DPDK patches and discussions
 help / color / mirror / Atom feed
* [PATCH] net/mlx5: fix integrity conversion scheme
@ 2021-11-11 16:16 Gregory Etelson
  2021-11-11 16:51 ` Slava Ovsiienko
  2021-11-14 11:19 ` Raslan Darawsheh
  0 siblings, 2 replies; 3+ messages in thread
From: Gregory Etelson @ 2021-11-11 16:16 UTC (permalink / raw)
  To: dev, getelson, viacheslavo; +Cc: stable, Matan Azrad

RTE flow integrity API provides top-level packet validations.
RTE integrity bits are not always translated one-to-one to
hardware integrity bits.
For example RTE l3_ok and l4_ok integrity bits require 2 hardware
integrity bits each.

The patch fixes RTE l3_ok and l4_ok bits translation to match
ConnectX-6 hardware.

Cc: stable@dpdk.org

Fixes: 79f8952783d0 ("net/mlx5: support integrity flow item")

Signed-off-by: Gregory Etelson <getelson@nvidia.com>
---
 drivers/net/mlx5/mlx5_flow_dv.c | 79 +++++++++++++++------------------
 1 file changed, 37 insertions(+), 42 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_flow_dv.c b/drivers/net/mlx5/mlx5_flow_dv.c
index 28911d6f0f..1b4e15dff1 100644
--- a/drivers/net/mlx5/mlx5_flow_dv.c
+++ b/drivers/net/mlx5/mlx5_flow_dv.c
@@ -12106,34 +12106,24 @@ flow_dv_translate_integrity_l4(const struct rte_flow_item_integrity *mask,
 			       void *headers_m, void *headers_v)
 {
 	if (mask->l4_ok) {
-		/* application l4_ok filter aggregates all hardware l4 filters
-		 * therefore hw l4_checksum_ok must be implicitly added here.
+		/* RTE l4_ok filter aggregates hardware l4_ok and
+		 * l4_checksum_ok filters.
+		 * Positive RTE l4_ok match requires hardware match on both L4
+		 * hardware integrity bits.
+		 * For negative match, check hardware l4_checksum_ok bit only,
+		 * because hardware sets that bit to 0 for all packets
+		 * with bad L4.
 		 */
-		struct rte_flow_item_integrity local_item;
-
-		local_item.l4_csum_ok = 1;
-		MLX5_SET(fte_match_set_lyr_2_4, headers_m, l4_checksum_ok,
-			 local_item.l4_csum_ok);
 		if (value->l4_ok) {
-			/* application l4_ok = 1 matches sets both hw flags
-			 * l4_ok and l4_checksum_ok flags to 1.
-			 */
-			MLX5_SET(fte_match_set_lyr_2_4, headers_v,
-				 l4_checksum_ok, local_item.l4_csum_ok);
-			MLX5_SET(fte_match_set_lyr_2_4, headers_m, l4_ok,
-				 mask->l4_ok);
-			MLX5_SET(fte_match_set_lyr_2_4, headers_v, l4_ok,
-				 value->l4_ok);
-		} else {
-			/* application l4_ok = 0 matches on hw flag
-			 * l4_checksum_ok = 0 only.
-			 */
-			MLX5_SET(fte_match_set_lyr_2_4, headers_v,
-				 l4_checksum_ok, 0);
+			MLX5_SET(fte_match_set_lyr_2_4, headers_m, l4_ok, 1);
+			MLX5_SET(fte_match_set_lyr_2_4, headers_v, l4_ok, 1);
 		}
-	} else if (mask->l4_csum_ok) {
-		MLX5_SET(fte_match_set_lyr_2_4, headers_m, l4_checksum_ok,
-			 mask->l4_csum_ok);
+		MLX5_SET(fte_match_set_lyr_2_4, headers_m, l4_checksum_ok, 1);
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v, l4_checksum_ok,
+			 !!value->l4_ok);
+	}
+	if (mask->l4_csum_ok) {
+		MLX5_SET(fte_match_set_lyr_2_4, headers_m, l4_checksum_ok, 1);
 		MLX5_SET(fte_match_set_lyr_2_4, headers_v, l4_checksum_ok,
 			 value->l4_csum_ok);
 	}
@@ -12145,28 +12135,33 @@ flow_dv_translate_integrity_l3(const struct rte_flow_item_integrity *mask,
 			       void *headers_m, void *headers_v, bool is_ipv4)
 {
 	if (mask->l3_ok) {
-		/* application l3_ok filter aggregates all hardware l3 filters
-		 * therefore hw ipv4_checksum_ok must be implicitly added here.
+		/* RTE l3_ok filter aggregates for IPv4 hardware l3_ok and
+		 * ipv4_csum_ok filters.
+		 * Positive RTE l3_ok match requires hardware match on both L3
+		 * hardware integrity bits.
+		 * For negative match, check hardware l3_csum_ok bit only,
+		 * because hardware sets that bit to 0 for all packets
+		 * with bad L3.
 		 */
-		struct rte_flow_item_integrity local_item;
-
-		local_item.ipv4_csum_ok = !!is_ipv4;
-		MLX5_SET(fte_match_set_lyr_2_4, headers_m, ipv4_checksum_ok,
-			 local_item.ipv4_csum_ok);
-		if (value->l3_ok) {
+		if (is_ipv4) {
+			if (value->l3_ok) {
+				MLX5_SET(fte_match_set_lyr_2_4, headers_m,
+					 l3_ok, 1);
+				MLX5_SET(fte_match_set_lyr_2_4, headers_v,
+					 l3_ok, 1);
+			}
+			MLX5_SET(fte_match_set_lyr_2_4, headers_m,
+				 ipv4_checksum_ok, 1);
 			MLX5_SET(fte_match_set_lyr_2_4, headers_v,
-				 ipv4_checksum_ok, local_item.ipv4_csum_ok);
-			MLX5_SET(fte_match_set_lyr_2_4, headers_m, l3_ok,
-				 mask->l3_ok);
+				 ipv4_checksum_ok, !!value->l3_ok);
+		} else {
+			MLX5_SET(fte_match_set_lyr_2_4, headers_m, l3_ok, 1);
 			MLX5_SET(fte_match_set_lyr_2_4, headers_v, l3_ok,
 				 value->l3_ok);
-		} else {
-			MLX5_SET(fte_match_set_lyr_2_4, headers_v,
-				 ipv4_checksum_ok, 0);
 		}
-	} else if (mask->ipv4_csum_ok) {
-		MLX5_SET(fte_match_set_lyr_2_4, headers_m, ipv4_checksum_ok,
-			 mask->ipv4_csum_ok);
+	}
+	if (mask->ipv4_csum_ok) {
+		MLX5_SET(fte_match_set_lyr_2_4, headers_m, ipv4_checksum_ok, 1);
 		MLX5_SET(fte_match_set_lyr_2_4, headers_v, ipv4_checksum_ok,
 			 value->ipv4_csum_ok);
 	}
-- 
2.33.1


