* [dts] [PATCH V1] tests/vhost_dequeue_zero_copy: update suite according to plan
@ 2019-12-24 21:53 lihong
2019-12-25 9:02 ` Ma, LihongX
2020-01-02 2:18 ` Tu, Lijuan
0 siblings, 2 replies; 4+ messages in thread
From: lihong @ 2019-12-24 21:53 UTC (permalink / raw)
To: dts; +Cc: yinan.wang, lihong
Signed-off-by: lihong <lihongx.ma@intel.com>
---
tests/TestSuite_vhost_dequeue_zero_copy.py | 393 +++++++++++++++++++----------
1 file changed, 261 insertions(+), 132 deletions(-)
diff --git a/tests/TestSuite_vhost_dequeue_zero_copy.py b/tests/TestSuite_vhost_dequeue_zero_copy.py
index 5705064..4052437 100644
--- a/tests/TestSuite_vhost_dequeue_zero_copy.py
+++ b/tests/TestSuite_vhost_dequeue_zero_copy.py
@@ -38,6 +38,7 @@ There are three topology test (PVP/VM2VM/VM2NIC) for this feature.
And this testsuite implement the topology of PVP.
Testsuite vm2vm_net_perf implement the topology VM2VM
Testsuite gso implement the topology VM2NIC
+To run this suite, the qemu version should support packed ring.
"""
import utils
import time
@@ -45,7 +46,8 @@ import re
from settings import HEADER_SIZE
from virt_common import VM
from test_case import TestCase
-from pktgen import PacketGeneratorHelper
+from packet import Packet
+from pktgen import TRANSMIT_CONT
class TestVhostDequeueZeroCopy(TestCase):
@@ -54,19 +56,19 @@ class TestVhostDequeueZeroCopy(TestCase):
"""
Run at the start of each test suite.
"""
+ self.verify(self.nic in ("fortville_spirit"), "the port can not run this suite")
self.tester.extend_external_packet_generator(TestVhostDequeueZeroCopy, self)
self.frame_sizes = [64, 128, 256, 512, 1024, 1518]
- self.queue_number = 1
- self.nb_cores = 1
self.dut_ports = self.dut.get_ports()
self.verify(len(self.dut_ports) >= 1, "Insufficient ports for testing")
self.ports_socket = self.dut.get_numa_id(self.dut_ports[0])
- self.cores_num = len([n for n in self.dut.cores if int(n['socket'])
- == self.ports_socket])
- self.mem_channels = self.dut.get_memory_channels()
self.dst_mac = self.dut.get_mac_address(self.dut_ports[0])
+ self.tx_port = self.tester.get_local_port(self.dut_ports[0])
+ self.port_pci = self.dut.ports_info[self.dut_ports[0]]['pci']
self.vm_dut = None
+ self.virtio_user = None
self.virtio1_mac = "52:54:00:00:00:01"
+ self.header_size = HEADER_SIZE['eth'] + HEADER_SIZE['ip'] + HEADER_SIZE['udp']
self.logger.info("you can config packet_size in file %s.cfg," % self.suite_name + \
"in region 'suite' like packet_sizes=[64, 128, 256]")
@@ -74,64 +76,93 @@ class TestVhostDequeueZeroCopy(TestCase):
if 'packet_sizes' in self.get_suite_cfg():
self.frame_sizes = self.get_suite_cfg()['packet_sizes']
- self.out_path = '/tmp'
- out = self.tester.send_expect('ls -d %s' % self.out_path, '# ')
- if 'No such file or directory' in out:
- self.tester.send_expect('mkdir -p %s' % self.out_path, '# ')
- # create an instance to set stream field setting
- self.pktgen_helper = PacketGeneratorHelper()
self.base_dir = self.dut.base_dir.replace('~', '/root')
+ self.vhost_user = self.dut.new_session(suite="vhost-user")
def set_up(self):
"""
Run before each test case.
"""
# Clean the execution ENV
- self.dut.send_expect("killall -s INT testpmd", "#")
- self.dut.send_expect("killall -s INT qemu-system-x86_64", "#")
self.dut.send_expect("rm -rf %s/vhost-net*" % self.base_dir, "#")
# Prepare the result table
self.table_header = ["FrameSize(B)", "Throughput(Mpps)",
"% linerate", "Queue number", "Cycle"]
self.result_table_create(self.table_header)
+ self.vm_dut = None
+ self.big_pkt_record = {}
- self.vhost = self.dut.new_session(suite="vhost-user")
-
- def get_core_mask(self):
+ def get_core_list(self):
"""
check whether the server has enough cores to run case
+ if want to get the best perf of the vhost, the vhost tesptmd at least
+ should have 3 cores to start testpmd
"""
- core_config = "1S/%dC/1T" % (self.nb_cores + 1)
- self.verify(self.cores_num >= (self.nb_cores + 1),
- "There has not enought cores to test this case %s" % self.running_case)
- core_list = self.dut.get_core_list(
+ if self.nb_cores == 1:
+ cores_num = 2
+ else:
+ cores_num = 1
+ core_config = "1S/%dC/1T" % (self.nb_cores + cores_num)
+ self.core_list = self.dut.get_core_list(
core_config, socket=self.ports_socket)
- self.core_mask = utils.create_mask(core_list)
+ self.verify(len(self.core_list) >= (self.nb_cores + cores_num),
+ "There has not enought cores to test this case %s" % self.running_case)
- def launch_testpmd_on_vhost(self, txfreet):
+ def launch_testpmd_as_vhost(self, txfreet, zero_copy=True, client_mode=False):
"""
launch testpmd on vhost
"""
- self.get_core_mask()
+ self.get_core_list()
+ mode_info = ""
+ if client_mode is True:
+ mode_info = ',client=1'
+ zero_copy_info = 1
+ if zero_copy is False:
+ zero_copy_info = 0
if txfreet == "normal":
- txfreet_args = "--txfreet=992"
+ txfreet_args = "--txd=1024 --rxd=1024 --txfreet=992"
elif txfreet == "maximum":
- txfreet_args = "--txfreet=1020 --txrs=4"
- command_client = self.dut.target + "/app/testpmd " + \
- " -n %d -c %s --socket-mem 1024,1024 " + \
- " --legacy-mem --file-prefix=vhost " + \
- " --vdev 'eth_vhost0,iface=%s/vhost-net,queues=%d,dequeue-zero-copy=1' " + \
- " -- -i --nb-cores=%d --rxq=%d --txq=%d " + \
- "--txd=1024 --rxd=1024 %s"
+ txfreet_args = "--txrs=4 --txd=992 --rxd=992 --txfreet=988"
+ elif txfreet == "vector_rx":
+ txfreet_args = "--txd=1024 --rxd=1024 --txfreet=992 --txrs=32"
+
+ eal_params = self.dut.create_eal_parameters(cores=self.core_list,
+ prefix='vhost', ports=[self.port_pci])
+ command_client = self.dut.target + "/app/testpmd %s " + \
+ " --socket-mem 1024,1024 --legacy-mem " + \
+ " --vdev 'eth_vhost0,iface=%s/vhost-net,queues=%d,dequeue-zero-copy=%d%s' " + \
+ " -- -i --nb-cores=%d --rxq=%d --txq=%d %s"
command_line_client = command_client % (
- self.mem_channels, self.core_mask, self.base_dir,
- self.queue_number, self.nb_cores,
+ eal_params, self.base_dir, self.queue_number,
+ zero_copy_info, mode_info, self.nb_cores,
self.queue_number, self.queue_number, txfreet_args)
- self.vhost.send_expect(command_line_client, "testpmd> ", 120)
- self.vhost.send_expect("set fwd mac", "testpmd> ", 120)
+ self.vhost_user.send_expect(command_line_client, "testpmd> ", 120)
+ self.vhost_user.send_expect("set fwd mac", "testpmd> ", 120)
- def launch_testpmd_on_vm(self):
+ def launch_testpmd_as_virtio_user(self, path_mode):
+ """
+ launch testpmd use vhost-net with path mode
+ """
+ # To get the best perf, the vhost and virtio testpmd should not use same cores,
+ # so get the additional 3 cores to start virtio testpmd
+ core_config = "1S/%dC/1T" % (len(self.core_list) + 3)
+ core_list = self.dut.get_core_list(
+ core_config, socket=self.ports_socket)
+ self.verify(len(core_list) >= (len(self.core_list) + 3),
+ "There has not enought cores to test this case %s" % self.running_case)
+ eal_params = self.dut.create_eal_parameters(cores=core_list[len(self.core_list):],
+ prefix='virtio', no_pci=True)
+ command_line = self.dut.target + "/app/testpmd %s " + \
+ "--socket-mem 1024,1024 --legacy-mem " + \
+ "--vdev=net_virtio_user0,mac=00:01:02:03:04:05,path=./vhost-net,queue_size=1024,%s " + \
+ "-- -i --tx-offloads=0x0 --nb-cores=%d --txd=1024 --rxd=1024"
+ command_line = command_line % (eal_params, path_mode, self.nb_cores)
+ self.virtio_user.send_expect(command_line, 'testpmd> ', 120)
+ self.virtio_user.send_expect('set fwd mac', 'testpmd> ', 120)
+ self.virtio_user.send_expect('start', 'testpmd> ', 120)
+
+ def start_testpmd_on_vm(self, fwd_mode='mac'):
"""
start testpmd in vm depend on different path
"""
@@ -142,12 +173,12 @@ class TestVhostDequeueZeroCopy(TestCase):
command_line = command % (self.nb_cores,
self.queue_number, self.queue_number)
self.vm_dut.send_expect(command_line, "testpmd> ", 30)
+ self.vm_dut.send_expect('set fwd %s' % fwd_mode, "testpmd> ", 30)
+ self.vm_dut.send_expect('start', "testpmd> ", 30)
- def relaunch_testpmd_on_vm(self):
+ def restart_testpmd_on_vm(self, fwd_mode):
self.vm_dut.send_expect("quit", "# ", 30)
- self.launch_testpmd_on_vm()
- self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
- self.vm_dut.send_expect("start", "testpmd> ", 30)
+ self.start_testpmd_on_vm(fwd_mode)
def set_vm_vcpu(self):
"""
@@ -166,7 +197,7 @@ class TestVhostDequeueZeroCopy(TestCase):
if 'cpupin' in self.vm.params[i]['cpu'][0].keys():
self.vm.params[i]['cpu'][0].pop('cpupin')
- def start_one_vm(self):
+ def start_one_vm(self, mode='client', packed=False):
"""
start qemu
"""
@@ -176,10 +207,14 @@ class TestVhostDequeueZeroCopy(TestCase):
vm_params['driver'] = 'vhost-user'
vm_params['opt_path'] = '%s/vhost-net' % self.base_dir
vm_params['opt_mac'] = self.virtio1_mac
+ if mode == 'server':
+ vm_params['opt_server'] = 'server'
opt_args = "mrg_rxbuf=on,rx_queue_size=1024,tx_queue_size=1024"
if self.queue_number > 1:
vm_params['opt_queue'] = self.queue_number
opt_args += ",mq=on,vectors=%d" % (2*self.queue_number + 2)
+ if packed is True:
+ opt_args += ',packed=on'
vm_params['opt_settings'] = opt_args
self.vm.set_vm_device(**vm_params)
self.set_vm_vcpu()
@@ -190,7 +225,26 @@ class TestVhostDequeueZeroCopy(TestCase):
if self.vm_dut is None:
raise Exception("Set up VM ENV failed")
except Exception as e:
- self.logger.error("ERROR: Failure for %s" % str(e))
+ self.logger.error("ERROR: Failure for %s, " % str(e) + \
+ "if 'packed not found' in output of start qemu log, " + \
+ "please use the qemu version which support packed ring")
+ raise e
+
+ def prepare_test_evn(self, vhost_txfreet_mode, vhost_zero_copy, vhost_client_mode,
+ vm_testpmd_fwd_mode, packed_mode):
+ """
+ start vhost testpmd and launch qemu, start testpmd on vm
+ """
+ if vhost_client_mode is True:
+ vm_mode = 'server'
+ else:
+ vm_mode = 'client'
+ self.launch_testpmd_as_vhost(txfreet=vhost_txfreet_mode, zero_copy=vhost_zero_copy,
+ client_mode=vhost_client_mode)
+ self.start_one_vm(mode=vm_mode, packed=packed_mode)
+ self.start_testpmd_on_vm(fwd_mode=vm_testpmd_fwd_mode)
+ # start testpmd at host side after VM and virtio-pmd launched
+ self.vhost_user.send_expect("start", "testpmd> ", 120)
def update_table_info(self, frame_size, Mpps, throughtput, cycle):
results_row = [frame_size]
@@ -200,50 +254,64 @@ class TestVhostDequeueZeroCopy(TestCase):
results_row.append(cycle)
self.result_table_add(results_row)
- def set_fields(self):
- """
- set ip protocol field behavior
- """
- fields_config = {'ip': {'dst': {'action': 'random'}, }, }
- return fields_config
+ # record the big pkt Mpps
+ if frame_size == 1518:
+ self.big_pkt_record[cycle] = Mpps
- def calculate_avg_throughput(self, frame_size, loopback):
+ def calculate_avg_throughput(self, frame_size, fwd_mode):
"""
start to send packet and get the throughput
"""
- payload = frame_size - HEADER_SIZE['eth'] - HEADER_SIZE['ip'] - HEADER_SIZE['udp']
- flow = '[Ether(dst="%s")/IP(src="192.168.4.1",proto=255)/UDP()/("X"*%d)]' % (
+ payload = frame_size - self.header_size
+ flow = 'Ether(dst="%s")/IP(src="192.168.4.1",proto=255)/UDP(sport=33,dport=34)/("X"*%d)' % (
self.dst_mac, payload)
- self.tester.scapy_append('wrpcap("%s/zero_copy.pcap", %s)' % (
- self.out_path, flow))
- self.tester.scapy_execute()
-
- tgenInput = []
- port = self.tester.get_local_port(self.dut_ports[0])
- tgenInput.append((port, port, "%s/zero_copy.pcap" % self.out_path))
- vm_config = self.set_fields()
+ pkt = Packet(pkt_str=flow)
+ pkt.save_pcapfile(self.tester, "%s/zero_copy.pcap" % self.tester.tmp_file)
+ stream_option = {
+ 'pcap': "%s/zero_copy.pcap" % self.tester.tmp_file,
+ 'fields_config': {
+ 'ip': {'src': {'action': 'random', 'start': '16.0.0.1', 'step': 1, 'end': '16.0.0.64'}}},
+ 'stream_config': {
+ 'rate': 100,
+ 'transmit_mode': TRANSMIT_CONT,
+ }
+ }
self.tester.pktgen.clear_streams()
- streams = self.pktgen_helper.prepare_stream_from_tginput(tgenInput, 100, vm_config, self.tester.pktgen)
- # set traffic option
- traffic_opt = {'delay': 5, 'duration': 20}
- _, pps = self.tester.pktgen.measure_throughput(stream_ids=streams, options=traffic_opt)
- Mpps = pps / 1000000.0
+ stream_id = self.tester.pktgen.add_stream(self.tx_port, self.tx_port,
