DPDK patches and discussions
 help / color / mirror / Atom feed
From: Rahul Bhansali <rbhansali@marvell.com>
To: <dev@dpdk.org>, Nithin Dabilpuram <ndabilpuram@marvell.com>,
	Kiran Kumar K <kirankumark@marvell.com>,
	Sunil Kumar Kori <skori@marvell.com>,
	Satha Rao <skoteshwar@marvell.com>,
	Harman Kalra <hkalra@marvell.com>,
	Pavan Nikhilesh <pbhagavatula@marvell.com>,
	Rahul Bhansali <rbhansali@marvell.com>,
	"Jerin Jacob" <jerinj@marvell.com>
Subject: [PATCH] net/cnxk: fix of NIX send header L3 type
Date: Mon, 27 Jan 2025 12:03:35 +0530	[thread overview]
Message-ID: <20250127063335.405019-1-rbhansali@marvell.com> (raw)

For small packets less than 55 bytes, SQ error interrupts are
observed.
When checksum offload flag is enabled and mbuf ol_flags are not
set, then default L3 type will be set to IPv6 in vector
processing. Based on this, HW will still validate for minimum
header size and generate send header error if mismatch.

To address this, will set default L3 type to none. Mbuf ol_flags
RTE_MBUF_F_TX_IPV6 will set with offload feature TSO or L4
checksum only, so handled in corresponding routine.

Fixes: f71b7dbbf04b ("net/cnxk: add vector Tx for CN10K")
Fixes: e829e60c6917 ("net/cnxk: support Tx burst vector for CN20K")

Signed-off-by: Rahul Bhansali <rbhansali@marvell.com>
---
 drivers/net/cnxk/cn10k_tx.h | 7 +++++--
 drivers/net/cnxk/cn20k_tx.h | 7 +++++--
 2 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/drivers/net/cnxk/cn10k_tx.h b/drivers/net/cnxk/cn10k_tx.h
index 7d9b259a5f..77f4e54cd8 100644
--- a/drivers/net/cnxk/cn10k_tx.h
+++ b/drivers/net/cnxk/cn10k_tx.h
@@ -1773,6 +1773,9 @@ cn10k_nix_prepare_tso(struct rte_mbuf *m, union nix_send_hdr_w1_u *w1,
 	w0->lso_mps = m->tso_segsz;
 	w0->lso_format = NIX_LSO_FORMAT_IDX_TSOV4 + !!(ol_flags & RTE_MBUF_F_TX_IPV6);
 	w1->ol4type = NIX_SENDL4TYPE_TCP_CKSUM;
+	w1->ol3type = ((!!(ol_flags & RTE_MBUF_F_TX_IPV4)) << 1) +
+		      ((!!(ol_flags & RTE_MBUF_F_TX_IPV6)) << 2) +
+		      !!(ol_flags & RTE_MBUF_F_TX_IP_CKSUM);
 
 	/* Handle tunnel tso */
 	if ((flags & NIX_TX_OFFLOAD_OL3_OL4_CSUM_F) &&
@@ -2477,7 +2480,7 @@ cn10k_nix_xmit_pkts_vector(void *tx_queue, uint64_t *ws,
 			 */
 			const uint8x16_t tbl = {
 				/* [0-15] = il4type:il3type */
-				0x04, /* none (IPv6 assumed) */
+				0x00, /* none */
 				0x14, /* RTE_MBUF_F_TX_TCP_CKSUM (IPv6 assumed) */
 				0x24, /* RTE_MBUF_F_TX_SCTP_CKSUM (IPv6 assumed) */
 				0x34, /* RTE_MBUF_F_TX_UDP_CKSUM (IPv6 assumed) */
@@ -2681,7 +2684,7 @@ cn10k_nix_xmit_pkts_vector(void *tx_queue, uint64_t *ws,
 			const uint8x16x2_t tbl = {{
 				{
 					/* [0-15] = il4type:il3type */
-					0x04, /* none (IPv6) */
+					0x00, /* none */
 					0x14, /* RTE_MBUF_F_TX_TCP_CKSUM (IPv6) */
 					0x24, /* RTE_MBUF_F_TX_SCTP_CKSUM (IPv6) */
 					0x34, /* RTE_MBUF_F_TX_UDP_CKSUM (IPv6) */
diff --git a/drivers/net/cnxk/cn20k_tx.h b/drivers/net/cnxk/cn20k_tx.h
index c731406529..dd28be1edd 100644
--- a/drivers/net/cnxk/cn20k_tx.h
+++ b/drivers/net/cnxk/cn20k_tx.h
@@ -1733,6 +1733,9 @@ cn20k_nix_prepare_tso(struct rte_mbuf *m, union nix_send_hdr_w1_u *w1, union nix
 	w0->lso_mps = m->tso_segsz;
 	w0->lso_format = NIX_LSO_FORMAT_IDX_TSOV4 + !!(ol_flags & RTE_MBUF_F_TX_IPV6);
 	w1->ol4type = NIX_SENDL4TYPE_TCP_CKSUM;
+	w1->ol3type = ((!!(ol_flags & RTE_MBUF_F_TX_IPV4)) << 1) +
+		      ((!!(ol_flags & RTE_MBUF_F_TX_IPV6)) << 2) +
+		      !!(ol_flags & RTE_MBUF_F_TX_IP_CKSUM);
 
 	/* Handle tunnel tso */
 	if ((flags & NIX_TX_OFFLOAD_OL3_OL4_CSUM_F) && (ol_flags & RTE_MBUF_F_TX_TUNNEL_MASK)) {
@@ -2395,7 +2398,7 @@ cn20k_nix_xmit_pkts_vector(void *tx_queue, uint64_t *ws, struct rte_mbuf **tx_pk
 			 */
 			const uint8x16_t tbl = {
 				/* [0-15] = il4type:il3type */
-				0x04, /* none (IPv6 assumed) */
+				0x04, /* none */
 				0x14, /* RTE_MBUF_F_TX_TCP_CKSUM (IPv6 assumed) */
 				0x24, /* RTE_MBUF_F_TX_SCTP_CKSUM (IPv6 assumed) */
 				0x34, /* RTE_MBUF_F_TX_UDP_CKSUM (IPv6 assumed) */
@@ -2595,7 +2598,7 @@ cn20k_nix_xmit_pkts_vector(void *tx_queue, uint64_t *ws, struct rte_mbuf **tx_pk
 			const uint8x16x2_t tbl = {{
 				{
 					/* [0-15] = il4type:il3type */
-					0x04, /* none (IPv6) */
+					0x04, /* none */
 					0x14, /* RTE_MBUF_F_TX_TCP_CKSUM (IPv6) */
 					0x24, /* RTE_MBUF_F_TX_SCTP_CKSUM (IPv6) */
 					0x34, /* RTE_MBUF_F_TX_UDP_CKSUM (IPv6) */
-- 
2.25.1


                 reply	other threads:[~2025-01-27  6:33 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=20250127063335.405019-1-rbhansali@marvell.com \
    --to=rbhansali@marvell.com \
    --cc=dev@dpdk.org \
    --cc=hkalra@marvell.com \
    --cc=jerinj@marvell.com \
    --cc=kirankumark@marvell.com \
    --cc=ndabilpuram@marvell.com \
    --cc=pbhagavatula@marvell.com \
    --cc=skori@marvell.com \
    --cc=skoteshwar@marvell.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).