From: Yingya Han <yingyax.han@intel.com>
To: dts@dpdk.org
Cc: Yingya Han <yingyax.han@intel.com>
Subject: [dts][PATCH V1 2/3]tests: add ice_header_split_perf test script
Date: Mon, 26 Dec 2022 06:55:41 +0000 [thread overview]
Message-ID: <20221226065541.10045-1-yingyax.han@intel.com> (raw)
Signed-off-by: Yingya Han <yingyax.han@intel.com>
---
| 416 +++++++++++++++++++++++
1 file changed, 416 insertions(+)
create mode 100644 tests/TestSuite_ice_header_split_perf.py
--git a/tests/TestSuite_ice_header_split_perf.py b/tests/TestSuite_ice_header_split_perf.py
new file mode 100644
index 00000000..e7ce11ca
--- /dev/null
+++ b/tests/TestSuite_ice_header_split_perf.py
@@ -0,0 +1,416 @@
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright(c) 2022 Intel Corporation
+#
+
+"""
+DPDK Test suite.
+"""
+
+import os
+import re
+import time
+from copy import deepcopy
+
+import framework.utils as utils
+from framework.exception import VerifyFailure
+from framework.packet import Packet
+from framework.pktgen import TRANSMIT_CONT
+from framework.settings import HEADER_SIZE
+from framework.test_case import TestCase
+from framework.utils import convert_int2ip, convert_ip2int
+
+
+class TestIceHeaderSplitPerf(TestCase):
+ #
+ # Test cases.
+ #
+ def set_up_all(self):
+ """
+ Run at the start of each test suite.
+ """
+ self.verify(
+ self.nic
+ in ["ICE_100G-E810C_QSFP", "ICE_25G-E810C_SFP", "ICE_25G-E810_XXV_SFP"],
+ "NIC Unsupported: " + str(self.nic),
+ )
+ self.dut_ports = self.dut.get_ports(self.nic)
+ self.verify(len(self.dut_ports) >= 1, "At least 1 port is required to test")
+ # Get socket and cores
+ self.socket = self.dut.get_numa_id(self.dut_ports[0])
+ cores = self.dut.get_core_list("1S/8C/1T", socket=self.socket)
+ self.verify(cores, "Requested 8 cores failed")
+ self.core_offset = 3
+ self.test_content = self.get_test_content_from_cfg(self.get_suite_cfg())
+
+ def set_up(self):
+ """
+ Run before each test case.
+ """
+ self.test_result = {"header": [], "data": []}
+
+ def flows(self):
+ """
+ Return a list of packets that implements the flows described.
+ """
+ return [
+ "198.18.0.0/24",
+ "198.18.1.0/24",
+ "198.18.2.0/24",
+ "198.18.3.0/24",
+ "198.18.4.0/24",
+ "198.18.5.0/24",
+ "198.18.6.0/24",
+ "198.18.7.0/24",
+ ]
+
+ def parse_test_config(self, config):
+ """
+ [n]C/[mT]-[i]Q
+ n: how many physical core use for polling.
+ m: how many cpu thread use for polling, if Hyper-threading disabled
+ in BIOS, m equals n, if enabled, m is 2 times as n.