+ "%s/zero_copy.pcap" % self.tester.tmp_file)
+ self.tester.pktgen.config_stream(stream_id, stream_option)
+ traffic_opt = {
+ 'method': 'throughput',
+ 'rate': 100,
+ 'interval': 6,
+ 'duration': 30}
+ stats = self.tester.pktgen.measure([stream_id], traffic_opt)
+
+ if isinstance(stats, list):
+ # if get multi result, ignore the first one, because it may not stable
+ num = len(stats)
+ Mpps = 0
+ for index in range(1, num):
+ Mpps += stats[index][1]
+ Mpps = Mpps / 1000000.0 / (num-1)
+ else:
+ Mpps = stats[1] / 1000000.0
# when the fwd mode is rxonly, we can not receive data, so should not verify it
- if loopback != "rxonly":
+ if fwd_mode != "rxonly":
self.verify(Mpps > 0, "can not receive packets of frame size %d" % (frame_size))
throughput = Mpps * 100 / \
float(self.wirespeed(self.nic, frame_size, 1))
return Mpps, throughput
- def check_packets_of_each_queue(self, frame_size, loopback):
+ def check_packets_of_each_queue(self, frame_size, fwd_mode):
"""
check each queue has receive packets
"""
- if loopback == "rxonly":
+ if fwd_mode == "rxonly":
verify_port = 1
else:
verify_port = 2
- out = self.vhost.send_expect("stop", "testpmd> ", 60)
+ out = self.vhost_user.send_expect("stop", "testpmd> ", 60)
for port_index in range(0, verify_port):
for queue_index in range(0, self.queue_number):
queue_info = re.findall("RX\s*Port=\s*%d/Queue=\s*%d" %
@@ -260,9 +328,9 @@ class TestVhostDequeueZeroCopy(TestCase):
"frame_size:%d, rx-packets:%d, tx-packets:%d" %
(frame_size, rx_packets, tx_packets))
- self.vhost.send_expect("start", "testpmd> ", 60)
+ self.vhost_user.send_expect("start", "testpmd> ", 60)
- def send_and_verify(self, cycle="", loopback=""):
+ def send_and_verify_throughput(self, cycle="", fwd_mode=""):
"""
start to send packets and verify it
"""
@@ -270,110 +338,171 @@ class TestVhostDequeueZeroCopy(TestCase):
info = "Running test %s, and %d frame size." % (self.running_case, frame_size)
self.logger.info(info)
- Mpps, throughput = self.calculate_avg_throughput(frame_size, loopback)
- if loopback != "rxonly":
+ Mpps, throughput = self.calculate_avg_throughput(frame_size, fwd_mode)
+ if fwd_mode != "rxonly":
self.update_table_info(frame_size, Mpps, throughput, cycle)
# when multi queues, check each queue can receive packets
if self.queue_number > 1:
- self.check_packets_of_each_queue(frame_size, loopback)
+ self.check_packets_of_each_queue(frame_size, fwd_mode)
+
+ def check_perf_drop_between_with_and_without_zero_copy(self):
+ """
+ for dequeue-zero-copy=0, about the small pkts we expect ~10% gain
+ compare to dequeue-zero-copy=1
+ """
+ value_with_zero_copy = 0
+ value_without_zero_copy = 0
+ if 'dequeue-zero-copy=1' in self.big_pkt_record.keys():
+ value_with_zero_copy = self.big_pkt_record['dequeue-zero-copy=1']
+ if 'dequeue-zero-copy=0' in self.big_pkt_record.keys():
+ value_without_zero_copy = self.big_pkt_record['dequeue-zero-copy=0']
+ self.verify(value_with_zero_copy != 0 and value_without_zero_copy != 0,
+ 'can not get the value of big pkts, please check self.frame_sizes')
+ self.verify(value_with_zero_copy - value_without_zero_copy >= value_with_zero_copy*0.1,
+ 'the drop with dequeue-zero-copy=0 is not as expected')
def close_all_testpmd_and_vm(self):
"""
close testpmd about vhost-user and vm_testpmd
"""
- self.vhost.send_expect("quit", "#", 60)
- self.vm_dut.send_expect("quit", "#", 60)
- self.dut.close_session(self.vhost)
- self.vm.stop()
+ if getattr(self, 'vhost_user', None):
+ self.vhost_user.send_expect("quit", "#", 60)
+ if getattr(self, 'virtio_user', None):
+ self.virtio_user.send_expect("quit", "#", 60)
+ self.dut.close_session(self.virtio_user)
+ self.virtio_user = None
+ if getattr(self, 'vm_dut', None):
+ self.vm_dut.send_expect("quit", "#", 60)
+ self.vm.stop()
- def test_perf_pvp_dequeue_zero_copy(self):
+ def test_perf_pvp_split_ring_dequeue_zero_copy(self):
"""
- performance of pvp zero-copy with [frame_sizes]
+ pvp split ring dequeue zero-copy test
"""
self.nb_cores = 1
self.queue_number = 1
- self.launch_testpmd_on_vhost(txfreet="normal")
- self.start_one_vm()
- self.launch_testpmd_on_vm()
- # set fwd mode on vm
- self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
- self.vm_dut.send_expect("start", "testpmd> ", 30)
- # start testpmd at host side after VM and virtio-pmd launched
- self.vhost.send_expect("start", "testpmd> ", 120)
- self.send_and_verify()
+ self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to test')
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=False)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
+
self.close_all_testpmd_and_vm()
+ self.logger.info('start vhost testpmd with dequeue-zero-copy=0 to test')
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=False,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=False)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=0')
self.result_table_print()
+ self.check_perf_drop_between_with_and_without_zero_copy()
- def test_perf_pvp_dequeue_zero_copy_with_2_queue(self):
+ def test_perf_pvp_packed_ring_dequeue_zero_copy(self):
"""
- pvp dequeue zero-copy test with 2 queues
+ pvp packed ring dequeue zero-copy test
+ """
+ self.nb_cores = 1
+ self.queue_number = 1
+ self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to test')
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=True)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
+
+ self.close_all_testpmd_and_vm()
+ self.logger.info('start vhost testpmd with dequeue-zero-copy=0 to test')
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=False,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=True)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=0')
+ self.result_table_print()
+ self.check_perf_drop_between_with_and_without_zero_copy()
+
+ def test_perf_pvp_split_ring_dequeue_zero_copy_with_2_queue(self):
+ """
+ pvp split ring dequeue zero-copy test with 2 queues
"""
self.nb_cores = 2
self.queue_number = 2
- self.launch_testpmd_on_vhost(txfreet="normal")
- self.start_one_vm()
- self.launch_testpmd_on_vm()
- # set fwd mode on vm
- self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
- self.vm_dut.send_expect("start", "testpmd> ", 30)
- # start testpmd at host side after VM and virtio-pmd launched
- self.vhost.send_expect("start", "testpmd> ", 120)
- # when multi queues, the function will check each queue can receive packets
- self.send_and_verify()
- self.close_all_testpmd_and_vm()
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=False)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
self.result_table_print()
- def test_perf_pvp_dequeue_zero_copy_with_driver_unload(self):
+ def test_perf_pvp_packed_ring_dequeue_zero_copy_with_2_queue(self):
"""
- pvp dequeue zero-copy test with driver unload test
+ pvp packed ring dequeue zero-copy test with 2 queues
+ """
+ self.nb_cores = 2
+ self.queue_number = 2
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=True)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
+ self.result_table_print()
+
+ def test_perf_pvp_split_ring_dequeue_zero_copy_with_driver_unload(self):
+ """
+ pvp split ring dequeue zero-copy test with driver reload test
"""
self.nb_cores = 4
self.queue_number = 16
- self.launch_testpmd_on_vhost(txfreet="normal")
- self.start_one_vm()
- self.launch_testpmd_on_vm()
- # set fwd mode on vm
- self.vm_dut.send_expect("set fwd rxonly", "testpmd> ", 30)
- self.vm_dut.send_expect("start", "testpmd> ", 30)
- # start testpmd at host side after VM and virtio-pmd launched
- self.vhost.send_expect("start", "testpmd> ", 120)
- # when multi queues, the function will check each queue can receive packets
- self.send_and_verify(cycle="befor relaunch", loopback="rxonly")
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=True, vm_testpmd_fwd_mode='rxonly', packed_mode=False)
+ self.send_and_verify_throughput(cycle="before relaunch", fwd_mode="rxonly")
# relaunch testpmd at virtio side in VM for driver reloading
- self.relaunch_testpmd_on_vm()
- self.send_and_verify(cycle="after relaunch")
- self.close_all_testpmd_and_vm()
+ self.restart_testpmd_on_vm(fwd_mode='mac')
+ self.send_and_verify_throughput(cycle="after relaunch")
self.result_table_print()
- def test_perf_pvp_dequeue_zero_copy_with_maximum_txfreet(self):
+ def test_perf_pvp_packed_ring_dequeue_zero_copy_with_driver_unload(self):
"""
- pvp dequeue zero-copy test with maximum txfreet
+ pvp packed ring dequeue zero-copy test with driver reload test
"""
self.nb_cores = 4
self.queue_number = 16
- self.launch_testpmd_on_vhost(txfreet="maximum")
- self.start_one_vm()
- self.launch_testpmd_on_vm()
- # set fwd mode on vm
- self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
- self.vm_dut.send_expect("start", "testpmd> ", 30)
- # start testpmd at host side after VM and virtio-pmd launched
- self.vhost.send_expect("start", "testpmd> ", 120)
- # when multi queues, the function will check each queue can receive packets
- self.send_and_verify()
- self.close_all_testpmd_and_vm()
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=True, vm_testpmd_fwd_mode='rxonly', packed_mode=True)
+ self.send_and_verify_throughput(cycle="before relaunch", fwd_mode="rxonly")
+
+ # relaunch testpmd at virtio side in VM for driver reloading
+ self.restart_testpmd_on_vm(fwd_mode='mac')
+ self.send_and_verify_throughput(cycle="after relaunch")
+ self.result_table_print()
+
+ def test_perf_pvp_split_ring_dequeue_zero_copy_with_maximum_txfreet(self):
+ """
+ pvp split ring dequeue zero-copy test with maximum txfreet
+ """
+ self.nb_cores = 4
+ self.queue_number = 16
+ self.prepare_test_evn(vhost_txfreet_mode='maximum', vhost_zero_copy=True,
+ vhost_client_mode=True, vm_testpmd_fwd_mode='mac', packed_mode=False)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
+ self.result_table_print()
+
+ def test_perf_pvp_split_ring_dequeue_zero_copy_with_vector_rx(self):
+ """
+ pvp split ring dequeue zero-copy test with vector_rx path
+ """
+ self.nb_cores = 1
+ self.queue_number = 1
+ path_mode = 'packed_vq=0,in_order=0,mrg_rxbuf=0'
+ self.virtio_user = self.dut.new_session(suite="virtio-user")
+
+ self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to test')
+ self.launch_testpmd_as_vhost(txfreet="vector_rx", zero_copy=True, client_mode=False)
+ self.vhost_user.send_expect("start", "testpmd> ", 120)
+ self.launch_testpmd_as_virtio_user(path_mode)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
self.result_table_print()
def tear_down(self):
"""
Run after each test case.
"""
- self.dut.send_expect("killall -s INT testpmd", "#")
- self.dut.send_expect("killall -s INT qemu-system-x86_64", "#")
+ self.close_all_testpmd_and_vm()
+ self.dut.kill_all()
time.sleep(2)
def tear_down_all(self):
"""
Run after each test suite.
"""
+ if getattr(self, 'vhost_user', None):
+ self.dut.close_session(self.vhost_user)
--
2.7.4
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [dts] [PATCH V1] tests/vhost_dequeue_zero_copy: update suite according to plan
2019-12-24 21:53 [dts] [PATCH V1] tests/vhost_dequeue_zero_copy: update suite according to plan lihong
@ 2019-12-25 9:02 ` Ma, LihongX
2019-12-27 8:30 ` Wang, Yinan
2020-01-02 2:18 ` Tu, Lijuan
1 sibling, 1 reply; 4+ messages in thread
From: Ma, LihongX @ 2019-12-25 9:02 UTC (permalink / raw)
To: dts; +Cc: Wang, Yinan
[-- Attachment #1: Type: text/plain, Size: 28760 bytes --]
Tested-by:ma,lihong<lihongx.ma@intel.com>
-----Original Message-----
From: Ma, LihongX
Sent: Wednesday, December 25, 2019 5:53 AM
To: dts@dpdk.org
Cc: Wang, Yinan <yinan.wang@intel.com>; Ma, LihongX <lihongx.ma@intel.com>
Subject: [dts][PATCH V1] tests/vhost_dequeue_zero_copy: update suite according to plan
Signed-off-by: lihong <lihongx.ma@intel.com>
---
tests/TestSuite_vhost_dequeue_zero_copy.py | 393 +++++++++++++++++++----------
1 file changed, 261 insertions(+), 132 deletions(-)
diff --git a/tests/TestSuite_vhost_dequeue_zero_copy.py b/tests/TestSuite_vhost_dequeue_zero_copy.py
index 5705064..4052437 100644
--- a/tests/TestSuite_vhost_dequeue_zero_copy.py
+++ b/tests/TestSuite_vhost_dequeue_zero_copy.py
@@ -38,6 +38,7 @@ There are three topology test (PVP/VM2VM/VM2NIC) for this feature.
