public inbox for dev@dpdk.org
 help / color / mirror / Atom feed
From: Dariusz Sosnowski <dsosnowski@nvidia.com>
To: Viacheslav Ovsiienko <viacheslavo@nvidia.com>,
	Bing Zhao <bingz@nvidia.com>, Ori Kam <orika@nvidia.com>,
	Suanming Mou <suanmingm@nvidia.com>,
	Matan Azrad <matan@nvidia.com>,
	Alexander Kozyrev <akozyrev@nvidia.com>
Cc: <dev@dpdk.org>, <stable@dpdk.org>
Subject: [PATCH] net/mlx5: fix VLAN strip info for CQE compression
Date: Fri, 20 Mar 2026 17:00:01 +0100	[thread overview]
Message-ID: <20260320160001.787223-1-dsosnowski@nvidia.com> (raw)

When L3/L4 mini CQE format is used with CQE compression
(rxq_cqe_comp_en set to 4 in lower 3 bits),
each mini CQE reports packet header info for relevant packet.
One bit in that header info is dedicated to CVLAN info,
used to determine whether CVLAN is present in the received packet.
This info is used to determine VLAN stripping info
when relevant offload is enabled.

Before this patch, users might have seen incorrectly set/unset
RTE_MBUF_F_RX_VLAN and RTE_MBUF_F_RX_VLAN_STRIPPED flags because:

- Struct for mlx5 mini CQE incorrectly defined the location
  of packet header info byte - it was swapped with neighboring
  reserved byte.
- CVLAN bit was checked incorrectly because of cast to uint16_t
  and big endian change. It should be checked as uint8_t.

This patch addresses the above.

Fixes: 54c2d46b160f ("net/mlx5: support flow tag and packet header miniCQEs")
Cc: akozyrev@nvidia.com
Cc: stable@dpdk.org

Signed-off-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
---
 drivers/common/mlx5/mlx5_prm.h | 2 +-
 drivers/net/mlx5/mlx5_rx.c     | 3 +--
 2 files changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/common/mlx5/mlx5_prm.h b/drivers/common/mlx5/mlx5_prm.h
index ba33336e58..11b14e4f9d 100644
--- a/drivers/common/mlx5/mlx5_prm.h
+++ b/drivers/common/mlx5/mlx5_prm.h
@@ -5516,8 +5516,8 @@ struct mlx5_mini_cqe8 {
 				uint16_t checksum;
 				uint16_t flow_tag_high;
 				struct {
-					uint8_t reserved;
 					uint8_t hdr_type;
+					uint8_t reserved;
 				};
 			};
 			uint16_t stride_idx;
diff --git a/drivers/net/mlx5/mlx5_rx.c b/drivers/net/mlx5/mlx5_rx.c
index bc0470e6af..da871fd18a 100644
--- a/drivers/net/mlx5/mlx5_rx.c
+++ b/drivers/net/mlx5/mlx5_rx.c
@@ -1000,8 +1000,7 @@ rxq_cq_to_mbuf(struct mlx5_rxq_data *rxq, struct rte_mbuf *pkt,
 			vlan_strip = cqe->hdr_type_etc &
 				     RTE_BE16(MLX5_CQE_VLAN_STRIPPED);
 		else
-			vlan_strip = mcqe->hdr_type &
-				     RTE_BE16(MLX5_CQE_VLAN_STRIPPED);
+			vlan_strip = mcqe->hdr_type & MLX5_CQE_VLAN_STRIPPED;
 		if (vlan_strip) {
 			pkt->ol_flags |= RTE_MBUF_F_RX_VLAN | RTE_MBUF_F_RX_VLAN_STRIPPED;
 			pkt->vlan_tci = rte_be_to_cpu_16(cqe->vlan_info);
-- 
2.47.3


             reply	other threads:[~2026-03-20 16:01 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-03-20 16:00 Dariusz Sosnowski [this message]
2026-03-24 13:03 ` [PATCH] net/mlx5: fix VLAN strip info for CQE compression Raslan Darawsheh

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=20260320160001.787223-1-dsosnowski@nvidia.com \
    --to=dsosnowski@nvidia.com \
    --cc=akozyrev@nvidia.com \
    --cc=bingz@nvidia.com \
    --cc=dev@dpdk.org \
    --cc=matan@nvidia.com \
    --cc=orika@nvidia.com \
    --cc=stable@dpdk.org \
    --cc=suanmingm@nvidia.com \
    --cc=viacheslavo@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