Patch Detail
get:
Show a patch.
patch:
Update a patch.
put:
Update a patch.
GET /api/patches/135704/?format=api
{ "id": 135704, "url": "http://patches.dpdk.org/api/patches/135704/?format=api", "web_url": "http://patches.dpdk.org/project/dpdk/patch/20240103101054.1330081-3-mingjinx.ye@intel.com/", "project": { "id": 1, "url": "http://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": "<20240103101054.1330081-3-mingjinx.ye@intel.com>", "list_archive_url": "https://inbox.dpdk.org/dev/20240103101054.1330081-3-mingjinx.ye@intel.com", "date": "2024-01-03T10:10:54", "name": "[v8,2/2] net/iavf: add diagnostic support in TX path", "commit_ref": null, "pull_url": null, "state": "superseded", "archived": true, "hash": "3a5063c943a4fdfe2503a4f15752a1874e8df6bf", "submitter": { "id": 2862, "url": "http://patches.dpdk.org/api/people/2862/?format=api", "name": "Mingjin Ye", "email": "mingjinx.ye@intel.com" }, "delegate": { "id": 1540, "url": "http://patches.dpdk.org/api/users/1540/?format=api", "username": "qzhan15", "first_name": "Qi", "last_name": "Zhang", "email": "qi.z.zhang@intel.com" }, "mbox": "http://patches.dpdk.org/project/dpdk/patch/20240103101054.1330081-3-mingjinx.ye@intel.com/mbox/", "series": [ { "id": 30710, "url": "http://patches.dpdk.org/api/series/30710/?format=api", "web_url": "http://patches.dpdk.org/project/dpdk/list/?series=30710", "date": "2024-01-03T10:10:52", "name": "net/iavf: fix Rx/Tx burst and add diagnostics", "version": 8, "mbox": "http://patches.dpdk.org/series/30710/mbox/" } ], "comments": "http://patches.dpdk.org/api/patches/135704/comments/", "check": "success", "checks": "http://patches.dpdk.org/api/patches/135704/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 434EB43809;\n\tWed, 3 Jan 2024 11:28:17 +0100 (CET)", "from mails.dpdk.org (localhost [127.0.0.1])\n\tby mails.dpdk.org (Postfix) with ESMTP id 2C5D140395;\n\tWed, 3 Jan 2024 11:28:17 +0100 (CET)", "from mgamail.intel.com (mgamail.intel.com [198.175.65.11])\n by mails.dpdk.org (Postfix) with ESMTP id D360440648\n for <dev@dpdk.org>; Wed, 3 Jan 2024 11:28:14 +0100 (CET)", "from fmsmga006.fm.intel.com ([10.253.24.20])\n by orvoesa103.