From: Hongbo Li <hongbox.li@intel.com>
To: dts@dpdk.org
Cc: Hongbo Li <hongbox.li@intel.com>
Subject: [dts][PATCH V1 5/7] tests/vxlan:Separated performance cases
Date: Thu, 22 Sep 2022 14:29:48 +0000 [thread overview]
Message-ID: <20220922142950.398902-5-hongbox.li@intel.com> (raw)
In-Reply-To: <20220922142950.398902-1-hongbox.li@intel.com>
Separated performance cases
Signed-off-by: Hongbo Li <hongbox.li@intel.com>
---
test_plans/perf_vxlan_test_plan.rst | 84 ++++
test_plans/vxlan_test_plan.rst | 57 ---
tests/TestSuite_perf_vxlan.py | 691 ++++++++++++++++++++++++++++
tests/TestSuite_vxlan.py | 213 ---------
4 files changed, 775 insertions(+), 270 deletions(-)
create mode 100644 test_plans/perf_vxlan_test_plan.rst
create mode 100644 tests/TestSuite_perf_vxlan.py
diff --git a/test_plans/perf_vxlan_test_plan.rst b/test_plans/perf_vxlan_test_plan.rst
new file mode 100644
index 00000000..18347a94
--- /dev/null
+++ b/test_plans/perf_vxlan_test_plan.rst
@@ -0,0 +1,84 @@
+.. SPDX-License-Identifier: BSD-3-Clause
+ Copyright(c) 2014-2017 Intel Corporation
+
+======================================
+Intel® Ethernet 700 Series Vxlan Tests
+======================================
+Cloud providers build virtual network overlays over existing network
+infrastructure that provide tenant isolation and scaling. Tunneling
+layers added to the packets carry the virtual networking frames over
+existing Layer 2 and IP networks. Conceptually, this is similar to
+creating virtual private networks over the Internet. Intel® Ethernet
+700 Series will process these tunneling layers by the hardware.
+
+This document provides test plan for Intel® Ethernet 700 Series vxlan
+packet detecting, checksum computing and filtering.
+
+Prerequisites
+=============
+1x Intel® X710 (Intel® Ethernet 700 Series) NICs (2x 40GbE full duplex
+optical ports per NIC) plugged into the available PCIe Gen3 8-lane slot.
+
+1x Intel® XL710-DA4 (Eagle Fountain) (1x 10GbE full duplex optical ports per NIC)
+plugged into the available PCIe Gen3 8-lane slot.
+
+DUT board must be two sockets system and each cpu have more than 8 lcores.
+
+
+Test Case: Vxlan Checksum Offload Performance Benchmarking
+==========================================================
+The throughput is measured for each of these cases for vxlan tx checksum
+offload of "all by software", "L3 offload by hardware", "L4 offload by
+hardware", "l3&l4 offload by hardware".
+
+The results are printed in the following table:
+
++----------------+--------+--------+------------+
+| Calculate Type | Queues | Mpps | % linerate |
++================+========+========+============+
+| SOFTWARE ALL | Single | | |
++----------------+--------+--------+------------+
+| HW L4 | Single | | |
++----------------+--------+--------+------------+
+| HW L3&L4 | Single | | |
++----------------+--------+--------+------------+
+| SOFTWARE ALL | Multi | | |
++----------------+--------+--------+------------+
+| HW L4 | Multi | | |
++----------------+--------+--------+------------+
+| HW L3&L4 | Multi | | |
++----------------+--------+--------+------------+
+
+Test Case: Vxlan Tunnel filter Performance Benchmarking
+=======================================================
+The throughput is measured for different Vxlan tunnel filter types.
+Queue single mean there's only one flow and forwarded to the first queue.
+Queue multi mean there are two flows and configure to different queues.
+
++--------+------------------+--------+--------+------------+
+| Packet | Filter | Queue | Mpps | % linerate |
++========+==================+========+========+============+
+| Normal | None | Single | | |
++--------+------------------+--------+--------+------------+
+| Vxlan | None | Single | | |
++--------+------------------+--------+--------+------------+
+| Vxlan | imac-ivlan | Single | | |
++--------+------------------+--------+--------+------------+
+| Vxlan | imac-ivlan-tenid | Single | | |
++--------+------------------+--------+--------+------------+
+| Vxlan | imac-tenid | Single | | |
++--------+------------------+--------+--------+------------+
+| Vxlan | imac | Single | | |
++--------+------------------+--------+--------+------------+
+| Vxlan | omac-imac-tenid | Single | | |
++--------+------------------+--------+--------+------------+
+| Vxlan | imac-ivlan | Multi | | |
++--------+------------------+--------+--------+------------+
+| Vxlan | imac-ivlan-tenid | Multi | | |
++--------+------------------+--------+--------+------------+
+| Vxlan | imac-tenid | Multi | | |
++--------+------------------+--------+--------+------------+
+| Vxlan | imac | Multi | | |
++--------+------------------+--------+--------+------------+
+| Vxlan | omac-imac-tenid | Multi | | |
++--------+------------------+--------+--------+------------+
diff --git a/test_plans/vxlan_test_plan.rst b/test_plans/vxlan_test_plan.rst
index d59cee31..96fdea0e 100644
--- a/test_plans/vxlan_test_plan.rst
+++ b/test_plans/vxlan_test_plan.rst
@@ -307,60 +307,3 @@ Add Cloud filter with invalid vni "16777216" will be failed.
