docs/tg_interface_driver


1
2
3
4
5
6
7
8
9
10
11
12
13
14

If using traffic scripts in test add "| Suite Setup | Setup all TGs before
traffic script" to test suite robot file, this bind TG interfaces to the kernel
driver specified in topology. Also add kernel driver name for TG interfaces to
topology YAML file. You can find driver name with following command where you
specify interface PCI address:
# lspci -vmmks 0000:00:07.0
Slot:   00:07.0
Class:  Ethernet controller
Vendor: Red Hat, Inc
Device: Virtio network device
SVendor:        Red Hat, Inc
SDevice:        Device 0001
PhySlot:        7
Driver: virtio-pci
#!/router/bin/python
from .stl_general_test import CStlGeneral_Test, CTRexScenario
from trex_stl_lib.api import *
import os, sys

ERROR_LATENCY_TOO_HIGH = 1

class STLRX_Test(CStlGeneral_Test):
    """Tests for RX feature"""

    def setUp(self):
        per_driver_params = {
                'net_vmxnet3': {
                        'rate_percent': 1,
                        'total_pkts': 50,
                        'rate_latency': 1,
                        'latency_9k_enable': False,
                        },
                'net_ixgbe': {
                        'rate_percent': 30,
                        'total_pkts': 1000,
                        'rate_latency': 1,
                        'latency_9k_enable': True,
                        'latency_9k_max_average': 300,
                        'latency_9k_max_latency': 400,
                        },
                'net_ixgbe_vf': {
                        'rate_percent': 30,
                        'total_pkts': 1000,
                        'rate_latency': 1,
                        'latency_9k_enable': False,
                        },

                'net_i40e': {
                        'rate_percent': 80,
                        'total_pkts': 1000,
                        'rate_latency': 1,
                        'latency_9k_enable': True,
                        'latency_9k_max_average': 100,
                        'latency_9k_max_latency': 250,
                        },
                'net_i40e_vf': {
                        'rate_percent': 80,
                        'total_pkts': 1000,
                        'rate_latency': 1,
                        'latency_9k_enable': False,
                        },
                'net_e1000_igb': {
                        'rate_percent': 80,
                        'total_pkts': 500,
                        'rate_latency': 1,
                        'latency_9k_enable': False,
                        },
                'net_e1000_em': {
                        'rate_percent': 1,
                        'total_pkts': 50,
                        'rate_latency': 1,
                        'latency_9k_enable': False,
                        },
                'net_virtio': {
                        'rate_percent': 1,
                        'total_pkts': 50,
                        'rate_latency': 1,
                        'latency_9k_enable': False,
                        'allow_packets_drop_num': 1, # allow 1 pkt drop
                        },

                 'net_mlx5': {
                        'rate_percent': 80,
                        'total_pkts': 1000,
                        'rate_latency': 1,
                        'latency_9k_enable': True,
                        'latency_9k_max_average': 100,
                        'latency_9k_max_latency': 250,
                        },

                 'net_enic': {
                        'rate_percent': 1,
                        'total_pkts': 50,
                        'rate_latency': 1,
                        'latency_9k_enable': False,
                        },

                  
                }

        CStlGeneral_Test.setUp(self)
        assert 'bi' in CTRexScenario.stl_ports_map

        self.c = CTRexScenario.stl_trex

        self.tx_port, self.rx_port = CTRexScenario.stl_ports_map['bi'][0]

        port_info = self.c.get_port_info(ports = self.rx_port)[0]
        self.speed = port_info['speed']

        cap = port_info['rx']['caps']
        if "flow_stats" not in cap or "latency" not in cap:
            self.skip('port {0} does not support RX'.format(self.rx_port))
        self.cap = cap

        drv_name = port_info['driver']
        if drv_name == 'net_ixgbe':
            self.ipv6_support = False
        else:
            self.ipv6_support = True
        self.rate_percent           = per_driver_params[drv_name]['rate_percent']
        self.total_pkts             = per_driver_params[drv_name]['total_pkts']
        self.rate_lat               = per_driver_params[drv_name].get('rate_latency', self.rate_percent)
        self.latency_9k_enable      = per_driver_params[drv_name]['latency_9k_enable']
        self.latency_9k_max_average = per_driver_params[drv_name].get('latency_9k_max_average')
        self.latency_9k_max_latency = per_driver_params[drv_name].get('latency_9k_max_latency')
        self.allow_drop             = per_driver_params[drv_name].get('allow_packets_drop_num', 0)

        self.lat_pps = 1000
        self.drops_expected = False
        self.c.reset(ports = [self.tx_port, self.rx_port])

        vm = STLScVmRaw( [ STLVmFlowVar ( "ip_src",  min_value="10.0.0.1",
                                          max_value="10.0.0.255", size=4, step=1,op="inc"),
                           STLVmWrFlowVar (fv_name="ip_src", pkt_offset= "IP.src" ), # write ip to packet IP.src
                           STLVmFixIpv4(offset = "IP")                                # fix checksum
                          ]
                         # Latency is bound to one core. We test that this option is not causing trouble
                         ,split_by_field = "ip_src"
                         ,cache_size =255 # Cache is ignored by latency flows. Need to test it is not crashing.
                         );

        self.pkt = STLPktBuilder(pkt = Ether()/IP(src="16.0.0.1",dst="48.0.0.1")/UDP(dport=12,sport=1025)/('Your_paylaod_comes_here'))
        self.ipv6pkt = STLPktBuilder(pkt = Ether()/IPv6(dst="2001:0:4137:9350:8000:f12a:b9c8:2815",src="2001:4860:0:2001::68")
                                     /UDP(dport=12,sport=1025)/('Your_paylaod_comes_here'))
        self.large_pkt = STLPktBuilder(pkt = Ether()/IP(src="16.0.0.1",dst="48.0.0.1")/UDP(dport=12,sport=1025)/('a'*1000))
        self.pkt_9k = STLPktBuilder(pkt = Ether()/IP(src="16.0.0.1",dst="48.0.0.1")/UDP(dport=12,sport=1025)/('a'*9000))
        self.vm_pkt = STLPktBuilder(pkt = Ether()/IP(src="16.0.0.1",dst="48.0.0.1")
                                    / UDP(dport=12,sport=1025)/('Your_paylaod_comes_here')
                                    , vm = vm)
        self.vm_large_pkt = STLPktBuilder(pkt = Ether()/IP(src="16.0.0.1",dst="48.0.0.1")/UDP(dport=12,sport=1025)/('a'*1000)
                                          , vm = vm)
        self.vm_9k_pkt = STLPktBuilder(pkt = Ether()/IP(src="16.0.0.1",dst="48.0.0.1")/UDP(dport=12,sport=1025)/('a'*9000)
                                       ,vm = vm)


    @classmethod
    def tearDownClass(cls):
        if CTRexScenario.stl_init_error:
            return
        # connect back at end of tests
        if not cls.is_connected():
            CTRexScenario.stl_trex.connect()


    def __verify_latency (self, latency_stats,max_latency,max_average):

        error=0;
        err_latency = latency_stats['err_cntrs']
        latency     = latency_stats['latency']

        for key in err_latency :
            error +=err_latency[key]
        if error !=0 :
            pprint.pprint(err_latency)
            tmp = 'RX pkts ERROR - one of the error is on'
            print(tmp)
            assert False, tmp

        if latency['average']> max_average:
            pprint.pprint(latency_stats)
            tmp = 'Average latency is too high {0} {1} '.format(latency['average'], max_average)
            print(tmp)
            return ERROR_LATENCY_TOO_HIGH

        if latency['total_max']> max_latency:
            pprint.pprint(latency_stats)
            tmp = 'Max latency is too high {0} {1} '.format(latency['total_max'], max_latency)
            print(tmp)
            return ERROR_LATENCY_TOO_HIGH

        return 0


    def __verify_flow (self, pg_id, total_pkts, pkt_len, stats):
        flow_stats = stats['flow_stats'].get(pg_id)
        latency_stats = stats['latency'].get(pg_id)

        if not flow_stats:
            assert False, "no flow stats available"

