NFVBENCH-190: Add a 'i40e_mixed' option, trex accepts other i40e driven ports to run in kernel mode on the same board

Change-Id: I33b1d848f409e0431bd4e247456b1d697d8b9d04
Signed-off-by: Pierrick Louin <pierrick.louin@orange.com>

4 files changed, 53 insertions, 3 deletions

diff --git a/docker/Dockerfile b/docker/Dockerfile
index 8676bbd..10b7c41 100644
--- a/docker/Dockerfile
+++ b/docker/Dockerfile
@@ -1,7 +1,7 @@
 # docker file for creating a container that has nfvbench installed and ready to use
 FROM ubuntu:20.04
 
-ENV TREX_VER "v2.88"
+ENV TREX_VER "v2.89"
 ENV VM_IMAGE_VER "0.12"
 ENV PYTHONIOENCODING "utf8"
 
diff --git a/nfvbench/cfg.default.yaml b/nfvbench/cfg.default.yaml
index 28f84f4..8e822e5 100644
--- a/nfvbench/cfg.default.yaml
+++ b/nfvbench/cfg.default.yaml
@@ -367,6 +367,19 @@ cache_size: 0
 # The cache size is actually limited by the number of 64B mbufs configured in the trex platform configuration (see Trex manual 6.2.2. Memory section configuration)
 # Note that the resulting value is finally capped to 10000, whatever the requested size is (by design limitation).
 
+# Specification of the TRex behaviour dealing with the i40e network card driver issue: Trex-528
+#   see https://trex-tgn.cisco.com/youtrack/issue/trex-528
+#       This issue states that if other ports, in the same card,
+#       are in kernel mode, they could impair traffic counting.      
+# Can be overridden by --i40e-mixed
+# Values can be:
+#  ignore - don't consider the case (default)
+#  exit   - should the case arise, exit (TRex default behaviour)
+#  unbind - unbind kernel bound ports (the former NFVbench behaviour)
+# The 'ignore' option might be OK as soon as the issue has been fixed in the driver.
+# The 'unbind' option should not be used! who knows the current use of other ports?
+i40e_mixed:
+
 # Trex will use 1 x 64B mbuf per pre-built cached packet, assuming 1 pre-built cached packet per flow, it means for very large number of flows, the number of configured mbuf_64 will need to be set accordingly.
 mbuf_64:
 
diff --git a/nfvbench/nfvbench.py b/nfvbench/nfvbench.py
index bd86810..598247a 100644
--- a/nfvbench/nfvbench.py
+++ b/nfvbench/nfvbench.py
@@ -530,6 +530,13 @@ def _parse_opts_from_cli():
                              'tagged with given VLAN id(s) or not (given \'no-tag\') '
                              '\'true\': use current vlans; \'false\': disable this mode.')
 
+    parser.add_argument('--i40e-mixed', dest='i40e_mixed',
+                        action='store',
+                        default=None,
+                        metavar='<ignore,check,unbind>',
+                        help='TRex behavior when dealing with a i40e network card driver'
+                             ' [ https://trex-tgn.cisco.com/youtrack/issue/trex-528 ]')
+
     parser.add_argument('--user-info', dest='user_info',
                         action='append',
                         metavar='<data>',
@@ -875,6 +882,8 @@ def main():
                                 config.service_chain_count = len(vlans[0])
             opts.l2_loopback = None
 
+        if config.i40e_mixed is None:
+            config.i40e_mixed = 'ignore'
         if config.use_sriov_middle_net is None:
             config.use_sriov_middle_net = False
         if opts.use_sriov_middle_net is not None:
diff --git a/nfvbench/traffic_server.py b/nfvbench/traffic_server.py
index 52ec2f1..629cb3d 100644
--- a/nfvbench/traffic_server.py
+++ b/nfvbench/traffic_server.py
@@ -34,12 +34,27 @@ class TRexTrafficServer(TrafficServer):
         assert len(contents) == 1
         self.trex_dir = os.path.join(trex_base_dir, contents[0])
 
