From: "Lin, Xueqin" <xueqin.lin@intel.com>
To: "Tang, Yaqi" <yaqi.tang@intel.com>, "dts@dpdk.org" <dts@dpdk.org>
Cc: "Tang, Yaqi" <yaqi.tang@intel.com>
Subject: RE: [dts][PATCH V3] test_plans/dcf_lifecycle_test_plan: add test plan for CVL DCF enable device reset API
Date: Tue, 18 Jan 2022 05:17:39 +0000 [thread overview]
Message-ID: <SJ0PR11MB5599B5C1C9CB205F942EB38694589@SJ0PR11MB5599.namprd11.prod.outlook.com> (raw)
In-Reply-To: <20220118105354.180850-1-yaqi.tang@intel.com>
Thanks, few comments as below.
> -----Original Message-----
> From: Yaqi Tang <yaqi.tang@intel.com>
> Sent: Tuesday, January 18, 2022 6:54 PM
> To: dts@dpdk.org
> Cc: Tang, Yaqi <yaqi.tang@intel.com>
> Subject: [dts][PATCH V3] test_plans/dcf_lifecycle_test_plan: add test plan for
> CVL DCF enable device reset API
>
> DCF PMD need to support rte_eth_dev_reset.The reason is that when a DCF
> instance be killed, all the flow rules still exist in hardware. when DCF gets to
> reconnect, It has already lost the flow context, and if the application wants to
> create new rules, they may fail because FW reports that the rules already
> exist.The current workaround assumes that the user can turn off / turn on VF
> trust mode, so kernel PF will reset DCF by removing all old rules.The
> rte_eth_dev_reset API provides a more elegant way for the application to reset
> DCF through the commmad "port reset all" when reconnecting.
>
> Signed-off-by: Yaqi Tang <yaqi.tang@intel.com>
> Tested-by: Yaqi Tang < yaqi.tang@intel.com>
> ---
> test_plans/dcf_lifecycle_test_plan.rst | 754 +++++++++++++++++++++++++
> 1 file changed, 754 insertions(+)
>
> diff --git a/test_plans/dcf_lifecycle_test_plan.rst
> b/test_plans/dcf_lifecycle_test_plan.rst
> index 4c010e76..a8701338 100644
> --- a/test_plans/dcf_lifecycle_test_plan.rst
> +++ b/test_plans/dcf_lifecycle_test_plan.rst
> @@ -1071,3 +1071,757 @@ TC34: ACL DCF mode is active, add ACL filters by
> way of host based configuration
>
> # ethtool -N enp24s0f1 flow-type tcp4 src-ip 192.168.10.0 m 0.255.255.255
> dst-port 8000 m 0x00ff action -1
> Added rule with ID 15871
> +
> +===============================
> +CVL DCF enable device reset API
> +===============================
> +
> +Description
> +===========
> +The test plan includes ten test cases, such as
> +two_testpmd_dcf_reset_port, two_testpmd_dcf_reset_device,
> two_testpmd_dcf_reset_port_detach, two_testpmd_dcf_reset_mtu,
> two_testpmd_dcf_reset_mac, one_testpmd_dcf_reset_port,
> one_testpmd_dcf_reset_device, one_testpmd_dcf_reset_port_detach,
> one_testpmd_dcf_reset_mtu, one_testpmd_dcf_reset_mac.
> +
> +DCF PMD need to support rte_eth_dev_reset.The reason is that when a DCF
> +instance be killed, all the flow rules still exist in hardware. when
> +DCF gets to reconnect, It has already lost the flow context, and if the
> +application wants to create new rules, they may fail because FW reports that
> the rules already exist.The current workaround assumes that the user can turn
> off / turn on VF trust mode, so kernel PF will reset DCF by removing all old
> rules.The rte_eth_dev_reset API provides a more elegant way for the
> application to reset DCF through the command "port reset all" when
> reconnecting.