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384;\n 03 Jan 2024 02:28:14 -0800", "from unknown (HELO localhost.localdomain) ([10.239.252.253])\n by fmsmga006-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384;\n 03 Jan 2024 02:28:10 -0800" ], "DKIM-Signature": "v=1; a=rsa-sha256; c=relaxed/simple;\n d=intel.com; i=@intel.com; q=dns/txt; s=Intel;\n t=1704277695; x=1735813695;\n h=from:to:cc:subject:date:message-id:in-reply-to:\n references:mime-version:content-transfer-encoding;\n bh=NVRWj1/J0Eg4g0bjW/a4sVrXWv/tdav8SkXsa28560A=;\n b=A0wsYjzQzl3LziG+UzBqx1xbP5YCEQKsvnA8FfB6+pbqp7h1q27c8jDv\n RolI276fZkBwAB/Owu0a2s1L6HDSuwSP1184I9debXPDzi+04Y7B6MALt\n hmI7Gm9XVIyuayB9yFN/Ly4/J2XFgrom3McpNcWEKoLZqrxYvizjCDb+q\n h5kI/5VN+njj1Ok2dcgjD1h4EG99dXG2WKmzFzvuqWPaNZ1xxJHxaUf2x\n UDpypxPsr5j3ME+SFWZ9wEs1zjiJN5Jeis7X1jpba2lw12Ho7lIUUfDtq\n psYrVZsClxolFEW/dLjn5Xgsci+OU/MZBedyM2cKEwLt9x33fFyoV4/ST Q==;", "X-IronPort-AV": [ "E=McAfee;i=\"6600,9927,10941\"; a=\"3761758\"", "E=Sophos;i=\"6.04,327,1695711600\";\n d=\"scan'208\";a=\"3761758\"", "E=McAfee;i=\"6600,9927,10941\"; a=\"1027036766\"", "E=Sophos;i=\"6.04,327,1695711600\"; d=\"scan'208\";a=\"1027036766\"" ], "X-ExtLoop1": "1", "From": "Mingjin Ye <mingjinx.ye@intel.com>", "To": "dev@dpdk.org", "Cc": "qiming.yang@intel.com, Mingjin Ye <mingjinx.ye@intel.com>,\n Simei Su <simei.su@intel.com>, Wenjun Wu <wenjun1.wu@intel.com>,\n Yuying Zhang <Yuying.Zhang@intel.com>, Beilei Xing <beilei.xing@intel.com>,\n Jingjing Wu <jingjing.wu@intel.com>", "Subject": "[PATCH v8 2/2] net/iavf: add diagnostic support in TX path", "Date": "Wed, 3 Jan 2024 10:10:54 +0000", "Message-Id": "<20240103101054.1330081-3-mingjinx.ye@intel.com>", "X-Mailer": "git-send-email 2.25.1", "In-Reply-To": "<20240103101054.1330081-1-mingjinx.ye@intel.com>", "References": "<20240102105211.788819-3-mingjinx.ye@intel.com>\n <20240103101054.1330081-1-mingjinx.ye@intel.com>", "MIME-Version": "1.0", "Content-Transfer-Encoding": "8bit", "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" }, "content": "The only way to enable diagnostics for TX paths is to modify the\napplication source code. Making it difficult to diagnose faults.\n\nIn this patch, the devarg option \"mbuf_check\" is introduced and the\nparameters are configured to enable the corresponding diagnostics.\n\nsupported cases: mbuf, size, segment, offload.\n 1. mbuf: check for corrupted mbuf.\n 2. size: check min/max packet length according to hw spec.\n 3. segment: check number of mbuf segments not exceed hw limitation.\n 4. offload: check any unsupported offload flag.\n\nparameter format: mbuf_check=[mbuf,<case1>,<case2>]\neg: dpdk-testpmd -a 0000:81:01.0,mbuf_check=[mbuf,size] -- -i\n\nSigned-off-by: Mingjin Ye <mingjinx.ye@intel.com>\n---\nv2: Remove call chain.\n---\nv3: Optimisation implementation.