From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by dpdk.org (Postfix) with ESMTP id 47F4F568F for ; Wed, 2 Sep 2015 08:07:33 +0200 (CEST) Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga102.jf.intel.com with ESMTP; 01 Sep 2015 23:07:32 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.17,451,1437462000"; d="scan'208";a="636872052" Received: from pgsmsx107.gar.corp.intel.com ([10.221.44.105]) by orsmga003.jf.intel.com with ESMTP; 01 Sep 2015 23:07:31 -0700 Received: from shsmsx101.ccr.corp.intel.com (10.239.4.153) by PGSMSX107.gar.corp.intel.com (10.221.44.105) with Microsoft SMTP Server (TLS) id 14.3.224.2; Wed, 2 Sep 2015 14:07:28 +0800 Received: from shsmsx103.ccr.corp.intel.com ([169.254.4.248]) by SHSMSX101.ccr.corp.intel.com ([169.254.1.171]) with mapi id 14.03.0224.002; Wed, 2 Sep 2015 14:07:27 +0800 From: "Liu, Yong" To: "Xu, HuilongX" , "dts@dpdk.org" Thread-Topic: [dts] [PATCH V1 1/4] add kni test plan Thread-Index: AQHQ5Sy8gS7+hn8CT0mAVRC+WHQJj54ouLNg Date: Wed, 2 Sep 2015 06:07:27 +0000 Message-ID: <86228AFD5BCD8E4EBFD2B90117B5E81E10EE0785@SHSMSX103.ccr.corp.intel.com> References: <1441163321-29448-1-git-send-email-huilongx.xu@intel.com> In-Reply-To: <1441163321-29448-1-git-send-email-huilongx.xu@intel.com> Accept-Language: zh-CN, en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [10.239.127.40] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Cc: "Tu, LijuanX A" Subject: Re: [dts] [PATCH V1 1/4] add kni test plan X-BeenThere: dts@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: test suite reviews and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 02 Sep 2015 06:07:34 -0000 Thanks huilong. Some comments in the plan. > -----Original Message----- > From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of Huilong Xu > Sent: Wednesday, September 02, 2015 11:09 AM > To: dts@dpdk.org > Cc: Tu, LijuanX A > Subject: [dts] [PATCH V1 1/4] add kni test plan >=20 > From: "huilongx,xu" >=20 > Signed-off-by: huilongx,xu > --- > test_plans/kni_test_plan.rst | 608 > +++++++++++++++++++++++++++++++++++++++++++ > 1 file changed, 608 insertions(+) > create mode 100644 test_plans/kni_test_plan.rst >=20 > diff --git a/test_plans/kni_test_plan.rst b/test_plans/kni_test_plan.rst > new file mode 100644 > index 0000000..94a0b76 > --- /dev/null > +++ b/test_plans/kni_test_plan.rst > @@ -0,0 +1,608 @@ > +.. Copyright (c) <2010,2011>, Intel Corporation Copyright to 2015:) > + 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. > + > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > +Kernel NIC Interface > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > + > +Description > +----------- > + > +This document provides the plan for testing the Kernel NIC Interface > +application with support of rte_kni kernel module. > +Kernel NIC Interface is a DPDK alternative solution to the existing linu= x > +tun-tap interface for the exception path. Kernel NIC Interface allows th= e > +standard Linux net tools(ethtool/ifconfig/tcpdump) to facilitate managin= g > the > +DPDK port. At the same time, it add an interface with the kernel net > stack. > +The test supports Multi-Thread KNI. > + > +The ``rte_kni`` kernel module can be installed by a ``lo_mode`` paramete= r. > + > +loopback disabled:: > + > + insmod rte_kni.ko > + insmod rte_kni.ko "lo_mode=3Dlo_mode_none" > + insmod rte_kni.ko "lo_mode=3Dunsupported string" > + > +loopback mode=3Dlo_mode_ring enabled:: > + > + insmod rte_kni.ko "lo_mode=3Dlo_mode_ring" > + > +loopback mode=3Dlo_mode_ring_skb enabled:: > + > + insmod rte_kni.ko "lo_mode=3Dlo_mode_ring_skb" > + Can you give us more information about the parameters? It can help us known= what's the difference of lo_mode_ring and lo_mode_ring_sk. > +The ``rte_kni`` kernel module can also be installed by a ``kthread_mode`= ` > +parameter. This parameter is ``single`` by