> +
> +Prerequisites
> +=============
> +
> +Hardware
> +--------
> +Supportted NICs: columbiaville_25g/columbiaville_100g
> +
> +Software
> +--------
> +dpdk: http://dpdk.org/git/dpdk
> +scapy: http://www.secdev.org/projects/scapy/
> +
> +General Set Up
> +--------------
> +1. Copy specific ice package to /lib/firmware/updates/intel/ice/ddp/ice.pkg,
> + then load driver::
> +
> + # cp ice_comms-1.3.31.0.pkg /lib/firmware/updates/intel/ice/ddp/ice.pkg
> + # rmmod ice
> + # insmod ice.ko
> +
> +2. Compile DPDK::
> +
> + # CC=gcc meson --werror -Denable_kmods=True -Dlibdir=lib --default-
> library=static x86_64-native-linuxapp-gcc
> + # ninja -C x86_64-native-linuxapp-gcc -j 110
> +
> +3. Get the pci device id and interface of DUT and tester::
> +
> + # ./usertools/dpdk-devbind.py -s
> +
> + 0000:3b:00.0 'Ethernet Controller E810-C for SFP 1593' if=ens785f0 drv=ice
> unused=vfio-pci
> + 0000:af:00.0 'Ethernet Controller XXV710 for 25GbE SFP28 158b'
> + if=ens260f0 drv=i40e unused=vfio-pci
> +
Above Prerequisites is common, not special for reset cases, suggest putting on top of the test plan if need.
> +4. Generate 4 VFs on PF0::
> +
> + # echo 4 > /sys/bus/pci/devices/0000:3b:00.0/sriov_numvfs
> +
> +5. Get VF pci device id of DUT.
> +
> + # ./usertools/dpdk-devbind.py -s
> +
> + 0000:3b:01.0 'Ethernet Adaptive Virtual Function 1889' if=ens785f0v0
> drv=iavf unused=vfio-pci
> + 0000:3b:01.1 'Ethernet Adaptive Virtual Function 1889'
> + if=ens785f0v1 drv=iavf unused=vfio-pci
> +
> +6. Set VF0 as trust::
> +
> + # ip link set ens785f0 vf 0 trust on
> +
> +7. Bind the VF ports of DUT to dpdk::
> +
> + # ./usertools/dpdk-devbind.py -b vfio-pci 3b:01.0 3b:01.1
> +
> +Test case
> +=========
> +
> +Common Steps
> +------------
> +The common steps launch two testpmds and kill DCF process, then relaunch
> two testpmds.
> +
> +1. Launch the userland ``testpmd`` application on DUT as follows::
> +
> + # ./x86_64-native-linuxapp-gcc/app/dpdk-testpmd -v -c 0x6 -n 4 -a
> + 0000:3b:01.0,cap=dcf --file-prefix=dcf -- -i
Could you follow template format for the whole test plan?
> +
> +2. Launch another ``testpmd`` application on the VF1 of DUT as follows::
> +
> + # ./x86_64-native-linuxapp-gcc/app/dpdk-testpmd -v -c 0x18 -n 4 -a
> + 0000:3b:01.1 --file-prefix=vf -- -i
> +
> +3. Set verbose in VF testpmd::
> +
> + testpmd> set verbose 1
> +
> +4. Set fwd engine and start in VF testpmd::
> +
> + testpmd> set fwd mac
> + testpmd> start
> +
> +5. Validate a switch rule to VF1 in DCF testpmd::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +6. Create a switch rule to VF1 in DCF testpmd and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +7. Send a matched packet in scapy to VF1, check the VF1 of DUT can receive
> the packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst='C6:44:32:0A:EC:E1')/IP(src='192.168.0.2',
> + dst='192.168.0.3')/("X"*64)],iface="ens260f0",count=1)
Could you follow template format for the whole test plan, same as below?
> +
> +the VF1 of DUT::
> +
> + testpmd> port 0/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 -
> + length=98 - nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN
> + L4_NONFRAG - sw ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 -
> + Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> + RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +8. Check the process ID and kill DCF process::
> +
> + # ps -ef |grep testpmd
> + # kill -9 #####
> +
> +9. Relaunch the userland ``testpmd`` application on DUT as follows::
> +
> + # ./x86_64-native-linuxapp-gcc/app/dpdk-testpmd -v -c 0x6 -n 4 -a
> + 0000:3b:01.0,cap=dcf --file-prefix=dcf -- -i
> +
> +10. Send a matched packet in scapy to VF1, DCF flow rule is still valid and
> check the VF1 of DUT can receive the packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +the VF1 of DUT::
The VF1 of DUT::
> +
> + testpmd> port 0/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 -
> + length=98 - nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN
> + L4_NONFRAG - sw ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 -
> + Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> + RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +Test Case 1: two_testpmd_dcf_reset_port
> +---------------------------------------
> +The test case resets DCF by resetting the port, and DCF should clean up all old
> rules.
