Patch Detail
get:
Show a patch.
patch:
Update a patch.
put:
Update a patch.
GET /api/patches/100028/?format=api
{ "id": 100028, "url": "https://patches.dpdk.org/api/patches/100028/?format=api", "web_url": "https://patches.dpdk.org/project/dpdk/patch/20210929205730.775-9-ivan.malov@oktetlabs.ru/", "project": { "id": 1, "url": "https://patches.dpdk.org/api/projects/1/?format=api", "name": "DPDK", "link_name": "dpdk", "list_id": "dev.dpdk.org", "list_email": "dev@dpdk.org", "web_url": "http://core.dpdk.org", "scm_url": "git://dpdk.org/dpdk", "webscm_url": "http://git.dpdk.org/dpdk", "list_archive_url": "https://inbox.dpdk.org/dev", "list_archive_url_format": "https://inbox.dpdk.org/dev/{}", "commit_url_format": "" }, "msgid": "<20210929205730.775-9-ivan.malov@oktetlabs.ru>", "list_archive_url": "https://inbox.dpdk.org/dev/20210929205730.775-9-ivan.malov@oktetlabs.ru", "date": "2021-09-29T20:57:28", "name": "[08/10] net/sfc: use action rules in tunnel offload JUMP rules", "commit_ref": null, "pull_url": null, "state": "superseded", "archived": true, "hash": "bc70bc4b9c490f7861fa8a84d9dcac26054c73fd", "submitter": { "id": 869, "url": "https://patches.dpdk.org/api/people/869/?format=api", "name": "Ivan Malov", "email": "Ivan.Malov@oktetlabs.ru" }, "delegate": { "id": 319, "url": "https://patches.dpdk.org/api/users/319/?format=api", "username": "fyigit", "first_name": "Ferruh", "last_name": "Yigit", "email": "ferruh.yigit@amd.com" }, "mbox": "https://patches.dpdk.org/project/dpdk/patch/20210929205730.775-9-ivan.malov@oktetlabs.ru/mbox/", "series": [ { "id": 19272, "url": "https://patches.dpdk.org/api/series/19272/?format=api", "web_url": "https://patches.dpdk.org/project/dpdk/list/?series=19272", "date": "2021-09-29T20:57:20", "name": "net/sfc: add support for tunnel offload", "version": 1, "mbox": "https://patches.dpdk.org/series/19272/mbox/" } ], "comments": "https://patches.dpdk.org/api/patches/100028/comments/", "check": "success", "checks": "https://patches.dpdk.org/api/patches/100028/checks/", "tags": {}, "related": [], "headers": { "Return-Path": "<dev-bounces@dpdk.org>", "X-Original-To": "patchwork@inbox.dpdk.org", "Delivered-To": "patchwork@inbox.dpdk.org", "Received": [ "from mails.dpdk.org (mails.dpdk.org [217.70.189.124])\n\tby inbox.dpdk.org (Postfix) with ESMTP id 1547DA0032;\n\tWed, 29 Sep 2021 22:58:37 +0200 (CEST)", "from [217.70.189.124] (localhost [127.0.0.1])\n\tby mails.dpdk.org (Postfix) with ESMTP id 0B48B4114B;\n\tWed, 29 Sep 2021 22:57:56 +0200 (CEST)", "from shelob.oktetlabs.ru (shelob.oktetlabs.ru [91.220.146.113])\n by mails.dpdk.org (Postfix) with ESMTP id 63754410F5\n for <dev@dpdk.org>; Wed, 29 Sep 2021 22:57:46 +0200 (CEST)", "from localhost.localdomain (unknown [5.144.122.192])\n (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)\n key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest\n SHA256)\n (No client