default. > + > +kthread single:: > + > + insmod rte_kni.ko > + insmod rte_kni.ko "kthread_mode=3Dsingle" > + > +kthread multiple:: > + > + insmod rte_kni.ko > + insmod rte_kni.ko "kthread_mode=3Dmultiple" > + Need more description of kthread_mode parameter. > + > +The ``kni`` application is run with EAL parameters and parameters for th= e > +application itself. For details about the EAL parameters, see the > relevant > +DPDK **Getting Started Guide**. This application supports two parameters > for > +itself. > + > + - ``--config=3D"(port id, rx lcore, tx lcore, kthread lcore, kthread > lcore, ...)"``: > + Port and cores selection. Kernel threads are ignored if > ``kthread_mode`` > + is not ``multiple``. > + > +ports cores:: > + > + e.g.: > + > + --config=3D"(0,1,2),(1,3,4)" No kernel thread > specified. > + --config=3D"(0,1,2,21),(1,3,4,23)" One kernel thread in u= se. > + --config=3D"(0,1,2,21,22),(1,3,4,23,25) Two kernel threads in = use. > + > + - ``-P``: Promiscuous mode. This is off by default. > + > +Prerequisites > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > + > +Support igb_uio and vfio driver, if used vfio, kernel need 3.6+ and > enable vt-d in bios. > +When used vfio , used "modprobe vfio" and "modprobe vfio-pci" insmod > vfiod driver, then used > +"./tools/dpdk_nic_bind.py --bind=3Dvfio-pci device_bus_id" to bind vfio > driver to test driver. > + > + > +The DUT has at least 2 DPDK supported IXGBE NIC ports. > + I think two NICs supported by DPDK are enough. Not only IXGBE NIC.=20 > +The DUT has to be able to install rte_kni kernel module and launch kni > +application with a default configuration (This configuration may change > form a > +system to another):: > + > + rmmod rte_kni > + rmmod igb_uio > + insmod ./x86_64-default-linuxapp-gcc/kmod/igb_uio.ko > + insmod ./x86_64-default-linuxapp-gcc/kmod/rte_kni.ko > + ./examples/kni/build/app/kni -c 0xa0001e -n 4 -- -P -p 0x3 -- > config=3D"(0,1,2,21),(1,3,4,23)" & > + What's meaning of lcore 21, please use "S0/C1/T1" to clarify that. > + > +Test Case: ifconfig testing > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D > + > +Launch the KNI application. Assume that ``port 2 and 3`` are used to thi= s > +application. Cores 1 and 2 are used to read from NIC, cores 2 and 4 are > used > +to write to NIC, threads 21 and 23 are used by the kernel. > + > +As the kernel module is installed using ``"kthread_mode=3Dsingle"`` the > core > +affinity is set using ``taskset``:: > + > + ./build/app/kni -c 0xa0001e -n 4 -- -P -p 0xc -- > config=3D"(2,1,2,21),(3,3,4,23)" > + This command use port2 and port3, thus mean this case need at least four po= rts? If so, please change test prerequisites.=20 > + > +Verify whether the interface has been added:: > + > + ifconfig -a > + > + > +If the application is launched successfully, it will add two interfaces > in > +kernel net stack named ``vEth2_0``, ``vEth3_0``. > + > +Interface name start with ``vEth`` followed by the port number and an > +additional incremental number depending on the number of kernel threads:= : > + > + vEth2_0: flags=3D4098 mtu 1500 > + ether 00:00:00:00:00:00 txqueuelen 1000 (Ethernet) > + RX packets 14 bytes 2098 (2.0 KiB) > + RX errors 0 dropped 10 overruns 0 frame 0 > + TX packets 0 bytes 0 (0.0 B) > + TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 > + > + vEth3_0: flags=3D4098 mtu 1500 > + ether 00:00:00:00:00:00 txqueuelen 1000 (Ethernet) > + RX packets 13 bytes 1756 (1.7 KiB) > + RX errors 0 dropped 10 overruns 0 frame 0 > + TX packets 0 bytes 0 (0.0 B) > + TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 > + > + > + > +Verify whether ifconfig can set Kernel NIC Interface up:: > + > + ifconfig vEth2_0 up > + > +Now ``vEth2_0`` is up and has IPv6 address:: > + > + vEth2_0: flags=3D4163 mtu 1500 > + inet6 fe80::92e2:baff:fe37:92f8 prefixlen 64 scopeid > 0x20 > + ether 