> +
> +Test Steps
> +~~~~~~~~~~
> +1. Reset port in DCF testpmd::
> +
> + testpmd> stop
> + testpmd> port stop all
> + testpmd> port reset all
> + testpmd> port start all
> + testpmd> start
> +
> +2. Validate a switch rule to VF1 in DCF testpmd::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +3. Create a switch rule to VF1 in DCF testpmd and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +4. Send a matched packet in scapy to VF1, and check the VF1 of DUT can
> receive the packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +the VF1 of DUT::
> +
> + testpmd> port 0/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 -
> + length=98 - nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN
> + L4_NONFRAG - sw ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 -
> + Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> + RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +Test Case 2: two_testpmd_dcf_reset_device
> +-----------------------------------------
> +The test case resets DCF by resetting the device, and DCF should clean up all
> old rules.
> +
> +Test Steps
> +~~~~~~~~~~
> +1. Reset DCF device::
> +
> + # echo 1 > /sys/bus/pci/devices/0000:3b:01.0/reset
> +
> +2. Reset port in DCF testpmd::
> +
> + testpmd> stop
> + testpmd> port stop all
> + testpmd> port reset all
> + testpmd> port start all
> + testpmd> start
> +
> +3. Validate a switch rule to VF1 in DCF testpmd::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +4. Create a switch rule to VF1 in DCF testpmd and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +5. Send a matched packet in scapy to VF1, and check the VF1 of DUT can
> receive the packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +the VF1 of DUT::
> +
> + testpmd> port 0/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 -
> + length=98 - nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN
> + L4_NONFRAG - sw ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 -
> + Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> + RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +Test Case 3: two_testpmd_dcf_reset_port_detach
> +----------------------------------------------
> +The test case resets DCF by detaching the port, and DCF should clean up all
> old rules.
> +
> +Test Steps
> +~~~~~~~~~~
> +1. Reset DCF device::
> +
> + # echo 1 > /sys/bus/pci/devices/0000:3b:01.0/reset
> +
> +2. Detach and reset port in DCF testpmd::
> +
> + testpmd> stop
> + testpmd> port stop 0
> + testpmd> port detach 0
> + testpmd> port attach 3b:01.0,cap=dcf
> + testpmd> port reset 0
> + testpmd> port start 0
> + testpmd> start
> +
> +3. Validate a switch rule to VF1 in DCF testpmd::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +4. Create a switch rule to VF1 in DCF testpmd and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +5. Send a matched packet in scapy to VF1, and check the VF1 of DUT can
> receive the packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +the VF1 of DUT::
> +
> + testpmd> port 0/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 - length=98 -
> nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN L4_NONFRAG - sw
> ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 - Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +Test Case 4: two_testpmd_dcf_reset_mtu
> +--------------------------------------
> +The test case resets DCF by reseting the mtu, and DCF should clean up all old
> rules.
> +
> +Test Steps
> +~~~~~~~~~~
> +1. Modify the value of mtu::
> +
> + # ifconfig ens785f0 mtu 3000
> +
> +2. Reset port in DCF testpmd::
> +
> + testpmd> stop
> + testpmd> port stop all
> + testpmd> port reset all
> + testpmd> port start all
> + testpmd> start
> +
> +3. Validate a switch rule to VF1 in DCF testpmd::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +4. Create a switch rule to VF1 in DCF testpmd and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +5. Send a matched packet in scapy to VF1, and check the VF1 of DUT can
> receive the packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +the VF1 of DUT::
> +
> + testpmd> port 0/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 -
> + length=98 - nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN
> + L4_NONFRAG - sw ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 -
> + Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> + RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +Test Case 5: two_testpmd_dcf_reset_mac
> +--------------------------------------
> +The test case resets DCF by reseting mac addr, and DCF should clean up all old
> rules.