        tx_pkts  = flow_stats['tx_pkts'].get(self.tx_port, 0)
        tx_bytes = flow_stats['tx_bytes'].get(self.tx_port, 0)
        rx_pkts  = flow_stats['rx_pkts'].get(self.rx_port, 0)
        if latency_stats is not None:
            drops = latency_stats['err_cntrs']['dropped']
            ooo = latency_stats['err_cntrs']['out_of_order']
            dup = latency_stats['err_cntrs']['dup']
            sth = latency_stats['err_cntrs']['seq_too_high']
            stl = latency_stats['err_cntrs']['seq_too_low']
            lat = latency_stats['latency']
            if ooo != 0 or dup != 0 or stl != 0:
                pprint.pprint(latency_stats)
                tmp='Error packets - dropped:{0}, ooo:{1} dup:{2} seq too high:{3} seq too low:{4}'.format(drops, ooo, dup, sth, stl)
                assert False, tmp

            if (drops > self.allow_drop or sth != 0) and not self.drops_expected:
                pprint.pprint(latency_stats)
                tmp='Error packets - dropped:{0}, ooo:{1} dup:{2} seq too high:{3} seq too low:{4}'.format(drops, ooo, dup, sth, stl)
                assert False, tmp

        if tx_pkts != total_pkts:
            pprint.pprint(flow_stats)
            tmp = 'TX pkts mismatch - got: {0}, expected: {1}'.format(tx_pkts, total_pkts)
            assert False, tmp

        if tx_bytes != (total_pkts * pkt_len):
            pprint.pprint(flow_stats)
            tmp = 'TX bytes mismatch - got: {0}, expected: {1}'.format(tx_bytes, (total_pkts * pkt_len))
            assert False, tmp

        if abs(total_pkts - rx_pkts) > self.allow_drop and not self.drops_expected:
            pprint.pprint(flow_stats)
            tmp = 'RX pkts mismatch - got: {0}, expected: {1}'.format(rx_pkts, total_pkts)
            assert False, tmp

        if "rx_bytes" in self.cap:
            rx_bytes = flow_stats['rx_bytes'].get(self.rx_port, 0)
            if abs(rx_bytes / pkt_len  - total_pkts ) > self.allow_drop and not self.drops_expected:
                pprint.pprint(flow_stats)
                tmp = 'RX bytes mismatch - got: {0}, expected: {1}'.format(rx_bytes, (total_pkts * pkt_len))
                assert False, tmp


    # RX itreation
    def __rx_iteration (self, exp_list):

        self.c.clear_stats()

        self.c.start(ports = [self.tx_port])
        self.c.wait_on_traffic(ports = [self.tx_port])
        stats = self.c.get_stats()

        for exp in exp_list:
            self.__verify_flow(exp['pg_id'], exp['total_pkts'], exp['pkt_len'], stats)


    # one stream on TX --> RX
    def test_one_stream(self):
        total_pkts = self.total_pkts

        try:
            s1 = STLStream(name = 'rx',
                           packet = self.pkt,
                           flow_stats = STLFlowLatencyStats(pg_id = 5),
                           mode = STLTXSingleBurst(total_pkts = total_pkts,
                                                   percentage = self.rate_lat
                                                   ))

            # add both streams to ports
            self.c.add_streams([s1], ports = [self.tx_port])

            print("\ninjecting {0} packets on port {1}\n".format(total_pkts, self.tx_port))

            exp = {'pg_id': 5, 'total_pkts': total_pkts, 'pkt_len': s1.get_pkt_len()}

            self.__rx_iteration( [exp] )


        except STLError as e:
            assert False , '{0}'.format(e)


    def test_multiple_streams(self):
        if self.is_virt_nics:
            self.skip('Skip this for virtual NICs')

        num_latency_streams = 128
        num_flow_stat_streams = 127
        total_pkts = int(self.total_pkts / (num_latency_streams + num_flow_stat_streams))
        if total_pkts == 0:
            total_pkts = 1
        percent = float(self.rate_lat) / (num_latency_streams + num_flow_stat_streams)