+    def __apply_trex_patches(self):
+        parent_dir = os.path.dirname(os.path.realpath(__file__))
+        patches_dir = os.path.join(parent_dir, "trex_patches")
+        patches = os.listdir(patches_dir)
+        for patch in patches:
+            patch = os.path.join(patches_dir, patch)
+            command = (
+                "patch --directory=" + self.trex_dir + " --strip=0"
+                " --forward --no-backup-if-mismatch --reject-file=-"
+                " --force --input=" + patch + " >&-")
+            os.system(command)
+
     def run_server(self, generator_config, filename='/etc/trex_cfg.yaml'):
         """Run TRex server for specified traffic profile.
 
         :param traffic_profile: traffic profile object based on config file
         :param filename: path where to save TRex config file
         """
+        # in order to allow for customized behaviors, let's apply some patches
+        # this scheme keeps acceptable since we have only simple modifications
+        self.__apply_trex_patches()
         cfg = self.__save_config(generator_config, filename)
         cores = generator_config.cores
         vtep_vlan = generator_config.gen_config.get('vtep_vlan')
@@ -53,10 +68,21 @@ class TRexTrafficServer(TrafficServer):
         # --unbind-unused-ports: for NIC that have more than 2 ports such as Intel X710
         # this will instruct trex to unbind all ports that are unused instead of
         # erroring out with an exception (i40e only)
+        # Try: --ignore-528-issue -> neither unbind nor exit with error,
+        #                            just proceed cause it might work!
+        # Note that force unbinding is probably a bad choice:
+        # we can't assume for sure that other ports are "unused".
+        # The default TRex behaviour - exit - is indeed a safer option;
+        # a message informs about the ports that should be unbound.
+        i40e_opt = ("--ignore-528-issue" if
+                    generator_config.config.i40e_mixed == 'ignore' else
+                    "--unbind-unused-ports" if
+                    generator_config.config.i40e_mixed == 'unbind' else "")
         cmd = ['nohup', '/bin/bash', '-c',
                './t-rex-64 -i -c {} --iom 0 --no-scapy-server '
-               '--unbind-unused-ports --close-at-end {} {} '
+               '--close-at-end {} {} {} '
                '{} {} --cfg {} &> /tmp/trex.log & disown'.format(cores, sw_mode,
+                                                                 i40e_opt,
                                                                  vlan_opt,
                                                                  hdrh_opt,
                                                                  mbuf_opt, cfg)]
@@ -91,7 +117,7 @@ class TRexTrafficServer(TrafficServer):
         # parameter, specified as one of the starting command line
         # arguments, has been modified since the last launch.
         # Hence we add some extra fields to the config file
-        # (nb_cores, use_vlan, mbuf_factor, hdrh)
+        # (nb_cores, use_vlan, mbuf_factor, i40e_mixed, hdrh)
         # which will serve as a memory between runs -
         # while being actually ignored by the T-Rex server.
 
@@ -108,6 +134,7 @@ class TRexTrafficServer(TrafficServer):
             hdrh       : {hdrh}
             nb_cores   : {nb_cores}
             use_vlan   : {use_vlan}
+            i40e_mixed : {i40e_mixed}
           interfaces   : [{ifs}]""".format(
             zmq_pub_port=generator_config.zmq_pub_port,
             zmq_rpc_port=generator_config.zmq_rpc_port,
@@ -119,6 +146,7 @@ class TRexTrafficServer(TrafficServer):
             nb_cores=generator_config.cores,
             use_vlan=generator_config.gen_config.get('vtep_vlan') or
             generator_config.vlan_tagging,
+            i40e_mixed=generator_config.config.i40e_mixed,
             ifs=ifs)
 
         if hasattr(generator_config, 'mbuf_64') and generator_config.mbuf_64: