-Reply-To: <20230417121903.46218-1-tariqt@nvidia.com>
Function mlx5e_rx_get_linear_stride_sz() returns PAGE_SIZE immediately
in case an XDP program is attached. The more accurate formula is
ALIGN(sz, PAGE_SIZE), to prevent two packets from residing on the same
page.
The assumption behind the current code is that sz <= PAGE_SIZE holds for
all cases with XDP program set.
This is true because it is being called from:
- 3 times from Striding RQ flows, in which XDP is not supported for such
large packets.
- 1 time from Legacy RQ flow, under the condition
mlx5e_rx_is_linear_skb().
No functional change here, just removing the implied assumption in
preparation for supporting XDP multi-buffer in Striding RQ.
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en/params.c | 11 +++++++----
1 file changed, 7 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 31f3c6e51d9e..196862e67af3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
@@ -253,17 +253,20 @@ static u32 mlx5e_rx_get_linear_stride_sz(struct mlx5_core_dev *mdev,
struct mlx5e_xsk_param *xsk,
bool mpwqe)
{
+ u32 sz;
+
/* XSK frames are mapped as individual pages, because frames may come in
* an arbitrary order from random locations in the UMEM.
*/
if (xsk)
return mpwqe ? 1 << mlx5e_mpwrq_page_shift(mdev, xsk) : PAGE_SIZE;
- /* XDP in mlx5e doesn't support multiple packets per page. */
- if (params->xdp_prog)
- return PAGE_SIZE;
+ sz = roundup_pow_of_two(mlx5e_rx_get_linear_sz_skb(params, false));
- return roundup_pow_of_two(mlx5e_rx_get_linear_sz_skb(params, false));
+ /* XDP in mlx5e doesn't support multiple packets per page.
+ * Do not assume sz <= PAGE_SIZE if params->xdp_prog is set.
+ */
+ return params->xdp_prog && sz < PAGE_SIZE ? PAGE_SIZE : sz;
}
static u8 mlx5e_mpwqe_log_pkts_per_wqe(struct mlx5_core_dev *mdev,
--
2.34.1
next prev parent reply other threads:[~2023-04-17 12:21 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-04-17 12:18 [PATCH net-next 00/15] net/mlx5e: Extend XDP multi-buffer capabilities Tariq Toukan
2023-04-17 12:18 ` [PATCH net-next 01/15] net/mlx5e: Move XDP struct and enum to XDP header Tariq Toukan
2023-04-17 12:18 ` [PATCH net-next 02/15] net/mlx5e: Move struct mlx5e_xmit_data to datapath header Tariq Toukan
2023-04-17 12:18 ` [PATCH net-next 03/15] net/mlx5e: Introduce extended version for mlx5e_xmit_data Tariq Toukan
2023-04-17 12:18 ` [PATCH net-next 04/15] net/mlx5e: XDP, Remove doubtful unlikely calls Tariq Toukan
2023-04-17 12:18 ` [PATCH net-next 05/15] net/mlx5e: XDP, Use multiple single-entry objects in xdpi_fifo Tariq Toukan
2023-04-17 12:18 ` [PATCH net-next 06/15] net/mlx5e: XDP, Add support for multi-buffer XDP redirect-in Tariq Toukan
2023-04-17 12:18 ` [PATCH net-next 07/15] net/mlx5e: XDP, Improve Striding RQ check with XDP Tariq Toukan
2023-04-17 12:18 ` [PATCH net-next 08/15] net/mlx5e: XDP, Let XDP checker function get the params as input Tariq Toukan
2023-04-17 12:18 ` Tariq Toukan [this message]
2023-04-17 12:18 ` [PATCH net-next 10/15] net/mlx5e: XDP, Remove un-established assumptions on XDP buffer Tariq Toukan
2023-04-17 12:18 ` [PATCH net-next 11/15] net/mlx5e: XDP, Allow non-linear single-segment frames in XDP TX MPWQE Tariq Toukan
2023-04-17 12:19 ` [PATCH net-next 12/15] net/mlx5e: RX, Take shared info fragment addition into a function Tariq Toukan
2023-04-17 12:19 ` [PATCH net-next 13/15] net/mlx5e: RX, Generalize mlx5e_fill_mxbuf() Tariq Toukan
2023-04-17 12:19 ` [PATCH net-next 14/15] net/mlx5e: RX, Prepare non-linear striding RQ for XDP multi-buffer support Tariq Toukan
2023-04-17 12:19 ` [PATCH net-next 15/15] net/mlx5e: RX, Add XDP multi-buffer support in Striding RQ Tariq Toukan
2023-04-20 0:47 ` [PATCH net-next 00/15] net/mlx5e: Extend XDP multi-buffer capabilities Jakub Kicinski
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=20230417121903.46218-10-tariqt@nvidia.com \
--to=tariqt@nvidia.com \
--cc=brouer@redhat.com \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=gal@nvidia.com \
--cc=henning.fehrmann@aei.mpg.de \
--cc=kuba@kernel.org \
--cc=lorenzo@kernel.org \
--cc=netdev@vger.kernel.org \
--cc=oliver.behnke@aei.mpg.de \
--cc=pabeni@redhat.com \
--cc=saeedm@nvidia.com \
--cc=toke@redhat.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).