From: Wei Ling <weix.ling@intel.com>
To: dts@dpdk.org
Cc: Wei Ling <weix.ling@intel.com>
Subject: [dts] [PATCH V1] tests/vswitch_sample_cbdma: fix code issue and sync with testplan
Date: Wed, 4 Aug 2021 11:06:13 +0800 [thread overview]
Message-ID: <20210804030613.3397183-1-weix.ling@intel.com> (raw)
1.Modify get_cbdma_ports_info_and_bind_to_dpdk method match regular
inorder to get the "0000:e1:00.01" device and retrun full pci name.
2.Fix the code not sync with testplan issue.
3.Add check the result steps in test case.
Signed-off-by: Wei Ling <weix.ling@intel.com>
---
tests/TestSuite_vswitch_sample_cbdma.py | 115 ++++++++++++------------
1 file changed, 60 insertions(+), 55 deletions(-)
diff --git a/tests/TestSuite_vswitch_sample_cbdma.py b/tests/TestSuite_vswitch_sample_cbdma.py
index 08422ea3..8a7f83db 100644
--- a/tests/TestSuite_vswitch_sample_cbdma.py
+++ b/tests/TestSuite_vswitch_sample_cbdma.py
@@ -81,8 +81,6 @@ class TestVswitchSampleCBDMA(TestCase):
self.vm_dst_mac0 = '52:54:00:00:00:01'
self.vm_dst_mac1 = '52:54:00:00:00:02'
self.vm_num = 2
- self.vm_dut = []
- self.vm = []
self.app_testpmd_path = self.dut.apps_name['test-pmd']
# create an instance to set stream field setting
self.pktgen_helper = PacketGeneratorHelper()
@@ -106,7 +104,11 @@ class TestVswitchSampleCBDMA(TestCase):
Run before each test case.
"""
self.dut.send_expect("rm -rf %s/vhost-net*" % self.base_dir, "#")
+ self.dut.send_expect("killall -I dpdk-vhost", '#', 20)
+ self.dut.send_expect("killall -I dpdk-testpmd", '#', 20)
self.dut.send_expect("killall -I qemu-system-x86_64", '#', 20)
+ self.vm_dut = []
+ self.vm = []
def set_max_queues(self, max_queues=512):
self.logger.info("Configure MAX_QUEUES to {}".format(max_queues))
@@ -172,22 +174,14 @@ class TestVswitchSampleCBDMA(TestCase):
pmd_session.start_testpmd(cores=cores, param=params, eal_param=eal_params, no_pci=True, ports=[], prefix=prefix,
fixed_prefix=True)
- def start_vms(self, mode=0, mergeable=True, server_mode=False, set_target=True):
+ def start_vms(self, mergeable=True, packed=False, server_mode=False, set_target=True, bind_dev=True, vm_diff_param=False):
"""
start two VM, each VM has one virtio device
"""
- if mode == 0:
- setting_args = "disable-modern=true"
- elif mode == 1:
- setting_args = "disable-modern=false"
- elif mode == 2:
- setting_args = "disable-modern=true,packed=on"
- if mergeable is True:
- setting_args += "," + "mrg_rxbuf=on"
- else:
- setting_args += "," + "mrg_rxbuf=off"
- setting_args += ",csum=on,gso=on,guest_csum=on,host_tso4=on,guest_tso4=on,guest_ecn=on"
-
+ mergeable = 'on' if mergeable else 'off'
+ setting_args = "disable-modern=true,mrg_rxbuf={0},csum=on,guest_csum=on,host_tso4=on,guest_tso4=on,guest_ecn=on".format(mergeable)
+ if packed:
+ setting_args = setting_args + ',packed=on'
for i in range(self.vm_num):
vm_dut = None
vm_info = VM(self.dut, 'vm%d' % i, 'vhost_sample')
@@ -198,11 +192,14 @@ class TestVswitchSampleCBDMA(TestCase):
else:
vm_params['opt_path'] = self.base_dir + '/vhost-net%d' % i
vm_params['opt_mac'] = "52:54:00:00:00:0%d" % (i + 1)
- vm_params['opt_settings'] = setting_args
+ if vm_diff_param and i > 0:
+ vm_params['opt_settings'] = setting_args + ',packed=on'
+ else:
+ vm_params['opt_settings'] = setting_args
vm_info.set_vm_device(**vm_params)
time.sleep(3)
try:
- vm_dut = vm_info.start(set_target=set_target)
+ vm_dut = vm_info.start(set_target=set_target, bind_dev=bind_dev)
if vm_dut is None:
raise Exception("Set up VM ENV failed")
except Exception as e:
@@ -238,7 +235,7 @@ class TestVswitchSampleCBDMA(TestCase):
out = self.dut.send_expect('./usertools/dpdk-devbind.py --status-dev misc', '# ', 30)
device_info = out.split('\n')
for device in device_info:
- pci_info = re.search('\s*(0000:\d*:\d*.\d*)', device)
+ pci_info = re.search('\s*(0000:\S*:\d*.\d*)', device)
if pci_info is not None:
dev_info = pci_info.group(1)
# the numa id of ioat dev, only add the device which on same socket with nic dev
@@ -254,7 +251,7 @@ class TestVswitchSampleCBDMA(TestCase):
dmas_info = ''
for dmas in used_cbdma:
number = used_cbdma.index(dmas)
- dmas = 'txd{}@{},'.format(number, dmas.replace('0000:', ''))
+ dmas = 'txd{}@{},'.format(number, dmas)
dmas_info += dmas
self.dmas_info = dmas_info[:-1]
self.device_str = ' '.join(used_cbdma)
@@ -281,7 +278,7 @@ class TestVswitchSampleCBDMA(TestCase):
self.dut.send_expect('./usertools/dpdk-devbind.py -u %s' % self.device_str, '# ', 30)
self.dut.send_expect('./usertools/dpdk-devbind.py --force --bind=ioatdma %s' % self.device_str, '# ', 60)
- def config_stream(self, frame_size, port_num, dst_mac_list):
+ def config_stream(self, frame_size, dst_mac_list):
tgen_input = []
rx_port = self.tester.get_local_port(self.dut_ports[0])
tx_port = self.tester.get_local_port(self.dut_ports[0])
@@ -303,7 +300,7 @@ class TestVswitchSampleCBDMA(TestCase):
test_result = {}
for frame_size in frame_sizes:
self.logger.info("Test running at parameters: " + "framesize: {}".format(frame_size))
- tgenInput = self.config_stream(frame_size, self.tester_tx_port_num, dst_mac_list)
+ tgenInput = self.config_stream(frame_size, dst_mac_list)
# clear streams before add new streams
self.tester.pktgen.clear_streams()
# run packet generator
@@ -311,9 +308,7 @@ class TestVswitchSampleCBDMA(TestCase):
# set traffic option
traffic_opt = {'duration': 5}
_, pps = self.tester.pktgen.measure_throughput(stream_ids=streams, options=traffic_opt)
- self.virtio_user0_pmd.execute_cmd('show port stats all')
throughput = pps / 1000000.0
- self.verify(throughput > 0, "No traffic detected")
test_result[frame_size] = throughput
self.result_table_add([frame_size, throughput])
self.result_table_print()
@@ -349,17 +344,16 @@ class TestVswitchSampleCBDMA(TestCase):
packed_ring_result = self.pvp_test_with_cbdma()
# relaunch split ring in_order vectorized with cbdma
+ self.virtio_user0_pmd.execute_cmd("quit", "# ")
self.mrg_rxbuf = 1
self.in_order = 1
self.vectorized = 1
self.packed_vq = 0
self.server = 1
- self.virtio_user0_pmd.execute_cmd("quit", "#")
+
self.start_virtio_testpmd(pmd_session=self.virtio_user0_pmd, dev_mac=self.virtio_dst_mac0, dev_id=0,
cores=self.vuser0_core_list, prefix='testpmd0', nb_cores=1, used_queues=1)
split_ring_reult = self.pvp_test_with_cbdma()
- self.virtio_user0_pmd.execute_cmd("quit", "#")
- self.vhost_user.send_expect("^C", "# ", 20)
self.table_header = ['Frame Size(Byte)', 'Mode', 'Throughput(Mpps)']
self.result_table_create(self.table_header)
@@ -370,10 +364,12 @@ class TestVswitchSampleCBDMA(TestCase):
for table_row in perf_result:
self.result_table_add(table_row)
self.result_table_print()
- self.virtio_user0_pmd.execute_cmd("quit", "#")
- self.vhost_user.send_expect("^C", "# ", 20)
+ for key in packed_ring_result.keys():
+ self.verify(packed_ring_result[key] > 1, "The perf test result is lower than 1 Mpps")
+ for key in split_ring_reult.keys():
+ self.verify(split_ring_reult[key] > 1, "The perf test result is lower than 1 Mpps")
- def config_stream_imix(self, frame_sizes, port_num, dst_mac_list):
+ def config_stream_imix(self, frame_sizes, dst_mac_list):
tgen_input = []
rx_port = self.tester.get_local_port(self.dut_ports[0])
tx_port = self.tester.get_local_port(self.dut_ports[0])
@@ -395,7 +391,7 @@ class TestVswitchSampleCBDMA(TestCase):
self.result_table_create(table_header)
# Begin test perf
test_result = {}
- tgenInput = self.config_stream_imix(frame_sizes, self.tester_tx_port_num, dst_mac_list)
+ tgenInput = self.config_stream_imix(frame_sizes, dst_mac_list)
fields_config = {'ip': {'src': {'action': 'random'}, }, }
# clear streams before add new streams
self.tester.pktgen.clear_streams()
@@ -404,10 +400,7 @@ class TestVswitchSampleCBDMA(TestCase):
# set traffic option
traffic_opt = {'delay': 5, 'duration': 5}
_, pps = self.tester.pktgen.measure_throughput(stream_ids=streams, options=traffic_opt)
- self.virtio_user0_pmd.execute_cmd("show port stats all")
- self.virtio_user1_pmd.execute_cmd("show port stats all")
throughput = pps / 1000000.0
- self.verify(throughput > 0, "traffic is too low: throughput=%s" % throughput)
test_result['imix'] = throughput
self.result_table_add(['imix', throughput])
self.result_table_print()
@@ -446,7 +439,7 @@ class TestVswitchSampleCBDMA(TestCase):
self.mrg_rxbuf = 0
self.in_order = 1
self.vectorized = 1
- self.packed_vq = 0
+ self.packed_vq = 1
self.server = 1
self.start_virtio_testpmd(pmd_session=self.virtio_user0_pmd, dev_mac=self.virtio_dst_mac0, dev_id=0,
cores=self.vuser0_core_list, prefix='testpmd0', nb_cores=1, used_queues=1)
@@ -473,9 +466,10 @@ class TestVswitchSampleCBDMA(TestCase):
for table_row in perf_result:
self.result_table_add(table_row)
self.result_table_print()
- self.virtio_user0_pmd.execute_cmd("quit", "#")
- self.virtio_user1_pmd.execute_cmd("quit", "#")
- self.vhost_user.send_expect("^C", "# ", 20)
+ for key in before_relunch.keys():
+ self.verify(before_relunch[key] > 1, "The perf test result is lower than 1 Mpps")
+ for key in after_relunch.keys():
+ self.verify(after_relunch[key] > 1, "The perf test result is lower than 1 Mpps")
def get_receive_throughput(self, pmd_session, count=10):
i = 0
@@ -513,9 +507,11 @@ class TestVswitchSampleCBDMA(TestCase):
frame_sizes = [64, 2000, 8000, 'imix']
if relaunch:
self.virtio_user0_pmd.execute_cmd('stop')
- self.virtio_user0_pmd.execute_cmd('clear port stats all')
self.virtio_user1_pmd.execute_cmd('stop')
+ self.virtio_user0_pmd.execute_cmd('clear port stats all')
self.virtio_user1_pmd.execute_cmd('clear port stats all')
+ self.virtio_user0_pmd.execute_cmd('show port stats all')
+ self.virtio_user1_pmd.execute_cmd('show port stats all')
self.set_testpmd0_param(self.virtio_user0_pmd, self.virtio_dst_mac1)
self.set_testpmd1_param(self.virtio_user1_pmd, self.virtio_dst_mac0)
@@ -541,20 +537,21 @@ class TestVswitchSampleCBDMA(TestCase):