And this testsuite implement the topology of PVP.
Testsuite vm2vm_net_perf implement the topology VM2VM Testsuite gso implement the topology VM2NIC
+To run this suite, the qemu version should support packed ring.
"""
import utils
import time
@@ -45,7 +46,8 @@ import re
from settings import HEADER_SIZE
from virt_common import VM
from test_case import TestCase
-from pktgen import PacketGeneratorHelper
+from packet import Packet
+from pktgen import TRANSMIT_CONT
class TestVhostDequeueZeroCopy(TestCase):
@@ -54,19 +56,19 @@ class TestVhostDequeueZeroCopy(TestCase):
"""
Run at the start of each test suite.
"""
+ self.verify(self.nic in ("fortville_spirit"), "the port can not
+ run this suite")
self.tester.extend_external_packet_generator(TestVhostDequeueZeroCopy, self)
self.frame_sizes = [64, 128, 256, 512, 1024, 1518]
- self.queue_number = 1
- self.nb_cores = 1
self.dut_ports = self.dut.get_ports()
self.verify(len(self.dut_ports) >= 1, "Insufficient ports for testing")
self.ports_socket = self.dut.get_numa_id(self.dut_ports[0])
- self.cores_num = len([n for n in self.dut.cores if int(n['socket'])
- == self.ports_socket])
- self.mem_channels = self.dut.get_memory_channels()
self.dst_mac = self.dut.get_mac_address(self.dut_ports[0])
+ self.tx_port = self.tester.get_local_port(self.dut_ports[0])
+ self.port_pci = self.dut.ports_info[self.dut_ports[0]]['pci']
self.vm_dut = None
+ self.virtio_user = None
self.virtio1_mac = "52:54:00:00:00:01"
+ self.header_size = HEADER_SIZE['eth'] + HEADER_SIZE['ip'] +
+ HEADER_SIZE['udp']
self.logger.info("you can config packet_size in file %s.cfg," % self.suite_name + \
"in region 'suite' like packet_sizes=[64, 128, 256]") @@ -74,64 +76,93 @@ class TestVhostDequeueZeroCopy(TestCase):
if 'packet_sizes' in self.get_suite_cfg():
self.frame_sizes = self.get_suite_cfg()['packet_sizes']
- self.out_path = '/tmp'
- out = self.tester.send_expect('ls -d %s' % self.out_path, '# ')
- if 'No such file or directory' in out:
- self.tester.send_expect('mkdir -p %s' % self.out_path, '# ')
- # create an instance to set stream field setting
- self.pktgen_helper = PacketGeneratorHelper()
self.base_dir = self.dut.base_dir.replace('~', '/root')
+ self.vhost_user = self.dut.new_session(suite="vhost-user")
def set_up(self):
"""
Run before each test case.
"""
# Clean the execution ENV
- self.dut.send_expect("killall -s INT testpmd", "#")
- self.dut.send_expect("killall -s INT qemu-system-x86_64", "#")
self.dut.send_expect("rm -rf %s/vhost-net*" % self.base_dir, "#")
# Prepare the result table
self.table_header = ["FrameSize(B)", "Throughput(Mpps)",
"% linerate", "Queue number", "Cycle"]
self.result_table_create(self.table_header)
+ self.vm_dut = None
+ self.big_pkt_record = {}
- self.vhost = self.dut.new_session(suite="vhost-user")
-
- def get_core_mask(self):
+ def get_core_list(self):
"""
check whether the server has enough cores to run case
+ if want to get the best perf of the vhost, the vhost tesptmd at least
+ should have 3 cores to start testpmd
"""
- core_config = "1S/%dC/1T" % (self.nb_cores + 1)
- self.verify(self.cores_num >= (self.nb_cores + 1),
- "There has not enought cores to test this case %s" % self.running_case)
- core_list = self.dut.get_core_list(
+ if self.nb_cores == 1:
+ cores_num = 2
+ else:
+ cores_num = 1
+ core_config = "1S/%dC/1T" % (self.nb_cores + cores_num)
+ self.core_list = self.dut.get_core_list(
core_config, socket=self.ports_socket)
- self.core_mask = utils.create_mask(core_list)
+ self.verify(len(self.core_list) >= (self.nb_cores + cores_num),
+ "There has not enought cores to test this case %s" %
+ self.running_case)
- def launch_testpmd_on_vhost(self, txfreet):
+ def launch_testpmd_as_vhost(self, txfreet, zero_copy=True, client_mode=False):
"""
launch testpmd on vhost
"""
- self.get_core_mask()
+ self.get_core_list()
+ mode_info = ""
+ if client_mode is True:
+ mode_info = ',client=1'
+ zero_copy_info = 1
+ if zero_copy is False:
+ zero_copy_info = 0
if txfreet == "normal":
- txfreet_args = "--txfreet=992"
+ txfreet_args = "--txd=1024 --rxd=1024 --txfreet=992"
elif txfreet == "maximum":
- txfreet_args = "--txfreet=1020 --txrs=4"
- command_client = self.dut.target + "/app/testpmd " + \
- " -n %d -c %s --socket-mem 1024,1024 " + \
- " --legacy-mem --file-prefix=vhost " + \
- " --vdev 'eth_vhost0,iface=%s/vhost-net,queues=%d,dequeue-zero-copy=1' " + \
- " -- -i --nb-cores=%d --rxq=%d --txq=%d " + \
- "--txd=1024 --rxd=1024 %s"
+ txfreet_args = "--txrs=4 --txd=992 --rxd=992 --txfreet=988"
+ elif txfreet == "vector_rx":
+ txfreet_args = "--txd=1024 --rxd=1024 --txfreet=992 --txrs=32"
+
+ eal_params = self.dut.create_eal_parameters(cores=self.core_list,
+ prefix='vhost', ports=[self.port_pci])
+ command_client = self.dut.target + "/app/testpmd %s " + \
+ " --socket-mem 1024,1024 --legacy-mem " + \
+ " --vdev 'eth_vhost0,iface=%s/vhost-net,queues=%d,dequeue-zero-copy=%d%s' " + \
+ " -- -i --nb-cores=%d --rxq=%d --txq=%d %s"
command_line_client = command_client % (
- self.mem_channels, self.core_mask, self.base_dir,
- self.queue_number, self.nb_cores,
+ eal_params, self.base_dir, self.queue_number,
+ zero_copy_info, mode_info, self.nb_cores,
self.queue_number, self.queue_number, txfreet_args)
- self.vhost.send_expect(command_line_client, "testpmd> ", 120)
- self.vhost.send_expect("set fwd mac", "testpmd> ", 120)
+ self.vhost_user.send_expect(command_line_client, "testpmd> ", 120)
+ self.vhost_user.send_expect("set fwd mac", "testpmd> ", 120)
- def launch_testpmd_on_vm(self):
+ def launch_testpmd_as_virtio_user(self, path_mode):
+ """
+ launch testpmd use vhost-net with path mode
+ """
+ # To get the best perf, the vhost and virtio testpmd should not use same cores,
+ # so get the additional 3 cores to start virtio testpmd
+ core_config = "1S/%dC/1T" % (len(self.core_list) + 3)
+ core_list = self.dut.get_core_list(
+ core_config, socket=self.ports_socket)
+ self.verify(len(core_list) >= (len(self.core_list) + 3),
+ "There has not enought cores to test this case %s" % self.running_case)
+ eal_params = self.dut.create_eal_parameters(cores=core_list[len(self.core_list):],
+ prefix='virtio', no_pci=True)
+ command_line = self.dut.target + "/app/testpmd %s " + \
+ "--socket-mem 1024,1024 --legacy-mem " + \
+ "--vdev=net_virtio_user0,mac=00:01:02:03:04:05,path=./vhost-net,queue_size=1024,%s " + \
+ "-- -i --tx-offloads=0x0 --nb-cores=%d --txd=1024 --rxd=1024"
+ command_line = command_line % (eal_params, path_mode, self.nb_cores)
+ self.virtio_user.send_expect(command_line, 'testpmd> ', 120)
+ self.virtio_user.send_expect('set fwd mac', 'testpmd> ', 120)
+ self.virtio_user.send_expect('start', 'testpmd> ', 120)
+
+ def start_testpmd_on_vm(self, fwd_mode='mac'):
"""
start testpmd in vm depend on different path
"""
@@ -142,12 +173,12 @@ class TestVhostDequeueZeroCopy(TestCase):
command_line = command % (self.nb_cores,
self.queue_number, self.queue_number)
self.vm_dut.send_expect(command_line, "testpmd> ", 30)
+ self.vm_dut.send_expect('set fwd %s' % fwd_mode, "testpmd> ", 30)
+ self.vm_dut.send_expect('start', "testpmd> ", 30)
- def relaunch_testpmd_on_vm(self):
+ def restart_testpmd_on_vm(self, fwd_mode):
self.vm_dut.send_expect("quit", "# ", 30)
- self.launch_testpmd_on_vm()
- self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
- self.vm_dut.send_expect("start", "testpmd> ", 30)
+ self.start_testpmd_on_vm(fwd_mode)
def set_vm_vcpu(self):
"""
@@ -166,7 +197,7 @@ class TestVhostDequeueZeroCopy(TestCase):
if 'cpupin' in self.vm.params[i]['cpu'][0].keys():
self.vm.params[i]['cpu'][0].pop('cpupin')
- def start_one_vm(self):
+ def start_one_vm(self, mode='client', packed=False):
"""
start qemu
"""
@@ -176,10 +207,14 @@ class TestVhostDequeueZeroCopy(TestCase):
vm_params['driver'] = 'vhost-user'
vm_params['opt_path'] = '%s/vhost-net' % self.base_dir
vm_params['opt_mac'] = self.virtio1_mac
+ if mode == 'server':
+ vm_params['opt_server'] = 'server'
opt_args = "mrg_rxbuf=on,rx_queue_size=1024,tx_queue_size=1024"
if self.queue_number > 1:
vm_params['opt_queue'] = self.queue_number
opt_args += ",mq=on,vectors=%d" % (2*self.queue_number + 2)
+ if packed is True:
+ opt_args += ',packed=on'
vm_params['opt_settings'] = opt_args
self.vm.set_vm_device(**vm_params)
self.set_vm_vcpu()
@@ -190,7 +225,26 @@ class TestVhostDequeueZeroCopy(TestCase):
if self.vm_dut is None:
raise Exception("Set up VM ENV failed")
except Exception as e:
- self.logger.error("ERROR: Failure for %s" % str(e))
+ self.logger.error("ERROR: Failure for %s, " % str(e) + \
+ "if 'packed not found' in output of start qemu log, " + \
+ "please use the qemu version which support packed ring")
+ raise e
+
+ def prepare_test_evn(self, vhost_txfreet_mode, vhost_zero_copy, vhost_client_mode,
+ vm_testpmd_fwd_mode, packed_mode):
+ """
+ start vhost testpmd and launch qemu, start testpmd on vm
+ """
+ if vhost_client_mode is True:
+ vm_mode = 'server'
+ else:
+ vm_mode = 'client'
+ self.launch_testpmd_as_vhost(txfreet=vhost_txfreet_mode, zero_copy=vhost_zero_copy,
+ client_mode=vhost_client_mode)
+ self.start_one_vm(mode=vm_mode, packed=packed_mode)
+ self.start_testpmd_on_vm(fwd_mode=vm_testpmd_fwd_mode)
+ # start testpmd at host side after VM and virtio-pmd launched
+ self.vhost_user.send_expect("start", "testpmd> ", 120)
def update_table_info(self, frame_size, Mpps, throughtput, cycle):
results_row = [frame_size]
@@ -200,50 +254,64 @@ class TestVhostDequeueZeroCopy(TestCase):
results_row.append(cycle)
self.result_table_add(results_row)
- def set_fields(self):
- """
- set ip protocol field behavior
- """
- fields_config = {'ip': {'dst': {'action': 'random'}, }, }
- return fields_config
+ # record the big pkt Mpps
+ if frame_size == 1518:
+ self.big_pkt_record[cycle] = Mpps
- def calculate_avg_throughput(self, frame_size, loopback):
+ def calculate_avg_throughput(self, frame_size, fwd_mode):
"""
start to send packet and get the throughput
"""
- payload = frame_size - HEADER_SIZE['eth'] - HEADER_SIZE['ip'] - HEADER_SIZE['udp']
- flow = '[Ether(dst="%s")/IP(src="192.168.4.1",proto=255)/UDP()/("X"*%d)]' % (
+ payload = frame_size - self.header_size
+ flow =
+ 'Ether(dst="%s")/IP(src="192.168.4.1",proto=255)/UDP(sport=33,dport=34
+ )/("X"*%d)' % (
self.dst_mac, payload)
- self.tester.scapy_append('wrpcap("%s/zero_copy.pcap", %s)' % (
- self.out_path, flow))
- self.tester.scapy_execute()
-
- tgenInput = []
- port = self.tester.get_local_port(self.dut_ports[0])
- tgenInput.append((port, port, "%s/zero_copy.pcap" % self.out_path))
- vm_config = self.set_fields()
+ pkt = Packet(pkt_str=flow)
+ pkt.save_pcapfile(self.tester, "%s/zero_copy.pcap" % self.tester.tmp_file)
+ stream_option = {
+ 'pcap': "%s/zero_copy.pcap" % self.tester.tmp_file,
+ 'fields_config': {
+ 'ip': {'src': {'action': 'random', 'start': '16.0.0.1', 'step': 1, 'end': '16.0.0.64'}}},
+ 'stream_config': {
+ 'rate': 100,
+ 'transmit_mode': TRANSMIT_CONT,
+ }
+ }
self.tester.pktgen.clear_streams()
- streams = self.pktgen_helper.prepare_stream_from_tginput(tgenInput, 100, vm_config, self.tester.pktgen)
- # set traffic option
- traffic_opt = {'delay': 5, 'duration': 20}
- _, pps = self.tester.pktgen.measure_throughput(stream_ids=streams, options=traffic_opt)
- Mpps = pps / 1000000.0
+ stream_id = self.tester.pktgen.add_stream(self.tx_port, self.tx_port,
+ "%s/zero_copy.pcap" % self.tester.tmp_file)
+ self.tester.pktgen.config_stream(stream_id, stream_option)
+ traffic_opt = {
+ 'method': 'throughput',
+ 'rate': 100,
+ 'interval': 6,
+ 'duration': 30}
+ stats = self.tester.pktgen.measure([stream_id], traffic_opt)
+
+ if isinstance(stats, list):
+ # if get multi result, ignore the first one, because it may not stable
+ num = len(stats)
+ Mpps = 0
+ for index in range(1, num):
+ Mpps += stats[index][1]
+ Mpps = Mpps / 1000000.0 / (num-1)
+ else:
+ Mpps = stats[1] / 1000000.0
# when the fwd mode is rxonly, we can not receive data, so should not verify it
- if loopback != "rxonly":