^ permalink raw reply	[flat|nested] 3+ messages in thread

* RE: [PATCH] net/mlx5: fix integrity conversion scheme
  2021-11-11 16:16 [PATCH] net/mlx5: fix integrity conversion scheme Gregory Etelson
@ 2021-11-11 16:51 ` Slava Ovsiienko
  2021-11-14 11:19 ` Raslan Darawsheh
  1 sibling, 0 replies; 3+ messages in thread
From: Slava Ovsiienko @ 2021-11-11 16:51 UTC (permalink / raw)
  To: Gregory Etelson, dev; +Cc: stable, Matan Azrad

> -----Original Message-----
> From: Gregory Etelson <getelson@nvidia.com>
> Sent: Thursday, November 11, 2021 18:16
> To: dev@dpdk.org; Gregory Etelson <getelson@nvidia.com>; Slava Ovsiienko
> <viacheslavo@nvidia.com>
> Cc: stable@dpdk.org; Matan Azrad <matan@nvidia.com>
> Subject: [PATCH] net/mlx5: fix integrity conversion scheme
> 
> RTE flow integrity API provides top-level packet validations.
> RTE integrity bits are not always translated one-to-one to hardware integrity
> bits.
> For example RTE l3_ok and l4_ok integrity bits require 2 hardware integrity
> bits each.
> 
> The patch fixes RTE l3_ok and l4_ok bits translation to match
> ConnectX-6 hardware.
> 
> Cc: stable@dpdk.org
> 
> Fixes: 79f8952783d0 ("net/mlx5: support integrity flow item")
> 
> Signed-off-by: Gregory Etelson <getelson@nvidia.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>


^ permalink raw reply	[flat|nested] 3+ messages in thread

* RE: [PATCH] net/mlx5: fix integrity conversion scheme
  2021-11-11 16:16 [PATCH] net/mlx5: fix integrity conversion scheme Gregory Etelson
  2021-11-11 16:51 ` Slava Ovsiienko
@ 2021-11-14 11:19 ` Raslan Darawsheh
  1 sibling, 0 replies; 3+ messages in thread
From: Raslan Darawsheh @ 2021-11-14 11:19 UTC (permalink / raw)
  To: Gregory Etelson, dev, Gregory Etelson, Slava Ovsiienko
  Cc: stable, Matan Azrad

Hi,

> -----Original Message-----
> From: Gregory Etelson <getelson@nvidia.com>
> Sent: Thursday, November 11, 2021 6:16 PM
> To: dev@dpdk.org; Gregory Etelson <getelson@nvidia.com>; Slava Ovsiienko
> <viacheslavo@nvidia.com>
> Cc: stable@dpdk.org; Matan Azrad <matan@nvidia.com>
> Subject: [PATCH] net/mlx5: fix integrity conversion scheme
> 
> RTE flow integrity API provides top-level packet validations.
> RTE integrity bits are not always translated one-to-one to hardware integrity
> bits.
> For example RTE l3_ok and l4_ok integrity bits require 2 hardware integrity
> bits each.
> 
> The patch fixes RTE l3_ok and l4_ok bits translation to match
> ConnectX-6 hardware.
> 
> Cc: stable@dpdk.org
> 
> Fixes: 79f8952783d0 ("net/mlx5: support integrity flow item")
> 
> Signed-off-by: Gregory Etelson <getelson@nvidia.com>

Patch applied to next-net-mlx,

Kindest regards,
Raslan Darawsheh

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2021-11-14 11:19 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-11-11 16:16 [PATCH] net/mlx5: fix integrity conversion scheme Gregory Etelson
2021-11-11 16:51 ` Slava Ovsiienko
2021-11-14 11:19 ` Raslan Darawsheh

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).