        try:
            streams = []
            exp = []
            # 10 identical streams
            for pg_id in range(1, num_latency_streams):

                streams.append(STLStream(name = 'rx {0}'.format(pg_id),
                                         packet = self.pkt,
                                         flow_stats = STLFlowLatencyStats(pg_id = pg_id),
                                         mode = STLTXSingleBurst(total_pkts = total_pkts+pg_id, percentage = percent)))

                exp.append({'pg_id': pg_id, 'total_pkts': total_pkts+pg_id, 'pkt_len': streams[-1].get_pkt_len()})

            for pg_id in range(num_latency_streams + 1, num_latency_streams + num_flow_stat_streams):

                streams.append(STLStream(name = 'rx {0}'.format(pg_id),
                                         packet = self.pkt,
                                         flow_stats = STLFlowStats(pg_id = pg_id),
                                         mode = STLTXSingleBurst(total_pkts = total_pkts+pg_id, percentage = percent)))

                exp.append({'pg_id': pg_id, 'total_pkts': total_pkts+pg_id, 'pkt_len': streams[-1].get_pkt_len()})

            # add both streams to ports
            self.c.add_streams(streams, ports = [self.tx_port])

            self.__rx_iteration(exp)


        except STLError as e:
            assert False , '{0}'.format(e)

    def test_1_stream_many_iterations (self):
        total_pkts = self.total_pkts

        try:
            streams_data = [
                {'name': 'Flow stat. No latency', 'pkt': self.pkt, 'lat': False},
                {'name': 'Latency, no field engine', 'pkt': self.pkt, 'lat': True},
                {'name': 'Latency, short packet with field engine', 'pkt': self.vm_pkt, 'lat': True},
                {'name': 'Latency, large packet field engine', 'pkt': self.vm_large_pkt, 'lat': True}
            ]
            if self.latency_9k_enable:
                streams_data.append({'name': 'Latency, 9k packet with field engine', 'pkt': self.vm_9k_pkt, 'lat': True})

            if self.ipv6_support:
                streams_data.append({'name': 'IPv6 flow stat. No latency', 'pkt': self.ipv6pkt, 'lat': False})
                streams_data.append({'name': 'IPv6 latency, no field engine', 'pkt': self.ipv6pkt, 'lat': True})

            streams = []
            for data in streams_data:
                if data['lat']:
                    flow_stats = STLFlowLatencyStats(pg_id = 5)
                    mode = STLTXSingleBurst(total_pkts = total_pkts, percentage = self.rate_percent)
                else:
                    flow_stats = STLFlowStats(pg_id = 5)
                    mode = STLTXSingleBurst(total_pkts = total_pkts, pps = self.lat_pps)

                s = STLStream(name = data['name'],
                              packet = data['pkt'],
                              flow_stats = flow_stats,
                              mode = mode
                              )
                streams.append(s)

            print("\ninjecting {0} packets on port {1}".format(total_pkts, self.tx_port))
            exp = {'pg_id': 5, 'total_pkts': total_pkts}

            for stream in streams:
                self.c.add_streams([stream], ports = [self.tx_port])
                print("Stream: {0}".format(stream.name))
                exp['pkt_len'] = stream.get_pkt_len()
                for i in range(0, 10):
                    print("Iteration {0}".format(i))
                    self.__rx_iteration( [exp] )
                self.c.remove_all_streams(ports = [self.tx_port])


        except STLError as e:
            assert False , '{0}'.format(e)


    def __9k_stream(self,pgid,ports,precet,max_latency,avg_latency,duration,pkt_size):
        my_pg_id=pgid
        s_ports=ports;
        all_ports=list(CTRexScenario.stl_ports_map['map'].keys());
        if ports == None:
            s_ports=all_ports
        assert( type(s_ports)==list)

        stream_pkt = STLPktBuilder(pkt = Ether()/IP(src="16.0.0.1",dst="48.0.0.1")/UDP(dport=12,sport=1025)/('a'*pkt_size))

        try:
            # reset all ports
            self.c.reset(ports = all_ports)


            for pid in s_ports:
                s1  = STLStream(name = 'rx',
                               packet = self.pkt,
                               flow_stats = STLFlowLatencyStats(pg_id = my_pg_id+pid),
                               mode = STLTXCont(pps = 1000))

                s2 = STLStream(name = 'bulk',
                               packet = stream_pkt,
                               mode = STLTXCont(percentage =precet))