+ i: how many queues use per port, so total queues = i x nb_port
+ """
+ pat = "(.*)/(.*)-(.*)"
+ result = re.findall(pat, config)
+ if not result:
+ msg = f"{config} is wrong format, please check"
+ raise VerifyFailure(msg)
+ cores, threads, queue = result[0]
+ _thread_num = int(int(threads[:-1]) // int(cores[:-1]))
+
+ _thread = str(_thread_num) + "T"
+ _cores = str(self.core_offset + int(cores[:-1])) + "C"
+ cores_config = "/".join(["1S", _cores, _thread])
+ queues_per_port = int(queue[:-1])
+ return cores_config, _thread_num, queues_per_port
+
+ def get_test_configs(self, test_parameters):
+ configs = []
+ frame_sizes_grp = []
+ for test_item, frame_sizes in sorted(test_parameters.items()):
+ _frame_sizes = [int(frame_size) for frame_size in frame_sizes]
+ frame_sizes_grp.extend([int(item) for item in _frame_sizes])
+ cores, thread_num, queues = self.parse_test_config(test_item)
+ corelist = self.dut.get_core_list(cores, self.socket)
+ core_list = corelist[(self.core_offset - 1) * thread_num :]
+ if "2T" in cores:
+ core_list = core_list[1:2] + core_list[0::2] + core_list[1::2][1:]
+ _core_list = core_list[thread_num - 1 :]
+ configs.append(
+ [
+ test_item,
+ _core_list,
+ [
+ " --txd=1024 --rxd=1024"
+ + " --rxq={0} --txq={0}".format(queues)
+ + " --nb-cores={}".format(len(core_list) - thread_num)
+ ],
+ ]
+ )
+ return configs, sorted(set(frame_sizes_grp))
+
+ def get_test_content_from_cfg(self, test_content):
+ test_content["flows"] = self.flows()
+ configs, frame_sizes = self.get_test_configs(test_content["test_parameters"])
+ test_content["configs"] = configs
+ test_content["frame_sizes"] = frame_sizes
+ return test_content
+
+ def get_mac_layer(self, port_id):
+ smac = "02:00:00:00:00:0%d" % port_id
+ dmac = "52:00:00:00:00:0%d" % port_id
+
+ layer = {
+ "ether": {
+ "dst": dmac,
+ "src": smac,
+ },
+ }
+ return layer
+
+ def get_ipv4_config(self, config):
+ netaddr, mask = config.split("/")
+ ip_range = int("1" * (32 - int(mask)), 2)
+ start_ip = convert_int2ip(convert_ip2int(netaddr) + 1)
+ end_ip = convert_int2ip(convert_ip2int(start_ip) + ip_range - 1)
+ layers = {
+ "ipv4": {
+ "src": start_ip,
+ },
+ }
+ fields_config = {
+ "ip": {
+ "src": {
+ "start": start_ip,
+ "end": end_ip,
+ "step": 1,
+ "action": "random",
+ },
+ },
+ }
+ return layers, fields_config
+
+ def preset_flows_configs(self):
+ flows = self.test_content.get("flows")
+ flows_configs = []
+ for index, config in enumerate(flows):
+ if index >= len(self.dut_ports):
+ break
+ port_id = self.dut_ports[index]
+ _layer = self.get_mac_layer(port_id)
+ _layer2, fields_config = self.get_ipv4_config(config)
+ _layer.update(_layer2)
+ flows_configs.append([_layer, fields_config])
+ return flows_configs
+
+ def preset_streams(self, pkt_type):
+ frame_sizes = self.test_content.get("frame_sizes")
+ test_streams = {}
+ frame_sizes_new = []
+ flows_configs = self.preset_flows_configs()
+ for frame_size in frame_sizes:
+ frame_size = 66 if "udp" == pkt_type and frame_size == 64 else frame_size
+ frame_sizes_new.append(frame_size)
+ for flow_config in flows_configs:
+ _layers, fields_config = flow_config
+ pkt = self.config_stream(_layers, frame_size, pkt_type)
+ test_streams.setdefault(frame_size, []).append([pkt, fields_config])
+ return test_streams, frame_sizes_new
+
+ def config_stream(self, layers, frame_size, pkt_mode):
+ """
+ Prepare traffic flow
+ """
+ if pkt_mode == "udp":
+ headers_size = sum([HEADER_SIZE[x] for x in ["eth", "ip", "udp"]])
+ payload_size = frame_size - headers_size
+ pkt_config = {
+ "type": "UDP",
+ "pkt_layers": {"raw": {"payload": ["58"] * payload_size}},
+ }
+ layers["udp"] = {"src": 53, "dst": 53}
+ else:
+ headers_size = sum([HEADER_SIZE[x] for x in ["eth", "ip"]])
+ payload_size = frame_size - headers_size
+ # Set streams for traffic
+ pkt_config = {
+ "type": "IP_RAW",
+ "pkt_layers": {"raw": {"payload": ["58"] * payload_size}},
+ }
+ pkt_config["pkt_layers"].update(layers)
+ pkt_type = pkt_config.get("type")
+ pkt_layers = pkt_config.get("pkt_layers")
+ pkt = Packet(pkt_type=pkt_type)
+ for layer in list(pkt_layers.keys()):
+ pkt.config_layer(layer, pkt_layers[layer])
+
+ return pkt.pktgen.pkt
+
+ def add_stream_to_pktgen(self, streams, option):
+ def port(index):
+ p = self.tester.get_local_port(self.dut_ports[index])
+ return p
+
+ topos = (
+ [
+ [port(index), port(index - 1)]
+ if index % 2
+ else [port(index), port(index + 1)]
+ for index, _ in enumerate(self.dut_ports)
+ ]
+ if len(self.dut_ports) > 1
+ else [[port(0), port(0)]]
+ )
+ stream_ids = []
+ step = int(len(streams) / len(self.dut_ports))
+ for cnt, stream in enumerate(streams):
+ pkt, fields_config = stream
+ index = cnt // step
+ txport, rxport = topos[index]
+ _option = deepcopy(option)
+ _option["pcap"] = pkt
+ if fields_config:
+ _option["fields_config"] = fields_config
+ stream_id = self.tester.pktgen.add_stream(txport, rxport, pkt)
+ self.tester.pktgen.config_stream(stream_id, _option)
+ stream_ids.append(stream_id)
+ return stream_ids
+
+ def start_testpmd(self, eal_pare, eal):
+ bin = os.path.join(self.dut.base_dir, self.dut.apps_name["test-pmd"])
+ command_line = (
+ "{bin} "
+ "{eal_para}"
+ " --force-max-simd-bitwidth=64 "
+ "-- -i "
+ "--portmask {port_mask} "
+ "{config} "
+ ""
+ ).format(
+ **{
+ "bin": bin,
+ "eal_para": eal_pare,
+ "port_mask": utils.create_mask(self.dut_ports),
+ "config": eal,
+ }
+ )
+ self.dut.send_expect(command_line, "testpmd>", 60)
+
+ def throughput(self, frame_size, fwd_mode):
+ streams = self.stream.get(frame_size)
+ # Set traffic option
+ duration = self.test_content.get("test_duration")
+ traffic_stop_wait_time = self.test_content.get("traffic_stop_wait_time", 0)
+ # Clear streams before adding new streams
+ self.tester.pktgen.clear_streams()
+ # Set stream into pktgen
+ stream_option = {
+ "stream_config": {
+ "txmode": {},
+ "transmit_mode": TRANSMIT_CONT,
+ "rate": 100,
+ }
+ }
+ traffic_option = {
+ "method": "throughput",
+ "duration": duration,
+ }
+ stream_ids = self.add_stream_to_pktgen(streams, stream_option)
+ # Run packet generator
+ result = self.tester.pktgen.measure(stream_ids, traffic_option)
+ time.sleep(traffic_stop_wait_time)
+ # Statistics result
+ if fwd_mode == "rxonly":
+ # Calculate rx throughput
+ out = self.dut.send_expect("show port stats all", "testpmd> ", 15)
+ pattern = re.compile(r"(?<=RX-packets: )\d+\.?\d*")
+ pps = int(pattern.findall(out)[0]) / duration
+ else:
+ _, pps = result
+ self.verify(pps > 0, "No traffic detected")
+ self.logger.info(
+ "Throughput of "
+ + "framesize: {}, is: {} Mpps".format(frame_size, pps / 1000000)
+ )
+ return pps
+
+ def get_port_allowlist(self, port_list=None):
+ allowlist = []
+ if port_list:
+ for port_id in port_list:
+ pci = self.dut.ports_info[port_id]["port"].pci
+ allowlist.append(pci)
+ return allowlist
+
+ def display_result(self, datas):
+ # Display result table
+ header_row = ["Fwd Core", "Frame Size", "Throughput", "Rate"]
+ self.test_result["header"] = header_row
+ self.result_table_create(header_row)
