From patchwork Thu Aug 11 11:16:57 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Yaqi Tang X-Patchwork-Id: 114810 Return-Path: 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]) by inbox.dpdk.org (Postfix) with ESMTP id D26E5A0545; Thu, 11 Aug 2022 05:18:54 +0200 (CEST) Received: from [217.70.189.124] (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id CCBF540DDE; Thu, 11 Aug 2022 05:18:54 +0200 (CEST) Received: from mga04.intel.com (mga04.intel.com [192.55.52.120]) by mails.dpdk.org (Postfix) with ESMTP id 6D7B44068E for ; Thu, 11 Aug 2022 05:18:52 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1660187932; x=1691723932; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=D5YE8JovFgqyy744jL6e+Gfd1EBtEqYKn/WJZEvNrbU=; b=bIqBYP7lEI6QbnEkzOSXl5pAsKWigV1Q/+bNoAQ7Aqkx82+TTwEMYddw zIyr6Qu03VK2P+z3FE+fChfsjTUdRG9KqTj+UgSrt4aBC22JfABgvRnvI i2QRRLekCbayU0otcMoAqw534/DZGaG7uWi3yJdwVaFOfsDF8mge1mhXl m5o2o3I5D5I72oFD75clw3Kb0be3VcTI+ehxypk7/ymcLTKafzjvNrD5y xKCUFKz9qGevEuzNaZ4z1QPYDpgfCxkJNCsVlSaKBNmJc/ZVl5HpOIZbP 7DnbxFk98yxTVkCwMMkaUiJ/izNuyb9G4WGF1k05DhKWkKG60paP4PNLI Q==; X-IronPort-AV: E=McAfee;i="6400,9594,10435"; a="290001139" X-IronPort-AV: E=Sophos;i="5.93,228,1654585200"; d="scan'208";a="290001139" Received: from orsmga006.jf.intel.com ([10.7.209.51]) by fmsmga104.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 10 Aug 2022 20:18:51 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.93,228,1654585200"; d="scan'208";a="581490808" Received: from dpdk-yaqi.sh.intel.com ([10.67.119.99]) by orsmga006.jf.intel.com with ESMTP; 10 Aug 2022 20:18:50 -0700 From: Yaqi Tang To: dts@dpdk.org Cc: Yaqi Tang Subject: [dts][PATCH V3 2/3] test_plans/ice_iavf_packet_pacing: ice enable packet pacing Date: Thu, 11 Aug 2022 11:16:57 +0000 Message-Id: <20220811111658.78579-3-yaqi.tang@intel.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220811111658.78579-1-yaqi.tang@intel.com> References: <20220811111658.78579-1-yaqi.tang@intel.com> MIME-Version: 1.0 X-BeenThere: dts@dpdk.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: test suite reviews and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dts-bounces@dpdk.org ICE support queue rate limitation configure, and parse the quanta_size for IGP pacing. Signed-off-by: Yaqi Tang --- .../ice_iavf_packet_pacing_test_plan.rst | 383 ++++++++++++++++++ 1 file changed, 383 insertions(+) create mode 100644 test_plans/ice_iavf_packet_pacing_test_plan.rst diff --git a/test_plans/ice_iavf_packet_pacing_test_plan.rst b/test_plans/ice_iavf_packet_pacing_test_plan.rst new file mode 100644 index 00000000..92969dac --- /dev/null +++ b/test_plans/ice_iavf_packet_pacing_test_plan.rst @@ -0,0 +1,383 @@ +.. SPDX-License-Identifier: BSD-3-Clause + Copyright(c) 2022 Intel Corporation + +============================= +ICE IAVF Enable Packet Pacing +============================= + +Description +=========== +1. Enable rte_tm to support queue rate limitation configure. +2. Enable devargs to parse the quanta_size for IGP pacing. + +The packet pacing quality is heavily rely on FlexRan application/OS setup. +For DPDK validation, we will focus on: +1. Make sure queue rate limitation configuration take effect by measure the bandwidth. +2. Should not exceed the bandwidth that be limited by the DCF. +3. Not to break anything else. + +Prerequisites +============= + +Topology +-------- +DUT port 0 <----> Tester port 0 + +Hardware +-------- +Supported NICs: Intel® Ethernet 800 Series E810-XXVDA4/E810-CQ + +Software +-------- +dpdk: http://dpdk.org/git/dpdk +runtime command: https://doc.dpdk.org/guides/testpmd_app_ug/testpmd_funcs.html + +General Set Up +-------------- +1. Compile DPDK:: + + # CC=gcc meson --werror -Denable_kmods=True -Dlibdir=lib --default-library=static + # ninja -C -j 110 + +2. Get the pci device id and interface of DUT and tester. + For example, 0000:3b:00.0 and 0000:3b:00.1 is pci device id, + ens785f0 and ens785f1 is interface:: + + # ./usertools/dpdk-devbind.py -s + + 0000:3b:00.0 'Device 159b' if=ens785f0 drv=ice unused=vfio-pci + 0000:3b:00.1 'Device 159b' if=ens785f1 