+ if fwd_mode != "rxonly":
self.verify(Mpps > 0, "can not receive packets of frame size %d" % (frame_size))
throughput = Mpps * 100 / \
float(self.wirespeed(self.nic, frame_size, 1))
return Mpps, throughput
- def check_packets_of_each_queue(self, frame_size, loopback):
+ def check_packets_of_each_queue(self, frame_size, fwd_mode):
"""
check each queue has receive packets
"""
- if loopback == "rxonly":
+ if fwd_mode == "rxonly":
verify_port = 1
else:
verify_port = 2
- out = self.vhost.send_expect("stop", "testpmd> ", 60)
+ out = self.vhost_user.send_expect("stop", "testpmd> ", 60)
for port_index in range(0, verify_port):
for queue_index in range(0, self.queue_number):
queue_info = re.findall("RX\s*Port=\s*%d/Queue=\s*%d" % @@ -260,9 +328,9 @@ class TestVhostDequeueZeroCopy(TestCase):
"frame_size:%d, rx-packets:%d, tx-packets:%d" %
(frame_size, rx_packets, tx_packets))
- self.vhost.send_expect("start", "testpmd> ", 60)
+ self.vhost_user.send_expect("start", "testpmd> ", 60)
- def send_and_verify(self, cycle="", loopback=""):
+ def send_and_verify_throughput(self, cycle="", fwd_mode=""):
"""
start to send packets and verify it
"""
@@ -270,110 +338,171 @@ class TestVhostDequeueZeroCopy(TestCase):
info = "Running test %s, and %d frame size." % (self.running_case, frame_size)
self.logger.info(info)
- Mpps, throughput = self.calculate_avg_throughput(frame_size, loopback)
- if loopback != "rxonly":
+ Mpps, throughput = self.calculate_avg_throughput(frame_size, fwd_mode)
+ if fwd_mode != "rxonly":
self.update_table_info(frame_size, Mpps, throughput, cycle)
# when multi queues, check each queue can receive packets
if self.queue_number > 1:
- self.check_packets_of_each_queue(frame_size, loopback)
+ self.check_packets_of_each_queue(frame_size, fwd_mode)
+
+ def check_perf_drop_between_with_and_without_zero_copy(self):
+ """
+ for dequeue-zero-copy=0, about the small pkts we expect ~10% gain
+ compare to dequeue-zero-copy=1
+ """
+ value_with_zero_copy = 0
+ value_without_zero_copy = 0
+ if 'dequeue-zero-copy=1' in self.big_pkt_record.keys():
+ value_with_zero_copy = self.big_pkt_record['dequeue-zero-copy=1']
+ if 'dequeue-zero-copy=0' in self.big_pkt_record.keys():
+ value_without_zero_copy = self.big_pkt_record['dequeue-zero-copy=0']
+ self.verify(value_with_zero_copy != 0 and value_without_zero_copy != 0,
+ 'can not get the value of big pkts, please check self.frame_sizes')
+ self.verify(value_with_zero_copy - value_without_zero_copy >= value_with_zero_copy*0.1,
+ 'the drop with dequeue-zero-copy=0 is not as expected')
def close_all_testpmd_and_vm(self):
"""
close testpmd about vhost-user and vm_testpmd
"""
- self.vhost.send_expect("quit", "#", 60)
- self.vm_dut.send_expect("quit", "#", 60)
- self.dut.close_session(self.vhost)
- self.vm.stop()
+ if getattr(self, 'vhost_user', None):
+ self.vhost_user.send_expect("quit", "#", 60)
+ if getattr(self, 'virtio_user', None):
+ self.virtio_user.send_expect("quit", "#", 60)
+ self.dut.close_session(self.virtio_user)
+ self.virtio_user = None
+ if getattr(self, 'vm_dut', None):
+ self.vm_dut.send_expect("quit", "#", 60)
+ self.vm.stop()
- def test_perf_pvp_dequeue_zero_copy(self):
+ def test_perf_pvp_split_ring_dequeue_zero_copy(self):
"""
- performance of pvp zero-copy with [frame_sizes]
+ pvp split ring dequeue zero-copy test
"""
self.nb_cores = 1
self.queue_number = 1
- self.launch_testpmd_on_vhost(txfreet="normal")
- self.start_one_vm()
- self.launch_testpmd_on_vm()
- # set fwd mode on vm
- self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
- self.vm_dut.send_expect("start", "testpmd> ", 30)
- # start testpmd at host side after VM and virtio-pmd launched
- self.vhost.send_expect("start", "testpmd> ", 120)
- self.send_and_verify()
+ self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to test')
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=False)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
+
self.close_all_testpmd_and_vm()
+ self.logger.info('start vhost testpmd with dequeue-zero-copy=0 to test')
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=False,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=False)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=0')
self.result_table_print()
+ self.check_perf_drop_between_with_and_without_zero_copy()
- def test_perf_pvp_dequeue_zero_copy_with_2_queue(self):
+ def test_perf_pvp_packed_ring_dequeue_zero_copy(self):
"""
- pvp dequeue zero-copy test with 2 queues
+ pvp packed ring dequeue zero-copy test
+ """
+ self.nb_cores = 1
+ self.queue_number = 1
+ self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to test')
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=True)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
+
+ self.close_all_testpmd_and_vm()
+ self.logger.info('start vhost testpmd with dequeue-zero-copy=0 to test')
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=False,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=True)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=0')
+ self.result_table_print()
+ self.check_perf_drop_between_with_and_without_zero_copy()
+
+ def test_perf_pvp_split_ring_dequeue_zero_copy_with_2_queue(self):
+ """
+ pvp split ring dequeue zero-copy test with 2 queues
"""
self.nb_cores = 2
self.queue_number = 2
- self.launch_testpmd_on_vhost(txfreet="normal")
- self.start_one_vm()
- self.launch_testpmd_on_vm()
- # set fwd mode on vm
- self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
- self.vm_dut.send_expect("start", "testpmd> ", 30)
- # start testpmd at host side after VM and virtio-pmd launched
- self.vhost.send_expect("start", "testpmd> ", 120)
- # when multi queues, the function will check each queue can receive packets
- self.send_and_verify()
- self.close_all_testpmd_and_vm()
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=False)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
self.result_table_print()
- def test_perf_pvp_dequeue_zero_copy_with_driver_unload(self):
+ def test_perf_pvp_packed_ring_dequeue_zero_copy_with_2_queue(self):
"""
- pvp dequeue zero-copy test with driver unload test
+ pvp packed ring dequeue zero-copy test with 2 queues
+ """
+ self.nb_cores = 2
+ self.queue_number = 2
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=False, vm_testpmd_fwd_mode='mac', packed_mode=True)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
+ self.result_table_print()
+
+ def test_perf_pvp_split_ring_dequeue_zero_copy_with_driver_unload(self):
+ """
+ pvp split ring dequeue zero-copy test with driver reload test
"""
self.nb_cores = 4
self.queue_number = 16
- self.launch_testpmd_on_vhost(txfreet="normal")
- self.start_one_vm()
- self.launch_testpmd_on_vm()
- # set fwd mode on vm
- self.vm_dut.send_expect("set fwd rxonly", "testpmd> ", 30)
- self.vm_dut.send_expect("start", "testpmd> ", 30)
- # start testpmd at host side after VM and virtio-pmd launched
- self.vhost.send_expect("start", "testpmd> ", 120)
- # when multi queues, the function will check each queue can receive packets
- self.send_and_verify(cycle="befor relaunch", loopback="rxonly")
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=True, vm_testpmd_fwd_mode='rxonly', packed_mode=False)
+ self.send_and_verify_throughput(cycle="before relaunch",
+ fwd_mode="rxonly")
# relaunch testpmd at virtio side in VM for driver reloading
- self.relaunch_testpmd_on_vm()
- self.send_and_verify(cycle="after relaunch")
- self.close_all_testpmd_and_vm()
+ self.restart_testpmd_on_vm(fwd_mode='mac')
+ self.send_and_verify_throughput(cycle="after relaunch")
self.result_table_print()
- def test_perf_pvp_dequeue_zero_copy_with_maximum_txfreet(self):
+ def test_perf_pvp_packed_ring_dequeue_zero_copy_with_driver_unload(self):
"""
- pvp dequeue zero-copy test with maximum txfreet
+ pvp packed ring dequeue zero-copy test with driver reload test
"""
self.nb_cores = 4
self.queue_number = 16
- self.launch_testpmd_on_vhost(txfreet="maximum")
- self.start_one_vm()
- self.launch_testpmd_on_vm()
- # set fwd mode on vm
- self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
- self.vm_dut.send_expect("start", "testpmd> ", 30)
- # start testpmd at host side after VM and virtio-pmd launched
- self.vhost.send_expect("start", "testpmd> ", 120)
- # when multi queues, the function will check each queue can receive packets
- self.send_and_verify()
- self.close_all_testpmd_and_vm()
+ self.prepare_test_evn(vhost_txfreet_mode='normal', vhost_zero_copy=True,
+ vhost_client_mode=True, vm_testpmd_fwd_mode='rxonly', packed_mode=True)
+ self.send_and_verify_throughput(cycle="before relaunch",
+ fwd_mode="rxonly")
+
+ # relaunch testpmd at virtio side in VM for driver reloading
+ self.restart_testpmd_on_vm(fwd_mode='mac')
+ self.send_and_verify_throughput(cycle="after relaunch")
+ self.result_table_print()
+
+ def test_perf_pvp_split_ring_dequeue_zero_copy_with_maximum_txfreet(self):
+ """
+ pvp split ring dequeue zero-copy test with maximum txfreet
+ """
+ self.nb_cores = 4
+ self.queue_number = 16
+ self.prepare_test_evn(vhost_txfreet_mode='maximum', vhost_zero_copy=True,
+ vhost_client_mode=True, vm_testpmd_fwd_mode='mac', packed_mode=False)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
+ self.result_table_print()
+
+ def test_perf_pvp_split_ring_dequeue_zero_copy_with_vector_rx(self):
+ """
+ pvp split ring dequeue zero-copy test with vector_rx path
+ """
+ self.nb_cores = 1
+ self.queue_number = 1
+ path_mode = 'packed_vq=0,in_order=0,mrg_rxbuf=0'
+ self.virtio_user = self.dut.new_session(suite="virtio-user")
+
+ self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to test')
+ self.launch_testpmd_as_vhost(txfreet="vector_rx", zero_copy=True, client_mode=False)
+ self.vhost_user.send_expect("start", "testpmd> ", 120)
+ self.launch_testpmd_as_virtio_user(path_mode)
+ self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
self.result_table_print()
def tear_down(self):
"""
Run after each test case.
"""
- self.dut.send_expect("killall -s INT testpmd", "#")
- self.dut.send_expect("killall -s INT qemu-system-x86_64", "#")
+ self.close_all_testpmd_and_vm()
+ self.dut.kill_all()
time.sleep(2)
def tear_down_all(self):
"""
Run after each test suite.
"""
+ if getattr(self, 'vhost_user', None):
+ self.dut.close_session(self.vhost_user)
--
2.7.4
[-- Attachment #2: TestVhostDequeueZeroCopy.log --]
[-- Type: application/octet-stream, Size: 659064 bytes --]
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [dts] [PATCH V1] tests/vhost_dequeue_zero_copy: update suite according to plan
2019-12-25 9:02 ` Ma, LihongX
@ 2019-12-27 8:30 ` Wang, Yinan
0 siblings, 0 replies; 4+ messages in thread
From: Wang, Yinan @ 2019-12-27 8:30 UTC (permalink / raw)
To: Ma, LihongX, dts
Acked-by: Wang, Yinan <yinan.wang@intel.com>
> -----Original Message-----
> From: Ma, LihongX <lihongx.ma@intel.com>
> Sent: 2019年12月25日 17:03
> To: dts@dpdk.org
> Cc: Wang, Yinan <yinan.wang@intel.com>
> Subject: RE: [dts][PATCH V1] tests/vhost_dequeue_zero_copy: update suite
> according to plan
>
> Tested-by:ma,lihong<lihongx.ma@intel.com>
>
> -----Original Message-----
> From: Ma, LihongX
> Sent: Wednesday, December 25, 2019 5:53 AM
> To: dts@dpdk.org
> Cc: Wang, Yinan <yinan.wang@intel.com>; Ma, LihongX
> <lihongx.ma@intel.com>
> Subject: [dts][PATCH V1] tests/vhost_dequeue_zero_copy: update suite
> according to plan
>
> Signed-off-by: lihong <lihongx.ma@intel.com>
> ---
> tests/TestSuite_vhost_dequeue_zero_copy.py | 393
> +++++++++++++++++++----------
> 1 file changed, 261 insertions(+), 132 deletions(-)
>
> diff --git a/tests/TestSuite_vhost_dequeue_zero_copy.py
> b/tests/TestSuite_vhost_dequeue_zero_copy.py
> index 5705064..4052437 100644
> --- a/tests/TestSuite_vhost_dequeue_zero_copy.py
> +++ b/tests/TestSuite_vhost_dequeue_zero_copy.py
> @@ -38,6 +38,7 @@ There are three topology test (PVP/VM2VM/VM2NIC) for
> this feature.
> And this testsuite implement the topology of PVP.
