DPDK patches and discussions
 help / color / mirror / Atom feed
From: Ivan Malov <ivan.malov@oktetlabs.ru>
To: dev@dpdk.org
Cc: stable@dpdk.org, Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>,
	Andy Moreton <amoreton@xilinx.com>
Subject: [dpdk-dev] [PATCH] net/sfc: fix outer match in MAE backend of RTE flow support
Date: Tue,  6 Jul 2021 15:37:26 +0300	[thread overview]
Message-ID: <20210706123726.12914-1-ivan.malov@oktetlabs.ru> (raw)

The current code doesn't use match on invalid outer rule ID
in action rules of non-encap. flows. This is wrong and must
be fixed. Do that and explain correct behaviour in the code.

Fixes: dadff137931c ("net/sfc: support encap flow items in transfer rules")
Cc: stable@dpdk.org

Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Reviewed-by: Andy Moreton <amoreton@xilinx.com>
---
 drivers/net/sfc/sfc_mae.c | 32 +++++++++++++++++++-------------
 1 file changed, 19 insertions(+), 13 deletions(-)

diff --git a/drivers/net/sfc/sfc_mae.c b/drivers/net/sfc/sfc_mae.c
index a2c0aa143..89f719648 100644
--- a/drivers/net/sfc/sfc_mae.c
+++ b/drivers/net/sfc/sfc_mae.c
@@ -1838,12 +1838,12 @@ sfc_mae_rule_process_outer(struct sfc_adapter *sa,
 			   struct sfc_mae_outer_rule **rulep,
 			   struct rte_flow_error *error)
 {
-	struct sfc_mae_outer_rule *rule;
+	efx_mae_rule_id_t invalid_rule_id = { .id = EFX_MAE_RSRC_ID_INVALID };
 	int rc;
 
 	if (ctx->encap_type == EFX_TUNNEL_PROTOCOL_NONE) {
 		*rulep = NULL;
-		return 0;
+		goto no_or_id;
 	}
 
 	SFC_ASSERT(ctx->match_spec_outer != NULL);
@@ -1871,21 +1871,27 @@ sfc_mae_rule_process_outer(struct sfc_adapter *sa,
 	/* The spec has now been tracked by the outer rule entry. */
 	ctx->match_spec_outer = NULL;
 
+no_or_id:
 	/*
-	 * Depending on whether we reuse an existing outer rule or create a
-	 * new one (see above), outer rule ID is either a valid value or
-	 * EFX_MAE_RSRC_ID_INVALID. Set it in the action rule match
-	 * specification (and the full mask, too) in order to have correct
-	 * class comparisons of the new rule with existing ones.
-	 * Also, action rule match specification will be validated shortly,
-	 * and having the full mask set for outer rule ID indicates that we
-	 * will use this field, and support for this field has to be checked.
+	 * In MAE, lookup sequence comprises outer parse, outer rule lookup,
+	 * inner parse (when some outer rule is hit) and action rule lookup.
+	 * If the currently processed flow does not come with an outer rule,
+	 * its action rule must be available only for packets which miss in
+	 * outer rule table. Set OR_ID match field to 0xffffffff/0xffffffff
+	 * in the action rule specification; this ensures correct behaviour.
+	 *
+	 * If, on the other hand, this flow does have an outer rule, its ID
+	 * may be unknown at the moment (not yet allocated), but OR_ID mask
+	 * has to be set to 0xffffffff anyway for correct class comparisons.
+	 * When the outer rule has been allocated, this match field will be
+	 * overridden by sfc_mae_outer_rule_enable() to use the right value.
 	 */
-	rule = *rulep;
 	rc = efx_mae_match_spec_outer_rule_id_set(ctx->match_spec_action,
-						  &rule->fw_rsrc.rule_id);
+						  &invalid_rule_id);
 	if (rc != 0) {
-		sfc_mae_outer_rule_del(sa, *rulep);
+		if (*rulep != NULL)
+			sfc_mae_outer_rule_del(sa, *rulep);
+
 		*rulep = NULL;
 
 		return rte_flow_error_set(error, rc,
-- 
2.20.1


             reply	other threads:[~2021-07-06 12:37 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-06 12:37 Ivan Malov [this message]
2021-07-07 10:03 ` [dpdk-dev] [dpdk-stable] " David Marchand

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=20210706123726.12914-1-ivan.malov@oktetlabs.ru \
    --to=ivan.malov@oktetlabs.ru \
    --cc=amoreton@xilinx.com \
    --cc=andrew.rybchenko@oktetlabs.ru \
    --cc=dev@dpdk.org \
    --cc=stable@dpdk.org \
    /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).