self.logger.info("Launch vhost app perf test")
self.start_vhost_app(with_cbdma=True, cbdma_num=2, socket_num=2, client_mode=True)
self.mrg_rxbuf = 1
- self.in_order = 1
+ self.in_order = 0
self.vectorized = 1
- self.packed_vq = 0
+ self.packed_vq = 1
self.server = 1
self.start_virtio_testpmd(pmd_session=self.virtio_user0_pmd, dev_mac=self.virtio_dst_mac0, dev_id=0,
cores=self.vuser0_core_list, prefix='testpmd0', nb_cores=1, used_queues=1)
self.mrg_rxbuf = 1
- self.in_order = 1
+ self.in_order = 0
self.vectorized = 1
self.packed_vq = 0
self.server = 1
self.start_virtio_testpmd(pmd_session=self.virtio_user1_pmd, dev_mac=self.virtio_dst_mac1, dev_id=1,
cores=self.vuser1_core_list, prefix='testpmd1', nb_cores=1, used_queues=1)
before_relunch_result = self.vm2vm_check_with_two_cbdma()
+
self.logger.info("Relaunch vhost app perf test")
self.vhost_user.send_expect("^C", "# ", 20)
self.start_vhost_app(with_cbdma=True, cbdma_num=2, socket_num=2, client_mode=True)
@@ -569,9 +566,16 @@ class TestVswitchSampleCBDMA(TestCase):
for table_row in perf_result:
self.result_table_add(table_row)
self.result_table_print()
- self.virtio_user0_pmd.execute_cmd("quit", "# ")
- self.virtio_user1_pmd.execute_cmd("quit", "# ")
- self.vhost_user.send_expect("^C", "# ", 20)
+ for key in before_relunch_result.keys():
+ if key == 64:
+ self.verify(before_relunch_result[key] > 1, "The perf test result is lower than 1 Mpps")
+ else:
+ self.verify(before_relunch_result[key] > 0.1, "The perf test result is lower than 0.1 Mpps")
+ for key in after_relunch_result.keys():
+ if key == 64:
+ self.verify(after_relunch_result[key] > 1, "The perf test result is lower than 1 Mpps")
+ else:
+ self.verify(after_relunch_result[key] > 0.1, "The perf test result is lower than 0.1 Mpps")
def vm2vm_check_with_two_vhost_device(self):
rx_throughput = {}
@@ -589,7 +593,7 @@ class TestVswitchSampleCBDMA(TestCase):
def start_vms_testpmd_and_test(self, need_start_vm=True):
if need_start_vm:
- self.start_vms(mode=2, mergeable=True, server_mode=True)
+ self.start_vms(mergeable=True, packed=False, server_mode=True, set_target=True, bind_dev=True, vm_diff_param=True)
self.vm0_pmd = PmdOutput(self.vm_dut[0])
self.vm1_pmd = PmdOutput(self.vm_dut[1])
self.start_vm_testpmd(self.vm0_pmd)
@@ -608,11 +612,11 @@ class TestVswitchSampleCBDMA(TestCase):
perf_result = []
self.get_cbdma_ports_info_and_bind_to_dpdk(cbdma_num=2)
- self.logger.info("Before rebind perf VM Driver test")
+ self.logger.info("Before rebind VM Driver perf test")
self.start_vhost_app(with_cbdma=True, cbdma_num=2, socket_num=2, client_mode=True)
before_rebind = self.start_vms_testpmd_and_test(need_start_vm=True)
- self.logger.info("After rebind perf VM Driver test")
+ self.logger.info("After rebind VM Driver perf test")
# repeat bind 50 time from virtio-pci to vfio-pci
self.repeat_bind_driver(dut=self.vm_dut[0], repeat_times=50)
self.repeat_bind_driver(dut=self.vm_dut[1], repeat_times=50)
@@ -622,9 +626,6 @@ class TestVswitchSampleCBDMA(TestCase):
# repeat bind 50 time from virtio-pci to vfio-pci
self.repeat_bind_driver(dut=self.vm_dut[0], repeat_times=50)
self.repeat_bind_driver(dut=self.vm_dut[1], repeat_times=50)
- for i in range(len(self.vm)):
- self.vm[i].stop()
- self.vhost_user.send_expect("^C", "# ", 20)