\n---\nv4: Fix Windows os compilation error.\n---\nv5: Split Patch.\n---\nv6: remove strict.\n---\nv7: Modify the description document.\n---\n doc/guides/nics/intel_vf.rst | 9 ++++\n drivers/net/iavf/iavf.h | 12 +++++\n drivers/net/iavf/iavf_ethdev.c | 76 ++++++++++++++++++++++++++\n drivers/net/iavf/iavf_rxtx.c | 98 ++++++++++++++++++++++++++++++++++\n drivers/net/iavf/iavf_rxtx.h | 2 +\n 5 files changed, 197 insertions(+)", "diff": "diff --git a/doc/guides/nics/intel_vf.rst b/doc/guides/nics/intel_vf.rst\nindex ad08198f0f..bda6648726 100644\n--- a/doc/guides/nics/intel_vf.rst\n+++ b/doc/guides/nics/intel_vf.rst\n@@ -111,6 +111,15 @@ For more detail on SR-IOV, please refer to the following documents:\n by setting the ``devargs`` parameter like ``-a 18:01.0,no-poll-on-link-down=1``\n when IAVF is backed by an Intel\\ |reg| E810 device or an Intel\\ |reg| 700 Series Ethernet device.\n \n+ When IAVF is backed by an Intel\\ |reg| E810 device or an Intel\\ |reg| 700 series Ethernet devices.\n+ Set the ``devargs`` parameter ``mbuf_check`` to enable TX diagnostics. For example,\n+ ``-a 18:01.0,mbuf_check=mbuf`` or ``-a 18:01.0,mbuf_check=[mbuf,size]``. Supported cases:\n+\n+ * mbuf: Check for corrupted mbuf.\n+ * size: Check min/max packet length according to hw spec.\n+ * segment: Check number of mbuf segments not exceed hw limitation.\n+ * offload: Check any unsupported offload flag.\n+\n The PCIE host-interface of Intel Ethernet Switch FM10000 Series VF infrastructure\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n \ndiff --git a/drivers/net/iavf/iavf.h b/drivers/net/iavf/iavf.h\nindex 73a089c199..6535b624cb 100644\n--- a/drivers/net/iavf/iavf.h\n+++ b/drivers/net/iavf/iavf.h\n@@ -113,9 +113,14 @@ struct iavf_ipsec_crypto_stats {\n \t} ierrors;\n };\n \n+struct iavf_mbuf_stats {\n+\tuint64_t tx_pkt_errors;\n+};\n+\n struct iavf_eth_xstats {\n \tstruct virtchnl_eth_stats eth_stats;\n \tstruct iavf_ipsec_crypto_stats ips_stats;\n+\tstruct iavf_mbuf_stats mbuf_stats;\n };\n \n /* Structure that defines a VSI, associated with a adapter. */\n@@ -309,6 +314,7 @@ struct iavf_devargs {\n \tuint32_t watchdog_period;\n \tint auto_reset;\n \tint no_poll_on_link_down;\n+\tint mbuf_check;\n };\n \n struct iavf_security_ctx;\n@@ -351,6 +357,11 @@ enum iavf_tx_burst_type {\n \tIAVF_TX_AVX512_CTX_OFFLOAD,\n };\n \n+#define IAVF_MBUF_CHECK_F_TX_MBUF (1ULL << 0)\n+#define IAVF_MBUF_CHECK_F_TX_SIZE (1ULL << 1)\n+#define IAVF_MBUF_CHECK_F_TX_SEGMENT (1ULL << 2)\n+#define IAVF_MBUF_CHECK_F_TX_OFFLOAD (1ULL << 3)\n+\n /* Structure to store private data for each VF instance. */\n struct iavf_adapter {\n \tstruct iavf_hw hw;\n@@ -368,6 +379,7 @@ struct iavf_adapter {\n \tbool no_poll;\n \tenum iavf_rx_burst_type rx_burst_type;\n \tenum iavf_tx_burst_type tx_burst_type;\n+\tuint64_t mc_flags; /* mbuf check flags. */\n \tuint16_t fdir_ref_cnt;\n \tstruct iavf_devargs devargs;\n };\ndiff --git a/drivers/net/iavf/iavf_ethdev.c b/drivers/net/iavf/iavf_ethdev.c\nindex d1edb0dd5c..7d1cd9050b 100644\n--- a/drivers/net/iavf/iavf_ethdev.c\n+++ b/drivers/net/iavf/iavf_ethdev.c\n@@ -13,6 +13,7 @@\n #include <inttypes.h>\n #include <rte_byteorder.h>\n #include <rte_common.h>\n+#include <rte_os_shim.h>\n \n #include <rte_interrupts.h>\n #include <rte_debug.h>\n@@ -39,6 +40,8 @@\n #define IAVF_RESET_WATCHDOG_ARG \"watchdog_period\"\n #define IAVF_ENABLE_AUTO_RESET_ARG \"auto_reset\"\n #define IAVF_NO_POLL_ON_LINK_DOWN_ARG \"no-poll-on-link-down\"\n+#define IAVF_MBUF_CHECK_ARG \"mbuf_check\"\n+\n uint64_t iavf_timestamp_dynflag;\n int iavf_timestamp_dynfield_offset = -1;\n \n@@ -48,6 +51,7 @@ static const char * const iavf_valid_args[] = {\n \tIAVF_RESET_WATCHDOG_ARG,\n \tIAVF_ENABLE_AUTO_RESET_ARG,\n \tIAVF_NO_POLL_ON_LINK_DOWN_ARG,\n+\tIAVF_MBUF_CHECK_ARG,\n \tNULL\n };\n \n@@ -174,6 +178,7 @@ static const struct rte_iavf_xstats_name_off rte_iavf_stats_strings[] = {\n \t{\"tx_broadcast_packets\", _OFF_OF(eth_stats.tx_broadcast)},\n \t{\"tx_dropped_packets\", _OFF_OF(eth_stats.tx_discards)},\n \t{\"tx_error_packets\", _OFF_OF(eth_stats.tx_errors)},\n+\t{\"tx_mbuf_error_packets\", _OFF_OF(mbuf_stats.tx_pkt_errors)},\n \n \t{\"inline_ipsec_crypto_ipackets\", _OFF_OF(ips_stats.icount)},\n \t{\"inline_ipsec_crypto_ibytes\", _OFF_OF(ips_stats.ibytes)},\n@@ -1837,6 +1842,9 @@ iavf_dev_xstats_reset(struct rte_eth_dev *dev)\n \tiavf_dev_stats_reset(dev);\n \tmemset(&vf->vsi.eth_stats_offset.ips_stats, 0,\n \t\t\tsizeof(struct iavf_ipsec_crypto_stats));\n+\tmemset(&vf->vsi.eth_stats_offset.mbuf_stats, 0,\n+\t\t\tsizeof(struct iavf_mbuf_stats));\n+\n \treturn 0;\n }\n \n@@ -1876,6 +1884,19 @@ iavf_dev_update_ipsec_xstats(struct rte_eth_dev *ethdev,\n \t}\n }\n \n+static void\n+iavf_dev_update_mbuf_stats(struct rte_eth_dev *ethdev,\n+\t\tstruct iavf_mbuf_stats *mbuf_stats)\n+{\n+\tuint16_t idx;\n+\tstruct iavf_tx_queue *txq;\n+\n+\tfor (idx = 0; idx < ethdev->data->nb_tx_queues; idx++) {\n+\t\ttxq = ethdev->data->tx_queues[idx];\n+\t\tmbuf_stats->tx_pkt_errors += txq->mbuf_errors;\n+\t}\n+}\n+\n