                # add both streams to ports
                self.c.add_streams([s1,s2], ports = [pid])

            self.c.clear_stats()

            self.c.start(ports = s_ports,duration = duration)
            self.c.wait_on_traffic(ports = s_ports,timeout = duration+10,rx_delay_ms = 100)
            stats = self.c.get_stats()

            for pid in s_ports:
                latency_stats = stats['latency'].get(my_pg_id+pid)
                #pprint.pprint(latency_stats)
                if self.__verify_latency (latency_stats,max_latency,avg_latency) !=0:
                    return (ERROR_LATENCY_TOO_HIGH);

            return 0

        except STLError as e:
            assert False , '{0}'.format(e)


    # check low latency when you have stream of 9K stream
    def test_9k_stream(self):
        if self.is_virt_nics:
            self.skip('Skip this for virtual NICs')

        if self.latency_9k_enable == False:
           print("SKIP")
           return

        for i in range(0,5):
            print("Iteration {0}".format(i));
            duration=random.randint(10, 70);
            pgid=random.randint(1, 65000);
            pkt_size=random.randint(1000, 9000);
            all_ports = list(CTRexScenario.stl_ports_map['map'].keys());


            s_port=random.sample(all_ports, random.randint(1, len(all_ports)) )
            s_port=sorted(s_port)

            if ((self.speed == 40) or (self.speed == 100)):
                # the NIC does not support all full rate in case both port works let's filter odd ports
                s_port=list(filter(lambda x: x % 2==0, s_port))
                if len(s_port)==0:
                    s_port=[0];


            error=1;
            for j in range(0,5):
               print(" {4} - duration {0} pgid {1} pkt_size {2} s_port {3} ".format(duration,pgid,pkt_size,s_port,j));
               if self.__9k_stream(pgid,
                                        s_port,90,
                                        self.latency_9k_max_latency,
                                        self.latency_9k_max_average,
                                        duration,
                                        pkt_size)==0:
                   error=0;
                   break;

            if error:
                assert False , "Latency too high"
            else:
                print("===>Iteration {0} PASS {1}".format(i,j));


    def  check_stats (self,stats,a,b,err):
        if a != b:
            tmp = 'ERROR field : {0}, read : {1} != expected : {2} '.format(err,a,b)
            pprint.pprint(stats)
            assert False,tmp


    def send_1_burst(self,from_port,is_latency,pkts):

        pid = from_port
        base_pkt =  Ether()/IP(src="16.0.0.1",dst="48.0.0.1")/UDP(dport=12,sport=1025)

        pad = (60 - len(base_pkt)) * 'x'

        stream_pkt = STLPktBuilder(pkt = base_pkt/pad)

        all_ports=list(CTRexScenario.stl_ports_map['map'].keys());

        dpid = CTRexScenario.stl_ports_map['map'][pid]

        s_ports =[pid]

        try:
            # reset all ports
            self.c.reset(ports = all_ports)


            for pid in s_ports:
                if is_latency:
                    s1  = STLStream(name = 'rx',
                               packet = stream_pkt,
                               flow_stats = STLFlowLatencyStats(pg_id = 5 + pid),
                               mode = STLTXSingleBurst(total_pkts = pkts,pps = 1000))
                else:
                    s1  = STLStream(name = 'rx',
                               packet = stream_pkt,
                               mode = STLTXSingleBurst(total_pkts = pkts,pps = 1000))