> +
> +Test Steps
> +~~~~~~~~~~
> +1. Reset VF0 by set mac addr::
> +
> + # ip link set ens785f0 vf 0 mac 00:01:02:03:04:05
> +
> +2. Reset port in DCF testpmd::
> +
> + testpmd> stop
> + testpmd> port stop all
> + testpmd> port reset all
> + testpmd> port start all
> + testpmd> start
> +
> +3. Validate a switch rule to VF1 in DCF testpmd::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +4. Create a switch rule to VF1 in DCF testpmd and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +5. Send a matched packet in scapy to VF1, and check the VF1 of DUT can
> receive the packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +the VF1 of DUT::
> +
> + testpmd> port 0/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 -
> + length=98 - nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN
> + L4_NONFRAG - sw ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 -
> + Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> + RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +Common Steps
> +------------
> +The common steps launch one testpmd and kill DCF process, then relaunch
> one testpmd.
> +
> +1. Launch the userland ``testpmd`` application on DUT as follows::
> +
> + # ./x86_64-native-linuxapp-gcc/app/dpdk-testpmd -v -c 0x6 -n 4 -a
> + 3b:01.0,cap=dcf -a 3b:01.1 --file-prefix=dcf -- -i
> +
> +2. Set verbose::
> +
> + testpmd> set verbose 1
> +
> +3. Set fwd engine and start::
> +
> + testpmd> set fwd mac
> + testpmd> start
> +
> +4. Validate a switch rule to VF1::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +5. Create a switch rule to VF1 and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +6. Send a matched packet in scapy to VF1, check DUT can receive the packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +DUT::
> +
> + testpmd> port 1/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 -
> + length=98 - nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN
> + L4_NONFRAG - sw ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 -
> + Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> + RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +7. Check the process ID and kill DCF process::
> +
> + # ps -ef |grep testpmd
> + # kill -9 #####
> +
> +8. Relaunch the userland ``testpmd`` application on DUT as follows::
> +
> + # ./x86_64-native-linuxapp-gcc/app/dpdk-testpmd -v -c 0x6 -n 4 -a
> + 3b:01.0,cap=dcf -a 3b:01.1 --file-prefix=dcf -- -i
> +
> +9. Set verbose::
> +
> + testpmd> set verbose 1
> +
> +10. Set fwd engine and start::
> +
> + testpmd> set fwd mac
> + testpmd> start
> +
> +11. Send a matched packet in scapy to VF1, DCF flow rule is cleared and check
> DUT can't receive the packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +DUT::
> +
> + testpmd>
> +
> +Test Case 6: one_testpmd_dcf_reset_port
> +---------------------------------------
> +The test case resets DCF by resetting the port, and DCF should clean up all old
> rules.
> +
> +Test Steps
> +~~~~~~~~~~
> +1. Reset port in testpmd::
> +
> + testpmd> stop
> + testpmd> port stop all
> + testpmd> port reset all
> + testpmd> port start all
> + testpmd> start
> +
> +2. Validate a switch rule to VF1::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +3. Create a switch rule to VF1 and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +4. Send a matched packet in scapy to VF1, and check DUT can receive the
> packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +DUT::
> +
> + testpmd> port 1/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 - length=98 -
> nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN L4_NONFRAG - sw
> ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 - Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +Test Case 7: one_testpmd_dcf_reset_device
> +-----------------------------------------
> +The test case resets DCF by resetting the device, and DCF should clean up all
> old rules.
> +
> +Test Steps
> +~~~~~~~~~~
> +1. Reset DCF device::
> +
> + # echo 1 > /sys/bus/pci/devices/0000:3b:01.0/reset
> +
> +2. Reset port in testpmd::
> +
> + testpmd> stop
> + testpmd> port stop all
> + testpmd> port reset all
> + testpmd> port start all
> + testpmd> start
> +
> +3. Validate a switch rule to VF1::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +4. Create a switch rule to VF1 and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +5. Send a matched packet in scapy to VF1, and check DUT can receive the
> packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +DUT::
> +
> + testpmd> port 1/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 -
> + length=98 - nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN
> + L4_NONFRAG - sw ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 -
> + Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> + RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +Test Case 8: one_testpmd_dcf_reset_port_detach
> +----------------------------------------------
> +The test case resets DCF by detaching the port, and DCF should clean up all
> old rules.