90:e2:ba:37:92:f8 txqueuelen 1000 (Ethernet) > + RX packets 30 bytes 4611 (4.5 KiB) > + RX errors 0 dropped 21 overruns 0 frame 0 > + TX packets 6 bytes 468 (468.0 B) > + TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 > + > + > +Verify whether ifconfig can add an ipv6 address:: > + > + ifconfig vEth2_0 add fe80::1 > + > +``vEth2_0`` has added ipv6 address:: > + > + 29: vEth2_0: mtu 1500 qlen 1000 > + inet6 fe80::1/128 scope link > + valid_lft forever preferred_lft forever > + inet6 fe80::92e2:baff:fe37:92f8/64 scope link > + valid_lft forever preferred_lft forever > + > + > +Delete the IPv6 address:: > + > + ifconfig vEth2_0 del fe80::1 > + > +The port deletes it:: > + > + 29: vEth2_0: mtu 1500 qlen 1000 > + inet6 fe80::92e2:baff:fe37:92f8/64 scope link > + valid_lft forever preferred_lft forever > + > +Set MTU parameter:: > + > + ifconfig vEth2_0 mtu 1300 > + > +``vEth2_0`` changes the mtu parameter:: > + > + 29: vEth2_0: mtu 1300 qdisc > pfifo_fast state UNKNOWN mode DEFAULT qlen 1000 > + link/ether 90:e2:ba:37:92:f8 brd ff:ff:ff:ff:ff:ff > + > +Verify whether ifconfig can set ip address:: > + > + ifconfig vEth2_0 192.168.2.1 netmask 255.255.255.192 > + ip -family inet address show dev vEth2_0 > + > +``vEth2_0`` has IP address and netmask now:: > + > + 29: vEth2_0: mtu 1300 qdisc > pfifo_fast state UNKNOWN qlen 1000 > + inet 192.168.2.1/26 brd 192.168.2.63 scope global vEth2_0 > + > +Verify whether ifconfig can set ``vEth2_0`` down:: > + > + ifconfig vEth2_0 down > + ifconfig vEth2_0 > + > +``vEth2_0`` is down and no ipv6 address:: > + > + vEth2_0: flags=3D4098 mtu 1300 > + inet 192.168.2.1 netmask 255.255.255.192 broadcast > 192.168.2.63 > + ether 90:e2:ba:37:92:f8 txqueuelen 1000 (Ethernet) > + RX packets 70 bytes 12373 (12.0 KiB) > + RX errors 0 dropped 43 overruns 0 frame 0 > + TX packets 25 bytes 4132 (4.0 KiB) > + TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 > + > + > +Repeat all the steps for interface ``vEth3_0`` > + > +Test Case: Ping and Ping6 testing > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D > + > +If the application is launched successfully, it will add two interfaces > in > +kernel net stack named ``vEth2_0``, ``vEth3_0``. > + > +Assume the link status of ``vEth2_0`` is up and set ip address is > ``192.168.2.1`` > +and ``vEth3_0`` is up and set ip address is ``192.168.3.1``. Verify the > +command ping:: > + > + ping -w 1 -I vEth2_0 192.168.2.1 > + > +it can receive all packets and no packet loss:: > + > + PING 192.168.2.1 (192.168.2.1) from 192.168.2.1 vEth2_0: 56(84) byte= s > of data. > + 64 bytes from 192.168.2.1: icmp_req=3D1 ttl=3D64 time=3D0.040 ms > + > + --- 192.168.2.1 ping statistics --- > + 1 packets transmitted, 1 received, 0% packet loss, time 0ms > + rtt min/avg/max/mdev =3D 0.040/0.040/0.040/0.000 ms > + > +Assume ``port A`` on tester is linked with ``port 2`` on DUT. Verify the > +command ping from tester:: > + > + ping -w 1 -I "port A" 192.168.2.1 > + > +it can receive all packets and no packet loss. > + > +Verify a wrong address:: > + > + ping -w 1 -I vEth2_0 192.168.0.123 > + > +no packets is received:: > + > + PING 192.168.0.123 (192.168.0.123) from 192.168.0.1 vEth2_0: 56(84) > bytes of data. > + > + --- 192.168.0.123 ping statistics --- > + 1 packets transmitted, 0 received, 100% packet loss, time 0ms > + > +Verify the command ping6:: > + > + ping6 -w 1 -I vEth2_0 "Eth2_0's ipv6 address" > + > +it can receive all packets and no packet loss:: > + > + PING fe80::92e2:baff:fe08:d6f0(fe80::92e2:baff:fe08:d6f0) from > fe80::92e2:baff:fe08:d6f0 vEth2_0: 56 data bytes > + 64 bytes from fe80::92e2:baff:fe08:d6f0: icmp_seq=3D1 ttl=3D64 time= =3D0.070 > ms > + > + --- fe80::92e2:baff:fe08:d6f0 ping statistics --- > + 1 packets transmitted, 1 received, 0% packet loss, time 0ms > + rtt min/avg/max/mdev =3D 0.070/0.070/0.070/0.000 ms > + > +Verify