                # add both streams to ports
                self.c.add_streams(s1, ports = [pid])

            self.c.clear_stats()

            self.c.start(ports = s_ports)
            self.c.wait_on_traffic(ports = s_ports)

            stats = self.c.get_stats()

            ips = stats[dpid]
            ops = stats[pid]
            tps = stats['total']
            tbytes = pkts*64

            self.check_stats (stats,ops["obytes"], tbytes,"ops[obytes]")
            self.check_stats (stats,ops["opackets"], pkts,"ops[opackets]")

            self.check_stats (stats,ips["ibytes"], tbytes,"ips[ibytes]")
            self.check_stats (stats,ips["ipackets"], pkts,"ips[ipackets]")

            self.check_stats (stats,tps['ibytes'], tbytes,"tps[ibytes]")
            self.check_stats (stats,tps['obytes'], tbytes,"tps[obytes]")
            self.check_stats (stats,tps['ipackets'], pkts,"tps[ipackets]")
            self.check_stats (stats,tps['opackets'], pkts,"tps[opackets]")

            if is_latency:
                ls=stats['flow_stats'][5+ pid]
                self.check_stats (stats,ls['rx_pkts']['total'], pkts,"ls['rx_pkts']['total']")
                self.check_stats (stats,ls['rx_pkts'][dpid], pkts,"ls['rx_pkts'][dpid]")

                self.check_stats (stats,ls['tx_pkts']['total'], pkts,"ls['tx_pkts']['total']")
                self.check_stats (stats,ls['tx_pkts'][pid], pkts,"ls['tx_pkts'][pid]")

                self.check_stats (stats,ls['tx_bytes']['total'], tbytes,"ls['tx_bytes']['total']")
                self.check_stats (stats,ls['tx_bytes'][pid], tbytes,"ls['tx_bytes'][pid]")


            return 0

        except STLError as e:
            assert False , '{0}'.format(e)

    def _run_fcs_stream (self,is_vm):
        """ this test send 1 64 byte packet with latency and check that all counters are reported as 64 bytes"""
        res=True
        try:
            all_ports=list(CTRexScenario.stl_ports_map['map'].keys());
            for port in all_ports:
                for l in [True,False]:
                    print(" test port {0} latency : {1} ".format(port,l))
                    self.send_1_burst(port,l,100)
        except Exception as e:
            if is_vm :
                res=False
            else:
                raise e
        return(res);


    def test_fcs_stream(self):
        """ this test send 1 64 byte packet with latency and check that all counters are reported as 64 bytes"""

        is_vm=self.is_virt_nics # in case of VM and vSwitch there are drop of packets in some cases, let retry number of times 
                                # in this case we just want to check functionality that packet of 64 is reported as 64 in all levels 
        retry=1
        if is_vm:
            retry=4
        for i in range(0,retry):
            if self._run_fcs_stream (is_vm):
                break;
            print("==> retry  %d .." %(i));


    # this test adds more and more latency streams and re-test with incremental
    def test_incremental_latency_streams (self):
        if self.is_virt_nics:
            self.skip('Skip this for virtual NICs')

        total_pkts = self.total_pkts
        percent = 0.5

        try:
            # We run till maximum streams allowed. At some point, expecting drops, because rate is too high.
            # then run with less streams again, to see that system is still working.
            for num_iter in [128, 5]:
                exp = []
                for i in range(1, num_iter):
                    # mix small and large packets
                    if i % 2 != 0:
                        my_pkt = self.pkt
                    else:
                        my_pkt = self.large_pkt
                    s1 = STLStream(name = 'rx',
                                   packet = my_pkt,
                                   flow_stats = STLFlowLatencyStats(pg_id = i),
                                   mode = STLTXSingleBurst(total_pkts = total_pkts,
                                                           percentage = percent
                                                       ))

                    # add both streams to ports
                    self.c.add_streams([s1], ports = [self.tx_port])
                    total_percent = i * percent
                    if total_percent > self.rate_lat:
                        self.drops_expected = True
                    else:
                        self.drops_expected = False

                    print("port {0} : {1} streams at {2}% of line rate\n".format(self.tx_port, i, total_percent))

                    exp.append({'pg_id': i, 'total_pkts': total_pkts, 'pkt_len': s1.get_pkt_len()})

                    self.__rx_iteration( exp )

                self.c.remove_all_streams(ports = [self.tx_port])

        except STLError as e:
            assert False , '{0}'.format(e)