Add Cloud filter with invalid queue id "64" will be failed.
-Test Case: Vxlan Checksum Offload Performance Benchmarking
-==========================================================
-The throughput is measured for each of these cases for vxlan tx checksum
-offload of "all by software", "L3 offload by hardware", "L4 offload by
-hardware", "l3&l4 offload by hardware".
-
-The results are printed in the following table:
-
-+----------------+--------+--------+------------+
-| Calculate Type | Queues | Mpps | % linerate |
-+================+========+========+============+
-| SOFTWARE ALL | Single | | |
-+----------------+--------+--------+------------+
-| HW L4 | Single | | |
-+----------------+--------+--------+------------+
-| HW L3&L4 | Single | | |
-+----------------+--------+--------+------------+
-| SOFTWARE ALL | Multi | | |
-+----------------+--------+--------+------------+
-| HW L4 | Multi | | |
-+----------------+--------+--------+------------+
-| HW L3&L4 | Multi | | |
-+----------------+--------+--------+------------+
-
-Test Case: Vxlan Tunnel filter Performance Benchmarking
-=======================================================
-The throughput is measured for different Vxlan tunnel filter types.
-Queue single mean there's only one flow and forwarded to the first queue.
-Queue multi mean there are two flows and configure to different queues.
-
-+--------+------------------+--------+--------+------------+
-| Packet | Filter | Queue | Mpps | % linerate |
-+========+==================+========+========+============+
-| Normal | None | Single | | |
-+--------+------------------+--------+--------+------------+
-| Vxlan | None | Single | | |
-+--------+------------------+--------+--------+------------+
-| Vxlan | imac-ivlan | Single | | |
-+--------+------------------+--------+--------+------------+
-| Vxlan | imac-ivlan-tenid | Single | | |
-+--------+------------------+--------+--------+------------+
-| Vxlan | imac-tenid | Single | | |
-+--------+------------------+--------+--------+------------+
-| Vxlan | imac | Single | | |
-+--------+------------------+--------+--------+------------+
-| Vxlan | omac-imac-tenid | Single | | |
-+--------+------------------+--------+--------+------------+
-| Vxlan | imac-ivlan | Multi | | |
-+--------+------------------+--------+--------+------------+
-| Vxlan | imac-ivlan-tenid | Multi | | |
-+--------+------------------+--------+--------+------------+
-| Vxlan | imac-tenid | Multi | | |
-+--------+------------------+--------+--------+------------+
-| Vxlan | imac | Multi | | |
-+--------+------------------+--------+--------+------------+
-| Vxlan | omac-imac-tenid | Multi | | |
-+--------+------------------+--------+--------+------------+
diff --git a/tests/TestSuite_perf_vxlan.py b/tests/TestSuite_perf_vxlan.py
new file mode 100644
index 00000000..4bd50c21
--- /dev/null
+++ b/tests/TestSuite_perf_vxlan.py
@@ -0,0 +1,691 @@
+import os
+import re
+import string
+import time
+from random import randint
+
+from scapy.config import conf
+from scapy.layers.inet import IP, TCP, UDP, Ether
+from scapy.layers.inet6 import IPv6
+from scapy.layers.l2 import Dot1Q
+from scapy.layers.sctp import SCTP, SCTPChunkData
+from scapy.layers.vxlan import VXLAN
+from scapy.route import *
+from scapy.sendrecv import sniff
+from scapy.utils import rdpcap, wrpcap
+
+import framework.packet as packet
+import framework.utils as utils
+from framework.packet import IncreaseIP, IncreaseIPv6
+from framework.pktgen import PacketGeneratorHelper
+from framework.pmd_output import PmdOutput
+from framework.settings import FOLDERS, HEADER_SIZE
+from framework.test_case import TestCase
+
+VXLAN_PORT = 4789
+PACKET_LEN = 128
+MAX_TXQ_RXQ = 4
+BIDIRECT = True
+
+
+class VxlanTestConfig(object):
+
+ """
+ Module for config/create/transmit vxlan packet
+ """
+
+ def __init__(self, test_case, **kwargs):
+ self.test_case = test_case
+ self.init()
+ for name in kwargs:
+ setattr(self, name, kwargs[name])
+ self.pkt_obj = packet.Packet()
+
+ def init(self):
+ self.packets_config()
+
+ def packets_config(self):
+ """
+ Default vxlan packet format
+ """
+ self.pcap_file = packet.TMP_PATH + "vxlan.pcap"
+ self.capture_file = packet.TMP_PATH + "vxlan_capture.pcap"
+ self.outer_mac_src = "00:00:10:00:00:00"
+ self.outer_mac_dst = "11:22:33:44:55:66"
+ self.outer_vlan = "N/A"
+ self.outer_ip_src = "192.168.1.1"
+ self.outer_ip_dst = "192.168.1.2"
+ self.outer_ip_invalid = 0
+ self.outer_ip6_src = "N/A"
+ self.outer_ip6_dst = "N/A"
+ self.outer_ip6_invalid = 0
+ self.outer_udp_src = 63
+ self.outer_udp_dst = VXLAN_PORT
+ self.outer_udp_invalid = 0
+ self.vni = 1
+ self.inner_mac_src = "00:00:20:00:00:00"
+ self.inner_mac_dst = "00:00:20:00:00:01"
+ self.inner_vlan = "N/A"
+ self.inner_ip_src = "192.168.2.1"
+ self.inner_ip_dst = "192.168.2.2"
+ self.inner_ip_invalid = 0
+ self.inner_ip6_src = "N/A"
+ self.inner_ip6_dst = "N/A"
+ self.inner_ip6_invalid = 0
+ self.payload_size = 18
+ self.inner_l4_type = "UDP"
+ self.inner_l4_invalid = 0
+
+ def packet_type(self):
+ """
+ Return vxlan packet type
+ """
+ if self.outer_udp_dst != VXLAN_PORT:
+ if self.outer_ip6_src != "N/A":
+ return "L3_IPV6_EXT_UNKNOWN"
+ else:
+ return "L3_IPV4_EXT_UNKNOWN"
+ else:
+ if self.inner_ip6_src != "N/A":
+ return "L3_IPV6_EXT_UNKNOWN"
+ else:
+ return "L3_IPV4_EXT_UNKNOWN"
+
+ def create_pcap(self):
+ """
+ Create pcap file and copy it to tester if configured
+ Return scapy packet object for later usage
+ """
+ if self.inner_l4_type == "SCTP":
+ self.inner_payload = SCTPChunkData(data="X" * 16)
+ else:
+ self.inner_payload = "X" * self.payload_size
+
+ if self.inner_l4_type == "TCP":
+ l4_pro = TCP()
+ elif self.inner_l4_type == "SCTP":
+ l4_pro = SCTP()
+ else:
+ l4_pro = UDP()
+
+ if self.inner_ip6_src != "N/A":
+ inner_l3 = IPv6()
+ else:
+ inner_l3 = IP()
+
+ if self.inner_vlan != "N/A":
+ inner = Ether() / Dot1Q() / inner_l3 / l4_pro / self.inner_payload
+ inner[Dot1Q].vlan = self.inner_vlan
+ else:
+ inner = Ether() / inner_l3 / l4_pro / self.inner_payload
+
+ if self.inner_ip6_src != "N/A":
+ inner[inner_l3.name].src = self.inner_ip6_src
+ inner[inner_l3.name].dst = self.inner_ip6_dst
+ else:
+ inner[inner_l3.name].src = self.inner_ip_src
+ inner[inner_l3.name].dst = self.inner_ip_dst
+
+ if self.inner_ip_invalid == 1:
+ inner[inner_l3.name].chksum = 0
+
+ # when udp checksum is 0, will skip checksum
+ if self.inner_l4_invalid == 1:
+ if self.inner_l4_type == "SCTP":
+ inner[SCTP].chksum = 0
+ else:
+ inner[self.inner_l4_type].chksum = 1
+
+ inner[Ether].src = self.inner_mac_src
+ inner[Ether].dst = self.inner_mac_dst
+
+ if self.outer_ip6_src != "N/A":
+ outer_l3 = IPv6()
+ else:
+ outer_l3 = IP()
+
+ if self.outer_vlan != "N/A":
+ outer = Ether() / Dot1Q() / outer_l3 / UDP()
+ outer[Dot1Q].vlan = self.outer_vlan
+ else:
+ outer = Ether() / outer_l3 / UDP()
+
+ outer[Ether].src = self.outer_mac_src
+ outer[Ether].dst = self.outer_mac_dst
+
+ if self.outer_ip6_src != "N/A":
+ outer[outer_l3.name].src = self.outer_ip6_src
+ outer[outer_l3.name].dst = self.outer_ip6_dst
+ else:
+ outer[outer_l3.name].src = self.outer_ip_src
+ outer[outer_l3.name].dst = self.outer_ip_dst
+
+ outer[UDP].sport = self.outer_udp_src
+ outer[UDP].dport = self.outer_udp_dst
+
+ if self.outer_ip_invalid == 1:
+ outer[outer_l3.name].chksum = 0
+ # when udp checksum is 0, will skip checksum
+ if self.outer_udp_invalid == 1:
+ outer[UDP].chksum = 1
+
+ if self.outer_udp_dst == VXLAN_PORT:
+ self.pkt = outer / VXLAN(vni=self.vni) / inner
+ else:
+ self.pkt = outer / ("X" * self.payload_size)
+
+ wrpcap(self.pcap_file, self.pkt)
+
+ return self.pkt
+
+ def get_chksums(self, pkt=None):
+ """
+ get chksum values of Outer and Inner packet L3&L4
+ Skip outer udp for it will be calculated by software
+ """
+ chk_sums = {}
+ if pkt is None:
+ pkt = rdpcap(self.pcap_file)
+ else:
+ pkt = pkt.pktgen.pkt
+
+ time.sleep(1)
+ if pkt[0].guess_payload_class(pkt[0]).name == "802.1Q":
+ payload = pkt[0][Dot1Q]
+ else:
+ payload = pkt[0]
+
+ if payload.guess_payload_class(payload).name == "IP":
+ chk_sums["outer_ip"] = hex(payload[IP].chksum)
+
+ if pkt[0].haslayer("VXLAN") == 1:
+ inner = pkt[0]["VXLAN"]
+ if inner.haslayer(IP) == 1:
+ chk_sums["inner_ip"] = hex(inner[IP].chksum)
+ if inner[IP].proto == 6:
+ chk_sums["inner_tcp"] = hex(inner[TCP].chksum)
+ if inner[IP].proto == 17:
+ chk_sums["inner_udp"] = hex(inner[UDP].chksum)
+ if inner[IP].proto == 132:
+ chk_sums["inner_sctp"] = hex(inner[SCTP].chksum)
+ elif inner.haslayer(IPv6) == 1:
+ if inner[IPv6].nh == 6:
+ chk_sums["inner_tcp"] = hex(inner[TCP].chksum)
+ if inner[IPv6].nh == 17:
+ chk_sums["inner_udp"] = hex(inner[UDP].chksum)
+ # scapy can not get sctp checksum, so extracted manually
+ if inner[IPv6].nh == 59:
+ load = str(inner[IPv6].payload)
+ chk_sums["inner_sctp"] = hex(
+ (ord(load[8]) << 24)
+ | (ord(load[9]) << 16)
+ | (ord(load[10]) << 8)
+ | (ord(load[11]))
+ )
+
+ return chk_sums
+
+ def send_pcap(self, iface=""):
+ """
+ Send vxlan pcap file by iface
+ """
+ del self.pkt_obj.pktgen.pkts[:]
+ self.pkt_obj.pktgen.assign_pkt(self.pkt)
+ self.pkt_obj.pktgen.update_pkts()
+ self.pkt_obj.send_pkt(crb=self.test_case.tester, tx_port=iface)
+
+ def pcap_len(self):
+ """
+ Return length of pcap packet, will plus 4 bytes crc
+ """
+ # add four bytes crc
+ return len(self.pkt) + 4
+
+
+class TestVxlan(TestCase):
+ def set_up_all(self):
+ """
+ vxlan Prerequisites
+ """
+ # this feature only enable in Intel® Ethernet 700 Series now
+ if self.nic in [
+ "I40E_10G-SFP_XL710",
+ "I40E_40G-QSFP_A",
+ "I40E_40G-QSFP_B",
+ "I40E_25G-25G_SFP28",
+ "I40E_10G-SFP_X722",
+ "I40E_10G-10G_BASE_T_X722",
+ "I40E_10G-10G_BASE_T_BC",
+ ]:
+ self.compile_switch = "CONFIG_RTE_LIBRTE_I40E_INC_VECTOR"
+ elif self.nic in ["IXGBE_10G-X550T", "IXGBE_10G-X550EM_X_10G_T"]:
+ self.compile_switch = "CONFIG_RTE_IXGBE_INC_VECTOR"
+ elif self.nic in ["ICE_25G-E810C_SFP", "ICE_100G-E810C_QSFP"]:
+ print("Intel® Ethernet 700 Series support default none VECTOR")
+ else:
+ self.verify(False, "%s not support this vxlan" % self.nic)
+ # Based on h/w type, choose how many ports to use
+ ports = self.dut.get_ports()
+
+ # Verify that enough ports are available
+ self.verify(len(ports) >= 2, "Insufficient ports for testing")
+ global valports
+ valports = [_ for _ in ports if self.tester.get_local_port(_) != -1]
+
+ self.portMask = utils.create_mask(valports[:2])
+
+ # Verify that enough threads are available
+ netdev = self.dut.ports_info[ports[0]]["port"]
+ self.ports_socket = netdev.socket
+
+ # start testpmd
+ self.pmdout = PmdOutput(self.dut)
+
+ # init port config
+ self.dut_port = valports[0]
+ self.dut_port_mac = self.dut.get_mac_address(self.dut_port)
+ tester_port = self.tester.get_local_port(self.dut_port)
+ self.tester_iface = self.tester.get_interface(tester_port)
+ self.recv_port = valports[1]
+ tester_recv_port = self.tester.get_local_port(self.recv_port)
+ self.recv_iface = self.tester.get_interface(tester_recv_port)
+
+ # invalid parameter
+ self.invalid_mac = "00:00:00:00:01"
+ self.invalid_ip = "192.168.1.256"
+ self.invalid_vlan = 4097
+ self.invalid_queue = 64
+ self.path = self.dut.apps_name["test-pmd"]
+
+ # vxlan payload length for performance test
+ # inner packet not contain crc, should need add four
+ self.vxlan_payload = (
+ PACKET_LEN
+ - HEADER_SIZE["eth"]
+ - HEADER_SIZE["ip"]
+ - HEADER_SIZE["udp"]
+ - HEADER_SIZE["vxlan"]
+ - HEADER_SIZE["eth"]
+ - HEADER_SIZE["ip"]
+ - HEADER_SIZE["udp"]
+ + 4
+ )
+
+ self.cal_type = [
+ {
+ "Type": "SOFTWARE ALL",
+ "csum": [],
+ "recvqueue": "Single",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Type": "HW L4",
+ "csum": ["udp"],
+ "recvqueue": "Single",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Type": "HW L3&L4",
+ "csum": ["ip", "udp", "outer-ip"],
+ "recvqueue": "Single",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Type": "SOFTWARE ALL",
+ "csum": [],
+ "recvqueue": "Multi",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Type": "HW L4",
+ "csum": ["udp"],
+ "recvqueue": "Multi",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Type": "HW L3&L4",
+ "csum": ["ip", "udp", "outer-ip"],
+ "recvqueue": "Multi",
+ "Mpps": {},
+ "pct": {},
+ },
+ ]
+
+ self.chksum_header = ["Calculate Type"]
+ self.chksum_header.append("Queues")
+ self.chksum_header.append("Mpps")
+ self.chksum_header.append("% linerate")
+
+ # tunnel filter performance test
+ self.default_vlan = 1
+ self.tunnel_multiqueue = 2
+ self.tunnel_header = ["Packet", "Filter", "Queue", "Mpps", "% linerate"]
+ self.tunnel_perf = [
+ {
+ "Packet": "Normal",
+ "tunnel_filter": "None",
+ "recvqueue": "Single",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "None",
+ "recvqueue": "Single",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "imac-ivlan",
+ "recvqueue": "Single",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "imac-ivlan-tenid",
+ "recvqueue": "Single",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "imac-tenid",
+ "recvqueue": "Single",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "imac",
+ "recvqueue": "Single",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "omac-imac-tenid",
+ "recvqueue": "Single",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "None",
+ "recvqueue": "Multi",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "imac-ivlan",
+ "recvqueue": "Multi",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "imac-ivlan-tenid",
+ "recvqueue": "Multi",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "imac-tenid",
+ "recvqueue": "Multi",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "imac",
+ "recvqueue": "Multi",
+ "Mpps": {},
+ "pct": {},
+ },
+ {
+ "Packet": "VXLAN",
+ "tunnel_filter": "omac-imac-tenid",
+ "recvqueue": "Multi",
+ },
+ ]
+
+ self.pktgen_helper = PacketGeneratorHelper()
+
+ def set_up(self):
+ """
+ Run before each test case.
+ """
+ pass
+
+ def test_perf_vxlan_tunnelfilter_performance_2ports(self):
+ self.result_table_create(self.tunnel_header)
+ core_list = self.dut.get_core_list(
+ "1S/%dC/1T" % (self.tunnel_multiqueue * 2 + 1), socket=self.ports_socket
+ )
+
+ pmd_temp = (
+ "./%s %s -- -i --disable-rss --rxq=2 --txq=2 --nb-cores=4 --portmask=%s"
+ )
+
+ for perf_config in self.tunnel_perf:
+ tun_filter = perf_config["tunnel_filter"]
+ recv_queue = perf_config["recvqueue"]
+ print(
+ (
+ utils.GREEN(
+ "Measure tunnel performance of [%s %s %s]"
+ % (perf_config["Packet"], tun_filter, recv_queue)
+ )
+ )
+ )
+
+ if tun_filter == "None" and recv_queue == "Multi":
+ pmd_temp = (
+ "./%s %s -- -i --rss-udp --rxq=2 --txq=2 --nb-cores=4 --portmask=%s"
+ )
+
+ self.eal_para = self.dut.create_eal_parameters(cores=core_list)
+ pmd_cmd = pmd_temp % (self.path, self.eal_para, self.portMask)
+ self.dut.send_expect(pmd_cmd, "testpmd> ", 100)
+
+ # config flow
+ self.config_tunnelfilter(
+ self.dut_port, self.recv_port, perf_config, "flow1.pcap"
+ )
+ # config the flows
+ tgen_input = []
+ tgen_input.append(
+ (
+ self.tester.get_local_port(self.dut_port),
+ self.tester.get_local_port(self.recv_port),
+ "flow1.pcap",
+ )
+ )
+
+ if BIDIRECT:
+ self.config_tunnelfilter(
+ self.recv_port, self.dut_port, perf_config, "flow2.pcap"
+ )
+ tgen_input.append(
+ (
+ self.tester.get_local_port(self.recv_port),
+ self.tester.get_local_port(self.dut_port),
+ "flow2.pcap",
+ )
+ )
+
+ self.dut.send_expect("set fwd io", "testpmd>", 10)
+ self.dut.send_expect("start", "testpmd>", 10)
+ self.pmdout.wait_link_status_up(self.dut_port)
+ if BIDIRECT:
+ wirespeed = self.wirespeed(self.nic, PACKET_LEN, 2)
+ else:
+ wirespeed = self.wirespeed(self.nic, PACKET_LEN, 1)
+
+ # run traffic generator
+ use_vm = True if recv_queue == "Multi" and tun_filter == "None" else False
+ _, pps = self.suite_measure_throughput(tgen_input, use_vm=use_vm)
+
+ pps /= 1000000.0
+ perf_config["Mpps"] = pps
+ perf_config["pct"] = pps * 100 / wirespeed
+
+ out = self.dut.send_expect("stop", "testpmd>", 10)