certificate requested)\n by shelob.oktetlabs.ru (Postfix) with ESMTPSA id 10D067F553;\n Wed, 29 Sep 2021 23:57:46 +0300 (MSK)" ], "DKIM-Filter": "OpenDKIM Filter v2.11.0 shelob.oktetlabs.ru 10D067F553", "DKIM-Signature": "v=1; a=rsa-sha256; c=simple/simple; d=oktetlabs.ru;\n s=default; t=1632949066;\n bh=FnoO6VMZKhJ/HQri2dLDih5KKfTTN1Y3iRRvtUfwYso=;\n h=From:To:Cc:Subject:Date:In-Reply-To:References;\n b=hq7OrGj/uLvRRjp+OJadwkgyNSr7V6MCo/yZ8pzPiwdeQxf+EWKbkHcHqHBrNb+MY\n m6XjQhnUiPkHAgPir9kgKRNoOKSD8QvCuFncU1ApwHtvkKhWnPPMLMxNrDdtiSATqv\n cHgPY5X79ujgvuyzRhn4sfSUL6d5v7T+ZnYi4Lg0=", "From": "Ivan Malov <ivan.malov@oktetlabs.ru>", "To": "dev@dpdk.org", "Cc": "Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>", "Date": "Wed, 29 Sep 2021 23:57:28 +0300", "Message-Id": "<20210929205730.775-9-ivan.malov@oktetlabs.ru>", "X-Mailer": "git-send-email 2.20.1", "In-Reply-To": "<20210929205730.775-1-ivan.malov@oktetlabs.ru>", "References": "<20210929205730.775-1-ivan.malov@oktetlabs.ru>", "MIME-Version": "1.0", "Content-Transfer-Encoding": "8bit", "Subject": "[dpdk-dev] [PATCH 08/10] net/sfc: use action rules in tunnel\n offload JUMP rules", "X-BeenThere": "dev@dpdk.org", "X-Mailman-Version": "2.1.29", "Precedence": "list", "List-Id": "DPDK patches and discussions <dev.dpdk.org>", "List-Unsubscribe": "<https://mails.dpdk.org/options/dev>,\n <mailto:dev-request@dpdk.org?subject=unsubscribe>", "List-Archive": "<http://mails.dpdk.org/archives/dev/>", "List-Post": "<mailto:dev@dpdk.org>", "List-Help": "<mailto:dev-request@dpdk.org?subject=help>", "List-Subscribe": "<https://mails.dpdk.org/listinfo/dev>,\n <mailto:dev-request@dpdk.org?subject=subscribe>", "Errors-To": "dev-bounces@dpdk.org", "Sender": "\"dev\" <dev-bounces@dpdk.org>" }, "content": "By design, JUMP flows should be represented solely by the outer rules. But\nthe HW/FW hasn't got support for setting Rx mark from RECIRC_ID on outer\nrule lookup yet. Neither does it support outer rule counters. As a\nworkaround, an action rule of lower priority is used to do the job.\n\nSigned-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>\nReviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>\n---\n drivers/net/sfc/sfc_flow.c | 11 +++++---\n drivers/net/sfc/sfc_mae.c | 55 ++++++++++++++++++++++++++++++++------\n 2 files changed, 54 insertions(+), 12 deletions(-)", "diff": "diff --git a/drivers/net/sfc/sfc_flow.c b/drivers/net/sfc/sfc_flow.c\nindex c3e75bae84..b0dd7d7b6c 100644\n--- a/drivers/net/sfc/sfc_flow.c\n+++ b/drivers/net/sfc/sfc_flow.c\n@@ -2561,17 +2561,20 @@ sfc_flow_parse_rte_to_mae(struct rte_eth_dev *dev,\n \n \tif (spec_mae->ft_rule_type == SFC_FT_RULE_JUMP) {\n \t\t/*\n-\t\t * This flow is represented solely by the outer rule.\n-\t\t * It is supposed to mark and count matching packets.