the command ping6 from tester:: > + > + ping6 -w 1 -I "port A" "Eth2_0's ipv6 address" > + > +it can receive all packets and no packet loss. > + > +Verify a wrong ipv6 address:: > + > + ping6 -w 1 -I vEth2_0 "random ipv6 address" > + > +no packets is received:: > + > + PING fe80::92e2:baff:fe08:d6f1(fe80::92e2:baff:fe08:d6f1) from > fe80::92e2:baff:fe08:d6f0 vEth2_0: 56 data bytes > + > + --- fe80::92e2:baff:fe08:d6f1 ping statistics --- > + 1 packets transmitted, 0 received, 100% packet loss, time 0ms > + > +Repeat all the steps for interface ``vEth3_0`` > + > +Test Case: Tcpdump testing > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D > + > +Assume ``port A and B`` on packet generator connects to NIC ``port 2 and > 3``. > +Trigger the packet generator of bursting packets from ``port A and B`, > then > +check if tcpdump can capture all packets. The packets should include > +``tcp`` packets, ``udp`` packets, ``icmp`` packets, ``ip`` packets, > +``ether+vlan tag+ip`` packets, ``ether`` packets. > + > +Verify whether tcpdump can capture packets:: > + > + tcpdump -i vEth2_0 > + tcpdump -i vEth3_0 > + > + > +Test Case: Ethtool testing > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D > + > +In this time, KNI can only support ethtool commands which is to get > information. > +So all belowing commands are to show information commands. > + > +Verify whether ethtool can show Kernel NIC Interface's standard > information:: > + > + ethtool vEth2_0 > + > +Verify whether ethtool can show Kernel NIC Interface's driver > information:: > + > + ethtool -i vEth2_0 > + > +Verify whether ethtool can show Kernel NIC Interface's statistics:: > + > + ethtool -S vEth2_0 > + > +Verify whether ethtool can show Kernel NIC Interface's pause parameters:= : > + > + ethtool -a vEth2_0 > + > +Verify whether ethtool can show Kernel NIC Interface's offload > parameters:: > + > + ethtool -k vEth2_0 > + > +Verify whether ethtool can show Kernel NIC Interface's RX/TX ring > parameters:: > + > + ethtool -g vEth2_0 > + > +Verify whether ethtool can show Kernel NIC Interface's Coalesce > parameters. > +It is not currently supported:: > + > + ethtool -c vEth2_0 > + > +Verify whether ethtool can show Kernel NIC Interface's MAC registers:: > + > + ethtool -d vEth2_0 > + > +Verify whether ethtool can show Kernel NIC Interface's EEPROM dump:: > + > + ethtool -e vEth2_0 > + > +Repeat all the steps for interface ``vEth3_0`` > + > +Test Case: Packets statistics testing > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > + > +Install the kernel module with loopback parameter > ``lo_mode=3Dlo_mode_ring_skb`` > +and launch the KNI application. > + > +Assume that ``port 2 and 3`` are used by this application:: > + > + rmmod kni > + insmod ./kmod/rte_kni.ko "lo_mode=3Dlo_mode_ring_skb" > + ./build/app/kni -c 0xff -n 3 -- -p 0xf -i 0xf -o 0xf0 > + > +Assume ``port A and B`` on tester connects to NIC ``port 2 and 3``. > + > +Get the RX packets count and TX packets count:: > + > + ifconfig vEth2_0 > + > +Send 5 packets from tester. And check whether both RX and TX packets of > +``vEth2_0`` have increased 5. > + > +Repeat for interface ``vEth3_0`` > + > +Test Case: Stress testing > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D > + > +Insert the rte_kni kernel module 50 times while changing the parameters. > +Iterate through lo_mode and kthread_mode values sequentially, include > wrong > +values. After each insertion check whether kni application can be > launched > +successfully. > + > +Insert the kernel module 50 times while changing randomly the parameters= . > +Iterate through lo_mode and kthread_mode values randomly, include wrong > +values. After each insertion check whether kni application can be > launched > +successfully:: > + > + rmmod rte_kni > + insmod ./kmod/rte_kni.ko > + ./build/app/kni -c 0xa0001e -n 4 -- -P -p 