self.table_header = ['Frame Size(Byte)', 'Before/After Bind VM Driver', 'Throughput(Mpps)']
self.result_table_create(self.table_header)
@@ -710,9 +711,9 @@ class TestVswitchSampleCBDMA(TestCase):
md5_revd = md5_revd[: md5_revd.find(' ')]
self.verify(md5_send == md5_revd, 'the received file is different with send file')
- def start_iperf_and_scp_test_in_vms(self, need_start_vm=True, mode=0, mergeable=False, server_mode=False):
+ def start_iperf_and_scp_test_in_vms(self, need_start_vm=True, mergeable=False, packed=False, server_mode=False):
if need_start_vm:
- self.start_vms(mode=mode, mergeable=mergeable, server_mode=server_mode, set_target=False)
+ self.start_vms(mergeable=mergeable, packed=packed, server_mode=server_mode, set_target=True, bind_dev=False)
self.vm0_pmd = PmdOutput(self.vm_dut[0])
self.vm1_pmd = PmdOutput(self.vm_dut[1])
self.config_vm_env()
@@ -730,7 +731,7 @@ class TestVswitchSampleCBDMA(TestCase):
self.logger.info("launch vhost")
self.start_vhost_app(with_cbdma=True, cbdma_num=2, socket_num=2, client_mode=True)
- before_rerun = self.start_iperf_and_scp_test_in_vms(need_start_vm=True, mode=0, mergeable=False, server_mode=True)
+ before_rerun = self.start_iperf_and_scp_test_in_vms(need_start_vm=True, mergeable=False, packed=False, server_mode=True)
self.logger.info("relaunch vhost")
self.vhost_user.send_expect("^C", "# ", 20)
@@ -762,7 +763,7 @@ class TestVswitchSampleCBDMA(TestCase):
self.get_cbdma_ports_info_and_bind_to_dpdk(cbdma_num=2)
self.start_vhost_app(with_cbdma=True, cbdma_num=2, socket_num=2, client_mode=False)
- before_rerun = self.start_iperf_and_scp_test_in_vms(need_start_vm=True, mode=2, mergeable=False, server_mode=False)
+ before_rerun = self.start_iperf_and_scp_test_in_vms(need_start_vm=True, mergeable=False, packed=True, server_mode=False)
self.logger.info("rerun scp and iperf test")
rerun_test_1 = self.start_iperf_and_scp_test_in_vms(need_start_vm=False)
@@ -770,6 +771,7 @@ class TestVswitchSampleCBDMA(TestCase):
rerun_test_3 = self.start_iperf_and_scp_test_in_vms(need_start_vm=False)
rerun_test_4 = self.start_iperf_and_scp_test_in_vms(need_start_vm=False)
rerun_test_5 = self.start_iperf_and_scp_test_in_vms(need_start_vm=False)
+
self.table_header = ['Path', 'Before/After rerun scp/iperf', 'Throughput(Mpps)']
self.result_table_create(self.table_header)
perf_result.append(['packed ring', 'Before rerun test', before_rerun])
@@ -794,6 +796,10 @@ class TestVswitchSampleCBDMA(TestCase):
"""
Run after each test case.
"""
+ self.dut.kill_all()
+ for i in range(len(self.vm)):
+ self.vm[i].stop()
+ self.vhost_user.send_expect("^C", "# ", 20)
self.bind_cbdma_device_to_kernel()
def tear_down_all(self):
@@ -801,5 +807,4 @@ class TestVswitchSampleCBDMA(TestCase):
Run after each test suite.
"""
self.set_max_queues(128)
- self.dut.build_install_dpdk(self.target)
self.close_all_session()
--
2.25.1
next reply other threads:[~2021-08-04 3:14 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-08-04 3:06 Wei Ling [this message]
2021-08-04 3:15 ` Ling, WeiX
2021-08-04 6:19 ` Wang, Yinan
2021-08-12 2:35 ` Wang, Yinan
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20210804030613.3397183-1-weix.ling@intel.com \
--to=weix.ling@intel.com \
--cc=dts@dpdk.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).