> +
> +Test Steps
> +~~~~~~~~~~
> +1. Reset DCF device::
> +
> + # echo 1 > /sys/bus/pci/devices/0000:3b:01.0/reset
> +
> +2. Detach and reset port in DCF testpmd::
> +
> + testpmd> stop
> + testpmd> port stop 0
> + testpmd> port detach 0
> + testpmd> port attach 3b:01.0,cap=dcf
> + testpmd> port reset 0
> + testpmd> port start 0
> + testpmd> start
> +
> +3. Validate a switch rule to VF1::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +4. Create a switch rule to VF1 and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +5. Send a matched packet in scapy to VF1, and check DUT can receive the
> packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +DUT::
> +
> + testpmd> port 1/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 - length=98 -
> nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN L4_NONFRAG - sw
> ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 - Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +Test Case 9: one_testpmd_dcf_reset_mtu
> +--------------------------------------
> +The test case resets DCF by resetting the mtu, and DCF should clean up all old
> rules.
> +
> +Test Steps
> +~~~~~~~~~~
> +1. Modify the value of mtu::
> +
> + # ifconfig ens785f0 mtu 3000
> +
> +2. Reset port in DCF testpmd::
> +
> + testpmd> stop
> + testpmd> port stop all
> + testpmd> port reset all
> + testpmd> port start all
> + testpmd> start
> +
> +3. Validate a switch rule to VF1::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +4. Create a switch rule to VF1 and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +5. Send a matched packet in scapy to VF1, and check DUT can receive the
> packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +DUT::
> +
> + testpmd> port 1/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 - length=98 -
> nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN L4_NONFRAG - sw
> ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 - Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> +
> +Test Case 10: one_testpmd_dcf_reset_mac
> +---------------------------------------
> +The test case resets DCF by resetting mac addr, and DCF should clean up all
> old rules.
> +
> +Test Steps
> +~~~~~~~~~~
> +1. Reset VF0 by set mac addr::
> +
> + # ip link set ens785f0 vf 0 mac 00:01:02:03:04:05
> +
> +2. Reset port in DCF testpmd::
> +
> + testpmd> stop
> + testpmd> port stop all
> + testpmd> port reset all
> + testpmd> port start all
> + testpmd> start
> +
> +3. Validate a switch rule to VF1::
> +
> + testpmd> flow validate 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule validated
> +
> +4. Create a switch rule to VF1 and list rules::
> +
> + testpmd> flow create 0 priority 0 ingress pattern eth / ipv4 src
> + is 192.168.0.2 dst is 192.168.0.3 / end actions vf id 1 / end
> +
> + Get the message::
> +
> + Flow rule #0 created
> +
> + Check the flow list::
> +
> + testpmd> flow list 0
> +
> + ID Group Prio Attr Rule
> + 0 0 0 i-- ETH IPV4 => VF
> +
> +5. Send a matched packet in scapy to VF1, and check DUT can receive the
> packet.
> +
> +Tester::
> +
> + >>> sendp([Ether(src="00:11:22:33:44:55",
> + dst="C6:44:32:0A:EC:E1")/IP(src="192.168.0.2",
> + dst="192.168.0.3")/("X"*64)],iface="ens260f0",count=1,inter=0,verbose=
> + False)
> +
> +DUT::
> +
> + testpmd> port 1/queue 0: received 1 packets
> + src=00:11:22:33:44:55 - dst=C6:44:32:0A:EC:E1 - type=0x0800 - length=98 -
> nb_segs=1 - hw ptype: L2_ETHER L3_IPV4_EXT_UNKNOWN L4_NONFRAG - sw
> ptype: L2_ETHER L3_IPV4 - l2_len=14 - l3_len=20 - Receive queue=0x0
> + ol_flags: RTE_MBUF_F_RX_L4_CKSUM_GOOD
> RTE_MBUF_F_RX_IP_CKSUM_GOOD
> RTE_MBUF_F_RX_OUTER_L4_CKSUM_UNKNOWN
> --
> 2.25.1
next prev parent reply other threads:[~2022-01-18 5:17 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-01-18 10:53 Yaqi Tang
2022-01-18 5:17 ` Lin, Xueqin [this message]
2022-01-19 3:39 ` Lin, Xueqin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=SJ0PR11MB5599B5C1C9CB205F942EB38694589@SJ0PR11MB5599.namprd11.prod.outlook.com \
--to=xueqin.lin@intel.com \
--cc=dts@dpdk.org \
--cc=yaqi.tang@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).