+ self.test_result["data"] = []
+ for data in datas:
+ config, frame_size, pps = data
+ pps /= 1000000.0
+ linerate = self.wirespeed(self.nic, frame_size, len(self.dut_ports))
+ percentage = pps * 100 / linerate
+ data_row = [
+ config,
+ frame_size,
+ "{:.3f} Mpps".format(pps),
+ "{:.3f}%".format(percentage),
+ ]
+ self.result_table_add(data_row)
+ self.test_result["data"].append(data_row)
+ self.result_table_print()
+
+ def config_header_split(self, split_type):
+ self.dut.send_expect("port stop all", "testpmd> ", 15)
+ for i in self.dut_ports:
+ self.dut.send_expect(
+ "port config {} rx_offload buffer_split on".format(i), "testpmd> ", 15
+ )
+ if split_type == "eth":
+ self.dut.send_expect("set rxhdrs eth", "testpmd> ", 15)
+ elif split_type == "udp":
+ self.dut.send_expect("set rxhdrs inner-ipv4-udp", "testpmd> ", 15)
+ self.dut.send_expect("port start all", "testpmd> ", 15)
+ self.check_port_stats()
+
+ def check_port_stats(self, times=60):
+ # Check port status up
+ for i in range(times):
+ out = self.dut.send_expect("show port info all", "testpmd> ")
+ pattern = r"(?<=%s)%s" % ("Link status: ", "\S+")
+ s = re.compile(pattern)
+ status = s.findall(out)
+ if "down" not in status:
+ break
+ time.sleep(1)
+
+ def perf_test(self, fwd_mode="", mbuf_size="", header_split="", pkt_type=""):
+ """
+ Benchmarking test
+ """
+ self.stream, self.frame_sizes = self.preset_streams(pkt_type)
+ # Ports allow list
+ port_allowlist = self.get_port_allowlist(self.dut_ports)
+ forward = " --forward={}".format(fwd_mode)
+ results = []
+ for config, core_list, eal in self.test_content["configs"]:
+ self.logger.info(
+ ("Executing Test Using cores: {0} of config {1}, ").format(
+ core_list, config
+ )
+ )
+ eal_pare = self.dut.create_eal_parameters(
+ cores=core_list, ports=port_allowlist, socket=self.socket
+ )
+ eal = eal[0] + forward
+ if "enable" == mbuf_size:
+ eal += " --mbuf-size=2048,2048"
+ self.start_testpmd(eal_pare, eal)
+ self.check_port_stats()
+ if header_split:
+ self.config_header_split(header_split)
+ self.dut.send_expect("start", "testpmd> ", 15)
+ for frame_size in self.frame_sizes:
+ self.logger.info("Test running at framesize: {}".format(frame_size))
+ self.dut.send_expect("clear port stats all", "testpmd> ", 15)
+ result = self.throughput(frame_size, fwd_mode)
+ if result:
+ results.append([config, frame_size, result])
+ self.dut.send_expect("stop", "testpmd> ", 15)
+ self.dut.send_expect("quit", "# ", 15)
+ self.display_result(results)
+
+ def test_perf_enable_header_split_rx_on(self):
+ self.perf_test(fwd_mode="rxonly", mbuf_size="enable", header_split="eth")
+
+ def test_perf_disable_header_split_rx_on(self):
+ self.perf_test(fwd_mode="rxonly")
+ self.perf_test(fwd_mode="rxonly", mbuf_size="enable")
+
+ def test_perf_enable_header_split(self):
+ self.perf_test(
+ fwd_mode="rxonly", mbuf_size="enable", header_split="udp", pkt_type="ip"
+ )
+ self.perf_test(
+ fwd_mode="rxonly", mbuf_size="enable", header_split="udp", pkt_type="udp"
+ )
+
+ def test_perf_disable_header_split(self):
+ self.perf_test(fwd_mode="io")
+ self.perf_test(fwd_mode="io", mbuf_size="enable")
+
+ def tear_down(self):
+ """
+ Run after each test case.
+ """
+ pass
+
+ def tear_down_all(self):
+ """
+ Run after each test suite.
+ """
+ self.dut.kill_all()
--
2.34.1
reply other threads:[~2022-12-26 6:55 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=20221226065541.10045-1-yingyax.han@intel.com \
--to=yingyax.han@intel.com \
--cc=dts@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).