[dts] [PATCH V1 1/3] add vf rss test plan

[dts] [PATCH V1 1/3] add vf rss test plan

[xu,huilong]

Signed-off-by: xu,huilong <huilongx.xu@intel.com>
---
 test_plans/vf_rss_test_plan.rst | 182 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 182 insertions(+)
 create mode 100644 test_plans/vf_rss_test_plan.rst

diff --git a/test_plans/vf_rss_test_plan.rst b/test_plans/vf_rss_test_plan.rst
new file mode 100644
index 0000000..f832924
--- /dev/null
+++ b/test_plans/vf_rss_test_plan.rst
@@ -0,0 +1,182 @@
+.. Copyright (c) <2016>, Intel Corporation
+   All rights reserved.
+
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+
+   - Redistributions of source code must retain the above copyright
+     notice, this list of conditions and the following disclaimer.
+
+   - Redistributions in binary form must reproduce the above copyright
+     notice, this list of conditions and the following disclaimer in
+     the documentation and/or other materials provided with the
+     distribution.
+
+   - Neither the name of Intel Corporation nor the names of its
+     contributors may be used to endorse or promote products derived
+     from this software without specific prior written permission.
+
+   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+   FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+   COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
+   INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+   (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+   SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+   HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+   STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+   ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+   OF THE POSSIBILITY OF SUCH DAMAGE.
+
+==================================================================
+Fortville RSS full support - Support configuring hash functions
+==================================================================
+
+This document provides test plan for testing the function of Fortville:
+Support configuring hash functions.
+
+Prerequisites
+-------------
+
+2x Intel? 82599 (Niantic) NICs (2x 10GbE full duplex optical ports per NIC)
+1x Fortville_eagle NIC (4x 10G)
+1x Fortville_spirit NIC (2x 40G) 
+2x Fortville_spirit_single NIC (1x 40G)
+
+The one port of the 82599 connect to the Fortville_eagle;
+The one port of Fortville_spirit connect to Fortville_spirit_single.
+The three kinds of NICs are the target NICs. the connected NICs can send packets
+to these three NICs using scapy.
+
+Network Traffic
+---------------
+
+The RSS feature is designed to improve networking performance by load balancing
+the packets received from a NIC port to multiple NIC RX queues, with each queue
+handled by a different logical core.
+
+#1. The receive packet is parsed into the header fields used by the hash
+operation (such as IP addresses, TCP port, etc.)
+
+#2. A hash calculation is performed. The Fortville  supports four hash function:
+Toeplitz, simple XOR and their Symmetric RSS.
+
+#3. The seven LSBs of the hash result are used as an index into a 128/512 entry
+'redirection table'. 
+
+The RSS RETA update feature is designed to make RSS more flexible by allowing
+users to define the correspondence between the seven LSBs of hash result and
+the queue id(RSS output index) by themself.
+
+
+Test Case:  test_rss_hash
+========================================
+
+The following RX Ports/Queues configurations have to be benchmarked:
+
+- 1 RX port / 4 RX queues (1P/4Q)
+
+
+Testpmd configuration - 4 RX/TX queues per port
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+::
+  
+  testpmd -c 1f -n 3  -- -i --rxq=4 --txq=4
+
+Testpmd Configuration Options
+-----------------------------
+
+By default, a single logical core runs the test.
+The CPU IDs and the number of logical cores running the test in parallel can
+be manually set with the ``set corelist X,Y`` and the ``set nbcore N``
+interactive commands of the ``testpmd`` application.
+#1. got the pci device id of DUT, for example,
+
+./dpdk_nic_bind.py --st
+
+0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0 drv=i40e unused=
+0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1 drv=i40e unused=
+
+#2. create 2 VFs from 2 PFs,
+
+echo 1 > /sys/bus/pci/devices/0000\:81\:00.0/sriov_numvfs
+echo 1 > /sys/bus/pci/devices/0000\:81\:00.1/sriov_numvfs
+./dpdk_nic_bind.py --st
+
+0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0 drv=i40e unused=
+0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1 drv=i40e unused=
+0000:81:02.0 'XL710/X710 Virtual Function' unused=
+0000:81:0a.0 'XL710/X710 Virtual Function' unused=
+
+#3. detach VFs from the host, bind them to pci-stub driver,
+
+/sbin/modprobe pci-stub
+
+using `lspci -nn|grep -i ethernet` got VF device id, for example "8086 154c",
+
+echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
+echo 0000:81:02.0 > /sys/bus/pci/devices/0000:08:02.0/driver/unbind
+echo 0000:81:02.0 > /sys/bus/pci/drivers/pci-stub/bind
+
+echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
+echo 0000:81:0a.0 > /sys/bus/pci/devices/0000:08:0a.0/driver/unbind
+echo 0000:81:0a.0 > /sys/bus/pci/drivers/pci-stub/bind
+
+  or using the following more easy way,
+
+  virsh nodedev-detach pci_0000_81_02_0; 
+  virsh nodedev-detach pci_0000_81_0a_0;
+
+  ./dpdk_nic_bind.py --st
+
+  0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0 drv=i40e unused=
+  0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1 drv=i40e unused=
+  0000:81:02.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
+  0000:81:0a.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
+
+  it can be seen that VFs 81:02.0 & 81:0a.0 's drv is pci-stub.
+
+#4. passthrough VFs 81:02.0 & 81:0a.0 to vm0, and start vm0,
+
+  /usr/bin/qemu-system-x86_64  -name vm0 -enable-kvm \
+  -cpu host -smp 4 -m 2048 -drive file=/home/image/sriov-fc20-1.img -vnc :1 \
+  -device pci-assign,host=81:02.0,id=pt_0 \
+  -device pci-assign,host=81:0a.0,id=pt_1
+
+#5. login vm0, got VFs pci device id in vm0, assume they are 00:06.0 & 00:07.0, bind them to igb_uio driver,
+and then start testpmd, set it in mac forward mode,
+
+./tools/dpdk_nic_bind.py --bind=igb_uio 00:06.0 00:07.0
+
+#6. Reta Configuration.  128 reta entries configuration::
+
+  testpmd command: port config 0 rss reta (hash_index,queue_id)
+
+#7. PMD fwd only receive the packets::
+
+  testpmd command: set fwd rxonly
+
+#8. rss recived package type configuration two received packet types configuration::
+
+  testpmd command: port config 0 rss ip/udp/tcp
+
+#9. verbose configuration::
+
+  testpmd command: set verbose 8
+
+#10. start packet receive::
+
+  testpmd command: start
+#11. send packet and check rx port received packet by different queue.
+  different hash type send different packet, example hash type is ip, packet src and dts ip not different
+  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.4", dst="192.168.0.5")], iface="eth3")
+  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.5", dst="192.168.0.4")], iface="eth3")   
+Test Case:  test_reta
+========================================
+   this case test hash reta table, the test steps same with test_rss_hash except config hash reta table
+#1 before send packet, config hash reta,512(niantic nic have 128 reta) reta entries configuration::
+
+     testpmd command: port config 0 rss reta (hash_index,queue_id)
-- 
1.9.3

[dts] [PATCH V1 2/3] add vf rss vm config file

[xu,huilong]

Signed-off-by: xu,huilong <huilongx.xu@intel.com>
---
 conf/vf_rss.cfg | 105 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 105 insertions(+)
 create mode 100644 conf/vf_rss.cfg

diff --git a/conf/vf_rss.cfg b/conf/vf_rss.cfg
new file mode 100644
index 0000000..fb88ad9
--- /dev/null
+++ b/conf/vf_rss.cfg
@@ -0,0 +1,105 @@
+# QEMU options
+# name
+#       name: vm0
+#
+# enable_kvm
+#       enable: [yes | no]
+#
+# cpu
+#       model: [host | core2duo | ...]
+#           usage:
+#               choose model value from the command
+#                   qemu-system-x86_64 -cpu help
+#       number: '4' #number of vcpus
+#       cpupin: '3 4 5 6' # host cpu list
+#
+# mem
+#       size: 1024
+#
+# disk
+#       file: /path/to/image/test.img
+#
+# net
+#        type: [nic | user | tap | bridge | ...]
+#           nic
+#               opt_vlan: 0
+#                   note: Default is 0.
+#               opt_macaddr: 00:00:00:00:01:01
+#                   note: if creating a nic, it`s better to specify a MAC,
+#                         else it will get a random number.
+#               opt_model:["e1000" | "virtio" | "i82551" | ...]
+#                   note: Default is e1000.
+#               opt_name: 'nic1'
+#               opt_addr: ''
+#                   note: PCI cards only.
+#               opt_vectors:
+#                   note: This option currently only affects virtio cards.
+#           user
+#               opt_vlan: 0
+#                   note: default is 0.
+#               opt_hostfwd: [tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport
+#                   note: If not specified, it will be setted automatically.
+#           tap
+#               opt_vlan: 0
+#                   note: default is 0.
+#               opt_br: br0
+#                   note: if choosing tap, need to specify bridge name,
+#                         else it will be br0.
+#               opt_script: QEMU_IFUP_PATH
+#                   note: if not specified, default is self.QEMU_IFUP_PATH.
+#               opt_downscript: QEMU_IFDOWN_PATH
+#                   note: if not specified, default is self.QEMU_IFDOWN_PATH.
+#
+# device
+#       driver: [pci-assign | virtio-net-pci | ...]
+#           pci-assign
+#               prop_host: 08:00.0
+#               prop_addr: 00:00:00:00:01:02
+#           virtio-net-pci
+#               prop_netdev: mynet1
+#               prop_id: net1
+#               prop_mac: 00:00:00:00:01:03
+#               prop_bus: pci.0
+#               prop_addr: 0x3
+#
+# monitor
+#       port: 6061   
+#           note: if adding monitor to vm, need to specicy
+#                 this port, else it will get a free port
+#                 on the host machine.
+#
+# qga
+#       enable: [yes | no]
+#
+# serial_port
+#       enable: [yes | no]
+#
+# vnc
+#       displayNum: 1
+#           note: you can choose a number not used on the host.
+#
+# daemon
+#       enable: 'yes'
+#           note:
+#               By default VM will start with the daemonize status.
+#               Not support starting it on the stdin now.
+
+# vm configuration for pmd sriov case
+[vm0]
+cpu =
+    model=host,number=8,cpupin=5 6 7 8 9 10 11 12 ;
+disk =
+    file=/home/image/fedora23-ok.img;
+login =
+    user=root,password=tester;
+net = 
+   type=nic,opt_vlan=0;
+   type=user,opt_vlan=0; 
+monitor = 
+    port=;
+qga = 
+    enable=yes;
+vnc = 
+    displayNum=1;
+daemon =
+    enable=yes;
-- 
1.9.3

[dts] [PATCH V1 3/3] add vf rss test script

[xu,huilong]

Signed-off-by: xu,huilong <huilongx.xu@intel.com>
---
 tests/TestSuite_vf_rss.py | 382 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 382 insertions(+)
 create mode 100644 tests/TestSuite_vf_rss.py

diff --git a/tests/TestSuite_vf_rss.py b/tests/TestSuite_vf_rss.py
new file mode 100644
index 0000000..a59516d
--- /dev/null
+++ b/tests/TestSuite_vf_rss.py
@@ -0,0 +1,382 @@
+# BSD LICENSE
+#
+# Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+# All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions
+# are met:
+#
+#   * Redistributions of source code must retain the above copyright
+#     notice, this list of conditions and the following disclaimer.
+#   * Redistributions in binary form must reproduce the above copyright
+#     notice, this list of conditions and the following disclaimer in
+#     the documentation and/or other materials provided with the
+#     distribution.
+#   * Neither the name of Intel Corporation nor the names of its
+#     contributors may be used to endorse or promote products derived
+#     from this software without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+"""
+DPDK Test suite.
+Test RSS reta (redirection table) update function.
+"""
+import time
+import random
+import re
+import dts
+testQueues = [4]
+reta_entries = []
+reta_lines = []
+
+# Use scapy to send packets with different source and dest ip.
+# and collect the hash result of five tuple and the queue id.
+from test_case import TestCase
+from pmd_output import PmdOutput
+from qemu_kvm import QEMUKvm
+
+class TestVfRss(TestCase):
+    def send_packet(self, itf, tran_type):
+        """
+        Sends packets.
+        """
+        global reta_lines
+        reta_lines = []
+        self.tester.scapy_foreground()
+        self.tester.scapy_append('sys.path.append("./")')
+        self.tester.scapy_append('from sctp import *')
+        self.vm_dut_0.send_expect("start", "testpmd>")
+        mac = self.vm_dut_0.get_mac_address(0)
+        # send packet with different source and dest ip
+        if tran_type == "ipv4-other":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IP(src="192.168.0.%d", dst="192.168.0.%d")], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv4-tcp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IP(src="192.168.0.%d", dst="192.168.0.%d")/TCP(sport=1024,dport=1024)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv4-udp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IP(src="192.168.0.%d", dst="192.168.0.%d")/UDP(sport=1024,dport=1024)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv4-sctp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s")/IP(src="192.168.0.%d", dst="192.168.0.%d")/SCTP(sport=1024,dport=1025,tag=1)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+                packet = r'sendp([Ether(dst="%s")/IP(src="192.168.0.%d", dst="192.168.0.%d")/SCTP(sport=1025,dport=1024,tag=1)], iface="%s")' % (
+                    mac, i + 2, i + 1, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "l2_payload":
+            for i in range(16):
+                packet = r'sendp([Ether(src="00:00:00:00:00:%02d",dst="%s")], iface="%s")' % (
+                    i + 1, mac, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+
+        elif tran_type == "ipv6-other":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IPv6(src="3ffe:2501:200:1fff::%d", dst="3ffe:2501:200:3::%d")], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv6-tcp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IPv6(src="3ffe:2501:200:1fff::%d", dst="3ffe:2501:200:3::%d")/TCP(sport=1024,dport=1024)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv6-udp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IPv6(src="3ffe:2501:200:1fff::%d", dst="3ffe:2501:200:3::%d")/UDP(sport=1024,dport=1024)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv6-sctp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s")/IPv6(src="3ffe:2501:200:1fff::%d", dst="3ffe:2501:200:3::%d")/SCTP(sport=1024,dport=1025,tag=1)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+                packet = r'sendp([Ether(dst="%s")/IPv6(src="3ffe:2501:200:1fff::%d", dst="3ffe:2501:200:3::%d")/SCTP(sport=1025,dport=1024,tag=1)], iface="%s")' % (
+                    mac, i + 2, i + 1, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+
+        else:
+            print "\ntran_type error!\n"
+
+        #out = self.vm_dut_0.send_expect("stop", "testpmd>")
+        out = self.vm_dut_0.get_session_output()
+        print '*******************************************'
+        print out
+        if  not reta_entries:
+           self.verify('RSS hash=' in out, 'rss faied')
+           return 
+        lines = out.split("\r\n")
+        out = ''
+        reta_line = {}
+
+        # collect the hash result of five tuple and the queue id
+        for line in lines:
+            line = line.strip()
+            if len(line) != 0 and line.startswith(("src=",)):
+                for item in line.split("-"):
+                    item = item.strip()
+                    if(item.startswith("RSS hash")):
+                        name, value = item.split("=", 1)
+                        print name + "-" + value
+
+                reta_line[name.strip()] = value.strip()
+                reta_lines.append(reta_line)
+                reta_line = {}
+            elif len(line) != 0 and line.strip().startswith("port "):
+                rexp = r"port (\d)/queue (\d{1,2}): received (\d) packets"
+                m = re.match(rexp, line.strip())
+                if m:
+                    reta_line["port"] = m.group(1)
+                    reta_line["queue"] = m.group(2)
+            elif len(line) != 0 and line.startswith("stop"):
+                break
+            else:
+                pass
+        self.verifyResult()
+             
+
+    def verifyResult(self):
+        """
+        Verify whether or not the result passes.
+        """
+
+        global reta_lines
+        result = []
+        dts.results_table_add_header(
+            ['packet index', 'hash value', 'hash index', 'queue id', 'actual queue id', 'pass '])
+
+        i = 0
+        for tmp_reta_line in reta_lines:
+            status = "false"
+            if(self.nic in ["niantic", "redrockcanyou"]):
+                # compute the hash result of five tuple into the 7 LSBs value.
+                hash_index = int(tmp_reta_line["RSS hash"], 16) % 128
+            else:
+                # compute the hash result of five tuple into the 7 LSBs value.
+                hash_index = int(tmp_reta_line["RSS hash"], 16) % 512
+            if(reta_entries[hash_index] == int(tmp_reta_line["queue"])):
+                status = "true"
+                result.insert(i, 0)
+            else:
+                status = "fail"
+                result.insert(i, 1)
+            dts.results_table_add_row(
+                [i, tmp_reta_line["RSS hash"], hash_index, reta_entries[hash_index], tmp_reta_line["queue"], status])
+            i = i + 1
+
+        dts.results_table_print()
+        reta_lines = []
+        self.verify(sum(result) == 0, "the reta update function failed!")
+
+    def set_up_all(self):
+        """
+        Run at the start of each test suite.
+        """
+
+        #self.verify(
+        #    self.nic in ["niantic", "fortville_eagle", "fortville_spirit", "fortville_spirit_single"],
+        #    "NIC Unsupported: " + str(self.nic))
+        self.dut_ports = self.dut.get_ports(self.nic)
+        self.ports_socket = self.dut.get_numa_id(self.dut_ports[0])
+        self.verify(len(self.dut_ports) >= 1, "Not enough ports available")
+
+        self.vm0 = None
+        self.setup_1pf_1vf_1vm_env_flag = 0
+        self.setup_1pf_1vf_1vm_env(driver='')
+        #self.vm0_dut_ports = self.vm_dut_0.get_ports('any')
+  
+        #self.pmdout = PmdOutput(self.dut)
+
+    def set_up(self):
+        """
+        Run before each test case.
+        """
+        #self.vm0_testpmd = PmdOutput(self.vm_dut_0)
+        #self.vm0_testpmd.start_testpmd(VM_CORES_MASK,param = "--rxq=4 --rxq=4")
+        pass
+    def setup_1pf_1vf_1vm_env(self, driver='default'):
+        
+        self.used_dut_port_0 = self.dut_ports[0]
+        self.dut.generate_sriov_vfs_by_port(self.used_dut_port_0, 1, driver=driver)
+        self.sriov_vfs_port_0 = self.dut.ports_info[self.used_dut_port_0]['vfs_port']
+
+        try:
+
+            for port in self.sriov_vfs_port_0:
+                port.bind_driver('pci-stub')
+
+            time.sleep(1)
+            vf0_prot = {'opt_host': self.sriov_vfs_port_0[0].pci}
+
+            if driver == 'igb_uio':
+                # start testpmd without the two VFs on the host
+                self.host_testpmd = PmdOutput(self.dut)
+                eal_param = '-b %(vf0)s -b %(vf1)s' % {'vf0': self.sriov_vfs_port_0[0].pci}
+                self.host_testpmd.start_testpmd("1S/2C/2T", eal_param=eal_param)
+
+            # set up VM0 ENV
+            self.vm0 = QEMUKvm(self.dut, 'vm0', 'vf_rss')
+            self.vm0.set_vm_device(driver='pci-assign', **vf0_prot)
+
+            self.vm_dut_0 = self.vm0.start()
+            self.vm0_testpmd = PmdOutput(self.vm_dut_0)
+            if self.vm_dut_0 is None:
+                raise Exception("Set up VM0 ENV failed!")
+
+            self.setup_1pf_1vf_1vm_env_flag = 1
+        except Exception as e:
+            self.destroy_1pf_1vf_1vm_env()
+            raise Exception(e)
+
+    def destroy_1pf_1vf_1vm_env(self):
+        if getattr(self, 'vm0', None):
+            #destroy testpmd in vm0
+            #self.vm0_testpmd.execute_cmd('stop')
+            self.vm0_testpmd.execute_cmd('quit', '# ')
+            self.vm0_testpmd = None
+            self.vm0_dut_ports = None
+            #destroy vm0
+            self.vm0.stop()
+            self.vm0 = None
+
+        if getattr(self, 'host_testpmd', None):
+            self.host_testpmd.execute_cmd('quit', '# ')
+            self.host_testpmd = None
+
+        if getattr(self, 'used_dut_port_0', None):
+            self.dut.destroy_sriov_vfs_by_port(self.used_dut_port_0)
+            port = self.dut.ports_info[self.used_dut_port_0]['port']
+            port.bind_driver()
+            self.used_dut_port_0 = None
+
+        for port_id in self.dut_ports:
+            port = self.dut.ports_info[port_id]['port']
+            port.bind_driver()
+
+        self.setup_1pf_2vf_1vm_env_flag = 0
+
+    def test_vf_pmdrss_reta(self):
+        
+        # niantic kernel host driver not support this case
+        if self.nic is 'niantic' and not self.host_testpmd:
+            return
+        vm0dutPorts = self.vm_dut_0.get_ports('any')
+        localPort = self.tester.get_local_port(vm0dutPorts[0])
+        itf = self.tester.get_interface(localPort)
+        iptypes = ['IPV4']
+
+        self.vm_dut_0.kill_all()
+
+        # test with different rss queues
+        for queue in testQueues:
+
+            self.vm0_testpmd.start_testpmd(
+                "all", "--rxq=%d --txq=%d" % (queue, queue), socket=self.ports_socket)
+
+            for iptype in iptypes:
+                self.vm_dut_0.send_expect("set verbose 8", "testpmd> ")
+                self.vm_dut_0.send_expect("set fwd rxonly", "testpmd> ")
+                self.vm_dut_0.send_expect(
+                    "set nbcore %d" % (queue + 1), "testpmd> ")
+
+                # configure the reta with specific mappings.
+                if(self.nic in ["niantic", "redrockcanyou"]):
+                    for i in range(128):
+                        reta_entries.insert(i, random.randint(0, queue - 1))
+                        self.vm_dut_0.send_expect(
+                            "port config 0 rss reta (%d,%d)" % (i, reta_entries[i]), "testpmd> ")
+                else:
+                    for i in range(512):
+                        reta_entries.insert(i, random.randint(0, queue - 1))
+                        self.vm_dut_0.send_expect(
+                            "port config 0 rss reta (%d,%d)" % (i, reta_entries[i]), "testpmd> ")
+
+                self.send_packet(itf, iptype)
+
+            self.vm_dut_0.send_expect("quit", "# ", 30)
+    def test_vf_pmdrss(self): 
+        vm0dutPorts = self.vm_dut_0.get_ports('any')
+        localPort = self.tester.get_local_port(vm0dutPorts[0])
+        itf = self.tester.get_interface(localPort)
+        iptypes = {'ipv4-sctp':'ip',
+                   'ipv4-other':'ip',
+                   'ipv4-udp':'udp',
+                   'ipv4-tcp':'tcp',
+                   'ipv4-sctp':'sctp',
+                   'ipv6-other':'ip',
+                   'ipv6-udp':'udp',
+                   'ipv6-tcp':'tcp',
+                   'ipv6-sctp':'sctp',
+                 #  'l2_payload':'ether'
+                  }
+
+        self.vm_dut_0.kill_all()
+
+        # test with different rss queues
+        for queue in testQueues:
+
+            self.vm0_testpmd.start_testpmd(
+                "all", "--rxq=%d --txq=%d" % (queue, queue), socket=self.ports_socket)
+
+            for iptype,rsstype in iptypes.items():
+                self.vm_dut_0.send_expect("set verbose 8", "testpmd> ")
+                self.vm_dut_0.send_expect("set fwd rxonly", "testpmd> ")
+                self.vm_dut_0.send_expect("port config all rss %s" % rsstype, "testpmd> ")
+                self.vm_dut_0.send_expect(
+                    "set nbcore %d" % (queue + 1), "testpmd> ")
+
+                self.send_packet(itf, iptype)
+
+            self.vm_dut_0.send_expect("quit", "# ", 30)
+    def tear_down(self):
+        """
+        Run after each test case.
+        """
+        pass
+
+    def tear_down_all(self):
+        """
+        Run after each test suite.
+        """
+        #self.vm_dut_0.kill_all()
+        self.destroy_1pf_1vf_1vm_env()
-- 
1.9.3

Re: [dts] [PATCH V1 1/3] add vf rss test plan

[Liu, Yong]

Huilong, one question about VF supported hash algorithm.

> -----Original Message-----
> From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of xu,huilong
> Sent: Tuesday, January 12, 2016 6:53 PM
> To: dts@dpdk.org
> Subject: [dts] [PATCH V1 1/3] add vf rss test plan
> 
> Signed-off-by: xu,huilong <huilongx.xu@intel.com>
> ---
>  test_plans/vf_rss_test_plan.rst | 182
> ++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 182 insertions(+)
>  create mode 100644 test_plans/vf_rss_test_plan.rst
> 
> diff --git a/test_plans/vf_rss_test_plan.rst
> b/test_plans/vf_rss_test_plan.rst
> new file mode 100644
> index 0000000..f832924
> --- /dev/null
> +++ b/test_plans/vf_rss_test_plan.rst
> @@ -0,0 +1,182 @@
> +.. Copyright (c) <2016>, Intel Corporation
> +   All rights reserved.
> +
> +   Redistribution and use in source and binary forms, with or without
> +   modification, are permitted provided that the following conditions
> +   are met:
> +
> +   - Redistributions of source code must retain the above copyright
> +     notice, this list of conditions and the following disclaimer.
> +
> +   - Redistributions in binary form must reproduce the above copyright
> +     notice, this list of conditions and the following disclaimer in
> +     the documentation and/or other materials provided with the
> +     distribution.
> +
> +   - Neither the name of Intel Corporation nor the names of its
> +     contributors may be used to endorse or promote products derived
> +     from this software without specific prior written permission.
> +
> +   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> +   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> +   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
> +   FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
> +   COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
> +   INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
> +   (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
> +   SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
> +   HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
> +   STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
> +   ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
> +   OF THE POSSIBILITY OF SUCH DAMAGE.
> +
> +==================================================================
> +Fortville RSS full support - Support configuring hash functions
> +==================================================================
> +
> +This document provides test plan for testing the function of Fortville:
> +Support configuring hash functions.
> +
> +Prerequisites
> +-------------
> +
> +2x Intel? 82599 (Niantic) NICs (2x 10GbE full duplex optical ports per
> NIC)
> +1x Fortville_eagle NIC (4x 10G)
> +1x Fortville_spirit NIC (2x 40G)
> +2x Fortville_spirit_single NIC (1x 40G)
> +
> +The one port of the 82599 connect to the Fortville_eagle;
> +The one port of Fortville_spirit connect to Fortville_spirit_single.
> +The three kinds of NICs are the target NICs. the connected NICs can send
> packets
> +to these three NICs using scapy.
> +
> +Network Traffic
> +---------------
> +
> +The RSS feature is designed to improve networking performance by load
> balancing
> +the packets received from a NIC port to multiple NIC RX queues, with each
> queue
> +handled by a different logical core.
> +
> +#1. The receive packet is parsed into the header fields used by the hash
> +operation (such as IP addresses, TCP port, etc.)
> +
> +#2. A hash calculation is performed. The Fortville  supports four hash
> function:
> +Toeplitz, simple XOR and their Symmetric RSS.
> +

Here mentioned that Fortville support three different types of algorithm. But in following plan, only default algorithm is validated.
My question is that whether VF can't change algorithm or here just tope?


> +#3. The seven LSBs of the hash result are used as an index into a 128/512
> entry
> +'redirection table'.
> +
> +The RSS RETA update feature is designed to make RSS more flexible by
> allowing
> +users to define the correspondence between the seven LSBs of hash result
> and
> +the queue id(RSS output index) by themself.
> +
> +
> +Test Case:  test_rss_hash
> +========================================
> +
> +The following RX Ports/Queues configurations have to be benchmarked:
> +
> +- 1 RX port / 4 RX queues (1P/4Q)
> +
> +
> +Testpmd configuration - 4 RX/TX queues per port
> +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> +
> +::
> +
> +  testpmd -c 1f -n 3  -- -i --rxq=4 --txq=4
> +
> +Testpmd Configuration Options
> +-----------------------------
> +
> +By default, a single logical core runs the test.
> +The CPU IDs and the number of logical cores running the test in parallel
> can
> +be manually set with the ``set corelist X,Y`` and the ``set nbcore N``
> +interactive commands of the ``testpmd`` application.
> +#1. got the pci device id of DUT, for example,
> +
> +./dpdk_nic_bind.py --st
> +
> +0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0
> drv=i40e unused=
> +0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1
> drv=i40e unused=
> +
> +#2. create 2 VFs from 2 PFs,
> +
> +echo 1 > /sys/bus/pci/devices/0000\:81\:00.0/sriov_numvfs
> +echo 1 > /sys/bus/pci/devices/0000\:81\:00.1/sriov_numvfs
> +./dpdk_nic_bind.py --st
> +
> +0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0
> drv=i40e unused=
> +0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1
> drv=i40e unused=
> +0000:81:02.0 'XL710/X710 Virtual Function' unused=
> +0000:81:0a.0 'XL710/X710 Virtual Function' unused=
> +
> +#3. detach VFs from the host, bind them to pci-stub driver,
> +
> +/sbin/modprobe pci-stub
> +
> +using `lspci -nn|grep -i ethernet` got VF device id, for example "8086
> 154c",
> +
> +echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
> +echo 0000:81:02.0 > /sys/bus/pci/devices/0000:08:02.0/driver/unbind
> +echo 0000:81:02.0 > /sys/bus/pci/drivers/pci-stub/bind
> +
> +echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
> +echo 0000:81:0a.0 > /sys/bus/pci/devices/0000:08:0a.0/driver/unbind
> +echo 0000:81:0a.0 > /sys/bus/pci/drivers/pci-stub/bind
> +
> +  or using the following more easy way,
> +
> +  virsh nodedev-detach pci_0000_81_02_0;
> +  virsh nodedev-detach pci_0000_81_0a_0;
> +
> +  ./dpdk_nic_bind.py --st
> +
> +  0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0
> drv=i40e unused=
> +  0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1
> drv=i40e unused=
> +  0000:81:02.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
> +  0000:81:0a.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
> +
> +  it can be seen that VFs 81:02.0 & 81:0a.0 's drv is pci-stub.
> +
> +#4. passthrough VFs 81:02.0 & 81:0a.0 to vm0, and start vm0,
> +
> +  /usr/bin/qemu-system-x86_64  -name vm0 -enable-kvm \
> +  -cpu host -smp 4 -m 2048 -drive file=/home/image/sriov-fc20-1.img -
> vnc :1 \
> +  -device pci-assign,host=81:02.0,id=pt_0 \
> +  -device pci-assign,host=81:0a.0,id=pt_1
> +
> +#5. login vm0, got VFs pci device id in vm0, assume they are 00:06.0 &
> 00:07.0, bind them to igb_uio driver,
> +and then start testpmd, set it in mac forward mode,
> +
> +./tools/dpdk_nic_bind.py --bind=igb_uio 00:06.0 00:07.0
> +
> +#6. Reta Configuration.  128 reta entries configuration::
> +
> +  testpmd command: port config 0 rss reta (hash_index,queue_id)
> +
> +#7. PMD fwd only receive the packets::
> +
> +  testpmd command: set fwd rxonly
> +
> +#8. rss recived package type configuration two received packet types
> configuration::
> +
> +  testpmd command: port config 0 rss ip/udp/tcp
> +
> +#9. verbose configuration::
> +
> +  testpmd command: set verbose 8
> +
> +#10. start packet receive::
> +
> +  testpmd command: start
> +#11. send packet and check rx port received packet by different queue.
> +  different hash type send different packet, example hash type is ip,
> packet src and dts ip not different
> +  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.4",
> dst="192.168.0.5")], iface="eth3")
> +  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.5",
> dst="192.168.0.4")], iface="eth3")
> +Test Case:  test_reta
> +========================================
> +   this case test hash reta table, the test steps same with test_rss_hash
> except config hash reta table
> +#1 before send packet, config hash reta,512(niantic nic have 128 reta)
> reta entries configuration::
> +
> +     testpmd command: port config 0 rss reta (hash_index,queue_id)
> --
> 1.9.3

Re: [dts] [PATCH V1 1/3] add vf rss test plan

[Xu, HuilongX]

> -----Original Message-----
> From: Liu, Yong
> Sent: Tuesday, January 12, 2016 9:34 PM
> To: Xu, HuilongX; dts@dpdk.org
> Subject: RE: [dts] [PATCH V1 1/3] add vf rss test plan
> 
> Huilong, one question about VF supported hash algorithm.
> 
> > -----Original Message-----
> > From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of xu,huilong
> > Sent: Tuesday, January 12, 2016 6:53 PM
> > To: dts@dpdk.org
> > Subject: [dts] [PATCH V1 1/3] add vf rss test plan
> >
> > Signed-off-by: xu,huilong <huilongx.xu@intel.com>
> > ---
> >  test_plans/vf_rss_test_plan.rst | 182
> > ++++++++++++++++++++++++++++++++++++++++
> >  1 file changed, 182 insertions(+)
> >  create mode 100644 test_plans/vf_rss_test_plan.rst
> >
> > diff --git a/test_plans/vf_rss_test_plan.rst
> > b/test_plans/vf_rss_test_plan.rst
> > new file mode 100644
> > index 0000000..f832924
> > --- /dev/null
> > +++ b/test_plans/vf_rss_test_plan.rst
> > @@ -0,0 +1,182 @@
> > +.. Copyright (c) <2016>, Intel Corporation
> > +   All rights reserved.
> > +
> > +   Redistribution and use in source and binary forms, with or without
> > +   modification, are permitted provided that the following conditions
> > +   are met:
> > +
> > +   - Redistributions of source code must retain the above copyright
> > +     notice, this list of conditions and the following disclaimer.
> > +
> > +   - Redistributions in binary form must reproduce the above copyright
> > +     notice, this list of conditions and the following disclaimer in
> > +     the documentation and/or other materials provided with the
> > +     distribution.
> > +
> > +   - Neither the name of Intel Corporation nor the names of its
> > +     contributors may be used to endorse or promote products derived
> > +     from this software without specific prior written permission.
> > +
> > +   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> > +   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> > +   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
> > +   FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
> > +   COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
> > +   INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
> > +   (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
> > +   SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
> > +   HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
> > +   STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
> > +   ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
> > +   OF THE POSSIBILITY OF SUCH DAMAGE.
> > +
> > +==================================================================
> > +Fortville RSS full support - Support configuring hash functions
> > +==================================================================
> > +
> > +This document provides test plan for testing the function of Fortville:
> > +Support configuring hash functions.
> > +
> > +Prerequisites
> > +-------------
> > +
> > +2x Intel? 82599 (Niantic) NICs (2x 10GbE full duplex optical ports per
> > NIC)
> > +1x Fortville_eagle NIC (4x 10G)
> > +1x Fortville_spirit NIC (2x 40G)
> > +2x Fortville_spirit_single NIC (1x 40G)
> > +
> > +The one port of the 82599 connect to the Fortville_eagle;
> > +The one port of Fortville_spirit connect to Fortville_spirit_single.
> > +The three kinds of NICs are the target NICs. the connected NICs can
> send
> > packets
> > +to these three NICs using scapy.
> > +
> > +Network Traffic
> > +---------------
> > +
> > +The RSS feature is designed to improve networking performance by load
> > balancing
> > +the packets received from a NIC port to multiple NIC RX queues, with
> each
> > queue
> > +handled by a different logical core.
> > +
> > +#1. The receive packet is parsed into the header fields used by the
> hash
> > +operation (such as IP addresses, TCP port, etc.)
> > +
> > +#2. A hash calculation is performed. The Fortville  supports four hash
> > function:
> > +Toeplitz, simple XOR and their Symmetric RSS.
> > +
> 
> Here mentioned that Fortville support three different types of algorithm.
> But in following plan, only default algorithm is validated.
> My question is that whether VF can't change algorithm or here just tope?
> 
FVL VF want change hash algorithm, host driver must is dpdk driver, because VF hash algorithm config is a global config.
In this plan, only used kernel driver. I will add new case for dpdk host driver.

> > +#3. The seven LSBs of the hash result are used as an index into a
> 128/512
> > entry
> > +'redirection table'.
> > +
> > +The RSS RETA update feature is designed to make RSS more flexible by
> > allowing
> > +users to define the correspondence between the seven LSBs of hash
> result
> > and
> > +the queue id(RSS output index) by themself.
> > +
> > +
> > +Test Case:  test_rss_hash
> > +========================================
> > +
> > +The following RX Ports/Queues configurations have to be benchmarked:
> > +
> > +- 1 RX port / 4 RX queues (1P/4Q)
> > +
> > +
> > +Testpmd configuration - 4 RX/TX queues per port
> > +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> > +
> > +::
> > +
> > +  testpmd -c 1f -n 3  -- -i --rxq=4 --txq=4
> > +
> > +Testpmd Configuration Options
> > +-----------------------------
> > +
> > +By default, a single logical core runs the test.
> > +The CPU IDs and the number of logical cores running the test in
> parallel
> > can
> > +be manually set with the ``set corelist X,Y`` and the ``set nbcore N``
> > +interactive commands of the ``testpmd`` application.
> > +#1. got the pci device id of DUT, for example,
> > +
> > +./dpdk_nic_bind.py --st
> > +
> > +0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0
> > drv=i40e unused=
> > +0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1
> > drv=i40e unused=
> > +
> > +#2. create 2 VFs from 2 PFs,
> > +
> > +echo 1 > /sys/bus/pci/devices/0000\:81\:00.0/sriov_numvfs
> > +echo 1 > /sys/bus/pci/devices/0000\:81\:00.1/sriov_numvfs
> > +./dpdk_nic_bind.py --st
> > +
> > +0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0
> > drv=i40e unused=
> > +0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1
> > drv=i40e unused=
> > +0000:81:02.0 'XL710/X710 Virtual Function' unused=
> > +0000:81:0a.0 'XL710/X710 Virtual Function' unused=
> > +
> > +#3. detach VFs from the host, bind them to pci-stub driver,
> > +
> > +/sbin/modprobe pci-stub
> > +
> > +using `lspci -nn|grep -i ethernet` got VF device id, for example "8086
> > 154c",
> > +
> > +echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
> > +echo 0000:81:02.0 > /sys/bus/pci/devices/0000:08:02.0/driver/unbind
> > +echo 0000:81:02.0 > /sys/bus/pci/drivers/pci-stub/bind
> > +
> > +echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
> > +echo 0000:81:0a.0 > /sys/bus/pci/devices/0000:08:0a.0/driver/unbind
> > +echo 0000:81:0a.0 > /sys/bus/pci/drivers/pci-stub/bind
> > +
> > +  or using the following more easy way,
> > +
> > +  virsh nodedev-detach pci_0000_81_02_0;
> > +  virsh nodedev-detach pci_0000_81_0a_0;
> > +
> > +  ./dpdk_nic_bind.py --st
> > +
> > +  0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0
> > drv=i40e unused=
> > +  0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1
> > drv=i40e unused=
> > +  0000:81:02.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
> > +  0000:81:0a.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
> > +
> > +  it can be seen that VFs 81:02.0 & 81:0a.0 's drv is pci-stub.
> > +
> > +#4. passthrough VFs 81:02.0 & 81:0a.0 to vm0, and start vm0,
> > +
> > +  /usr/bin/qemu-system-x86_64  -name vm0 -enable-kvm \
> > +  -cpu host -smp 4 -m 2048 -drive file=/home/image/sriov-fc20-1.img -
> > vnc :1 \
> > +  -device pci-assign,host=81:02.0,id=pt_0 \
> > +  -device pci-assign,host=81:0a.0,id=pt_1
> > +
> > +#5. login vm0, got VFs pci device id in vm0, assume they are 00:06.0 &
> > 00:07.0, bind them to igb_uio driver,
> > +and then start testpmd, set it in mac forward mode,
> > +
> > +./tools/dpdk_nic_bind.py --bind=igb_uio 00:06.0 00:07.0
> > +
> > +#6. Reta Configuration.  128 reta entries configuration::
> > +
> > +  testpmd command: port config 0 rss reta (hash_index,queue_id)
> > +
> > +#7. PMD fwd only receive the packets::
> > +
> > +  testpmd command: set fwd rxonly
> > +
> > +#8. rss recived package type configuration two received packet types
> > configuration::
> > +
> > +  testpmd command: port config 0 rss ip/udp/tcp
> > +
> > +#9. verbose configuration::
> > +
> > +  testpmd command: set verbose 8
> > +
> > +#10. start packet receive::
> > +
> > +  testpmd command: start
> > +#11. send packet and check rx port received packet by different queue.
> > +  different hash type send different packet, example hash type is ip,
> > packet src and dts ip not different
> > +  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.4",
> > dst="192.168.0.5")], iface="eth3")
> > +  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.5",
> > dst="192.168.0.4")], iface="eth3")
> > +Test Case:  test_reta
> > +========================================
> > +   this case test hash reta table, the test steps same with
> test_rss_hash
> > except config hash reta table
> > +#1 before send packet, config hash reta,512(niantic nic have 128 reta)
> > reta entries configuration::
> > +
> > +     testpmd command: port config 0 rss reta (hash_index,queue_id)
> > --
> > 1.9.3

Re: [dts] [PATCH V1 1/3] add vf rss test plan

[Pei, Yulong]

Hi Huilong,

It seems that you also did not change the plan according to our discussion by telephone.

Best Regards
Yulong Pei

-----Original Message-----
From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of xu,huilong
Sent: Tuesday, January 12, 2016 6:53 PM
To: dts@dpdk.org
Subject: [dts] [PATCH V1 1/3] add vf rss test plan

Signed-off-by: xu,huilong <huilongx.xu@intel.com>
---
 test_plans/vf_rss_test_plan.rst | 182 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 182 insertions(+)
 create mode 100644 test_plans/vf_rss_test_plan.rst

diff --git a/test_plans/vf_rss_test_plan.rst b/test_plans/vf_rss_test_plan.rst new file mode 100644 index 0000000..f832924
--- /dev/null
+++ b/test_plans/vf_rss_test_plan.rst
@@ -0,0 +1,182 @@
+.. Copyright (c) <2016>, Intel Corporation
+   All rights reserved.
+
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+
+   - Redistributions of source code must retain the above copyright
+     notice, this list of conditions and the following disclaimer.
+
+   - Redistributions in binary form must reproduce the above copyright
+     notice, this list of conditions and the following disclaimer in
+     the documentation and/or other materials provided with the
+     distribution.
+
+   - Neither the name of Intel Corporation nor the names of its
+     contributors may be used to endorse or promote products derived
+     from this software without specific prior written permission.
+
+   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+   FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+   COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
+   INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+   (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+   SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+   HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+   STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+   ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+   OF THE POSSIBILITY OF SUCH DAMAGE.
+
+==================================================================
+Fortville RSS full support - Support configuring hash functions 
+==================================================================
+
+This document provides test plan for testing the function of Fortville:
+Support configuring hash functions.
+
+Prerequisites
+-------------
+
+2x Intel? 82599 (Niantic) NICs (2x 10GbE full duplex optical ports per 
+NIC) 1x Fortville_eagle NIC (4x 10G) 1x Fortville_spirit NIC (2x 40G) 
+2x Fortville_spirit_single NIC (1x 40G)
+
+The one port of the 82599 connect to the Fortville_eagle; The one port 
+of Fortville_spirit connect to Fortville_spirit_single.
+The three kinds of NICs are the target NICs. the connected NICs can 
+send packets to these three NICs using scapy.
+
+Network Traffic
+---------------
+
+The RSS feature is designed to improve networking performance by load 
+balancing the packets received from a NIC port to multiple NIC RX 
+queues, with each queue handled by a different logical core.
+
+#1. The receive packet is parsed into the header fields used by the 
+hash operation (such as IP addresses, TCP port, etc.)
+
+#2. A hash calculation is performed. The Fortville  supports four hash function:
+Toeplitz, simple XOR and their Symmetric RSS.
+
+#3. The seven LSBs of the hash result are used as an index into a 
+128/512 entry 'redirection table'.
+
+The RSS RETA update feature is designed to make RSS more flexible by 
+allowing users to define the correspondence between the seven LSBs of 
+hash result and the queue id(RSS output index) by themself.
+
+
+Test Case:  test_rss_hash
+========================================
+
+The following RX Ports/Queues configurations have to be benchmarked:
+
+- 1 RX port / 4 RX queues (1P/4Q)
+
+
+Testpmd configuration - 4 RX/TX queues per port 
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+::
+  
+  testpmd -c 1f -n 3  -- -i --rxq=4 --txq=4
+
+Testpmd Configuration Options
+-----------------------------
+
+By default, a single logical core runs the test.
+The CPU IDs and the number of logical cores running the test in 
+parallel can be manually set with the ``set corelist X,Y`` and the 
+``set nbcore N`` interactive commands of the ``testpmd`` application.
+#1. got the pci device id of DUT, for example,
+
+./dpdk_nic_bind.py --st
+
+0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0 
+drv=i40e unused=
+0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1 
+drv=i40e unused=
+
+#2. create 2 VFs from 2 PFs,
+
+echo 1 > /sys/bus/pci/devices/0000\:81\:00.0/sriov_numvfs
+echo 1 > /sys/bus/pci/devices/0000\:81\:00.1/sriov_numvfs
+./dpdk_nic_bind.py --st
+
+0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0 
+drv=i40e unused=
+0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1 
+drv=i40e unused=
+0000:81:02.0 'XL710/X710 Virtual Function' unused=
+0000:81:0a.0 'XL710/X710 Virtual Function' unused=
+
+#3. detach VFs from the host, bind them to pci-stub driver,
+
+/sbin/modprobe pci-stub
+
+using `lspci -nn|grep -i ethernet` got VF device id, for example "8086 
+154c",
+
+echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
+echo 0000:81:02.0 > /sys/bus/pci/devices/0000:08:02.0/driver/unbind
+echo 0000:81:02.0 > /sys/bus/pci/drivers/pci-stub/bind
+
+echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
+echo 0000:81:0a.0 > /sys/bus/pci/devices/0000:08:0a.0/driver/unbind
+echo 0000:81:0a.0 > /sys/bus/pci/drivers/pci-stub/bind
+
+  or using the following more easy way,
+
+  virsh nodedev-detach pci_0000_81_02_0;  virsh nodedev-detach 
+ pci_0000_81_0a_0;
+
+  ./dpdk_nic_bind.py --st
+
+  0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0 
+ drv=i40e unused=
+  0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1 
+ drv=i40e unused=
+  0000:81:02.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
+  0000:81:0a.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
+
+  it can be seen that VFs 81:02.0 & 81:0a.0 's drv is pci-stub.
+
+#4. passthrough VFs 81:02.0 & 81:0a.0 to vm0, and start vm0,
+
+  /usr/bin/qemu-system-x86_64  -name vm0 -enable-kvm \  -cpu host -smp 
+ 4 -m 2048 -drive file=/home/image/sriov-fc20-1.img -vnc :1 \  -device 
+ pci-assign,host=81:02.0,id=pt_0 \  -device 
+ pci-assign,host=81:0a.0,id=pt_1
+
+#5. login vm0, got VFs pci device id in vm0, assume they are 00:06.0 & 
+00:07.0, bind them to igb_uio driver, and then start testpmd, set it in 
+mac forward mode,
+
+./tools/dpdk_nic_bind.py --bind=igb_uio 00:06.0 00:07.0
+
+#6. Reta Configuration.  128 reta entries configuration::
+
+  testpmd command: port config 0 rss reta (hash_index,queue_id)
+
+#7. PMD fwd only receive the packets::
+
+  testpmd command: set fwd rxonly
+
+#8. rss recived package type configuration two received packet types configuration::
+
+  testpmd command: port config 0 rss ip/udp/tcp
+
+#9. verbose configuration::
+
+  testpmd command: set verbose 8
+
+#10. start packet receive::
+
+  testpmd command: start
+#11. send packet and check rx port received packet by different queue.
+  different hash type send different packet, example hash type is ip, 
+packet src and dts ip not different
+  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.4", dst="192.168.0.5")], iface="eth3")
+  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.5", dst="192.168.0.4")], iface="eth3")   
+Test Case:  test_reta
+========================================
+   this case test hash reta table, the test steps same with 
+test_rss_hash except config hash reta table
+#1 before send packet, config hash reta,512(niantic nic have 128 reta) reta entries configuration::
+
+     testpmd command: port config 0 rss reta (hash_index,queue_id)
--
1.9.3

Re: [dts] [PATCH V1 1/3] add vf rss test plan

[Xu, HuilongX]

Hi yulong,
Which changed loss in last check by telephone?

> -----Original Message-----
> From: Pei, Yulong
> Sent: Wednesday, January 13, 2016 2:25 PM
> To: Xu, HuilongX; dts@dpdk.org
> Subject: RE: [dts] [PATCH V1 1/3] add vf rss test plan
> 
> Hi Huilong,
> 
> It seems that you also did not change the plan according to our discussion
> by telephone.
> 
> Best Regards
> Yulong Pei
> 
> -----Original Message-----
> From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of xu,huilong
> Sent: Tuesday, January 12, 2016 6:53 PM
> To: dts@dpdk.org
> Subject: [dts] [PATCH V1 1/3] add vf rss test plan
> 
> Signed-off-by: xu,huilong <huilongx.xu@intel.com>
> ---
>  test_plans/vf_rss_test_plan.rst | 182
> ++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 182 insertions(+)
>  create mode 100644 test_plans/vf_rss_test_plan.rst
> 
> diff --git a/test_plans/vf_rss_test_plan.rst
> b/test_plans/vf_rss_test_plan.rst new file mode 100644 index
> 0000000..f832924
> --- /dev/null
> +++ b/test_plans/vf_rss_test_plan.rst
> @@ -0,0 +1,182 @@
> +.. Copyright (c) <2016>, Intel Corporation
> +   All rights reserved.
> +
> +   Redistribution and use in source and binary forms, with or without
> +   modification, are permitted provided that the following conditions
> +   are met:
> +
> +   - Redistributions of source code must retain the above copyright
> +     notice, this list of conditions and the following disclaimer.
> +
> +   - Redistributions in binary form must reproduce the above copyright
> +     notice, this list of conditions and the following disclaimer in
> +     the documentation and/or other materials provided with the
> +     distribution.
> +
> +   - Neither the name of Intel Corporation nor the names of its
> +     contributors may be used to endorse or promote products derived
> +     from this software without specific prior written permission.
> +
> +   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> +   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> +   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
> +   FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
> +   COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
> +   INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
> +   (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
> +   SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
> +   HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
> +   STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
> +   ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
> +   OF THE POSSIBILITY OF SUCH DAMAGE.
> +
> +==================================================================
> +Fortville RSS full support - Support configuring hash functions
> +==================================================================
> +
> +This document provides test plan for testing the function of Fortville:
> +Support configuring hash functions.
> +
> +Prerequisites
> +-------------
> +
> +2x Intel? 82599 (Niantic) NICs (2x 10GbE full duplex optical ports per
> +NIC) 1x Fortville_eagle NIC (4x 10G) 1x Fortville_spirit NIC (2x 40G)
> +2x Fortville_spirit_single NIC (1x 40G)
> +
> +The one port of the 82599 connect to the Fortville_eagle; The one port
> +of Fortville_spirit connect to Fortville_spirit_single.
> +The three kinds of NICs are the target NICs. the connected NICs can
> +send packets to these three NICs using scapy.
> +
> +Network Traffic
> +---------------
> +
> +The RSS feature is designed to improve networking performance by load
> +balancing the packets received from a NIC port to multiple NIC RX
> +queues, with each queue handled by a different logical core.
> +
> +#1. The receive packet is parsed into the header fields used by the
> +hash operation (such as IP addresses, TCP port, etc.)
> +
> +#2. A hash calculation is performed. The Fortville  supports four hash
> function:
> +Toeplitz, simple XOR and their Symmetric RSS.
> +
> +#3. The seven LSBs of the hash result are used as an index into a
> +128/512 entry 'redirection table'.
> +
> +The RSS RETA update feature is designed to make RSS more flexible by
> +allowing users to define the correspondence between the seven LSBs of
> +hash result and the queue id(RSS output index) by themself.
> +
> +
> +Test Case:  test_rss_hash
> +========================================
> +
> +The following RX Ports/Queues configurations have to be benchmarked:
> +
> +- 1 RX port / 4 RX queues (1P/4Q)
> +
> +
> +Testpmd configuration - 4 RX/TX queues per port
> +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> +
> +::
> +
> +  testpmd -c 1f -n 3  -- -i --rxq=4 --txq=4
> +
> +Testpmd Configuration Options
> +-----------------------------
> +
> +By default, a single logical core runs the test.
> +The CPU IDs and the number of logical cores running the test in
> +parallel can be manually set with the ``set corelist X,Y`` and the
> +``set nbcore N`` interactive commands of the ``testpmd`` application.
> +#1. got the pci device id of DUT, for example,
> +
> +./dpdk_nic_bind.py --st
> +
> +0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0
> +drv=i40e unused=
> +0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1
> +drv=i40e unused=
> +
> +#2. create 2 VFs from 2 PFs,
> +
> +echo 1 > /sys/bus/pci/devices/0000\:81\:00.0/sriov_numvfs
> +echo 1 > /sys/bus/pci/devices/0000\:81\:00.1/sriov_numvfs
> +./dpdk_nic_bind.py --st
> +
> +0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0
> +drv=i40e unused=
> +0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1
> +drv=i40e unused=
> +0000:81:02.0 'XL710/X710 Virtual Function' unused=
> +0000:81:0a.0 'XL710/X710 Virtual Function' unused=
> +
> +#3. detach VFs from the host, bind them to pci-stub driver,
> +
> +/sbin/modprobe pci-stub
> +
> +using `lspci -nn|grep -i ethernet` got VF device id, for example "8086
> +154c",
> +
> +echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
> +echo 0000:81:02.0 > /sys/bus/pci/devices/0000:08:02.0/driver/unbind
> +echo 0000:81:02.0 > /sys/bus/pci/drivers/pci-stub/bind
> +
> +echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
> +echo 0000:81:0a.0 > /sys/bus/pci/devices/0000:08:0a.0/driver/unbind
> +echo 0000:81:0a.0 > /sys/bus/pci/drivers/pci-stub/bind
> +
> +  or using the following more easy way,
> +
> +  virsh nodedev-detach pci_0000_81_02_0;  virsh nodedev-detach
> + pci_0000_81_0a_0;
> +
> +  ./dpdk_nic_bind.py --st
> +
> +  0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0
> + drv=i40e unused=
> +  0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1
> + drv=i40e unused=
> +  0000:81:02.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
> +  0000:81:0a.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
> +
> +  it can be seen that VFs 81:02.0 & 81:0a.0 's drv is pci-stub.
> +
> +#4. passthrough VFs 81:02.0 & 81:0a.0 to vm0, and start vm0,
> +
> +  /usr/bin/qemu-system-x86_64  -name vm0 -enable-kvm \  -cpu host -smp
> + 4 -m 2048 -drive file=/home/image/sriov-fc20-1.img -vnc :1 \  -device
> + pci-assign,host=81:02.0,id=pt_0 \  -device
> + pci-assign,host=81:0a.0,id=pt_1
> +
> +#5. login vm0, got VFs pci device id in vm0, assume they are 00:06.0 &
> +00:07.0, bind them to igb_uio driver, and then start testpmd, set it in
> +mac forward mode,
> +
> +./tools/dpdk_nic_bind.py --bind=igb_uio 00:06.0 00:07.0
> +
> +#6. Reta Configuration.  128 reta entries configuration::
> +
> +  testpmd command: port config 0 rss reta (hash_index,queue_id)
> +
> +#7. PMD fwd only receive the packets::
> +
> +  testpmd command: set fwd rxonly
> +
> +#8. rss recived package type configuration two received packet types
> configuration::
> +
> +  testpmd command: port config 0 rss ip/udp/tcp
> +
> +#9. verbose configuration::
> +
> +  testpmd command: set verbose 8
> +
> +#10. start packet receive::
> +
> +  testpmd command: start
> +#11. send packet and check rx port received packet by different queue.
> +  different hash type send different packet, example hash type is ip,
> +packet src and dts ip not different
> +  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.4",
> dst="192.168.0.5")], iface="eth3")
> +  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.5",
> dst="192.168.0.4")], iface="eth3")
> +Test Case:  test_reta
> +========================================
> +   this case test hash reta table, the test steps same with
> +test_rss_hash except config hash reta table
> +#1 before send packet, config hash reta,512(niantic nic have 128 reta)
> reta entries configuration::
> +
> +     testpmd command: port config 0 rss reta (hash_index,queue_id)
> --
> 1.9.3

Re: [dts] [PATCH V1 1/3] add vf rss test plan

[Pei, Yulong]

In your test plan:

> +#2. A hash calculation is performed. The Fortville  supports four 
> +hash
> function:
> +Toeplitz, simple XOR and their Symmetric RSS.

Should be three.

> +#3. The seven LSBs of the hash result are used as an index into a
> +128/512 entry 'redirection table'.

seven LSBs also not correct for Fortville and Fortville VF

> +2x Intel? 82599 (Niantic) NICs (2x 10GbE full duplex optical ports 
> +per
> +NIC) 1x Fortville_eagle NIC (4x 10G) 1x Fortville_spirit NIC (2x 40G) 
> +2x Fortville_spirit_single NIC (1x 40G)
> +
> +The one port of the 82599 connect to the Fortville_eagle; The one 
> +port of Fortville_spirit connect to Fortville_spirit_single.
> +The three kinds of NICs are the target NICs. the connected NICs can 
> +send packets to these three NICs using scapy.
> +

We come to agreement to simply it.

Best Regards
Yulong Pei

-----Original Message-----
From: Xu, HuilongX 
Sent: Wednesday, January 13, 2016 4:34 PM
To: Pei, Yulong <yulong.pei@intel.com>; dts@dpdk.org
Subject: RE: [dts] [PATCH V1 1/3] add vf rss test plan

Hi yulong,
Which changed loss in last check by telephone?

> -----Original Message-----
> From: Pei, Yulong
> Sent: Wednesday, January 13, 2016 2:25 PM
> To: Xu, HuilongX; dts@dpdk.org
> Subject: RE: [dts] [PATCH V1 1/3] add vf rss test plan
> 
> Hi Huilong,
> 
> It seems that you also did not change the plan according to our 
> discussion by telephone.
> 
> Best Regards
> Yulong Pei
> 
> -----Original Message-----
> From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of xu,huilong
> Sent: Tuesday, January 12, 2016 6:53 PM
> To: dts@dpdk.org
> Subject: [dts] [PATCH V1 1/3] add vf rss test plan
> 
> Signed-off-by: xu,huilong <huilongx.xu@intel.com>
> ---
>  test_plans/vf_rss_test_plan.rst | 182
> ++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 182 insertions(+)
>  create mode 100644 test_plans/vf_rss_test_plan.rst
> 
> diff --git a/test_plans/vf_rss_test_plan.rst 
> b/test_plans/vf_rss_test_plan.rst new file mode 100644 index
> 0000000..f832924
> --- /dev/null
> +++ b/test_plans/vf_rss_test_plan.rst
> @@ -0,0 +1,182 @@
> +.. Copyright (c) <2016>, Intel Corporation
> +   All rights reserved.
> +
> +   Redistribution and use in source and binary forms, with or without
> +   modification, are permitted provided that the following conditions
> +   are met:
> +
> +   - Redistributions of source code must retain the above copyright
> +     notice, this list of conditions and the following disclaimer.
> +
> +   - Redistributions in binary form must reproduce the above copyright
> +     notice, this list of conditions and the following disclaimer in
> +     the documentation and/or other materials provided with the
> +     distribution.
> +
> +   - Neither the name of Intel Corporation nor the names of its
> +     contributors may be used to endorse or promote products derived
> +     from this software without specific prior written permission.
> +
> +   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> +   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> +   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
> +   FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
> +   COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
> +   INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
> +   (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
> +   SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
> +   HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
> +   STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
> +   ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
> +   OF THE POSSIBILITY OF SUCH DAMAGE.
> +
> +==================================================================
> +Fortville RSS full support - Support configuring hash functions 
> +==================================================================
> +
> +This document provides test plan for testing the function of Fortville:
> +Support configuring hash functions.
> +
> +Prerequisites
> +-------------
> +
> +2x Intel? 82599 (Niantic) NICs (2x 10GbE full duplex optical ports 
> +per
> +NIC) 1x Fortville_eagle NIC (4x 10G) 1x Fortville_spirit NIC (2x 40G) 
> +2x Fortville_spirit_single NIC (1x 40G)
> +
> +The one port of the 82599 connect to the Fortville_eagle; The one 
> +port of Fortville_spirit connect to Fortville_spirit_single.
> +The three kinds of NICs are the target NICs. the connected NICs can 
> +send packets to these three NICs using scapy.
> +
> +Network Traffic
> +---------------
> +
> +The RSS feature is designed to improve networking performance by load 
> +balancing the packets received from a NIC port to multiple NIC RX 
> +queues, with each queue handled by a different logical core.
> +
> +#1. The receive packet is parsed into the header fields used by the 
> +hash operation (such as IP addresses, TCP port, etc.)
> +
> +#2. A hash calculation is performed. The Fortville  supports four 
> +hash
> function:
> +Toeplitz, simple XOR and their Symmetric RSS.
> +
> +#3. The seven LSBs of the hash result are used as an index into a
> +128/512 entry 'redirection table'.
> +
> +The RSS RETA update feature is designed to make RSS more flexible by 
> +allowing users to define the correspondence between the seven LSBs of 
> +hash result and the queue id(RSS output index) by themself.
> +
> +
> +Test Case:  test_rss_hash
> +========================================
> +
> +The following RX Ports/Queues configurations have to be benchmarked:
> +
> +- 1 RX port / 4 RX queues (1P/4Q)
> +
> +
> +Testpmd configuration - 4 RX/TX queues per port 
> +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> +
> +::
> +
> +  testpmd -c 1f -n 3  -- -i --rxq=4 --txq=4
> +
> +Testpmd Configuration Options
> +-----------------------------
> +
> +By default, a single logical core runs the test.
> +The CPU IDs and the number of logical cores running the test in 
> +parallel can be manually set with the ``set corelist X,Y`` and the 
> +``set nbcore N`` interactive commands of the ``testpmd`` application.
> +#1. got the pci device id of DUT, for example,
> +
> +./dpdk_nic_bind.py --st
> +
> +0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0 
> +drv=i40e unused=
> +0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1 
> +drv=i40e unused=
> +
> +#2. create 2 VFs from 2 PFs,
> +
> +echo 1 > /sys/bus/pci/devices/0000\:81\:00.0/sriov_numvfs
> +echo 1 > /sys/bus/pci/devices/0000\:81\:00.1/sriov_numvfs
> +./dpdk_nic_bind.py --st
> +
> +0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0 
> +drv=i40e unused=
> +0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1 
> +drv=i40e unused=
> +0000:81:02.0 'XL710/X710 Virtual Function' unused=
> +0000:81:0a.0 'XL710/X710 Virtual Function' unused=
> +
> +#3. detach VFs from the host, bind them to pci-stub driver,
> +
> +/sbin/modprobe pci-stub
> +
> +using `lspci -nn|grep -i ethernet` got VF device id, for example 
> +"8086 154c",
> +
> +echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
> +echo 0000:81:02.0 > /sys/bus/pci/devices/0000:08:02.0/driver/unbind
> +echo 0000:81:02.0 > /sys/bus/pci/drivers/pci-stub/bind
> +
> +echo "8086 154c" > /sys/bus/pci/drivers/pci-stub/new_id
> +echo 0000:81:0a.0 > /sys/bus/pci/devices/0000:08:0a.0/driver/unbind
> +echo 0000:81:0a.0 > /sys/bus/pci/drivers/pci-stub/bind
> +
> +  or using the following more easy way,
> +
> +  virsh nodedev-detach pci_0000_81_02_0;  virsh nodedev-detach 
> + pci_0000_81_0a_0;
> +
> +  ./dpdk_nic_bind.py --st
> +
> +  0000:81:00.0 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f0 
> + drv=i40e unused=
> +  0000:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' if=ens259f1 
> + drv=i40e unused=
> +  0000:81:02.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
> +  0000:81:0a.0 'XL710/X710 Virtual Function' if= drv=pci-stub unused=
> +
> +  it can be seen that VFs 81:02.0 & 81:0a.0 's drv is pci-stub.
> +
> +#4. passthrough VFs 81:02.0 & 81:0a.0 to vm0, and start vm0,
> +
> +  /usr/bin/qemu-system-x86_64  -name vm0 -enable-kvm \  -cpu host 
> + -smp
> + 4 -m 2048 -drive file=/home/image/sriov-fc20-1.img -vnc :1 \  
> + -device
> + pci-assign,host=81:02.0,id=pt_0 \  -device
> + pci-assign,host=81:0a.0,id=pt_1
> +
> +#5. login vm0, got VFs pci device id in vm0, assume they are 00:06.0 
> +& 00:07.0, bind them to igb_uio driver, and then start testpmd, set 
> +it in mac forward mode,
> +
> +./tools/dpdk_nic_bind.py --bind=igb_uio 00:06.0 00:07.0
> +
> +#6. Reta Configuration.  128 reta entries configuration::
> +
> +  testpmd command: port config 0 rss reta (hash_index,queue_id)
> +
> +#7. PMD fwd only receive the packets::
> +
> +  testpmd command: set fwd rxonly
> +
> +#8. rss recived package type configuration two received packet types
> configuration::
> +
> +  testpmd command: port config 0 rss ip/udp/tcp
> +
> +#9. verbose configuration::
> +
> +  testpmd command: set verbose 8
> +
> +#10. start packet receive::
> +
> +  testpmd command: start
> +#11. send packet and check rx port received packet by different queue.
> +  different hash type send different packet, example hash type is ip, 
> +packet src and dts ip not different
> +  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.4",
> dst="192.168.0.5")], iface="eth3")
> +  sendp([Ether(dst="90:e2:ba:36:99:3c")/IP(src="192.168.0.5",
> dst="192.168.0.4")], iface="eth3")
> +Test Case:  test_reta
> +========================================
> +   this case test hash reta table, the test steps same with 
> +test_rss_hash except config hash reta table
> +#1 before send packet, config hash reta,512(niantic nic have 128 
> +reta)
> reta entries configuration::
> +
> +     testpmd command: port config 0 rss reta (hash_index,queue_id)
> --
> 1.9.3

Re: [dts] [PATCH V1 3/3] add vf rss test script

[Pei, Yulong]

Hi,

-----Original Message-----
From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of xu,huilong
Sent: Tuesday, January 12, 2016 6:53 PM
To: dts@dpdk.org
Subject: [dts] [PATCH V1 3/3] add vf rss test script

Signed-off-by: xu,huilong <huilongx.xu@intel.com>
---
 tests/TestSuite_vf_rss.py | 382 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 382 insertions(+)
 create mode 100644 tests/TestSuite_vf_rss.py

diff --git a/tests/TestSuite_vf_rss.py b/tests/TestSuite_vf_rss.py new file mode 100644 index 0000000..a59516d
--- /dev/null
+++ b/tests/TestSuite_vf_rss.py
@@ -0,0 +1,382 @@
+# BSD LICENSE
+#
+# Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+# All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without # 
+modification, are permitted provided that the following conditions # 
+are met:
+#
+#   * Redistributions of source code must retain the above copyright
+#     notice, this list of conditions and the following disclaimer.
+#   * Redistributions in binary form must reproduce the above copyright
+#     notice, this list of conditions and the following disclaimer in
+#     the documentation and/or other materials provided with the
+#     distribution.
+#   * Neither the name of Intel Corporation nor the names of its
+#     contributors may be used to endorse or promote products derived
+#     from this software without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # 
+"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # 
+LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR # 
+A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT # 
+OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, # 
+SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT # 
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, # 
+DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY # 
+THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT # 
+(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE # 
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+"""
+DPDK Test suite.
+Test RSS reta (redirection table) update function.
+"""
+import time
+import random
+import re
+import dts
+testQueues = [4]
+reta_entries = []
+reta_lines = []
+
+# Use scapy to send packets with different source and dest ip.
+# and collect the hash result of five tuple and the queue id.
+from test_case import TestCase
+from pmd_output import PmdOutput
+from qemu_kvm import QEMUKvm
+
+class TestVfRss(TestCase):
+    def send_packet(self, itf, tran_type):
+        """
+        Sends packets.
+        """
+        global reta_lines
+        reta_lines = []
+        self.tester.scapy_foreground()
+        self.tester.scapy_append('sys.path.append("./")')
+        self.tester.scapy_append('from sctp import *')
+        self.vm_dut_0.send_expect("start", "testpmd>")
+        mac = self.vm_dut_0.get_mac_address(0)
+        # send packet with different source and dest ip
+        if tran_type == "ipv4-other":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IP(src="192.168.0.%d", dst="192.168.0.%d")], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv4-tcp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IP(src="192.168.0.%d", dst="192.168.0.%d")/TCP(sport=1024,dport=1024)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv4-udp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IP(src="192.168.0.%d", dst="192.168.0.%d")/UDP(sport=1024,dport=1024)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv4-sctp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s")/IP(src="192.168.0.%d", dst="192.168.0.%d")/SCTP(sport=1024,dport=1025,tag=1)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+                packet = r'sendp([Ether(dst="%s")/IP(src="192.168.0.%d", dst="192.168.0.%d")/SCTP(sport=1025,dport=1024,tag=1)], iface="%s")' % (
+                    mac, i + 2, i + 1, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "l2_payload":
+            for i in range(16):
+                packet = r'sendp([Ether(src="00:00:00:00:00:%02d",dst="%s")], iface="%s")' % (
+                    i + 1, mac, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+
+        elif tran_type == "ipv6-other":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IPv6(src="3ffe:2501:200:1fff::%d", dst="3ffe:2501:200:3::%d")], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv6-tcp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IPv6(src="3ffe:2501:200:1fff::%d", dst="3ffe:2501:200:3::%d")/TCP(sport=1024,dport=1024)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv6-udp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s", src="02:00:00:00:00:00")/IPv6(src="3ffe:2501:200:1fff::%d", dst="3ffe:2501:200:3::%d")/UDP(sport=1024,dport=1024)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+        elif tran_type == "ipv6-sctp":
+            for i in range(16):
+                packet = r'sendp([Ether(dst="%s")/IPv6(src="3ffe:2501:200:1fff::%d", dst="3ffe:2501:200:3::%d")/SCTP(sport=1024,dport=1025,tag=1)], iface="%s")' % (
+                    mac, i + 1, i + 2, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+                packet = r'sendp([Ether(dst="%s")/IPv6(src="3ffe:2501:200:1fff::%d", dst="3ffe:2501:200:3::%d")/SCTP(sport=1025,dport=1024,tag=1)], iface="%s")' % (
+                    mac, i + 2, i + 1, itf)
+                self.tester.scapy_append(packet)
+                self.tester.scapy_execute()
+                time.sleep(.5)
+
+        else:
+            print "\ntran_type error!\n"
+
+        #out = self.vm_dut_0.send_expect("stop", "testpmd>")
+        out = self.vm_dut_0.get_session_output()
+        print '*******************************************'
+        print out
+        if  not reta_entries:
+           self.verify('RSS hash=' in out, 'rss faied')
+           return 
+        lines = out.split("\r\n")
+        out = ''
+        reta_line = {}
+
+        # collect the hash result of five tuple and the queue id
+        for line in lines:
+            line = line.strip()
+            if len(line) != 0 and line.startswith(("src=",)):
+                for item in line.split("-"):
+                    item = item.strip()
+                    if(item.startswith("RSS hash")):
+                        name, value = item.split("=", 1)
+                        print name + "-" + value
+
+                reta_line[name.strip()] = value.strip()
+                reta_lines.append(reta_line)
+                reta_line = {}
+            elif len(line) != 0 and line.strip().startswith("port "):
+                rexp = r"port (\d)/queue (\d{1,2}): received (\d) packets"
+                m = re.match(rexp, line.strip())
+                if m:
+                    reta_line["port"] = m.group(1)
+                    reta_line["queue"] = m.group(2)
+            elif len(line) != 0 and line.startswith("stop"):
+                break
+            else:
+                pass
+        self.verifyResult()
+             
+
+    def verifyResult(self):
+        """
+        Verify whether or not the result passes.
+        """
+
+        global reta_lines
+        result = []
+        dts.results_table_add_header(
+            ['packet index', 'hash value', 'hash index', 'queue id', 
+ 'actual queue id', 'pass '])
+
+        i = 0
+        for tmp_reta_line in reta_lines:
+            status = "false"
+            if(self.nic in ["niantic", "redrockcanyou"]):
+                # compute the hash result of five tuple into the 7 LSBs value.
+                hash_index = int(tmp_reta_line["RSS hash"], 16) % 128
+            else:
+                # compute the hash result of five tuple into the 7 LSBs value.
+                hash_index = int(tmp_reta_line["RSS hash"], 16) % 512
+            if(reta_entries[hash_index] == int(tmp_reta_line["queue"])):
+                status = "true"
+                result.insert(i, 0)
+            else:
+                status = "fail"
+                result.insert(i, 1)
+            dts.results_table_add_row(
+                [i, tmp_reta_line["RSS hash"], hash_index, reta_entries[hash_index], tmp_reta_line["queue"], status])
+            i = i + 1
+
+        dts.results_table_print()
+        reta_lines = []
+        self.verify(sum(result) == 0, "the reta update function 
+ failed!")
+
+    def set_up_all(self):
+        """
+        Run at the start of each test suite.
+        """
+
+        #self.verify(
+        #    self.nic in ["niantic", "fortville_eagle", "fortville_spirit", "fortville_spirit_single"],
+        #    "NIC Unsupported: " + str(self.nic))

Can you remove above lines if it is useless?

+        self.dut_ports = self.dut.get_ports(self.nic)
+        self.ports_socket = self.dut.get_numa_id(self.dut_ports[0])
+        self.verify(len(self.dut_ports) >= 1, "Not enough ports 
+ available")
+
+        self.vm0 = None
+        self.setup_1pf_1vf_1vm_env_flag = 0
+        self.setup_1pf_1vf_1vm_env(driver='')
+        #self.vm0_dut_ports = self.vm_dut_0.get_ports('any')
+  
+        #self.pmdout = PmdOutput(self.dut)
+

Can you remove above lines if it is useless?

+    def set_up(self):
+        """
+        Run before each test case.
+        """
+        #self.vm0_testpmd = PmdOutput(self.vm_dut_0)
+        #self.vm0_testpmd.start_testpmd(VM_CORES_MASK,param = "--rxq=4 --rxq=4")

Can you remove above lines if it is useless?

+        pass
+    def setup_1pf_1vf_1vm_env(self, driver='default'):
+        
+        self.used_dut_port_0 = self.dut_ports[0]
+        self.dut.generate_sriov_vfs_by_port(self.used_dut_port_0, 1, driver=driver)
+        self.sriov_vfs_port_0 = 
+ self.dut.ports_info[self.used_dut_port_0]['vfs_port']
+
+        try:
+
+            for port in self.sriov_vfs_port_0:
+                port.bind_driver('pci-stub')
+
+            time.sleep(1)
+            vf0_prot = {'opt_host': self.sriov_vfs_port_0[0].pci}
+
+            if driver == 'igb_uio':
+                # start testpmd without the two VFs on the host
+                self.host_testpmd = PmdOutput(self.dut)
+                eal_param = '-b %(vf0)s -b %(vf1)s' % {'vf0': self.sriov_vfs_port_0[0].pci}
+                self.host_testpmd.start_testpmd("1S/2C/2T", 
+ eal_param=eal_param)
+
+            # set up VM0 ENV
+            self.vm0 = QEMUKvm(self.dut, 'vm0', 'vf_rss')
+            self.vm0.set_vm_device(driver='pci-assign', **vf0_prot)
+
+            self.vm_dut_0 = self.vm0.start()
+            self.vm0_testpmd = PmdOutput(self.vm_dut_0)
+            if self.vm_dut_0 is None:
+                raise Exception("Set up VM0 ENV failed!")
+
+            self.setup_1pf_1vf_1vm_env_flag = 1
+        except Exception as e:
+            self.destroy_1pf_1vf_1vm_env()
+            raise Exception(e)
+
+    def destroy_1pf_1vf_1vm_env(self):
+        if getattr(self, 'vm0', None):
+            #destroy testpmd in vm0
+            #self.vm0_testpmd.execute_cmd('stop')

Can you remove above lines if it is useless?

+            self.vm0_testpmd.execute_cmd('quit', '# ')
+            self.vm0_testpmd = None
+            self.vm0_dut_ports = None
+            #destroy vm0
+            self.vm0.stop()
+            self.vm0 = None
+
+        if getattr(self, 'host_testpmd', None):
+            self.host_testpmd.execute_cmd('quit', '# ')
+            self.host_testpmd = None
+
+        if getattr(self, 'used_dut_port_0', None):
+            self.dut.destroy_sriov_vfs_by_port(self.used_dut_port_0)
+            port = self.dut.ports_info[self.used_dut_port_0]['port']
+            port.bind_driver()
+            self.used_dut_port_0 = None
+
+        for port_id in self.dut_ports:
+            port = self.dut.ports_info[port_id]['port']
+            port.bind_driver()
+
+        self.setup_1pf_2vf_1vm_env_flag = 0
+
+    def test_vf_pmdrss_reta(self):
+        
+        # niantic kernel host driver not support this case
+        if self.nic is 'niantic' and not self.host_testpmd:
+            return
+        vm0dutPorts = self.vm_dut_0.get_ports('any')
+        localPort = self.tester.get_local_port(vm0dutPorts[0])
+        itf = self.tester.get_interface(localPort)
+        iptypes = ['IPV4']
+
+        self.vm_dut_0.kill_all()
+
+        # test with different rss queues
+        for queue in testQueues:
+
+            self.vm0_testpmd.start_testpmd(
+                "all", "--rxq=%d --txq=%d" % (queue, queue), 
+ socket=self.ports_socket)
+
+            for iptype in iptypes:
+                self.vm_dut_0.send_expect("set verbose 8", "testpmd> ")
+                self.vm_dut_0.send_expect("set fwd rxonly", "testpmd> ")
+                self.vm_dut_0.send_expect(
+                    "set nbcore %d" % (queue + 1), "testpmd> ")
+
+                # configure the reta with specific mappings.
+                if(self.nic in ["niantic", "redrockcanyou"]):
+                    for i in range(128):
+                        reta_entries.insert(i, random.randint(0, queue - 1))
+                        self.vm_dut_0.send_expect(
+                            "port config 0 rss reta (%d,%d)" % (i, reta_entries[i]), "testpmd> ")
+                else:
+                    for i in range(512):
+                        reta_entries.insert(i, random.randint(0, queue - 1))
+                        self.vm_dut_0.send_expect(
+                            "port config 0 rss reta (%d,%d)" % (i, 
+ reta_entries[i]), "testpmd> ")
+
+                self.send_packet(itf, iptype)
+
+            self.vm_dut_0.send_expect("quit", "# ", 30)
+    def test_vf_pmdrss(self): 
+        vm0dutPorts = self.vm_dut_0.get_ports('any')
+        localPort = self.tester.get_local_port(vm0dutPorts[0])
+        itf = self.tester.get_interface(localPort)
+        iptypes = {'ipv4-sctp':'ip',
+                   'ipv4-other':'ip',
+                   'ipv4-udp':'udp',
+                   'ipv4-tcp':'tcp',
+                   'ipv4-sctp':'sctp',
+                   'ipv6-other':'ip',
+                   'ipv6-udp':'udp',
+                   'ipv6-tcp':'tcp',
+                   'ipv6-sctp':'sctp',
+                 #  'l2_payload':'ether'

Why did not include l2_payload type ?

+                  }
+
+        self.vm_dut_0.kill_all()
+
+        # test with different rss queues
+        for queue in testQueues:
+
+            self.vm0_testpmd.start_testpmd(
+                "all", "--rxq=%d --txq=%d" % (queue, queue), 
+ socket=self.ports_socket)
+
+            for iptype,rsstype in iptypes.items():
+                self.vm_dut_0.send_expect("set verbose 8", "testpmd> ")
+                self.vm_dut_0.send_expect("set fwd rxonly", "testpmd> ")
+                self.vm_dut_0.send_expect("port config all rss %s" % rsstype, "testpmd> ")
+                self.vm_dut_0.send_expect(
+                    "set nbcore %d" % (queue + 1), "testpmd> ")
+
+                self.send_packet(itf, iptype)
+
+            self.vm_dut_0.send_expect("quit", "# ", 30)
+    def tear_down(self):
+        """
+        Run after each test case.
+        """
+        pass
+
+    def tear_down_all(self):
+        """
+        Run after each test suite.
+        """
+        #self.vm_dut_0.kill_all()

Can you remove above lines if it is useless?

Best Regards
Yulong Pei

+        self.destroy_1pf_1vf_1vm_env()
--
1.9.3

Re: [dts] [PATCH V1 3/3] add vf rss test script

[Xu, HuilongX]

Hi yulong,
Thanks a lot for you review my patch, pls check my comments.
thanks

> -----Original Message-----
> From: Pei, Yulong
> Sent: Thursday, January 14, 2016 7:52 PM
> To: Xu, HuilongX; dts@dpdk.org
> Subject: RE: [dts] [PATCH V1 3/3] add vf rss test script
> 
> Hi,
> 
> -----Original Message-----
> From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of xu,huilong
> Sent: Tuesday, January 12, 2016 6:53 PM
> To: dts@dpdk.org
> Subject: [dts] [PATCH V1 3/3] add vf rss test script
> 
> Signed-off-by: xu,huilong <huilongx.xu@intel.com>
> ---
>  tests/TestSuite_vf_rss.py | 382
> ++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 382 insertions(+)
>  create mode 100644 tests/TestSuite_vf_rss.py
> 
> diff --git a/tests/TestSuite_vf_rss.py b/tests/TestSuite_vf_rss.py new
> file mode 100644 index 0000000..a59516d
> --- /dev/null
> +++ b/tests/TestSuite_vf_rss.py
> @@ -0,0 +1,382 @@
> +# BSD LICENSE
> +#
> +# Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
> +# All rights reserved.
> +#
> +# Redistribution and use in source and binary forms, with or without #
> +modification, are permitted provided that the following conditions #
> +are met:
> +#
> +#   * Redistributions of source code must retain the above copyright
> +#     notice, this list of conditions and the following disclaimer.
> +#   * Redistributions in binary form must reproduce the above copyright
> +#     notice, this list of conditions and the following disclaimer in
> +#     the documentation and/or other materials provided with the
> +#     distribution.
> +#   * Neither the name of Intel Corporation nor the names of its
> +#     contributors may be used to endorse or promote products derived
> +#     from this software without specific prior written permission.
> +#
> +# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS #
> +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT #
> +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR #
> +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT #
> +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, #
> +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT #
> +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, #
> +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY #
> +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT #
> +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE #
> +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
> +
> +"""
> +DPDK Test suite.
> +Test RSS reta (redirection table) update function.
> +"""
> +import time
> +import random
> +import re
> +import dts
> +testQueues = [4]
> +reta_entries = []
> +reta_lines = []
> +
> +# Use scapy to send packets with different source and dest ip.
> +# and collect the hash result of five tuple and the queue id.
> +from test_case import TestCase
> +from pmd_output import PmdOutput
> +from qemu_kvm import QEMUKvm
> +
> +class TestVfRss(TestCase):
> +    def send_packet(self, itf, tran_type):
> +        """
> +        Sends packets.
> +        """
> +        global reta_lines
> +        reta_lines = []
> +        self.tester.scapy_foreground()
> +        self.tester.scapy_append('sys.path.append("./")')
> +        self.tester.scapy_append('from sctp import *')
> +        self.vm_dut_0.send_expect("start", "testpmd>")
> +        mac = self.vm_dut_0.get_mac_address(0)
> +        # send packet with different source and dest ip
> +        if tran_type == "ipv4-other":
> +            for i in range(16):
> +                packet = r'sendp([Ether(dst="%s",
> src="02:00:00:00:00:00")/IP(src="192.168.0.%d", dst="192.168.0.%d")],
> iface="%s")' % (
> +                    mac, i + 1, i + 2, itf)
> +                self.tester.scapy_append(packet)
> +                self.tester.scapy_execute()
> +                time.sleep(.5)
> +        elif tran_type == "ipv4-tcp":
> +            for i in range(16):
> +                packet = r'sendp([Ether(dst="%s",
> src="02:00:00:00:00:00")/IP(src="192.168.0.%d",
> dst="192.168.0.%d")/TCP(sport=1024,dport=1024)], iface="%s")' % (
> +                    mac, i + 1, i + 2, itf)
> +                self.tester.scapy_append(packet)
> +                self.tester.scapy_execute()
> +                time.sleep(.5)
> +        elif tran_type == "ipv4-udp":
> +            for i in range(16):
> +                packet = r'sendp([Ether(dst="%s",
> src="02:00:00:00:00:00")/IP(src="192.168.0.%d",
> dst="192.168.0.%d")/UDP(sport=1024,dport=1024)], iface="%s")' % (
> +                    mac, i + 1, i + 2, itf)
> +                self.tester.scapy_append(packet)
> +                self.tester.scapy_execute()
> +                time.sleep(.5)
> +        elif tran_type == "ipv4-sctp":
> +            for i in range(16):
> +                packet = r'sendp([Ether(dst="%s")/IP(src="192.168.0.%d",
> dst="192.168.0.%d")/SCTP(sport=1024,dport=1025,tag=1)], iface="%s")' % (
> +                    mac, i + 1, i + 2, itf)
> +                self.tester.scapy_append(packet)
> +                self.tester.scapy_execute()
> +                time.sleep(.5)
> +                packet = r'sendp([Ether(dst="%s")/IP(src="192.168.0.%d",
> dst="192.168.0.%d")/SCTP(sport=1025,dport=1024,tag=1)], iface="%s")' % (
> +                    mac, i + 2, i + 1, itf)
> +                self.tester.scapy_append(packet)
> +                self.tester.scapy_execute()
> +                time.sleep(.5)
> +        elif tran_type == "l2_payload":
> +            for i in range(16):
> +                packet =
> r'sendp([Ether(src="00:00:00:00:00:%02d",dst="%s")], iface="%s")' % (
> +                    i + 1, mac, itf)
> +                self.tester.scapy_append(packet)
> +                self.tester.scapy_execute()
> +                time.sleep(.5)
> +
> +        elif tran_type == "ipv6-other":
> +            for i in range(16):
> +                packet = r'sendp([Ether(dst="%s",
> src="02:00:00:00:00:00")/IPv6(src="3ffe:2501:200:1fff::%d",
> dst="3ffe:2501:200:3::%d")], iface="%s")' % (
> +                    mac, i + 1, i + 2, itf)
> +                self.tester.scapy_append(packet)
> +                self.tester.scapy_execute()
> +                time.sleep(.5)
> +        elif tran_type == "ipv6-tcp":
> +            for i in range(16):
> +                packet = r'sendp([Ether(dst="%s",
> src="02:00:00:00:00:00")/IPv6(src="3ffe:2501:200:1fff::%d",
> dst="3ffe:2501:200:3::%d")/TCP(sport=1024,dport=1024)], iface="%s")' % (
> +                    mac, i + 1, i + 2, itf)
> +                self.tester.scapy_append(packet)
> +                self.tester.scapy_execute()
> +                time.sleep(.5)
> +        elif tran_type == "ipv6-udp":
> +            for i in range(16):
> +                packet = r'sendp([Ether(dst="%s",
> src="02:00:00:00:00:00")/IPv6(src="3ffe:2501:200:1fff::%d",
> dst="3ffe:2501:200:3::%d")/UDP(sport=1024,dport=1024)], iface="%s")' % (
> +                    mac, i + 1, i + 2, itf)
> +                self.tester.scapy_append(packet)
> +                self.tester.scapy_execute()
> +                time.sleep(.5)
> +        elif tran_type == "ipv6-sctp":
> +            for i in range(16):
> +                packet =
> r'sendp([Ether(dst="%s")/IPv6(src="3ffe:2501:200:1fff::%d",
> dst="3ffe:2501:200:3::%d")/SCTP(sport=1024,dport=1025,tag=1)],
> iface="%s")' % (
> +                    mac, i + 1, i + 2, itf)
> +                self.tester.scapy_append(packet)
> +                self.tester.scapy_execute()
> +                time.sleep(.5)
> +                packet =
> r'sendp([Ether(dst="%s")/IPv6(src="3ffe:2501:200:1fff::%d",
> dst="3ffe:2501:200:3::%d")/SCTP(sport=1025,dport=1024,tag=1)],
> iface="%s")' % (
> +                    mac, i + 2, i + 1, itf)
> +                self.tester.scapy_append(packet)
> +                self.tester.scapy_execute()
> +                time.sleep(.5)
> +
> +        else:
> +            print "\ntran_type error!\n"
> +
> +        #out = self.vm_dut_0.send_expect("stop", "testpmd>")
> +        out = self.vm_dut_0.get_session_output()
> +        print '*******************************************'
> +        print out
> +        if  not reta_entries:
> +           self.verify('RSS hash=' in out, 'rss faied')
> +           return
> +        lines = out.split("\r\n")
> +        out = ''
> +        reta_line = {}
> +
> +        # collect the hash result of five tuple and the queue id
> +        for line in lines:
> +            line = line.strip()
> +            if len(line) != 0 and line.startswith(("src=",)):
> +                for item in line.split("-"):
> +                    item = item.strip()
> +                    if(item.startswith("RSS hash")):
> +                        name, value = item.split("=", 1)
> +                        print name + "-" + value
> +
> +                reta_line[name.strip()] = value.strip()
> +                reta_lines.append(reta_line)
> +                reta_line = {}
> +            elif len(line) != 0 and line.strip().startswith("port "):
> +                rexp = r"port (\d)/queue (\d{1,2}): received (\d)
> packets"
> +                m = re.match(rexp, line.strip())
> +                if m:
> +                    reta_line["port"] = m.group(1)
> +                    reta_line["queue"] = m.group(2)
> +            elif len(line) != 0 and line.startswith("stop"):
> +                break
> +            else:
> +                pass
> +        self.verifyResult()
> +
> +
> +    def verifyResult(self):
> +        """
> +        Verify whether or not the result passes.
> +        """
> +
> +        global reta_lines
> +        result = []
> +        dts.results_table_add_header(
> +            ['packet index', 'hash value', 'hash index', 'queue id',
> + 'actual queue id', 'pass '])
> +
> +        i = 0
> +        for tmp_reta_line in reta_lines:
> +            status = "false"
> +            if(self.nic in ["niantic", "redrockcanyou"]):
> +                # compute the hash result of five tuple into the 7 LSBs
> value.
> +                hash_index = int(tmp_reta_line["RSS hash"], 16) % 128
> +            else:
> +                # compute the hash result of five tuple into the 7 LSBs
> value.
> +                hash_index = int(tmp_reta_line["RSS hash"], 16) % 512
> +            if(reta_entries[hash_index] == int(tmp_reta_line["queue"])):
> +                status = "true"
> +                result.insert(i, 0)
> +            else:
> +                status = "fail"
> +                result.insert(i, 1)
> +            dts.results_table_add_row(
> +                [i, tmp_reta_line["RSS hash"], hash_index,
> reta_entries[hash_index], tmp_reta_line["queue"], status])
> +            i = i + 1
> +
> +        dts.results_table_print()
> +        reta_lines = []
> +        self.verify(sum(result) == 0, "the reta update function
> + failed!")
> +
> +    def set_up_all(self):
> +        """
> +        Run at the start of each test suite.
> +        """
> +
> +        #self.verify(
> +        #    self.nic in ["niantic", "fortville_eagle",
> "fortville_spirit", "fortville_spirit_single"],
> +        #    "NIC Unsupported: " + str(self.nic))
> 
> Can you remove above lines if it is useless?
> 
> +        self.dut_ports = self.dut.get_ports(self.nic)
> +        self.ports_socket = self.dut.get_numa_id(self.dut_ports[0])
> +        self.verify(len(self.dut_ports) >= 1, "Not enough ports
> + available")
> +
> +        self.vm0 = None
> +        self.setup_1pf_1vf_1vm_env_flag = 0
> +        self.setup_1pf_1vf_1vm_env(driver='')
> +        #self.vm0_dut_ports = self.vm_dut_0.get_ports('any')
> +
> +        #self.pmdout = PmdOutput(self.dut)
> +
> 
> Can you remove above lines if it is useless?
> yes, I will remove it  next version.
> +    def set_up(self):
> +        """
> +        Run before each test case.
> +        """
> +        #self.vm0_testpmd = PmdOutput(self.vm_dut_0)
> +        #self.vm0_testpmd.start_testpmd(VM_CORES_MASK,param = "--rxq=4 --
> rxq=4")
> 
> Can you remove above lines if it is useless?
> yes, I will remove it next version
> +        pass
> +    def setup_1pf_1vf_1vm_env(self, driver='default'):
> +
> +        self.used_dut_port_0 = self.dut_ports[0]
> +        self.dut.generate_sriov_vfs_by_port(self.used_dut_port_0, 1,
> driver=driver)
> +        self.sriov_vfs_port_0 =
> + self.dut.ports_info[self.used_dut_port_0]['vfs_port']
> +
> +        try:
> +
> +            for port in self.sriov_vfs_port_0:
> +                port.bind_driver('pci-stub')
> +
> +            time.sleep(1)
> +            vf0_prot = {'opt_host': self.sriov_vfs_port_0[0].pci}
> +
> +            if driver == 'igb_uio':
> +                # start testpmd without the two VFs on the host
> +                self.host_testpmd = PmdOutput(self.dut)
> +                eal_param = '-b %(vf0)s -b %(vf1)s' % {'vf0':
> self.sriov_vfs_port_0[0].pci}
> +                self.host_testpmd.start_testpmd("1S/2C/2T",
> + eal_param=eal_param)
> +
> +            # set up VM0 ENV
> +            self.vm0 = QEMUKvm(self.dut, 'vm0', 'vf_rss')
> +            self.vm0.set_vm_device(driver='pci-assign', **vf0_prot)
> +
> +            self.vm_dut_0 = self.vm0.start()
> +            self.vm0_testpmd = PmdOutput(self.vm_dut_0)
> +            if self.vm_dut_0 is None:
> +                raise Exception("Set up VM0 ENV failed!")
> +
> +            self.setup_1pf_1vf_1vm_env_flag = 1
> +        except Exception as e:
> +            self.destroy_1pf_1vf_1vm_env()
> +            raise Exception(e)
> +
> +    def destroy_1pf_1vf_1vm_env(self):
> +        if getattr(self, 'vm0', None):
> +            #destroy testpmd in vm0
> +            #self.vm0_testpmd.execute_cmd('stop')
> 
> Can you remove above lines if it is useless?
> yes I will remove it next version
> +            self.vm0_testpmd.execute_cmd('quit', '# ')
> +            self.vm0_testpmd = None
> +            self.vm0_dut_ports = None
> +            #destroy vm0
> +            self.vm0.stop()
> +            self.vm0 = None
> +
> +        if getattr(self, 'host_testpmd', None):
> +            self.host_testpmd.execute_cmd('quit', '# ')
> +            self.host_testpmd = None
> +
> +        if getattr(self, 'used_dut_port_0', None):
> +            self.dut.destroy_sriov_vfs_by_port(self.used_dut_port_0)
> +            port = self.dut.ports_info[self.used_dut_port_0]['port']
> +            port.bind_driver()
> +            self.used_dut_port_0 = None
> +
> +        for port_id in self.dut_ports:
> +            port = self.dut.ports_info[port_id]['port']
> +            port.bind_driver()
> +
> +        self.setup_1pf_2vf_1vm_env_flag = 0
> +
> +    def test_vf_pmdrss_reta(self):
> +
> +        # niantic kernel host driver not support this case
> +        if self.nic is 'niantic' and not self.host_testpmd:
> +            return
> +        vm0dutPorts = self.vm_dut_0.get_ports('any')
> +        localPort = self.tester.get_local_port(vm0dutPorts[0])
> +        itf = self.tester.get_interface(localPort)
> +        iptypes = ['IPV4']
> +
> +        self.vm_dut_0.kill_all()
> +
> +        # test with different rss queues
> +        for queue in testQueues:
> +
> +            self.vm0_testpmd.start_testpmd(
> +                "all", "--rxq=%d --txq=%d" % (queue, queue),
> + socket=self.ports_socket)
> +
> +            for iptype in iptypes:
> +                self.vm_dut_0.send_expect("set verbose 8", "testpmd> ")
> +                self.vm_dut_0.send_expect("set fwd rxonly", "testpmd> ")
> +                self.vm_dut_0.send_expect(
> +                    "set nbcore %d" % (queue + 1), "testpmd> ")
> +
> +                # configure the reta with specific mappings.
> +                if(self.nic in ["niantic", "redrockcanyou"]):
> +                    for i in range(128):
> +                        reta_entries.insert(i, random.randint(0, queue -
> 1))
> +                        self.vm_dut_0.send_expect(
> +                            "port config 0 rss reta (%d,%d)" % (i,
> reta_entries[i]), "testpmd> ")
> +                else:
> +                    for i in range(512):
> +                        reta_entries.insert(i, random.randint(0, queue -
> 1))
> +                        self.vm_dut_0.send_expect(
> +                            "port config 0 rss reta (%d,%d)" % (i,
> + reta_entries[i]), "testpmd> ")
> +
> +                self.send_packet(itf, iptype)
> +
> +            self.vm_dut_0.send_expect("quit", "# ", 30)
> +    def test_vf_pmdrss(self):
> +        vm0dutPorts = self.vm_dut_0.get_ports('any')
> +        localPort = self.tester.get_local_port(vm0dutPorts[0])
> +        itf = self.tester.get_interface(localPort)
> +        iptypes = {'ipv4-sctp':'ip',
> +                   'ipv4-other':'ip',
> +                   'ipv4-udp':'udp',
> +                   'ipv4-tcp':'tcp',
> +                   'ipv4-sctp':'sctp',
> +                   'ipv6-other':'ip',
> +                   'ipv6-udp':'udp',
> +                   'ipv6-tcp':'tcp',
> +                   'ipv6-sctp':'sctp',
> +                 #  'l2_payload':'ether'
> 
> Why did not include l2_payload type ?
> l2_payload hash type, it looks not support on niantic vf NIC, I will check it with development.
> +                  }
> +
> +        self.vm_dut_0.kill_all()
> +
> +        # test with different rss queues
> +        for queue in testQueues:
> +
> +            self.vm0_testpmd.start_testpmd(
> +                "all", "--rxq=%d --txq=%d" % (queue, queue),
> + socket=self.ports_socket)
> +
> +            for iptype,rsstype in iptypes.items():
> +                self.vm_dut_0.send_expect("set verbose 8", "testpmd> ")
> +                self.vm_dut_0.send_expect("set fwd rxonly", "testpmd> ")
> +                self.vm_dut_0.send_expect("port config all rss %s" %
> rsstype, "testpmd> ")
> +                self.vm_dut_0.send_expect(
> +                    "set nbcore %d" % (queue + 1), "testpmd> ")
> +
> +                self.send_packet(itf, iptype)
> +
> +            self.vm_dut_0.send_expect("quit", "# ", 30)
> +    def tear_down(self):
> +        """
> +        Run after each test case.
> +        """
> +        pass
> +
> +    def tear_down_all(self):
> +        """
> +        Run after each test suite.
> +        """
> +        #self.vm_dut_0.kill_all()
> 
> Can you remove above lines if it is useless?
> yes I will remove it next version
> Best Regards
> Yulong Pei
> 
> +        self.destroy_1pf_1vf_1vm_env()
> --
> 1.9.3