+ self.dut.send_expect("quit", "# ", 10)
+
+ # verify every queue work fine
+ check_queue = 0
+ if recv_queue == "Multi":
+ for queue in range(check_queue):
+ self.verify(
+ "Queue= %d -> TX Port" % (queue) in out,
+ "Queue %d no traffic" % queue,
+ )
+
+ table_row = [
+ perf_config["Packet"],
+ tun_filter,
+ recv_queue,
+ perf_config["Mpps"],
+ perf_config["pct"],
+ ]
+
+ self.result_table_add(table_row)
+
+ self.result_table_print()
+
+ def test_perf_vxlan_checksum_performance_2ports(self):
+ self.result_table_create(self.chksum_header)
+ vxlan = VxlanTestConfig(self, payload_size=self.vxlan_payload)
+ vxlan.outer_mac_dst = self.dut.get_mac_address(self.dut_port)
+ vxlan.pcap_file = "vxlan1.pcap"
+ vxlan.inner_mac_dst = "00:00:20:00:00:01"
+ vxlan.create_pcap()
+
+ vxlan_queue = VxlanTestConfig(self, payload_size=self.vxlan_payload)
+ vxlan_queue.outer_mac_dst = self.dut.get_mac_address(self.dut_port)
+ vxlan_queue.pcap_file = "vxlan1_1.pcap"
+ vxlan_queue.inner_mac_dst = "00:00:20:00:00:02"
+ vxlan_queue.create_pcap()
+
+ # socket/core/thread
+ core_list = self.dut.get_core_list(
+ "1S/%dC/1T" % (self.tunnel_multiqueue * 2 + 1), socket=self.ports_socket
+ )
+ core_mask = utils.create_mask(core_list)
+
+ self.dut_ports = self.dut.get_ports_performance(force_different_nic=False)
+ tx_port = self.tester.get_local_port(self.dut_ports[0])
+ rx_port = self.tester.get_local_port(self.dut_ports[1])
+
+ for cal in self.cal_type:
+ recv_queue = cal["recvqueue"]
+ print(
+ (
+ utils.GREEN(
+ "Measure checksum performance of [%s %s %s]"
+ % (cal["Type"], recv_queue, cal["csum"])
+ )
+ )
+ )
+
+ # configure flows
+ tgen_input = []
+ if recv_queue == "Multi":
+ tgen_input.append((tx_port, rx_port, "vxlan1.pcap"))
+ tgen_input.append((tx_port, rx_port, "vxlan1_1.pcap"))
+ else:
+ tgen_input.append((tx_port, rx_port, "vxlan1.pcap"))
+
+ # multi queue and signle queue commands
+ if recv_queue == "Multi":
+ pmd_temp = "./%s %s -- -i --disable-rss --rxq=2 --txq=2 --nb-cores=4 --portmask=%s"
+ else:
+ pmd_temp = "./%s %s -- -i --nb-cores=2 --portmask=%s"
+
+ self.eal_para = self.dut.create_eal_parameters(cores=core_list)
+ pmd_cmd = pmd_temp % (self.path, self.eal_para, self.portMask)
+
+ self.dut.send_expect(pmd_cmd, "testpmd> ", 100)
+ self.dut.send_expect("set fwd csum", "testpmd>", 10)
+ self.enable_vxlan(self.dut_port)
+ self.enable_vxlan(self.recv_port)
+ self.pmdout.wait_link_status_up(self.dut_port)
+
+ # redirect flow to another queue by tunnel filter
+ rule_config = {
+ "dut_port": self.dut_port,
+ "outer_mac_dst": vxlan.outer_mac_dst,
+ "inner_mac_dst": vxlan.inner_mac_dst,
+ "inner_ip_dst": vxlan.inner_ip_dst,
+ "inner_vlan": 0,
+ "tun_filter": "imac",
+ "vni": vxlan.vni,
+ "queue": 0,
+ }
+ self.perf_tunnel_filter_set_rule(rule_config)
+
+ if recv_queue == "Multi":
+ rule_config = {
+ "dut_port": self.dut_port,
+ "outer_mac_dst": vxlan_queue.outer_mac_dst,
+ "inner_mac_dst": vxlan_queue.inner_mac_dst,
+ "inner_ip_dst": vxlan_queue.inner_ip_dst,
+ "inner_vlan": 0,
+ "tun_filter": "imac",
+ "vni": vxlan.vni,
+ "queue": 1,
+ }
+ self.perf_tunnel_filter_set_rule(rule_config)
+
+ for pro in cal["csum"]:
+ self.csum_set_type(pro, self.dut_port)
+ self.csum_set_type(pro, self.recv_port)
+
+ self.dut.send_expect("start", "testpmd>", 10)
+
+ wirespeed = self.wirespeed(self.nic, PACKET_LEN, 1)
+
+ # run traffic generator
+ _, pps = self.suite_measure_throughput(tgen_input)
+
+ pps /= 1000000.0
+ cal["Mpps"] = pps
+ cal["pct"] = pps * 100 / wirespeed
+
+ out = self.dut.send_expect("stop", "testpmd>", 10)
+ self.dut.send_expect("quit", "# ", 10)
+
+ # verify every queue work fine
+ check_queue = 1
+ if recv_queue == "Multi":
+ for queue in range(check_queue):
+ self.verify(
+ "Queue= %d -> TX Port" % (queue) in out,
+ "Queue %d no traffic" % queue,
+ )
+
+ table_row = [cal["Type"], recv_queue, cal["Mpps"], cal["pct"]]
+ self.result_table_add(table_row)
+
+ self.result_table_print()
+
+ def tear_down(self):
+ """
+ Run after each test case.