0xc -- > config=3D"(2,1,2,21),(3,3,4,23)" > + > + > +Using ``dmesg`` to check whether kernel module is loaded with the > specified > +parameters. Some permutations, those with wrong values, must fail to > +success. For permutations with valid parameter values, verify the > application can be > +successfully launched and then close the application using CTRL+C. > + > +Test Case: loopback mode performance testing > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > + > +Compare performance results for loopback mode using: > + > + - lo_mode: lo_mode_fifo and lo_mode_fifo_skb. > + - kthread_mode: single and multiple. > + - Number of ports: 1 and 2. > + - Number of virtual interfaces per port: 1 and 2 > + - Frame sizes: 64 and 256. > + - Cores combinations: > + > + - Different cores for Rx, Tx and Kernel. > + - Shared core between Rx and Kernel. > + - Shared cores between Rx and Tx. > + - Shared cores between Rx, Tx and Kernel. > + - Multiple cores for Kernel, implies multiple virtual interfaces > per port. > + > +:: > + > + insmod ./x86_64-default-linuxapp-gcc/kmod/igb_uio.ko > + insmod ./x86_64-default-linuxapp-gcc/kmod/rte_kni.ko kthread_mode parameters> > + ./examples/kni/build/app/kni -c -n 4 -- -P -p > --config=3D"" & > + > + > +At this point, the throughput is measured and recorded for the different > +frame sizes. After this, the application is closed using CTRL+C. > + > +The measurements are presented in a table format. > + > ++------------------+--------------+-------+-----------------+--------+--= - > -----+ > +| lo_mode | kthread_mode | Ports | Config | 64 | > 256 | > ++=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D+=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D+=3D=3D=3D=3D=3D=3D=3D+=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D+=3D=3D=3D=3D=3D=3D=3D=3D+=3D=3D=3D > =3D=3D=3D=3D=3D+ > +| | | | | | > | > ++------------------+--------------+-------+-----------------+--------+--= - > -----+ > + > + > +Test Case: bridge mode performance testing > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > + > +Compare performance results for bridge mode using: > + > + - kthread_mode: single and multiple. > + - Number of ports: 2 > + - Number of ports: 1 and 2. > + - Number of flows per port: 1 and 2 > + - Number of virtual interfaces per port: 1 and 2 > + - Frame size: 64. > + - Cores combinations: > + > + - Different cores for Rx, Tx and Kernel. > + - Shared core between Rx and Kernel. > + - Shared cores between Rx and Tx. > + - Shared cores between Rx, Tx and Kernel. > + - Multiple cores for Kernel, implies multiple virtual interfaces > per port. > + > +The application is launched and the bridge is setup using the commands > below:: > + > + insmod ./x86_64-default-linuxapp-gcc/kmod/rte_kni.ko parameter> > + ./build/app/kni -c -n 4 -- -P -p -- > config=3D"" & > + > + ifconfig vEth2_0 up > + ifconfig vEth3_0 up > + brctl addbr "br_kni" > + brctl addif br_kni vEth2_0 > + brctl addif br_kni vEth3_0 > + ifconfig br_kni up > + > + > +At this point, the throughput is measured and recorded. After this, the > +application is closed using CTRL+C and the bridge deleted:: > + > + ifconfig br_kni down > + brctl delbr br_kni > + > + > +The measurements are presented in a table format. > + > ++--------------+-------+-----------------------------+-------+ > +| kthread_mode | Flows | Config | 64 | > ++=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D+=3D=3D=3D=3D=3D=3D=3D+=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D+=3D=3D=3D=3D=3D=3D=3D+ > +| | | | | > ++--------------+-------+-----------------------------+-------+ > + Please extend the table with detail configurations with Rx/Tx/Kernel core s= ettings. >>From this table, we have no idea that what core combination the case is pro= filing. > +Test Case: bridge mode without KNI performance testing > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D > + > +Compare performance results for bridge mode using only Kernel bridge, no > DPDK > +support. Use: > + > + - Number of ports: 2 > + - Number of flows per port: 1 and 2 > + - Frame size: 64. > + > +Set up the interfaces and the bridge:: > + > + rmmod rte_kni > + ifconfig vEth2_0 up > + ifconfig vEth3_0 up > + brctl addbr "br1" > + brctl addif br1 vEth2_0 > + brctl addif br1 vEth3_0 > + ifconfig br1 up > + > + Some questions here, if rmmod kni module and kill kni process, how does the= virtual interface like vEth2_0 receive packets? > +At this point, the throughput is measured and recorded. After this, the > +application is closed using CTRL+C and the bridge deleted:: > + > + ifconfig br1 down > + brctl delbr br1 > + > + > +The measurements are presented in a table format. > + > ++-------+-------+ > +| Flows | 64 | > ++=3D=3D=3D=3D=3D=3D=3D+=3D=3D=3D=3D=3D=3D=3D+ > +| 1 | | > ++-------+-------+ > +| 2 | | > ++-------+-------+ > + Same comment in performance table. > +Test Case: routing mode performance testing > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > + > +Compare performance results for routing mode using: > + > + - kthread_mode: single and multiple. > + - Number of ports: 2 > + - Number of ports: 1 and 2. > + - Number of virtual interfaces per port: 1 and 2 > + - Frame size: 64 and 256. > + - Cores combinations: > + > + - Different cores for Rx, Tx and Kernel. > + - Shared core between Rx and Kernel. > + - Shared cores between Rx and Tx. > + - Shared cores between Rx, Tx and Kernel. > + - Multiple cores for Kernel, implies multiple virtual interfaces > per port. > + > +The application is launched and the bridge is setup using the commands > below:: > + > + echo 1 > /proc/sys/net/ipv4/ip_forward > + > + insmod ./x86_64-default-linuxapp-gcc/kmod/rte_kni.ko parameter> > + ./build/app/kni -c -n 4 -- -P -p -- > config=3D"" & > + > + ifconfig vEth2_0 192.170.2.1 > + ifconfig vEth2_0 192.170.3.1 > + route add -net 192.170.2.0 netmask 255.255.255.0 gw 192.170.2.1 > + route add -net 192.170.3.0 netmask 255.255.255.0 gw 192.170.3.1 > + arp -s 192.170.2.2 vEth2_0 > + arp -s 192.170.3.2 vEth3_0 > + > +At this point, the throughput is measured and recorded. After this, the > +application is closed using CTRL+C. > + > +The measurements are presented in a table format. > + > ++--------------+-------+-----------------------------+-------+-------+ > +| kthread_mode | Ports | Config | 64 | 256 | > ++=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D+=3D=3D=3D=3D=3D=3D=3D+=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D+=3D=3D=3D=3D=3D=3D=3D+=3D=3D=3D=3D=3D=3D=3D+ > +| | | | | | > ++--------------+-------+-----------------------------+-------+-------+ > + Same comment in performance table. > + > +Test Case: routing mode without KNI performance testing > +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D > + > +Compare performance results for routing mode using only Kernel, no DPDK > +support. Use: > + > + - Number of ports: 2 > + - Frame size: 64 and 256 > + > +Set up the interfaces and the bridge:: > + > + > + echo 1 > /proc/sys/net/ipv4/ip_forward > + rmmod rte_kni > + ifconfig vEth2_0 192.170.2.1 > + ifconfig vEth2_0 192.170.3.1 > + route add -net 192.170.2.0 netmask 255.255.255.0 gw 192.170.2.1 > + route add -net 192.170.3.0 netmask 255.255.255.0 gw 192.170.3.1 > + arp -s 192.170.2.2 vEth2_0 > + arp -s 192.170.3.2 vEth3_0 > + > +At this point, the throughput is measured and recorded. After this, the > +application is closed using CTRL+C. > + > +The measurements are presented in a table format. > + > ++-------+-------+-------+ > +| Ports | 64 | 256 | > ++=3D=3D=3D=3D=3D=3D=3D+=3D=3D=3D=3D=3D=3D=3D+=3D=3D=3D=3D=3D=3D=3D+ > +| 1 | | | > ++-------+-------+-------+ > +| 2 | | | > ++-------+-------+-------+ > -- Same comment in performance table. > 1.9.3