drv=ice unused=vfio-pci + +3. If vf_num >= 2, Generate 2 VF on PF0, set mac address for VF1:: + + # echo 2 > /sys/bus/pci/devices/0000:3b:00.0/sriov_numvfs + # ip link set dev ens785f0 vf 0 trust on + # ip link set ens785f0 vf 1 mac 00:11:22:33:44:55 + + Else, Generate 1 VF on PF0, set mac address for VF0:: + + # echo 1 > /sys/bus/pci/devices/0000:3b:00.0/sriov_numvfs + # ip link set ens785f0 vf 0 mac 00:11:22:33:44:55 + +4. Bind the DUT port to dpdk:: + + # ./usertools/dpdk-devbind.py -b vfio-pci + +Test Case +========= +Common Steps +------------ +TREX sends one stream with frame size 64, mac address is "00:11:22:33:44:55", 100% max rate. +When check the throughput ratio of each queue, stop the forward and check the TX-packets ratio of queues. +The TX-packets ratio of queues is same as TX throughput ratio of queues. + +Test Case 1: Without quanta size, check peak_tb_rate +---------------------------------------------------- +Launch testpmd without quanta size, check the TX throughput. + +Test Steps +~~~~~~~~~~ +1. Launch testpmd:: + + # .//app/dpdk-testpmd -l 1-4 -n 4 -a 0000:3b:01.0 --file-prefix=tx -- -i + +2. Configure rate limit:: + + testpmd> port stop all + testpmd> add port tm node shaper profile 0 1 1000000 0 612980769 0 0 0 + testpmd> add port tm nonleaf node 0 1000 -1 0 1 0 -1 1 0 0 + testpmd> add port tm nonleaf node 0 900 1000 0 1 1 -1 1 0 0 + testpmd> add port tm leaf node 0 0 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> port tm hierarchy commit 0 no + testpmd> port start all + testpmd> set fwd mac + testpmd> start + +3. Send one stream from TREX. + +4. Check the TX throughput of port 0:: + + testpmd> show port stats 0 + + Check the TX throughput is 612MBps. + +Test Case 2: Single queue, check peak_tb_rate +--------------------------------------------- +Launch testpmd with quanta_size=1024, check the TX throughput. + +Test Steps +~~~~~~~~~~ +1. Launch testpmd:: + + # .//app/dpdk-testpmd -l 1-4 -n 4 -a 0000:3b:01.0,quanta_size=1024 --file-prefix=tx -- -i + +2. Configure rate limit:: + + testpmd> port stop all + testpmd> add port tm node shaper profile 0 1 1000000 0 612980769 0 0 0 + testpmd> add port tm nonleaf node 0 1000 -1 0 1 0 -1 1 0 0 + testpmd> add port tm nonleaf node 0 900 1000 0 1 1 -1 1 0 0 + testpmd> add port tm leaf node 0 0 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> port tm hierarchy commit 0 no + testpmd> port start all + testpmd> set fwd mac + testpmd> start + +3. Send one stream from TREX. + +4. Check the TX throughput of port 0:: + + testpmd> show port stats 0 + + Check the TX throughput is 612MBps. + +Test Case 3: Multi queues, check peak_tb_rate +--------------------------------------------- +Launch testpmd with quanta_size=1024 and multi queues, check the TX throughput and the throughput ratio of each queue. + +Test Steps +~~~~~~~~~~ +1. Launch testpmd with 8 queues:: + + # .//app/dpdk-testpmd -l 1-4 -n 4 -a 0000:3b:01.0,quanta_size=1024 --file-prefix=tx -- -i --txq=8 --rxq=8 + +2. Configure each queue with same rate limit:: + + testpmd> port stop all + testpmd> add port tm node shaper profile 0 1 1000000 0 100000000 0 0 0 + testpmd> add port tm nonleaf node 0 1000 -1 0 1 0 -1 1 0 0 + testpmd> add port tm nonleaf node 0 900 1000 0 1 1 -1 1 0 0 + testpmd> add port tm leaf node 0 0 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 1 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 2 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 3 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 4 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 5 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 6 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 7 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> port tm hierarchy commit 0 no + testpmd> port start all + testpmd> set fwd mac + testpmd> start + +3. Send one stream from TREX. + +4. Check the TX throughput of port 0:: + + testpmd> show port stats 0 + + Check the TX throughput is 800MBps. + + Check the throughput ratio of each queue:: + + testpmd> stop + + Check the TX throughput of queue 0-7 is 1:1:1:1:1:1:1:1. + +Test Case 4: Modify quanta_size, check peak_tb_rate +--------------------------------------------------- +Launch testpmd with quanta_size=4096 and multi queues, check the TX throughput and the throughput ratio of each queue. + +Test Steps +~~~~~~~~~~ +1. Launch testpmd with 8 queues:: + + # .//app/dpdk-testpmd -l 1-4 -n 4 -a 0000:3b:01.0,quanta_size=4096 --file-prefix=tx -- -i --txq=8 --rxq=8 + +2. Configure each queue with same rate limit:: + + testpmd> port stop all + testpmd> add port tm node shaper profile 0 1 1000000 0 100000000 0 0 0 + testpmd> add port tm nonleaf node 0 1000 -1 0 1 0 -1 1 0 0 + testpmd> add port tm nonleaf node 0 900 1000 0 1 1 -1 1 0 0 + testpmd> add port tm leaf node 0 0 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 1 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 2 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 3 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 4 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 5 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 6 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 7 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> port tm hierarchy commit 0 no + testpmd> port start all + testpmd> set fwd mac + testpmd> start + +3. Send one stream from TREX. + +4. Check the TX throughput of port 0:: + + testpmd> show port stats 0 + + Check the TX throughput is 800MBps. + + Check the throughput ratio of each queue:: + + testpmd> stop + + Check the TX throughput of queue 0-7 is 1:1:1:1:1:1:1:1. + +Test Case 5: Invalid quanta_size, check peak_tb_rate +---------------------------------------------------- +The scope of quanta size is [256, 4096], and it should be the product of 64. +Launch testpmd with quanta_size=1000, check it shows invalid quanta size. + +Test Steps +~~~~~~~~~~ +1. Launch testpmd:: + + # .//app/dpdk-testpmd -l 1-4 -n 4 -a 0000:3b:01.0,quanta_size=1000 --file-prefix=tx -- -i + +2. check it shows invalid quanta size. + +Test Case 6: Multi queues with different rate limit, check peak_tb_rate +----------------------------------------------------------------------- +Launch testpmd with quanta_size=1024 and multi queues, then configure each queue with diff rate limit, +check the TX throughput and the throughput ratio of each queue. + +Test Steps +~~~~~~~~~~ +1. Launch testpmd with 8 queues:: + + # .//app/dpdk-testpmd -l 1-4 -n 4 -a 0000:3b:01.0,quanta_size=1024 --file-prefix=tx -- -i --txq=8 --rxq=8 + +2. Configure each queue with diff rate limit:: + + testpmd> port stop all + testpmd> add port tm node shaper profile 0 1 1000000 0 10000000 0 0 0 + testpmd> add port tm node shaper profile 0 2 1000000 0 20000000 0 0 0 + testpmd> add port tm node shaper profile 0 3 1000000 0 30000000 0 0 0 + testpmd> add port tm node shaper profile 0 4 1000000 0 40000000 0 0 0 + testpmd> add port tm node shaper profile 0 5 1000000 0 50000000 0 0 0 + testpmd> add port tm node shaper profile 0 6 1000000 0 60000000 0 0 0 + testpmd> add port tm node shaper profile 0 7 1000000 0 70000000 0 0 0 + testpmd> add port tm node shaper profile 0 8 1000000 0 80000000 0 0 0 + testpmd> add port tm nonleaf node 0 1000 -1 0 1 0 -1 1 0 0 + testpmd> add port tm nonleaf node 0 900 1000 0 1 1 -1 1 0 0 + testpmd> add port tm leaf node 0 0 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 1 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 2 900 0 1 2 3 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 3 900 0 1 2 4 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 4 900 0 1 2 5 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 5 900 0 1 2 6 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 6 900 0 1 2 7 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 7 900 0 1 2 8 0 0xffffffff 0 0 + testpmd> port tm hierarchy commit 0 no + testpmd> port start all + testpmd> set fwd mac + testpmd> start + +3. Send one stream from TREX. + +4. Check the TX throughput of port 0:: + + testpmd> show port stats 0 + + Check the TX throughput is 360MBps. + + Check the throughput ratio of each queue:: + + testpmd> stop + + Check the TX throughput of queue 0-7 is 1:2:3:4:5:6:7:8. + +Test Case 7: Port rate limit less than queue rate limit +------------------------------------------------------- +Launch dcf testpmd with quanta_size=1024 and multi queues, then configure port rate limit less than queue rate limit, +check the TX throughput and the throughput ratio of each queue. + +Test Steps +~~~~~~~~~~ +1. Launch dcf testpmd with 8 queues:: + + # .//app/dpdk-testpmd -l 1-4 -n 4 -a 0000:3b:01.0,cap=dcf -a 3b:01.1,quanta_size=1024 --file-prefix=tx -- -i --txq=8 --rxq=8 --port-topology=loop + +2. Configure port rate limit less than queue rate limit:: + + testpmd> port stop all + testpmd> add port tm node shaper profile 0 1 1000000 0 100000000 0 0 0 + testpmd> add port tm node shaper profile 1 2 1000000 0 612980769 0 0 0 + testpmd> add port tm nonleaf node 0 1000 -1 0 1 0 -1 1 0 0 + testpmd> add port tm nonleaf node 0 900 1000 0 1 1 -1 1 0 0 + testpmd> add port tm leaf node 0 0 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 1 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm nonleaf node 1 1000 -1 0 1 0 -1 1 0 0 + testpmd> add port tm nonleaf node 1 900 1000 0 1 1 -1 1 0 0 + testpmd> add port tm leaf node 1 0 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 1 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 2 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 3 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 4 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 5 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 6 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 7 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> port tm hierarchy commit 0 no + testpmd> port tm hierarchy commit 1 no + testpmd> port start all + testpmd> set fwd mac + testpmd> start + +3. Send one stream from TREX. + +4. Check the TX throughput of port 1:: + + testpmd> show port stats 1 + + Check the TX throughput is 100MBps. + + Check the throughput ratio of each queue:: + + testpmd> stop + + Check the TX throughput of port1 queue 0-7 is 1:1:1:1:1:1:1:1. + +Test Case 8: Port rate limit more than queue rate limit +------------------------------------------------------- +Launch dcf testpmd with quanta_size=1024 and multi queues, then configure port rate limit more than queue rate limit, +check the TX throughput and the throughput ratio of each queue. + +Test Steps +~~~~~~~~~~ +1. Launch dcf testpmd with 8 queues:: + + # .//app/dpdk-testpmd -l 1-4 -n 4 -a 0000:3b:01.0,cap=dcf -a 3b:01.1,quanta_size=1024 --file-prefix=tx -- -i --txq=8 --rxq=8 --port-topology=loop + +2. Configure port rate limit more than queue rate limit:: + + testpmd> port stop all + testpmd> add port tm node shaper profile 0 1 1000000 0 200000000 0 0 0 + testpmd> add port tm node shaper profile 1 2 1000000 0 10000000 0 0 0 + testpmd> add port tm nonleaf node 0 1000 -1 0 1 0 -1 1 0 0 + testpmd> add port tm nonleaf node 0 900 1000 0 1 1 -1 1 0 0 + testpmd> add port tm leaf node 0 0 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm leaf node 0 1 900 0 1 2 1 0 0xffffffff 0 0 + testpmd> add port tm nonleaf node 1 1000 -1 0 1 0 -1 1 0 0 + testpmd> add port tm nonleaf node 1 900 1000 0 1 1 -1 1 0 0 + testpmd> add port tm leaf node 1 0 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 1 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 2 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 3 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 4 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 5 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 6 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> add port tm leaf node 1 7 900 0 1 2 2 0 0xffffffff 0 0 + testpmd> port tm hierarchy commit 0 no + testpmd> port tm hierarchy commit 1 no + testpmd> port start all + testpmd> set fwd mac + testpmd> start + +3. Send one stream from TREX. + +4. Check the TX throughput of port 1:: + + testpmd> show port stats 1 + + Check the TX throughput is 80MBps. + + Check the throughput ratio of each queue:: + + testpmd> stop + + Check the TX throughput of port1 queue 0-7 is 1:1:1:1:1:1:1:1. +