+ """
+ self.dut.kill_all()
+
+ def tear_down_all(self):
+ """
+ Run after each test suite.
+ """
+ pass
diff --git a/tests/TestSuite_vxlan.py b/tests/TestSuite_vxlan.py
index c69d7903..5dd49ecd 100644
--- a/tests/TestSuite_vxlan.py
+++ b/tests/TestSuite_vxlan.py
@@ -1163,219 +1163,6 @@ class TestVxlan(TestCase):
wrpcap(dest_pcap, pkts)
- def test_perf_vxlan_tunnelfilter_performance_2ports(self):
- self.result_table_create(self.tunnel_header)
- core_list = self.dut.get_core_list(
- "1S/%dC/1T" % (self.tunnel_multiqueue * 2 + 1), socket=self.ports_socket
- )
-
- pmd_temp = (
- "./%s %s -- -i --disable-rss --rxq=2 --txq=2 --nb-cores=4 --portmask=%s"
- )
-
- for perf_config in self.tunnel_perf:
- tun_filter = perf_config["tunnel_filter"]
- recv_queue = perf_config["recvqueue"]
- print(
- (
- utils.GREEN(
- "Measure tunnel performance of [%s %s %s]"
- % (perf_config["Packet"], tun_filter, recv_queue)
- )
- )
- )
-
- if tun_filter == "None" and recv_queue == "Multi":
- pmd_temp = (
- "./%s %s -- -i --rss-udp --rxq=2 --txq=2 --nb-cores=4 --portmask=%s"
- )
-
- self.eal_para = self.dut.create_eal_parameters(cores=core_list)
- pmd_cmd = pmd_temp % (self.path, self.eal_para, self.portMask)
- self.dut.send_expect(pmd_cmd, "testpmd> ", 100)
-
- # config flow
- self.config_tunnelfilter(
- self.dut_port, self.recv_port, perf_config, "flow1.pcap"
- )
- # config the flows
- tgen_input = []
- tgen_input.append(
- (
- self.tester.get_local_port(self.dut_port),
- self.tester.get_local_port(self.recv_port),
- "flow1.pcap",
- )
- )
-
- if BIDIRECT:
- self.config_tunnelfilter(
- self.recv_port, self.dut_port, perf_config, "flow2.pcap"
- )
- tgen_input.append(
- (
- self.tester.get_local_port(self.recv_port),
- self.tester.get_local_port(self.dut_port),
- "flow2.pcap",
- )
- )
-
- self.dut.send_expect("set fwd io", "testpmd>", 10)
- self.dut.send_expect("start", "testpmd>", 10)
- self.pmdout.wait_link_status_up(self.dut_port)
- if BIDIRECT:
- wirespeed = self.wirespeed(self.nic, PACKET_LEN, 2)
- else:
- wirespeed = self.wirespeed(self.nic, PACKET_LEN, 1)
-
- # run traffic generator
- use_vm = True if recv_queue == "Multi" and tun_filter == "None" else False
- _, pps = self.suite_measure_throughput(tgen_input, use_vm=use_vm)
-
- pps /= 1000000.0
- perf_config["Mpps"] = pps
- perf_config["pct"] = pps * 100 / wirespeed
-
- out = self.dut.send_expect("stop", "testpmd>", 10)
- self.dut.send_expect("quit", "# ", 10)
-
- # verify every queue work fine
- check_queue = 0
- if recv_queue == "Multi":
- for queue in range(check_queue):
- self.verify(
- "Queue= %d -> TX Port" % (queue) in out,
- "Queue %d no traffic" % queue,
- )
-
- table_row = [
- perf_config["Packet"],
- tun_filter,
- recv_queue,
- perf_config["Mpps"],
- perf_config["pct"],
- ]
-
- self.result_table_add(table_row)
-
- self.result_table_print()
-
- def test_perf_vxlan_checksum_performance_2ports(self):
- self.result_table_create(self.chksum_header)
- vxlan = VxlanTestConfig(self, payload_size=self.vxlan_payload)
- vxlan.outer_mac_dst = self.dut.get_mac_address(self.dut_port)
- vxlan.pcap_file = "vxlan1.pcap"
- vxlan.inner_mac_dst = "00:00:20:00:00:01"
- vxlan.create_pcap()
-
- vxlan_queue = VxlanTestConfig(self, payload_size=self.vxlan_payload)
- vxlan_queue.outer_mac_dst = self.dut.get_mac_address(self.dut_port)
- vxlan_queue.pcap_file = "vxlan1_1.pcap"
- vxlan_queue.inner_mac_dst = "00:00:20:00:00:02"
- vxlan_queue.create_pcap()
-
- # socket/core/thread
- core_list = self.dut.get_core_list(
- "1S/%dC/1T" % (self.tunnel_multiqueue * 2 + 1), socket=self.ports_socket
- )
- core_mask = utils.create_mask(core_list)