\n+\t\t * By design, this flow should be represented solely by the\n+\t\t * outer rule. But the HW/FW hasn't got support for setting\n+\t\t * Rx mark from RECIRC_ID on outer rule lookup yet. Neither\n+\t\t * does it support outer rule counters. As a workaround, an\n+\t\t * action rule of lower priority is used to do the job.\n+\t\t *\n+\t\t * So don't skip sfc_mae_rule_parse_actions() below.\n \t\t */\n-\t\tgoto skip_action_rule;\n \t}\n \n \trc = sfc_mae_rule_parse_actions(sa, actions, spec_mae, error);\n \tif (rc != 0)\n \t\tgoto fail;\n \n-skip_action_rule:\n \tif (spec_mae->ft != NULL) {\n \t\tif (spec_mae->ft_rule_type == SFC_FT_RULE_JUMP)\n \t\t\tspec_mae->ft->jump_rule_is_set = B_TRUE;\ndiff --git a/drivers/net/sfc/sfc_mae.c b/drivers/net/sfc/sfc_mae.c\nindex 374ef29d71..faf3be522d 100644\n--- a/drivers/net/sfc/sfc_mae.c\n+++ b/drivers/net/sfc/sfc_mae.c\n@@ -2467,6 +2467,7 @@ sfc_mae_rule_parse_pattern(struct sfc_adapter *sa,\n \t\t\t struct rte_flow_error *error)\n {\n \tstruct sfc_mae_parse_ctx ctx_mae;\n+\tunsigned int priority_shift = 0;\n \tstruct sfc_flow_parse_ctx ctx;\n \tint rc;\n \n@@ -2478,13 +2479,32 @@ sfc_mae_rule_parse_pattern(struct sfc_adapter *sa,\n \n \tswitch (ctx_mae.ft_rule_type) {\n \tcase SFC_FT_RULE_JUMP:\n-\t\t/* No action rule */\n-\t\tbreak;\n+\t\t/*\n+\t\t * By design, this flow should be represented solely by the\n+\t\t * outer rule. But the HW/FW hasn't got support for setting\n+\t\t * Rx mark from RECIRC_ID on outer rule lookup yet. Neither\n+\t\t * does it support outer rule counters. As a workaround, an\n+\t\t * action rule of lower priority is used to do the job.\n+\t\t */\n+\t\tpriority_shift = 1;\n+\n+\t\t/* FALLTHROUGH */\n \tcase SFC_FT_RULE_GROUP:\n+\t\tif (ctx_mae.priority != 0) {\n+\t\t\t/*\n+\t\t\t * Because of the above workaround, deny the\n+\t\t\t * use of priorities to JUMP and GROUP rules.\n+\t\t\t */\n+\t\t\trc = rte_flow_error_set(error, ENOTSUP,\n+\t\t\t\tRTE_FLOW_ERROR_TYPE_ATTR_PRIORITY, NULL,\n+\t\t\t\t\"tunnel offload: priorities are not supported\");\n+\t\t\tgoto fail_priority_check;\n+\t\t}\n+\n \t\t/* FALLTHROUGH */\n \tcase SFC_FT_RULE_NONE:\n \t\trc = efx_mae_match_spec_init(sa->nic, EFX_MAE_RULE_ACTION,\n-\t\t\t\t\t spec->priority,\n+\t\t\t\t\t spec->priority + priority_shift,\n \t\t\t\t\t &ctx_mae.match_spec_action);\n \t\tif (rc != 0) {\n \t\t\trc = rte_flow_error_set(error, rc,\n@@ -2559,6 +2579,7 @@ sfc_mae_rule_parse_pattern(struct sfc_adapter *sa,\n \t\tefx_mae_match_spec_fini(sa->nic, ctx_mae.match_spec_action);\n \n fail_init_match_spec_action:\n+fail_priority_check:\n \treturn rc;\n }\n \n@@ -3008,11 +3029,14 @@ sfc_mae_rule_parse_action_vxlan_encap(\n static int\n sfc_mae_rule_parse_action_mark(struct sfc_adapter *sa,\n \t\t\t const struct rte_flow_action_mark *conf,\n+\t\t\t const struct sfc_flow_spec_mae *spec_mae,\n \t\t\t efx_mae_actions_t *spec)\n {\n \tint rc;\n \n-\tif (conf->id > SFC_FT_USER_MARK_MASK) {\n+\tif (spec_mae->ft_rule_type == SFC_FT_RULE_JUMP) {\n+\t\t/* Workaround. See