> Testsuite vm2vm_net_perf implement the topology VM2VM Testsuite gso
> implement the topology VM2NIC
> +To run this suite, the qemu version should support packed ring.
> """
> import utils
> import time
> @@ -45,7 +46,8 @@ import re
> from settings import HEADER_SIZE
> from virt_common import VM
> from test_case import TestCase
> -from pktgen import PacketGeneratorHelper
> +from packet import Packet
> +from pktgen import TRANSMIT_CONT
>
>
> class TestVhostDequeueZeroCopy(TestCase):
> @@ -54,19 +56,19 @@ class TestVhostDequeueZeroCopy(TestCase):
> """
> Run at the start of each test suite.
> """
> + self.verify(self.nic in ("fortville_spirit"), "the port can not
> + run this suite")
>
> self.tester.extend_external_packet_generator(TestVhostDequeueZeroCopy, self)
> self.frame_sizes = [64, 128, 256, 512, 1024, 1518]
> - self.queue_number = 1
> - self.nb_cores = 1
> self.dut_ports = self.dut.get_ports()
> self.verify(len(self.dut_ports) >= 1, "Insufficient ports for testing")
> self.ports_socket = self.dut.get_numa_id(self.dut_ports[0])
> - self.cores_num = len([n for n in self.dut.cores if int(n['socket'])
> - == self.ports_socket])
> - self.mem_channels = self.dut.get_memory_channels()
> self.dst_mac = self.dut.get_mac_address(self.dut_ports[0])
> + self.tx_port = self.tester.get_local_port(self.dut_ports[0])
> + self.port_pci = self.dut.ports_info[self.dut_ports[0]]['pci']
> self.vm_dut = None
> + self.virtio_user = None
> self.virtio1_mac = "52:54:00:00:00:01"
> + self.header_size = HEADER_SIZE['eth'] + HEADER_SIZE['ip'] +
> + HEADER_SIZE['udp']
>
> self.logger.info("you can config packet_size in file %s.cfg," %
> self.suite_name + \
> "in region 'suite' like packet_sizes=[64, 128, 256]")
> @@ -74,64 +76,93 @@ class TestVhostDequeueZeroCopy(TestCase):
> if 'packet_sizes' in self.get_suite_cfg():
> self.frame_sizes = self.get_suite_cfg()['packet_sizes']
>
> - self.out_path = '/tmp'
> - out = self.tester.send_expect('ls -d %s' % self.out_path, '# ')
> - if 'No such file or directory' in out:
> - self.tester.send_expect('mkdir -p %s' % self.out_path, '# ')
> - # create an instance to set stream field setting
> - self.pktgen_helper = PacketGeneratorHelper()
> self.base_dir = self.dut.base_dir.replace('~', '/root')
> + self.vhost_user = self.dut.new_session(suite="vhost-user")
>
> def set_up(self):
> """
> Run before each test case.
> """
> # Clean the execution ENV
> - self.dut.send_expect("killall -s INT testpmd", "#")
> - self.dut.send_expect("killall -s INT qemu-system-x86_64", "#")
> self.dut.send_expect("rm -rf %s/vhost-net*" % self.base_dir, "#")
> # Prepare the result table
> self.table_header = ["FrameSize(B)", "Throughput(Mpps)",
> "% linerate", "Queue number", "Cycle"]
> self.result_table_create(self.table_header)
> + self.vm_dut = None
> + self.big_pkt_record = {}
>
> - self.vhost = self.dut.new_session(suite="vhost-user")
> -
> - def get_core_mask(self):
> + def get_core_list(self):
> """
> check whether the server has enough cores to run case
> + if want to get the best perf of the vhost, the vhost tesptmd at least
> + should have 3 cores to start testpmd
> """
> - core_config = "1S/%dC/1T" % (self.nb_cores + 1)
> - self.verify(self.cores_num >= (self.nb_cores + 1),
> - "There has not enought cores to test this case %s" %
> self.running_case)
> - core_list = self.dut.get_core_list(
> + if self.nb_cores == 1:
> + cores_num = 2
> + else:
> + cores_num = 1
> + core_config = "1S/%dC/1T" % (self.nb_cores + cores_num)
> + self.core_list = self.dut.get_core_list(
> core_config, socket=self.ports_socket)
> - self.core_mask = utils.create_mask(core_list)
> + self.verify(len(self.core_list) >= (self.nb_cores + cores_num),
> + "There has not enought cores to test this case %s" %
> + self.running_case)
>
> - def launch_testpmd_on_vhost(self, txfreet):
> + def launch_testpmd_as_vhost(self, txfreet, zero_copy=True,
> client_mode=False):
> """
> launch testpmd on vhost
> """
> - self.get_core_mask()
> + self.get_core_list()
>
> + mode_info = ""
> + if client_mode is True:
> + mode_info = ',client=1'
> + zero_copy_info = 1
> + if zero_copy is False:
> + zero_copy_info = 0
> if txfreet == "normal":
> - txfreet_args = "--txfreet=992"
> + txfreet_args = "--txd=1024 --rxd=1024 --txfreet=992"
> elif txfreet == "maximum":
> - txfreet_args = "--txfreet=1020 --txrs=4"
> - command_client = self.dut.target + "/app/testpmd " + \
> - " -n %d -c %s --socket-mem 1024,1024 " + \
> - " --legacy-mem --file-prefix=vhost " + \
> - " --vdev
> 'eth_vhost0,iface=%s/vhost-net,queues=%d,dequeue-zero-copy=1' " + \
> - " -- -i --nb-cores=%d --rxq=%d --txq=%d " + \
> - "--txd=1024 --rxd=1024 %s"
> + txfreet_args = "--txrs=4 --txd=992 --rxd=992 --txfreet=988"
> + elif txfreet == "vector_rx":
> + txfreet_args = "--txd=1024 --rxd=1024 --txfreet=992 --txrs=32"
> +
> + eal_params = self.dut.create_eal_parameters(cores=self.core_list,
> + prefix='vhost', ports=[self.port_pci])
> + command_client = self.dut.target + "/app/testpmd %s " + \
> + " --socket-mem 1024,1024 --legacy-mem " + \
> + " --vdev
> 'eth_vhost0,iface=%s/vhost-net,queues=%d,dequeue-zero-copy=%d%s' " + \
> + " -- -i --nb-cores=%d --rxq=%d --txq=%d %s"
> command_line_client = command_client % (
> - self.mem_channels, self.core_mask, self.base_dir,
> - self.queue_number, self.nb_cores,
> + eal_params, self.base_dir, self.queue_number,
> + zero_copy_info, mode_info, self.nb_cores,
> self.queue_number, self.queue_number, txfreet_args)
> - self.vhost.send_expect(command_line_client, "testpmd> ", 120)
> - self.vhost.send_expect("set fwd mac", "testpmd> ", 120)
> + self.vhost_user.send_expect(command_line_client, "testpmd> ", 120)
> + self.vhost_user.send_expect("set fwd mac", "testpmd> ", 120)
>
> - def launch_testpmd_on_vm(self):
> + def launch_testpmd_as_virtio_user(self, path_mode):
> + """
> + launch testpmd use vhost-net with path mode
> + """
> + # To get the best perf, the vhost and virtio testpmd should not use
> same cores,
> + # so get the additional 3 cores to start virtio testpmd
> + core_config = "1S/%dC/1T" % (len(self.core_list) + 3)
> + core_list = self.dut.get_core_list(
> + core_config, socket=self.ports_socket)
> + self.verify(len(core_list) >= (len(self.core_list) + 3),
> + "There has not enought cores to test this case %s" %
> self.running_case)
> + eal_params =
> self.dut.create_eal_parameters(cores=core_list[len(self.core_list):],
> + prefix='virtio', no_pci=True)
> + command_line = self.dut.target + "/app/testpmd %s " + \
> + "--socket-mem 1024,1024 --legacy-mem " + \
> +
> "--vdev=net_virtio_user0,mac=00:01:02:03:04:05,path=./vhost-net,queue_size=
> 1024,%s " + \
> + "-- -i --tx-offloads=0x0 --nb-cores=%d --txd=1024
> --rxd=1024"
> + command_line = command_line % (eal_params, path_mode,
> self.nb_cores)
> + self.virtio_user.send_expect(command_line, 'testpmd> ', 120)
> + self.virtio_user.send_expect('set fwd mac', 'testpmd> ', 120)
> + self.virtio_user.send_expect('start', 'testpmd> ', 120)
> +
> + def start_testpmd_on_vm(self, fwd_mode='mac'):
> """
> start testpmd in vm depend on different path
> """
> @@ -142,12 +173,12 @@ class TestVhostDequeueZeroCopy(TestCase):
> command_line = command % (self.nb_cores,
> self.queue_number,
> self.queue_number)
> self.vm_dut.send_expect(command_line, "testpmd> ", 30)
> + self.vm_dut.send_expect('set fwd %s' % fwd_mode, "testpmd> ", 30)
> + self.vm_dut.send_expect('start', "testpmd> ", 30)
>
> - def relaunch_testpmd_on_vm(self):
> + def restart_testpmd_on_vm(self, fwd_mode):
> self.vm_dut.send_expect("quit", "# ", 30)
> - self.launch_testpmd_on_vm()
> - self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
> - self.vm_dut.send_expect("start", "testpmd> ", 30)
> + self.start_testpmd_on_vm(fwd_mode)
>
> def set_vm_vcpu(self):
> """
> @@ -166,7 +197,7 @@ class TestVhostDequeueZeroCopy(TestCase):
> if 'cpupin' in self.vm.params[i]['cpu'][0].keys():
> self.vm.params[i]['cpu'][0].pop('cpupin')
>
> - def start_one_vm(self):
> + def start_one_vm(self, mode='client', packed=False):
> """
> start qemu
> """
> @@ -176,10 +207,14 @@ class TestVhostDequeueZeroCopy(TestCase):
> vm_params['driver'] = 'vhost-user'
> vm_params['opt_path'] = '%s/vhost-net' % self.base_dir
> vm_params['opt_mac'] = self.virtio1_mac
> + if mode == 'server':
> + vm_params['opt_server'] = 'server'
> opt_args =
> "mrg_rxbuf=on,rx_queue_size=1024,tx_queue_size=1024"
> if self.queue_number > 1:
> vm_params['opt_queue'] = self.queue_number
> opt_args += ",mq=on,vectors=%d" % (2*self.queue_number + 2)
> + if packed is True:
> + opt_args += ',packed=on'
> vm_params['opt_settings'] = opt_args
> self.vm.set_vm_device(**vm_params)
> self.set_vm_vcpu()
> @@ -190,7 +225,26 @@ class TestVhostDequeueZeroCopy(TestCase):
> if self.vm_dut is None:
> raise Exception("Set up VM ENV failed")
> except Exception as e:
> - self.logger.error("ERROR: Failure for %s" % str(e))
> + self.logger.error("ERROR: Failure for %s, " % str(e) + \
> + "if 'packed not found' in output of start qemu log, " + \
> + "please use the qemu version which support packed
> ring")
> + raise e
> +
> + def prepare_test_evn(self, vhost_txfreet_mode, vhost_zero_copy,
> vhost_client_mode,
> + vm_testpmd_fwd_mode, packed_mode):
> + """
> + start vhost testpmd and launch qemu, start testpmd on vm
> + """
> + if vhost_client_mode is True:
> + vm_mode = 'server'
> + else:
> + vm_mode = 'client'
> + self.launch_testpmd_as_vhost(txfreet=vhost_txfreet_mode,
> zero_copy=vhost_zero_copy,
> + client_mode=vhost_client_mode)
> + self.start_one_vm(mode=vm_mode, packed=packed_mode)
> + self.start_testpmd_on_vm(fwd_mode=vm_testpmd_fwd_mode)
> + # start testpmd at host side after VM and virtio-pmd launched
> + self.vhost_user.send_expect("start", "testpmd> ", 120)
>
> def update_table_info(self, frame_size, Mpps, throughtput, cycle):
> results_row = [frame_size]
> @@ -200,50 +254,64 @@ class TestVhostDequeueZeroCopy(TestCase):
> results_row.append(cycle)
> self.result_table_add(results_row)
>
> - def set_fields(self):
> - """
> - set ip protocol field behavior
> - """
> - fields_config = {'ip': {'dst': {'action': 'random'}, }, }
> - return fields_config
> + # record the big pkt Mpps
> + if frame_size == 1518:
> + self.big_pkt_record[cycle] = Mpps
>
> - def calculate_avg_throughput(self, frame_size, loopback):
> + def calculate_avg_throughput(self, frame_size, fwd_mode):
> """
> start to send packet and get the throughput
> """
> - payload = frame_size - HEADER_SIZE['eth'] - HEADER_SIZE['ip'] -
> HEADER_SIZE['udp']
> - flow =
> '[Ether(dst="%s")/IP(src="192.168.4.1",proto=255)/UDP()/("X"*%d)]' % (
> + payload = frame_size - self.header_size
> + flow =
> + 'Ether(dst="%s")/IP(src="192.168.4.1",proto=255)/UDP(sport=33,dport=34
> + )/("X"*%d)' % (
> self.dst_mac, payload)
> - self.tester.scapy_append('wrpcap("%s/zero_copy.pcap", %s)' % (
> - self.out_path, flow))
> - self.tester.scapy_execute()
> -
> - tgenInput = []
> - port = self.tester.get_local_port(self.dut_ports[0])
> - tgenInput.append((port, port, "%s/zero_copy.pcap" % self.out_path))
> - vm_config = self.set_fields()
> + pkt = Packet(pkt_str=flow)
> + pkt.save_pcapfile(self.tester, "%s/zero_copy.pcap" %
> self.tester.tmp_file)
> + stream_option = {
> + 'pcap': "%s/zero_copy.pcap" % self.tester.tmp_file,
> + 'fields_config': {
> + 'ip': {'src': {'action': 'random', 'start': '16.0.0.1', 'step': 1,
> 'end': '16.0.0.64'}}},
> + 'stream_config': {
> + 'rate': 100,
> + 'transmit_mode': TRANSMIT_CONT,
> + }
> + }
> self.tester.pktgen.clear_streams()
> - streams =
> self.pktgen_helper.prepare_stream_from_tginput(tgenInput, 100, vm_config,
> self.tester.pktgen)
> - # set traffic option
> - traffic_opt = {'delay': 5, 'duration': 20}
> - _, pps = self.tester.pktgen.measure_throughput(stream_ids=streams,
> options=traffic_opt)
> - Mpps = pps / 1000000.0
> + stream_id = self.tester.pktgen.add_stream(self.tx_port, self.tx_port,
> + "%s/zero_copy.pcap" %
> self.tester.tmp_file)
> + self.tester.pktgen.config_stream(stream_id, stream_option)
> + traffic_opt = {
> + 'method': 'throughput',
> + 'rate': 100,
> + 'interval': 6,
> + 'duration': 30}
> + stats = self.tester.pktgen.measure([stream_id], traffic_opt)
> +
> + if isinstance(stats, list):
> + # if get multi result, ignore the first one, because it may not
> stable
> + num = len(stats)
> + Mpps = 0
> + for index in range(1, num):
> + Mpps += stats[index][1]
> + Mpps = Mpps / 1000000.0 / (num-1)
> + else:
> + Mpps = stats[1] / 1000000.0
> # when the fwd mode is rxonly, we can not receive data, so should
> not verify it
> - if loopback != "rxonly":
> + if fwd_mode != "rxonly":
> self.verify(Mpps > 0, "can not receive packets of frame size %d" %
> (frame_size))
> throughput = Mpps * 100 / \
> float(self.wirespeed(self.nic, frame_size, 1))
> return Mpps, throughput
>
> - def check_packets_of_each_queue(self, frame_size, loopback):
> + def check_packets_of_each_queue(self, frame_size, fwd_mode):
> """
> check each queue has receive packets
> """
> - if loopback == "rxonly":
> + if fwd_mode == "rxonly":
> verify_port = 1
> else:
> verify_port = 2
> - out = self.vhost.send_expect("stop", "testpmd> ", 60)
> + out = self.vhost_user.send_expect("stop", "testpmd> ", 60)
> for port_index in range(0, verify_port):
> for queue_index in range(0, self.queue_number):
> queue_info = re.findall("RX\s*Port=\s*%d/Queue=\s*%d" %
> @@ -260,9 +328,9 @@ class TestVhostDequeueZeroCopy(TestCase):
> "frame_size:%d, rx-packets:%d, tx-packets:%d" %
> (frame_size, rx_packets, tx_packets))
>
> - self.vhost.send_expect("start", "testpmd> ", 60)
> + self.vhost_user.send_expect("start", "testpmd> ", 60)
>
> - def send_and_verify(self, cycle="", loopback=""):
> + def send_and_verify_throughput(self, cycle="", fwd_mode=""):
> """
> start to send packets and verify it
> """
> @@ -270,110 +338,171 @@ class TestVhostDequeueZeroCopy(TestCase):
> info = "Running test %s, and %d frame size." %
> (self.running_case, frame_size)
> self.logger.info(info)
>
> - Mpps, throughput = self.calculate_avg_throughput(frame_size,
> loopback)
> - if loopback != "rxonly":
> + Mpps, throughput = self.calculate_avg_throughput(frame_size,
> fwd_mode)
> + if fwd_mode != "rxonly":
> self.update_table_info(frame_size, Mpps, throughput,
> cycle)
> # when multi queues, check each queue can receive packets
> if self.queue_number > 1:
> - self.check_packets_of_each_queue(frame_size, loopback)
> + self.check_packets_of_each_queue(frame_size,
> fwd_mode)
> +
> + def check_perf_drop_between_with_and_without_zero_copy(self):
> + """
> + for dequeue-zero-copy=0, about the small pkts we expect ~10% gain
> + compare to dequeue-zero-copy=1
> + """
> + value_with_zero_copy = 0
> + value_without_zero_copy = 0
> + if 'dequeue-zero-copy=1' in self.big_pkt_record.keys():
> + value_with_zero_copy =
> self.big_pkt_record['dequeue-zero-copy=1']
> + if 'dequeue-zero-copy=0' in self.big_pkt_record.keys():
> + value_without_zero_copy =
> self.big_pkt_record['dequeue-zero-copy=0']
> + self.verify(value_with_zero_copy != 0 and
> value_without_zero_copy != 0,
> + 'can not get the value of big pkts, please check
> self.frame_sizes')
> + self.verify(value_with_zero_copy - value_without_zero_copy >=
> value_with_zero_copy*0.1,
> + 'the drop with dequeue-zero-copy=0 is not as expected')
>
> def close_all_testpmd_and_vm(self):
> """
> close testpmd about vhost-user and vm_testpmd
> """
> - self.vhost.send_expect("quit", "#", 60)
> - self.vm_dut.send_expect("quit", "#", 60)
> - self.dut.close_session(self.vhost)
> - self.vm.stop()
> + if getattr(self, 'vhost_user', None):
> + self.vhost_user.send_expect("quit", "#", 60)
> + if getattr(self, 'virtio_user', None):
> + self.virtio_user.send_expect("quit", "#", 60)
> + self.dut.close_session(self.virtio_user)
> + self.virtio_user = None
> + if getattr(self, 'vm_dut', None):
> + self.vm_dut.send_expect("quit", "#", 60)
> + self.vm.stop()
>
> - def test_perf_pvp_dequeue_zero_copy(self):
> + def test_perf_pvp_split_ring_dequeue_zero_copy(self):
> """
> - performance of pvp zero-copy with [frame_sizes]
> + pvp split ring dequeue zero-copy test
> """
> self.nb_cores = 1
> self.queue_number = 1
> - self.launch_testpmd_on_vhost(txfreet="normal")
> - self.start_one_vm()
> - self.launch_testpmd_on_vm()
> - # set fwd mode on vm
> - self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
> - self.vm_dut.send_expect("start", "testpmd> ", 30)
> - # start testpmd at host side after VM and virtio-pmd launched
> - self.vhost.send_expect("start", "testpmd> ", 120)
> - self.send_and_verify()
> + self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to
> test')
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=False,
> vm_testpmd_fwd_mode='mac', packed_mode=False)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> +
> self.close_all_testpmd_and_vm()
> + self.logger.info('start vhost testpmd with dequeue-zero-copy=0 to
> test')
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=False,
> + vhost_client_mode=False,
> vm_testpmd_fwd_mode='mac', packed_mode=False)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=0')
> self.result_table_print()
> + self.check_perf_drop_between_with_and_without_zero_copy()
>
> - def test_perf_pvp_dequeue_zero_copy_with_2_queue(self):
> + def test_perf_pvp_packed_ring_dequeue_zero_copy(self):
> """
> - pvp dequeue zero-copy test with 2 queues
> + pvp packed ring dequeue zero-copy test
> + """
> + self.nb_cores = 1
> + self.queue_number = 1
> + self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to
> test')
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=False,
> vm_testpmd_fwd_mode='mac', packed_mode=True)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> +
> + self.close_all_testpmd_and_vm()
> + self.logger.info('start vhost testpmd with dequeue-zero-copy=0 to
> test')
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=False,
> + vhost_client_mode=False,
> vm_testpmd_fwd_mode='mac', packed_mode=True)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=0')
> + self.result_table_print()
> + self.check_perf_drop_between_with_and_without_zero_copy()
> +
> + def test_perf_pvp_split_ring_dequeue_zero_copy_with_2_queue(self):
> + """
> + pvp split ring dequeue zero-copy test with 2 queues
> """
> self.nb_cores = 2
> self.queue_number = 2
> - self.launch_testpmd_on_vhost(txfreet="normal")
> - self.start_one_vm()
> - self.launch_testpmd_on_vm()
> - # set fwd mode on vm
> - self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
> - self.vm_dut.send_expect("start", "testpmd> ", 30)
> - # start testpmd at host side after VM and virtio-pmd launched
> - self.vhost.send_expect("start", "testpmd> ", 120)
> - # when multi queues, the function will check each queue can receive
> packets
> - self.send_and_verify()
> - self.close_all_testpmd_and_vm()
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=False,
> vm_testpmd_fwd_mode='mac', packed_mode=False)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> self.result_table_print()
>
> - def test_perf_pvp_dequeue_zero_copy_with_driver_unload(self):
> + def
> test_perf_pvp_packed_ring_dequeue_zero_copy_with_2_queue(self):
> """
> - pvp dequeue zero-copy test with driver unload test
> + pvp packed ring dequeue zero-copy test with 2 queues
> + """
> + self.nb_cores = 2
> + self.queue_number = 2
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=False,
> vm_testpmd_fwd_mode='mac', packed_mode=True)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> + self.result_table_print()
> +
> + def
> test_perf_pvp_split_ring_dequeue_zero_copy_with_driver_unload(self):
> + """
> + pvp split ring dequeue zero-copy test with driver reload test
> """
> self.nb_cores = 4
> self.queue_number = 16
> - self.launch_testpmd_on_vhost(txfreet="normal")
> - self.start_one_vm()
> - self.launch_testpmd_on_vm()
> - # set fwd mode on vm
> - self.vm_dut.send_expect("set fwd rxonly", "testpmd> ", 30)
> - self.vm_dut.send_expect("start", "testpmd> ", 30)
> - # start testpmd at host side after VM and virtio-pmd launched
> - self.vhost.send_expect("start", "testpmd> ", 120)
> - # when multi queues, the function will check each queue can receive
> packets
> - self.send_and_verify(cycle="befor relaunch", loopback="rxonly")
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=True,
> vm_testpmd_fwd_mode='rxonly', packed_mode=False)
> + self.send_and_verify_throughput(cycle="before relaunch",
> + fwd_mode="rxonly")
>
> # relaunch testpmd at virtio side in VM for driver reloading
> - self.relaunch_testpmd_on_vm()
> - self.send_and_verify(cycle="after relaunch")
> - self.close_all_testpmd_and_vm()
> + self.restart_testpmd_on_vm(fwd_mode='mac')
> + self.send_and_verify_throughput(cycle="after relaunch")
> self.result_table_print()
>
> - def test_perf_pvp_dequeue_zero_copy_with_maximum_txfreet(self):
> + def
> test_perf_pvp_packed_ring_dequeue_zero_copy_with_driver_unload(self):
> """
> - pvp dequeue zero-copy test with maximum txfreet
> + pvp packed ring dequeue zero-copy test with driver reload test
> """
> self.nb_cores = 4
> self.queue_number = 16
> - self.launch_testpmd_on_vhost(txfreet="maximum")
> - self.start_one_vm()
> - self.launch_testpmd_on_vm()
> - # set fwd mode on vm
> - self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
> - self.vm_dut.send_expect("start", "testpmd> ", 30)
> - # start testpmd at host side after VM and virtio-pmd launched
> - self.vhost.send_expect("start", "testpmd> ", 120)
> - # when multi queues, the function will check each queue can receive
> packets
> - self.send_and_verify()
> - self.close_all_testpmd_and_vm()
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=True,
> vm_testpmd_fwd_mode='rxonly', packed_mode=True)
> + self.send_and_verify_throughput(cycle="before relaunch",
> + fwd_mode="rxonly")
> +
> + # relaunch testpmd at virtio side in VM for driver reloading
> + self.restart_testpmd_on_vm(fwd_mode='mac')
> + self.send_and_verify_throughput(cycle="after relaunch")
> + self.result_table_print()
> +
> + def
> test_perf_pvp_split_ring_dequeue_zero_copy_with_maximum_txfreet(self):
> + """
> + pvp split ring dequeue zero-copy test with maximum txfreet
> + """
> + self.nb_cores = 4
> + self.queue_number = 16
> + self.prepare_test_evn(vhost_txfreet_mode='maximum',
> vhost_zero_copy=True,
> + vhost_client_mode=True,
> vm_testpmd_fwd_mode='mac', packed_mode=False)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> + self.result_table_print()
> +
> + def test_perf_pvp_split_ring_dequeue_zero_copy_with_vector_rx(self):
> + """
> + pvp split ring dequeue zero-copy test with vector_rx path
> + """
> + self.nb_cores = 1
> + self.queue_number = 1
> + path_mode = 'packed_vq=0,in_order=0,mrg_rxbuf=0'
> + self.virtio_user = self.dut.new_session(suite="virtio-user")
> +
> + self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to
> test')
> + self.launch_testpmd_as_vhost(txfreet="vector_rx", zero_copy=True,
> client_mode=False)
> + self.vhost_user.send_expect("start", "testpmd> ", 120)
> + self.launch_testpmd_as_virtio_user(path_mode)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> self.result_table_print()
>
> def tear_down(self):
> """
> Run after each test case.
> """
> - self.dut.send_expect("killall -s INT testpmd", "#")
> - self.dut.send_expect("killall -s INT qemu-system-x86_64", "#")
> + self.close_all_testpmd_and_vm()
> + self.dut.kill_all()
> time.sleep(2)
>
> def tear_down_all(self):
> """
> Run after each test suite.
> """
> + if getattr(self, 'vhost_user', None):
> + self.dut.close_session(self.vhost_user)
> --
> 2.7.4
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [dts] [PATCH V1] tests/vhost_dequeue_zero_copy: update suite according to plan
2019-12-24 21:53 [dts] [PATCH V1] tests/vhost_dequeue_zero_copy: update suite according to plan lihong
2019-12-25 9:02 ` Ma, LihongX
@ 2020-01-02 2:18 ` Tu, Lijuan
1 sibling, 0 replies; 4+ messages in thread
From: Tu, Lijuan @ 2020-01-02 2:18 UTC (permalink / raw)
To: Ma, LihongX, dts; +Cc: Wang, Yinan, Ma, LihongX
applied
> -----Original Message-----
> From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of lihong
> Sent: Wednesday, December 25, 2019 5:53 AM
> To: dts@dpdk.org
> Cc: Wang, Yinan <yinan.wang@intel.com>; Ma, LihongX
> <lihongx.ma@intel.com>
> Subject: [dts] [PATCH V1] tests/vhost_dequeue_zero_copy: update suite
> according to plan
>
> Signed-off-by: lihong <lihongx.ma@intel.com>
> ---
> tests/TestSuite_vhost_dequeue_zero_copy.py | 393 +++++++++++++++++++-
> ---------
> 1 file changed, 261 insertions(+), 132 deletions(-)
>
> diff --git a/tests/TestSuite_vhost_dequeue_zero_copy.py
> b/tests/TestSuite_vhost_dequeue_zero_copy.py
> index 5705064..4052437 100644
> --- a/tests/TestSuite_vhost_dequeue_zero_copy.py
> +++ b/tests/TestSuite_vhost_dequeue_zero_copy.py
> @@ -38,6 +38,7 @@ There are three topology test (PVP/VM2VM/VM2NIC)
> for this feature.
> And this testsuite implement the topology of PVP.
> Testsuite vm2vm_net_perf implement the topology VM2VM Testsuite gso
> implement the topology VM2NIC
> +To run this suite, the qemu version should support packed ring.
> """
> import utils
> import time
> @@ -45,7 +46,8 @@ import re
> from settings import HEADER_SIZE
> from virt_common import VM
> from test_case import TestCase
> -from pktgen import PacketGeneratorHelper
> +from packet import Packet
> +from pktgen import TRANSMIT_CONT
>
>
> class TestVhostDequeueZeroCopy(TestCase):
> @@ -54,19 +56,19 @@ class TestVhostDequeueZeroCopy(TestCase):
> """
> Run at the start of each test suite.
> """
> + self.verify(self.nic in ("fortville_spirit"), "the port can not
> + run this suite")
>
> self.tester.extend_external_packet_generator(TestVhostDequeueZeroCopy,
> self)
> self.frame_sizes = [64, 128, 256, 512, 1024, 1518]
> - self.queue_number = 1
> - self.nb_cores = 1
> self.dut_ports = self.dut.get_ports()
> self.verify(len(self.dut_ports) >= 1, "Insufficient ports for testing")
> self.ports_socket = self.dut.get_numa_id(self.dut_ports[0])
> - self.cores_num = len([n for n in self.dut.cores if int(n['socket'])
> - == self.ports_socket])
> - self.mem_channels = self.dut.get_memory_channels()
> self.dst_mac = self.dut.get_mac_address(self.dut_ports[0])
> + self.tx_port = self.tester.get_local_port(self.dut_ports[0])
> + self.port_pci = self.dut.ports_info[self.dut_ports[0]]['pci']
> self.vm_dut = None
> + self.virtio_user = None
> self.virtio1_mac = "52:54:00:00:00:01"
> + self.header_size = HEADER_SIZE['eth'] + HEADER_SIZE['ip'] +
> + HEADER_SIZE['udp']
>
> self.logger.info("you can config packet_size in file %s.cfg," %
> self.suite_name + \
> "in region 'suite' like packet_sizes=[64, 128, 256]") @@ -74,64
> +76,93 @@ class TestVhostDequeueZeroCopy(TestCase):
> if 'packet_sizes' in self.get_suite_cfg():
> self.frame_sizes = self.get_suite_cfg()['packet_sizes']
>
> - self.out_path = '/tmp'
> - out = self.tester.send_expect('ls -d %s' % self.out_path, '# ')
> - if 'No such file or directory' in out:
> - self.tester.send_expect('mkdir -p %s' % self.out_path, '# ')
> - # create an instance to set stream field setting
> - self.pktgen_helper = PacketGeneratorHelper()
> self.base_dir = self.dut.base_dir.replace('~', '/root')
> + self.vhost_user = self.dut.new_session(suite="vhost-user")
>
> def set_up(self):
> """
> Run before each test case.
> """
> # Clean the execution ENV
> - self.dut.send_expect("killall -s INT testpmd", "#")
> - self.dut.send_expect("killall -s INT qemu-system-x86_64", "#")
> self.dut.send_expect("rm -rf %s/vhost-net*" % self.base_dir, "#")
> # Prepare the result table
> self.table_header = ["FrameSize(B)", "Throughput(Mpps)",
> "% linerate", "Queue number", "Cycle"]
> self.result_table_create(self.table_header)
> + self.vm_dut = None
> + self.big_pkt_record = {}
>
> - self.vhost = self.dut.new_session(suite="vhost-user")
> -
> - def get_core_mask(self):
> + def get_core_list(self):
> """
> check whether the server has enough cores to run case
> + if want to get the best perf of the vhost, the vhost tesptmd at least
> + should have 3 cores to start testpmd
> """
> - core_config = "1S/%dC/1T" % (self.nb_cores + 1)
> - self.verify(self.cores_num >= (self.nb_cores + 1),
> - "There has not enought cores to test this case %s" %
> self.running_case)
> - core_list = self.dut.get_core_list(
> + if self.nb_cores == 1:
> + cores_num = 2
> + else:
> + cores_num = 1
> + core_config = "1S/%dC/1T" % (self.nb_cores + cores_num)
> + self.core_list = self.dut.get_core_list(
> core_config, socket=self.ports_socket)
> - self.core_mask = utils.create_mask(core_list)
> + self.verify(len(self.core_list) >= (self.nb_cores + cores_num),
> + "There has not enought cores to test this case %s" %
> + self.running_case)
>
> - def launch_testpmd_on_vhost(self, txfreet):
> + def launch_testpmd_as_vhost(self, txfreet, zero_copy=True,
> client_mode=False):
> """
> launch testpmd on vhost
> """
> - self.get_core_mask()
> + self.get_core_list()
>
> + mode_info = ""
> + if client_mode is True:
> + mode_info = ',client=1'
> + zero_copy_info = 1
> + if zero_copy is False:
> + zero_copy_info = 0
> if txfreet == "normal":
> - txfreet_args = "--txfreet=992"
> + txfreet_args = "--txd=1024 --rxd=1024 --txfreet=992"
> elif txfreet == "maximum":
> - txfreet_args = "--txfreet=1020 --txrs=4"
> - command_client = self.dut.target + "/app/testpmd " + \
> - " -n %d -c %s --socket-mem 1024,1024 " + \
> - " --legacy-mem --file-prefix=vhost " + \
> - " --vdev 'eth_vhost0,iface=%s/vhost-net,queues=%d,dequeue-
> zero-copy=1' " + \
> - " -- -i --nb-cores=%d --rxq=%d --txq=%d " + \
> - "--txd=1024 --rxd=1024 %s"
> + txfreet_args = "--txrs=4 --txd=992 --rxd=992 --txfreet=988"
> + elif txfreet == "vector_rx":
> + txfreet_args = "--txd=1024 --rxd=1024 --txfreet=992 --txrs=32"
> +
> + eal_params = self.dut.create_eal_parameters(cores=self.core_list,
> + prefix='vhost', ports=[self.port_pci])
> + command_client = self.dut.target + "/app/testpmd %s " + \
> + " --socket-mem 1024,1024 --legacy-mem " + \
> + " --vdev 'eth_vhost0,iface=%s/vhost-
> net,queues=%d,dequeue-zero-copy=%d%s' " + \
> + " -- -i --nb-cores=%d --rxq=%d --txq=%d %s"
> command_line_client = command_client % (
> - self.mem_channels, self.core_mask, self.base_dir,
> - self.queue_number, self.nb_cores,
> + eal_params, self.base_dir, self.queue_number,
> + zero_copy_info, mode_info, self.nb_cores,
> self.queue_number, self.queue_number, txfreet_args)
> - self.vhost.send_expect(command_line_client, "testpmd> ", 120)
> - self.vhost.send_expect("set fwd mac", "testpmd> ", 120)
> + self.vhost_user.send_expect(command_line_client, "testpmd> ", 120)
> + self.vhost_user.send_expect("set fwd mac", "testpmd> ", 120)
>
> - def launch_testpmd_on_vm(self):
> + def launch_testpmd_as_virtio_user(self, path_mode):
> + """
> + launch testpmd use vhost-net with path mode
> + """
> + # To get the best perf, the vhost and virtio testpmd should not use
> same cores,
> + # so get the additional 3 cores to start virtio testpmd
> + core_config = "1S/%dC/1T" % (len(self.core_list) + 3)
> + core_list = self.dut.get_core_list(
> + core_config, socket=self.ports_socket)
> + self.verify(len(core_list) >= (len(self.core_list) + 3),
> + "There has not enought cores to test this case %s" %
> self.running_case)
> + eal_params =
> self.dut.create_eal_parameters(cores=core_list[len(self.core_list):],
> + prefix='virtio', no_pci=True)
> + command_line = self.dut.target + "/app/testpmd %s " + \
> + "--socket-mem 1024,1024 --legacy-mem " + \
> + "--
> vdev=net_virtio_user0,mac=00:01:02:03:04:05,path=./vhost-
> net,queue_size=1024,%s " + \
> + "-- -i --tx-offloads=0x0 --nb-cores=%d --txd=1024 --rxd=1024"
> + command_line = command_line % (eal_params, path_mode,
> self.nb_cores)
> + self.virtio_user.send_expect(command_line, 'testpmd> ', 120)
> + self.virtio_user.send_expect('set fwd mac', 'testpmd> ', 120)
> + self.virtio_user.send_expect('start', 'testpmd> ', 120)
> +
> + def start_testpmd_on_vm(self, fwd_mode='mac'):
> """
> start testpmd in vm depend on different path
> """
> @@ -142,12 +173,12 @@ class TestVhostDequeueZeroCopy(TestCase):
> command_line = command % (self.nb_cores,
> self.queue_number, self.queue_number)
> self.vm_dut.send_expect(command_line, "testpmd> ", 30)
> + self.vm_dut.send_expect('set fwd %s' % fwd_mode, "testpmd> ", 30)
> + self.vm_dut.send_expect('start', "testpmd> ", 30)
>
> - def relaunch_testpmd_on_vm(self):
> + def restart_testpmd_on_vm(self, fwd_mode):
> self.vm_dut.send_expect("quit", "# ", 30)
> - self.launch_testpmd_on_vm()
> - self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
> - self.vm_dut.send_expect("start", "testpmd> ", 30)
> + self.start_testpmd_on_vm(fwd_mode)
>
> def set_vm_vcpu(self):
> """
> @@ -166,7 +197,7 @@ class TestVhostDequeueZeroCopy(TestCase):
> if 'cpupin' in self.vm.params[i]['cpu'][0].keys():
> self.vm.params[i]['cpu'][0].pop('cpupin')
>
> - def start_one_vm(self):
> + def start_one_vm(self, mode='client', packed=False):
> """
> start qemu
> """
> @@ -176,10 +207,14 @@ class TestVhostDequeueZeroCopy(TestCase):
> vm_params['driver'] = 'vhost-user'
> vm_params['opt_path'] = '%s/vhost-net' % self.base_dir
> vm_params['opt_mac'] = self.virtio1_mac
> + if mode == 'server':
> + vm_params['opt_server'] = 'server'
> opt_args = "mrg_rxbuf=on,rx_queue_size=1024,tx_queue_size=1024"
> if self.queue_number > 1:
> vm_params['opt_queue'] = self.queue_number
> opt_args += ",mq=on,vectors=%d" % (2*self.queue_number + 2)
> + if packed is True:
> + opt_args += ',packed=on'
> vm_params['opt_settings'] = opt_args
> self.vm.set_vm_device(**vm_params)
> self.set_vm_vcpu()
> @@ -190,7 +225,26 @@ class TestVhostDequeueZeroCopy(TestCase):
> if self.vm_dut is None:
> raise Exception("Set up VM ENV failed")
> except Exception as e:
> - self.logger.error("ERROR: Failure for %s" % str(e))
> + self.logger.error("ERROR: Failure for %s, " % str(e) + \
> + "if 'packed not found' in output of start qemu log, " + \
> + "please use the qemu version which support packed ring")
> + raise e
> +
> + def prepare_test_evn(self, vhost_txfreet_mode, vhost_zero_copy,
> vhost_client_mode,
> + vm_testpmd_fwd_mode, packed_mode):
> + """
> + start vhost testpmd and launch qemu, start testpmd on vm
> + """
> + if vhost_client_mode is True:
> + vm_mode = 'server'
> + else:
> + vm_mode = 'client'
> + self.launch_testpmd_as_vhost(txfreet=vhost_txfreet_mode,
> zero_copy=vhost_zero_copy,
> + client_mode=vhost_client_mode)
> + self.start_one_vm(mode=vm_mode, packed=packed_mode)
> + self.start_testpmd_on_vm(fwd_mode=vm_testpmd_fwd_mode)
> + # start testpmd at host side after VM and virtio-pmd launched
> + self.vhost_user.send_expect("start", "testpmd> ", 120)
>
> def update_table_info(self, frame_size, Mpps, throughtput, cycle):
> results_row = [frame_size]
> @@ -200,50 +254,64 @@ class TestVhostDequeueZeroCopy(TestCase):
> results_row.append(cycle)
> self.result_table_add(results_row)
>
> - def set_fields(self):
> - """
> - set ip protocol field behavior
> - """
> - fields_config = {'ip': {'dst': {'action': 'random'}, }, }
> - return fields_config
> + # record the big pkt Mpps
> + if frame_size == 1518:
> + self.big_pkt_record[cycle] = Mpps
>
> - def calculate_avg_throughput(self, frame_size, loopback):
> + def calculate_avg_throughput(self, frame_size, fwd_mode):
> """
> start to send packet and get the throughput
> """
> - payload = frame_size - HEADER_SIZE['eth'] - HEADER_SIZE['ip'] -
> HEADER_SIZE['udp']
> - flow =
> '[Ether(dst="%s")/IP(src="192.168.4.1",proto=255)/UDP()/("X"*%d)]' % (
> + payload = frame_size - self.header_size
> + flow =
> + 'Ether(dst="%s")/IP(src="192.168.4.1",proto=255)/UDP(sport=33,dport=34
> + )/("X"*%d)' % (
> self.dst_mac, payload)
> - self.tester.scapy_append('wrpcap("%s/zero_copy.pcap", %s)' % (
> - self.out_path, flow))
> - self.tester.scapy_execute()
> -
> - tgenInput = []
> - port = self.tester.get_local_port(self.dut_ports[0])
> - tgenInput.append((port, port, "%s/zero_copy.pcap" % self.out_path))
> - vm_config = self.set_fields()
> + pkt = Packet(pkt_str=flow)
> + pkt.save_pcapfile(self.tester, "%s/zero_copy.pcap" %
> self.tester.tmp_file)
> + stream_option = {
> + 'pcap': "%s/zero_copy.pcap" % self.tester.tmp_file,
> + 'fields_config': {
> + 'ip': {'src': {'action': 'random', 'start': '16.0.0.1', 'step': 1, 'end':
> '16.0.0.64'}}},
> + 'stream_config': {
> + 'rate': 100,
> + 'transmit_mode': TRANSMIT_CONT,
> + }
> + }
> self.tester.pktgen.clear_streams()
> - streams = self.pktgen_helper.prepare_stream_from_tginput(tgenInput,
> 100, vm_config, self.tester.pktgen)
> - # set traffic option
> - traffic_opt = {'delay': 5, 'duration': 20}
> - _, pps = self.tester.pktgen.measure_throughput(stream_ids=streams,
> options=traffic_opt)
> - Mpps = pps / 1000000.0
> + stream_id = self.tester.pktgen.add_stream(self.tx_port, self.tx_port,
> + "%s/zero_copy.pcap" % self.tester.tmp_file)