static int iavf_dev_xstats_get(struct rte_eth_dev *dev,\n \t\t\t\t struct rte_eth_xstat *xstats, unsigned int n)\n {\n@@ -1904,6 +1925,9 @@ static int iavf_dev_xstats_get(struct rte_eth_dev *dev,\n \tif (iavf_ipsec_crypto_supported(adapter))\n \t\tiavf_dev_update_ipsec_xstats(dev, &iavf_xtats.ips_stats);\n \n+\tif (adapter->devargs.mbuf_check)\n+\t\tiavf_dev_update_mbuf_stats(dev, &iavf_xtats.mbuf_stats);\n+\n \t/* loop over xstats array and values from pstats */\n \tfor (i = 0; i < IAVF_NB_XSTATS; i++) {\n \t\txstats[i].id = i;\n@@ -2286,6 +2310,50 @@ iavf_parse_watchdog_period(__rte_unused const char *key, const char *value, void\n \treturn 0;\n }\n \n+static int\n+iavf_parse_mbuf_check(__rte_unused const char *key, const char *value, void *args)\n+{\n+\tchar *cur;\n+\tchar *tmp;\n+\tint str_len;\n+\tint valid_len;\n+\n+\tint ret = 0;\n+\tuint64_t *mc_flags = args;\n+\tchar *str2 = strdup(value);\n+\tif (str2 == NULL)\n+\t\treturn -1;\n+\n+\tstr_len = strlen(str2);\n+\tif (str2[0] == '[' && str2[str_len - 1] == ']') {\n+\t\tif (str_len < 3) {\n+\t\t\tret = -1;\n+\t\t\tgoto mdd_end;\n+\t\t}\n+\t\tvalid_len = str_len - 2;\n+\t\tmemmove(str2, str2 + 1, valid_len);\n+\t\tmemset(str2 + valid_len, '\\0', 2);\n+\t}\n+\tcur = strtok_r(str2, \",\", &tmp);\n+\twhile (cur != NULL) {\n+\t\tif (!strcmp(cur, \"mbuf\"))\n+\t\t\t*mc_flags |= IAVF_MBUF_CHECK_F_TX_MBUF;\n+\t\telse if (!strcmp(cur, \"size\"))\n+\t\t\t*mc_flags |= IAVF_MBUF_CHECK_F_TX_SIZE;\n+\t\telse if (!strcmp(cur, \"segment\"))\n+\t\t\t*mc_flags |= IAVF_MBUF_CHECK_F_TX_SEGMENT;\n+\t\telse if (!strcmp(cur, \"offload\"))\n+\t\t\t*mc_flags |= IAVF_MBUF_CHECK_F_TX_OFFLOAD;\n+\t\telse\n+\t\t\tPMD_DRV_LOG(ERR, \"Unsupported mdd check type: %s\", cur);\n+\t\tcur = strtok_r(NULL, \",\", &tmp);\n+\t}\n+\n+mdd_end:\n+\tfree(str2);\n+\treturn ret;\n+}\n+\n static int iavf_parse_devargs(struct rte_eth_dev *dev)\n {\n \tstruct iavf_adapter *ad =\n@@ -2340,6 +2408,14 @@ static int iavf_parse_devargs(struct rte_eth_dev *dev)\n \t\tgoto bail;\n \t}\n \n+\tret = rte_kvargs_process(kvlist, IAVF_MBUF_CHECK_ARG,\n+\t\t\t\t &iavf_parse_mbuf_check, &ad->mc_flags);\n+\tif (ret)\n+\t\tgoto bail;\n+\n+\tif (ad->mc_flags)\n+\t\tad->devargs.mbuf_check = 1;\n+\n \tret = rte_kvargs_process(kvlist, IAVF_ENABLE_AUTO_RESET_ARG,\n \t\t\t\t &parse_bool, &ad->devargs.auto_reset);\n \tif (ret)\ndiff --git a/drivers/net/iavf/iavf_rxtx.c b/drivers/net/iavf/iavf_rxtx.c\nindex 89db82c694..74dd72ce68 100644\n--- a/drivers/net/iavf/iavf_rxtx.c\n+++ b/drivers/net/iavf/iavf_rxtx.c\n@@ -3796,6 +3796,97 @@ iavf_xmit_pkts_no_poll(void *tx_queue, struct rte_mbuf **tx_pkts,\n \t\t\t\t\t\t\t\ttx_pkts, nb_pkts);\n }\n \n+/* Tx mbuf check */\n+static uint16_t\n+iavf_xmit_pkts_check(void *tx_queue, struct rte_mbuf **tx_pkts,\n+\t uint16_t nb_pkts)\n+{\n+\tuint16_t idx;\n+\tuint64_t ol_flags;\n+\tstruct rte_mbuf *mb;\n+\tuint16_t good_pkts = nb_pkts;\n+\tconst char *reason = NULL;\n+\tbool pkt_error = false;\n+\tstruct iavf_tx_queue *txq = tx_queue;\n+\tstruct iavf_adapter *adapter = txq->vsi->adapter;\n+\tenum iavf_tx_burst_type tx_burst_type =\n+\t\ttxq->vsi->adapter->tx_burst_type;\n+\n+\tfor (idx = 0; idx < nb_pkts; idx++) {\n+\t\tmb = tx_pkts[idx];\n+\t\tol_flags = mb->ol_flags;\n+\n+\t\tif ((adapter->mc_flags & IAVF_MBUF_CHECK_F_TX_MBUF) &&\n+\t\t\t(rte_mbuf_check(mb, 1, &reason) != 0)) {\n+\t\t\tPMD_TX_LOG(ERR, \"INVALID mbuf: %s\\n\", reason);\n+\t\t\tpkt_error = true;\n+\t\t\tbreak;\n+\t\t}\n+\n+\t\tif ((adapter->mc_flags & IAVF_MBUF_CHECK_F_TX_SIZE) &&\n+\t\t\t(mb->data_len < IAVF_TX_MIN_PKT_LEN ||\n+\t\t\tmb->data_len > adapter->vf.max_pkt_len)) {\n+\t\t\tPMD_TX_LOG(ERR, \"INVALID mbuf: data_len (%u) is out \"\n+\t\t\t\"of range, reasonable range (%d - %u)\\n\", mb->data_len,\n+\t\t\tIAVF_TX_MIN_PKT_LEN, adapter->vf.max_pkt_len);\n+\t\t\tpkt_error = true;\n+\t\t\tbreak;\n+\t\t}\n+\n+\t\tif (adapter->mc_flags & IAVF_MBUF_CHECK_F_TX_SEGMENT) {\n+\t\t\t/* Check condition for nb_segs > IAVF_TX_MAX_MTU_SEG. */\n+\t\t\tif (!(ol_flags & (RTE_MBUF_F_TX_TCP_SEG | RTE_MBUF_F_TX_UDP_SEG))) {\n+\t\t\t\tif (mb->nb_segs > IAVF_TX_MAX_MTU_SEG) {\n+\t\t\t\t\tPMD_TX_LOG(ERR, \"INVALID mbuf: nb_segs (%d) exceeds \"\n+\t\t\t\t\t\"HW limit, maximum allowed value is %d\\n\", mb->nb_segs,\n+\t\t\t\t\tIAVF_TX_MAX_MTU_SEG);\n+\t\t\t\t\tpkt_error = true;\n+\t\t\t\t\tbreak;\n+\t\t\t\t}\n+\t\t\t} else if ((mb->tso_segsz < IAVF_MIN_TSO_MSS) ||\n+\t\t\t\t(mb->tso_segsz > IAVF_MAX_TSO_MSS)) {\n+\t\t\t\t/* MSS outside the range are considered malicious */\n+\t\t\t\tPMD_TX_LOG(ERR, \"INVALID mbuf: tso_segsz (%u) is out \"\n+\t\t\t\t\"of range, reasonable range (%d - %u)\\n\", mb->tso_segsz,\n+\t\t\t\tIAVF_MIN_TSO_MSS, IAVF_MAX_TSO_MSS);\n+\t\t\t\tpkt_error = true;\n+\t\t\t\tbreak;\n+\t\t\t} else if (mb->nb_segs > txq->nb_tx_desc) {\n+\t\t\t\tPMD_TX_LOG(ERR, \"INVALID mbuf: nb_segs out \"\n+\t\t\t\t\"of ring length\\n\");\n+\t\t\t\tpkt_error = true;\n+\t\t\t\tbreak;\n+\t\t\t}\n+\t\t}\n+\n+\t\tif (adapter->mc_flags & IAVF_MBUF_CHECK_F_TX_OFFLOAD) {\n+\t\t\tif (ol_flags & IAVF_TX_OFFLOAD_NOTSUP_MASK) {\n+\t\t\t\tPMD_TX_LOG(ERR, \"INVALID mbuf: TX offload \"\n+\t\t\t\t\"is not supported\\n\");\n+\t\t\t\tpkt_error = true;\n+\t\t\t\tbreak;\n+\t\t\t}\n+\n+\t\t\tif (!rte_validate_tx_offload(mb)) {\n+\t\t\t\tPMD_TX_LOG(ERR, \"INVALID mbuf: TX offload \"\n+\t\t\t\t\"setup error\\n\");\n+\t\t\t\tpkt_error = true;\n+\t\t\t\tbreak;\n+\t\t\t}\n+\t\t}\n+\t}\n+\n+\tif (pkt_error) {\n+\t\ttxq->mbuf_errors++;\n+\t\tgood_pkts = idx;\n+\t\tif (good_pkts == 0)\n+\t\t\treturn 0;\n+\t}\n+\n+\treturn iavf_tx_pkt_burst_ops[tx_burst_type](tx_queue,\n+\t\t\t\t\t\t\t\ttx_pkts, good_pkts);\n+}\n+\n /* choose rx function*/\n void\n iavf_set_rx_function(struct rte_eth_dev *dev)\n@@ -4041,6 +4132,7 @@ iavf_set_tx_function(struct rte_eth_dev *dev)\n \tstruct iavf_adapter *adapter =\n \t\tIAVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);\n \tenum iavf_tx_burst_type tx_burst_type;\n+\tint mbuf_check = adapter->devargs.mbuf_check;\n \tint no_poll_on_link_down = adapter->devargs.no_poll_on_link_down;\n #ifdef RTE_ARCH_X86\n \tstruct iavf_tx_queue *txq;\n@@ -4131,6 +4223,9 @@ iavf_set_tx_function(struct rte_eth_dev *dev)\n \t\tif (no_poll_on_link_down) {\n \t\t\tadapter->tx_burst_type = tx_burst_type;\n \t\t\tdev->tx_pkt_burst = iavf_xmit_pkts_no_poll;\n+\t\t} else if (mbuf_check) {\n+\t\t\tadapter->tx_burst_type = tx_burst_type;\n+\t\t\tdev->tx_pkt_burst = iavf_xmit_pkts_check;\n \t\t} else {\n \t\t\tdev->tx_pkt_burst = iavf_tx_pkt_burst_ops[tx_burst_type];\n \t\t}\n@@ -4147,6 +4242,9 @@ iavf_set_tx_function(struct rte_eth_dev *dev)\n \tif (no_poll_on_link_down) {\n \t\tadapter->tx_burst_type = tx_burst_type;\n \t\tdev->tx_pkt_burst = iavf_xmit_pkts_no_poll;\n+\t} else if (mbuf_check) {\n+\t\tadapter->tx_burst_type = tx_burst_type;\n+\t\tdev->tx_pkt_burst = iavf_xmit_pkts_check;\n \t} else {\n \t\tdev->tx_pkt_burst = iavf_tx_pkt_burst_ops[tx_burst_type];\n \t}\ndiff --git a/drivers/net/iavf/iavf_rxtx.h b/drivers/net/iavf/iavf_rxtx.h\nindex f432f9d956..90e7291928 100644\n--- a/drivers/net/iavf/iavf_rxtx.h\n+++ b/drivers/net/iavf/iavf_rxtx.h\n@@ -297,6 +297,8 @@ struct iavf_tx_queue {\n \tuint16_t next_rs; /* next to check DD, for VPMD */\n \tuint16_t ipsec_crypto_pkt_md_offset;\n \n+\tuint64_t mbuf_errors;\n+\n \tbool q_set; /* if rx queue has been configured */\n \tbool tx_deferred_start; /* don't start this queue in dev start */\n \tconst struct iavf_txq_ops *ops;\n", "prefixes": [ "v8", "2/2" ] }