sfc_flow_parse_rte_to_mae() */\n+\t} else if (conf->id > SFC_FT_USER_MARK_MASK) {\n \t\tsfc_err(sa, \"the mark value is too large\");\n \t\treturn EINVAL;\n \t}\n@@ -3182,11 +3206,12 @@ static const char * const action_names[] = {\n static int\n sfc_mae_rule_parse_action(struct sfc_adapter *sa,\n \t\t\t const struct rte_flow_action *action,\n-\t\t\t const struct sfc_mae_outer_rule *outer_rule,\n+\t\t\t const struct sfc_flow_spec_mae *spec_mae,\n \t\t\t struct sfc_mae_actions_bundle *bundle,\n \t\t\t efx_mae_actions_t *spec,\n \t\t\t struct rte_flow_error *error)\n {\n+\tconst struct sfc_mae_outer_rule *outer_rule = spec_mae->outer_rule;\n \tconst uint64_t rx_meta = sa->negotiated_rx_meta;\n \tbool custom_error = B_FALSE;\n \tint rc = 0;\n@@ -3250,9 +3275,10 @@ sfc_mae_rule_parse_action(struct sfc_adapter *sa,\n \tcase RTE_FLOW_ACTION_TYPE_MARK:\n \t\tSFC_BUILD_SET_OVERFLOW(RTE_FLOW_ACTION_TYPE_MARK,\n \t\t\t\t bundle->actions_mask);\n-\t\tif ((rx_meta & RTE_ETH_RX_META_USER_MARK) != 0) {\n+\t\tif ((rx_meta & RTE_ETH_RX_META_USER_MARK) != 0 ||\n+\t\t spec_mae->ft_rule_type == SFC_FT_RULE_JUMP) {\n \t\t\trc = sfc_mae_rule_parse_action_mark(sa, action->conf,\n-\t\t\t\t\t\t\t spec);\n+\t\t\t\t\t\t\t spec_mae, spec);\n \t\t} else {\n \t\t\trc = rte_flow_error_set(error, ENOTSUP,\n \t\t\t\t\t\tRTE_FLOW_ERROR_TYPE_ACTION,\n@@ -3286,6 +3312,12 @@ sfc_mae_rule_parse_action(struct sfc_adapter *sa,\n \t\t\t\t bundle->actions_mask);\n \t\trc = efx_mae_action_set_populate_drop(spec);\n \t\tbreak;\n+\tcase RTE_FLOW_ACTION_TYPE_JUMP:\n+\t\tif (spec_mae->ft_rule_type == SFC_FT_RULE_JUMP) {\n+\t\t\t/* Workaround. See sfc_flow_parse_rte_to_mae() */\n+\t\t\tbreak;\n+\t\t}\n+\t\t/* FALLTHROUGH */\n \tdefault:\n \t\treturn rte_flow_error_set(error, ENOTSUP,\n \t\t\t\tRTE_FLOW_ERROR_TYPE_ACTION, NULL,\n@@ -3375,7 +3407,7 @@ sfc_mae_rule_parse_actions(struct sfc_adapter *sa,\n \t\tif (rc != 0)\n \t\t\tgoto fail_rule_parse_action;\n \n-\t\trc = sfc_mae_rule_parse_action(sa, action, spec_mae->outer_rule,\n+\t\trc = sfc_mae_rule_parse_action(sa, action, spec_mae,\n \t\t\t\t\t &bundle, spec, error);\n \t\tif (rc != 0)\n \t\t\tgoto fail_rule_parse_action;\n@@ -3399,6 +3431,12 @@ sfc_mae_rule_parse_actions(struct sfc_adapter *sa,\n \tswitch (spec_mae->ft_rule_type) {\n \tcase SFC_FT_RULE_NONE:\n \t\tbreak;\n+\tcase SFC_FT_RULE_JUMP:\n+\t\t/* Workaround. See sfc_flow_parse_rte_to_mae() */\n+\t\trc = sfc_mae_rule_parse_action_pf_vf(sa, NULL, spec);\n+\t\tif (rc != 0)\n+\t\t\tgoto fail_workaround_jump_delivery;\n+\t\tbreak;\n \tcase SFC_FT_RULE_GROUP:\n \t\t/*\n \t\t * Packets that go to the rule's AR have FT mark set (from the\n@@ -3428,6 +3466,7 @@ sfc_mae_rule_parse_actions(struct sfc_adapter *sa,\n \treturn 0;\n \n fail_action_set_add:\n+fail_workaround_jump_delivery:\n fail_nb_count:\n \tsfc_mae_encap_header_del(sa, encap_header);\n \n", "prefixes": [ "08/10" ] }