> + self.tester.pktgen.config_stream(stream_id, stream_option)
> + traffic_opt = {
> + 'method': 'throughput',
> + 'rate': 100,
> + 'interval': 6,
> + 'duration': 30}
> + stats = self.tester.pktgen.measure([stream_id], traffic_opt)
> +
> + if isinstance(stats, list):
> + # if get multi result, ignore the first one, because it may not stable
> + num = len(stats)
> + Mpps = 0
> + for index in range(1, num):
> + Mpps += stats[index][1]
> + Mpps = Mpps / 1000000.0 / (num-1)
> + else:
> + Mpps = stats[1] / 1000000.0
> # when the fwd mode is rxonly, we can not receive data, so should not
> verify it
> - if loopback != "rxonly":
> + if fwd_mode != "rxonly":
> self.verify(Mpps > 0, "can not receive packets of frame size %d" %
> (frame_size))
> throughput = Mpps * 100 / \
> float(self.wirespeed(self.nic, frame_size, 1))
> return Mpps, throughput
>
> - def check_packets_of_each_queue(self, frame_size, loopback):
> + def check_packets_of_each_queue(self, frame_size, fwd_mode):
> """
> check each queue has receive packets
> """
> - if loopback == "rxonly":
> + if fwd_mode == "rxonly":
> verify_port = 1
> else:
> verify_port = 2
> - out = self.vhost.send_expect("stop", "testpmd> ", 60)
> + out = self.vhost_user.send_expect("stop", "testpmd> ", 60)
> for port_index in range(0, verify_port):
> for queue_index in range(0, self.queue_number):
> queue_info = re.findall("RX\s*Port=\s*%d/Queue=\s*%d" % @@ -
> 260,9 +328,9 @@ class TestVhostDequeueZeroCopy(TestCase):
> "frame_size:%d, rx-packets:%d, tx-packets:%d" %
> (frame_size, rx_packets, tx_packets))
>
> - self.vhost.send_expect("start", "testpmd> ", 60)
> + self.vhost_user.send_expect("start", "testpmd> ", 60)
>
> - def send_and_verify(self, cycle="", loopback=""):
> + def send_and_verify_throughput(self, cycle="", fwd_mode=""):
> """
> start to send packets and verify it
> """
> @@ -270,110 +338,171 @@ class TestVhostDequeueZeroCopy(TestCase):
> info = "Running test %s, and %d frame size." % (self.running_case,
> frame_size)
> self.logger.info(info)
>
> - Mpps, throughput = self.calculate_avg_throughput(frame_size,
> loopback)
> - if loopback != "rxonly":
> + Mpps, throughput = self.calculate_avg_throughput(frame_size,
> fwd_mode)
> + if fwd_mode != "rxonly":
> self.update_table_info(frame_size, Mpps, throughput, cycle)
> # when multi queues, check each queue can receive packets
> if self.queue_number > 1:
> - self.check_packets_of_each_queue(frame_size, loopback)
> + self.check_packets_of_each_queue(frame_size, fwd_mode)
> +
> + def check_perf_drop_between_with_and_without_zero_copy(self):
> + """
> + for dequeue-zero-copy=0, about the small pkts we expect ~10% gain
> + compare to dequeue-zero-copy=1
> + """
> + value_with_zero_copy = 0
> + value_without_zero_copy = 0
> + if 'dequeue-zero-copy=1' in self.big_pkt_record.keys():
> + value_with_zero_copy = self.big_pkt_record['dequeue-zero-copy=1']
> + if 'dequeue-zero-copy=0' in self.big_pkt_record.keys():
> + value_without_zero_copy = self.big_pkt_record['dequeue-zero-
> copy=0']
> + self.verify(value_with_zero_copy != 0 and value_without_zero_copy !=
> 0,
> + 'can not get the value of big pkts, please check self.frame_sizes')
> + self.verify(value_with_zero_copy - value_without_zero_copy >=
> value_with_zero_copy*0.1,
> + 'the drop with dequeue-zero-copy=0 is not as expected')
>
> def close_all_testpmd_and_vm(self):
> """
> close testpmd about vhost-user and vm_testpmd
> """
> - self.vhost.send_expect("quit", "#", 60)
> - self.vm_dut.send_expect("quit", "#", 60)
> - self.dut.close_session(self.vhost)
> - self.vm.stop()
> + if getattr(self, 'vhost_user', None):
> + self.vhost_user.send_expect("quit", "#", 60)
> + if getattr(self, 'virtio_user', None):
> + self.virtio_user.send_expect("quit", "#", 60)
> + self.dut.close_session(self.virtio_user)
> + self.virtio_user = None
> + if getattr(self, 'vm_dut', None):
> + self.vm_dut.send_expect("quit", "#", 60)
> + self.vm.stop()
>
> - def test_perf_pvp_dequeue_zero_copy(self):
> + def test_perf_pvp_split_ring_dequeue_zero_copy(self):
> """
> - performance of pvp zero-copy with [frame_sizes]
> + pvp split ring dequeue zero-copy test
> """
> self.nb_cores = 1
> self.queue_number = 1
> - self.launch_testpmd_on_vhost(txfreet="normal")
> - self.start_one_vm()
> - self.launch_testpmd_on_vm()
> - # set fwd mode on vm
> - self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
> - self.vm_dut.send_expect("start", "testpmd> ", 30)
> - # start testpmd at host side after VM and virtio-pmd launched
> - self.vhost.send_expect("start", "testpmd> ", 120)
> - self.send_and_verify()
> + self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to test')
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=False, vm_testpmd_fwd_mode='mac',
> packed_mode=False)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> +
> self.close_all_testpmd_and_vm()
> + self.logger.info('start vhost testpmd with dequeue-zero-copy=0 to test')
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=False,
> + vhost_client_mode=False, vm_testpmd_fwd_mode='mac',
> packed_mode=False)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=0')
> self.result_table_print()
> + self.check_perf_drop_between_with_and_without_zero_copy()
>
> - def test_perf_pvp_dequeue_zero_copy_with_2_queue(self):
> + def test_perf_pvp_packed_ring_dequeue_zero_copy(self):
> """
> - pvp dequeue zero-copy test with 2 queues
> + pvp packed ring dequeue zero-copy test
> + """
> + self.nb_cores = 1
> + self.queue_number = 1
> + self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to test')
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=False, vm_testpmd_fwd_mode='mac',
> packed_mode=True)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> +
> + self.close_all_testpmd_and_vm()
> + self.logger.info('start vhost testpmd with dequeue-zero-copy=0 to test')
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=False,
> + vhost_client_mode=False, vm_testpmd_fwd_mode='mac',
> packed_mode=True)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=0')
> + self.result_table_print()
> + self.check_perf_drop_between_with_and_without_zero_copy()
> +
> + def test_perf_pvp_split_ring_dequeue_zero_copy_with_2_queue(self):
> + """
> + pvp split ring dequeue zero-copy test with 2 queues
> """
> self.nb_cores = 2
> self.queue_number = 2
> - self.launch_testpmd_on_vhost(txfreet="normal")
> - self.start_one_vm()
> - self.launch_testpmd_on_vm()
> - # set fwd mode on vm
> - self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
> - self.vm_dut.send_expect("start", "testpmd> ", 30)
> - # start testpmd at host side after VM and virtio-pmd launched
> - self.vhost.send_expect("start", "testpmd> ", 120)
> - # when multi queues, the function will check each queue can receive
> packets
> - self.send_and_verify()
> - self.close_all_testpmd_and_vm()
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=False, vm_testpmd_fwd_mode='mac',
> packed_mode=False)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> self.result_table_print()
>
> - def test_perf_pvp_dequeue_zero_copy_with_driver_unload(self):
> + def test_perf_pvp_packed_ring_dequeue_zero_copy_with_2_queue(self):
> """
> - pvp dequeue zero-copy test with driver unload test
> + pvp packed ring dequeue zero-copy test with 2 queues
> + """
> + self.nb_cores = 2
> + self.queue_number = 2
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=False, vm_testpmd_fwd_mode='mac',
> packed_mode=True)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> + self.result_table_print()
> +
> + def
> test_perf_pvp_split_ring_dequeue_zero_copy_with_driver_unload(self):
> + """
> + pvp split ring dequeue zero-copy test with driver reload test
> """
> self.nb_cores = 4
> self.queue_number = 16
> - self.launch_testpmd_on_vhost(txfreet="normal")
> - self.start_one_vm()
> - self.launch_testpmd_on_vm()
> - # set fwd mode on vm
> - self.vm_dut.send_expect("set fwd rxonly", "testpmd> ", 30)
> - self.vm_dut.send_expect("start", "testpmd> ", 30)
> - # start testpmd at host side after VM and virtio-pmd launched
> - self.vhost.send_expect("start", "testpmd> ", 120)
> - # when multi queues, the function will check each queue can receive
> packets
> - self.send_and_verify(cycle="befor relaunch", loopback="rxonly")
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=True, vm_testpmd_fwd_mode='rxonly',
> packed_mode=False)
> + self.send_and_verify_throughput(cycle="before relaunch",
> + fwd_mode="rxonly")
>
> # relaunch testpmd at virtio side in VM for driver reloading
> - self.relaunch_testpmd_on_vm()
> - self.send_and_verify(cycle="after relaunch")
> - self.close_all_testpmd_and_vm()
> + self.restart_testpmd_on_vm(fwd_mode='mac')
> + self.send_and_verify_throughput(cycle="after relaunch")
> self.result_table_print()
>
> - def test_perf_pvp_dequeue_zero_copy_with_maximum_txfreet(self):
> + def
> test_perf_pvp_packed_ring_dequeue_zero_copy_with_driver_unload(self):
> """
> - pvp dequeue zero-copy test with maximum txfreet
> + pvp packed ring dequeue zero-copy test with driver reload test
> """
> self.nb_cores = 4
> self.queue_number = 16
> - self.launch_testpmd_on_vhost(txfreet="maximum")
> - self.start_one_vm()
> - self.launch_testpmd_on_vm()
> - # set fwd mode on vm
> - self.vm_dut.send_expect("set fwd mac", "testpmd> ", 30)
> - self.vm_dut.send_expect("start", "testpmd> ", 30)
> - # start testpmd at host side after VM and virtio-pmd launched
> - self.vhost.send_expect("start", "testpmd> ", 120)
> - # when multi queues, the function will check each queue can receive
> packets
> - self.send_and_verify()
> - self.close_all_testpmd_and_vm()
> + self.prepare_test_evn(vhost_txfreet_mode='normal',
> vhost_zero_copy=True,
> + vhost_client_mode=True, vm_testpmd_fwd_mode='rxonly',
> packed_mode=True)
> + self.send_and_verify_throughput(cycle="before relaunch",
> + fwd_mode="rxonly")
> +
> + # relaunch testpmd at virtio side in VM for driver reloading
> + self.restart_testpmd_on_vm(fwd_mode='mac')
> + self.send_and_verify_throughput(cycle="after relaunch")
> + self.result_table_print()
> +
> + def
> test_perf_pvp_split_ring_dequeue_zero_copy_with_maximum_txfreet(self):
> + """
> + pvp split ring dequeue zero-copy test with maximum txfreet
> + """
> + self.nb_cores = 4
> + self.queue_number = 16
> + self.prepare_test_evn(vhost_txfreet_mode='maximum',
> vhost_zero_copy=True,
> + vhost_client_mode=True, vm_testpmd_fwd_mode='mac',
> packed_mode=False)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> + self.result_table_print()
> +
> + def test_perf_pvp_split_ring_dequeue_zero_copy_with_vector_rx(self):
> + """
> + pvp split ring dequeue zero-copy test with vector_rx path
> + """
> + self.nb_cores = 1
> + self.queue_number = 1
> + path_mode = 'packed_vq=0,in_order=0,mrg_rxbuf=0'
> + self.virtio_user = self.dut.new_session(suite="virtio-user")
> +
> + self.logger.info('start vhost testpmd with dequeue-zero-copy=1 to test')
> + self.launch_testpmd_as_vhost(txfreet="vector_rx", zero_copy=True,
> client_mode=False)
> + self.vhost_user.send_expect("start", "testpmd> ", 120)
> + self.launch_testpmd_as_virtio_user(path_mode)
> + self.send_and_verify_throughput(cycle='dequeue-zero-copy=1')
> self.result_table_print()
>
> def tear_down(self):
> """
> Run after each test case.
> """
> - self.dut.send_expect("killall -s INT testpmd", "#")
> - self.dut.send_expect("killall -s INT qemu-system-x86_64", "#")
> + self.close_all_testpmd_and_vm()
> + self.dut.kill_all()
> time.sleep(2)
>
> def tear_down_all(self):
> """
> Run after each test suite.
> """
> + if getattr(self, 'vhost_user', None):
> + self.dut.close_session(self.vhost_user)
> --
> 2.7.4
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2020-01-02 2:18 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-12-24 21:53 [dts] [PATCH V1] tests/vhost_dequeue_zero_copy: update suite according to plan lihong
2019-12-25 9:02 ` Ma, LihongX
2019-12-27 8:30 ` Wang, Yinan
2020-01-02 2:18 ` Tu, Lijuan
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).