-
- self.dut_ports = self.dut.get_ports_performance(force_different_nic=False)
- tx_port = self.tester.get_local_port(self.dut_ports[0])
- rx_port = self.tester.get_local_port(self.dut_ports[1])
-
- for cal in self.cal_type:
- recv_queue = cal["recvqueue"]
- print(
- (
- utils.GREEN(
- "Measure checksum performance of [%s %s %s]"
- % (cal["Type"], recv_queue, cal["csum"])
- )
- )
- )
-
- # configure flows
- tgen_input = []
- if recv_queue == "Multi":
- tgen_input.append((tx_port, rx_port, "vxlan1.pcap"))
- tgen_input.append((tx_port, rx_port, "vxlan1_1.pcap"))
- else:
- tgen_input.append((tx_port, rx_port, "vxlan1.pcap"))
-
- # multi queue and signle queue commands
- if recv_queue == "Multi":
- pmd_temp = "./%s %s -- -i --disable-rss --rxq=2 --txq=2 --nb-cores=4 --portmask=%s"
- else:
- pmd_temp = "./%s %s -- -i --nb-cores=2 --portmask=%s"
-
- self.eal_para = self.dut.create_eal_parameters(cores=core_list)
- pmd_cmd = pmd_temp % (self.path, self.eal_para, self.portMask)
-
- self.dut.send_expect(pmd_cmd, "testpmd> ", 100)
- self.dut.send_expect("set fwd csum", "testpmd>", 10)
- self.enable_vxlan(self.dut_port)
- self.enable_vxlan(self.recv_port)
- self.pmdout.wait_link_status_up(self.dut_port)
-
- # redirect flow to another queue by tunnel filter
- rule_config = {
- "dut_port": self.dut_port,
- "outer_mac_dst": vxlan.outer_mac_dst,
- "inner_mac_dst": vxlan.inner_mac_dst,
- "inner_ip_dst": vxlan.inner_ip_dst,
- "inner_vlan": 0,
- "tun_filter": "imac",
- "vni": vxlan.vni,
- "queue": 0,
- }
- self.perf_tunnel_filter_set_rule(rule_config)
-
- if recv_queue == "Multi":
- rule_config = {
- "dut_port": self.dut_port,
- "outer_mac_dst": vxlan_queue.outer_mac_dst,
- "inner_mac_dst": vxlan_queue.inner_mac_dst,
- "inner_ip_dst": vxlan_queue.inner_ip_dst,
- "inner_vlan": 0,
- "tun_filter": "imac",
- "vni": vxlan.vni,
- "queue": 1,
- }
- self.perf_tunnel_filter_set_rule(rule_config)
-
- for pro in cal["csum"]:
- self.csum_set_type(pro, self.dut_port)
- self.csum_set_type(pro, self.recv_port)
-
- self.dut.send_expect("start", "testpmd>", 10)
-
- wirespeed = self.wirespeed(self.nic, PACKET_LEN, 1)
-
- # run traffic generator
- _, pps = self.suite_measure_throughput(tgen_input)
-
- pps /= 1000000.0
- cal["Mpps"] = pps
- cal["pct"] = pps * 100 / wirespeed
-
- out = self.dut.send_expect("stop", "testpmd>", 10)
- self.dut.send_expect("quit", "# ", 10)
-
- # verify every queue work fine
- check_queue = 1
- if recv_queue == "Multi":
- for queue in range(check_queue):
- self.verify(
- "Queue= %d -> TX Port" % (queue) in out,
- "Queue %d no traffic" % queue,
- )
-
- table_row = [cal["Type"], recv_queue, cal["Mpps"], cal["pct"]]
- self.result_table_add(table_row)
-
- self.result_table_print()
-
def enable_vxlan(self, port):
self.dut.send_expect(
"rx_vxlan_port add %d %d" % (VXLAN_PORT, port), "testpmd>", 10
--
2.25.1
next prev parent reply other threads:[~2022-09-22 6:25 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-22 14:29 [dts][PATCH V1 1/7] tests/efd:Separated " Hongbo Li
2022-09-22 14:29 ` [dts][PATCH V1 2/7] tests/kni:Separated " Hongbo Li
2022-09-22 14:29 ` [dts][PATCH V1 3/7] tests/l2fwd:Separated " Hongbo Li
2022-09-22 14:29 ` [dts][PATCH V1 4/7] tests/tso:Separated " Hongbo Li
2022-09-22 14:29 ` Hongbo Li [this message]
2022-09-22 14:29 ` [dts][PATCH V1 6/7] tests/ipfrag:Separated " Hongbo Li
2022-09-22 14:29 ` [PATCH V1 7/7] tests/multiprocess:Separated " Hongbo Li
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=20220922142950.398902-5-hongbox.li@intel.com \
--to=hongbox.li@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).