From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga04.intel.com (mga04.intel.com [192.55.52.120]) by dpdk.org (Postfix) with ESMTP id BA31393FE for ; Fri, 5 Feb 2016 03:46:19 +0100 (CET) Received: from orsmga001.jf.intel.com ([10.7.209.18]) by fmsmga104.fm.intel.com with ESMTP; 04 Feb 2016 18:46:19 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.22,398,1449561600"; d="scan'208";a="877436401" Received: from shvmail01.sh.intel.com ([10.239.29.42]) by orsmga001.jf.intel.com with ESMTP; 04 Feb 2016 18:46:18 -0800 Received: from shecgisg004.sh.intel.com (shecgisg004.sh.intel.com [10.239.29.89]) by shvmail01.sh.intel.com with ESMTP id u152kGOi030719; Fri, 5 Feb 2016 10:46:16 +0800 Received: from shecgisg004.sh.intel.com (localhost [127.0.0.1]) by shecgisg004.sh.intel.com (8.13.6/8.13.6/SuSE Linux 0.8) with ESMTP id u152kDeP021637; Fri, 5 Feb 2016 10:46:15 +0800 Received: (from heshaope@localhost) by shecgisg004.sh.intel.com (8.13.6/8.13.6/Submit) id u152kDa4021633; Fri, 5 Feb 2016 10:46:13 +0800 From: Shaopeng He To: dev@dpdk.org Date: Fri, 5 Feb 2016 10:46:11 +0800 Message-Id: <1454640371-21600-1-git-send-email-shaopeng.he@intel.com> X-Mailer: git-send-email 1.7.4.1 In-Reply-To: <1454589920-801-1-git-send-email-shaopeng.he@intel.com> References: <1454589920-801-1-git-send-email-shaopeng.he@intel.com> Subject: [dpdk-dev] [PATCH v3] fm10k: fix switch manager high CPU usage X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 05 Feb 2016 02:46:20 -0000 fm10k switch core uses source MAC + VID + SGLORT to do look up in MAC table. If no match, an exception interrupt will be sent to the switch manager. Too much of this kind of exception interrupts cause switch manager side high CPU usage. To reproduce this issue, one DPDK testpmd runs on a server with one fm10k NIC, mac forwards test traffic from one of fm10k ports to another port. The CPU usage for the switch manager will go up to about 20% for test traffic rate at 10G bps, comparing to near 0% for no test traffic. This patch fixes this issue. A default SGLORT is assigned to each TX queue. This default value works for non-VMDq mode and current VMDq example. For advanced VMDq usage, e.g. different source MAC address for different TX queue, FTAG forwarding function could be used to change this default SGLORT value. Fixes: 9ae6068c ("fm10k: add dev start/stop") Signed-off-by: Shaopeng He --- v3: - add fixes line v2: - reword the commit log message --- drivers/net/fm10k/fm10k_ethdev.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/fm10k/fm10k_ethdev.c b/drivers/net/fm10k/fm10k_ethdev.c index e4aed94..f6eb05d 100644 --- a/drivers/net/fm10k/fm10k_ethdev.c +++ b/drivers/net/fm10k/fm10k_ethdev.c @@ -675,6 +675,9 @@ fm10k_dev_tx_init(struct rte_eth_dev *dev) FM10K_WRITE_REG(hw, FM10K_TDBAH(i), base_addr >> (CHAR_BIT * sizeof(uint32_t))); FM10K_WRITE_REG(hw, FM10K_TDLEN(i), size); + + /* assign default SGLORT for each TX queue */ + FM10K_WRITE_REG(hw, FM10K_TX_SGLORT(i), hw->mac.dglort_map); } /* set up vector or scalar TX function as appropriate */ -- 1.9.3