From: "Liu, Yong" <yong.liu@intel.com>
To: "Peng, Yuan" <yuan.peng@intel.com>, "dts@dpdk.org" <dts@dpdk.org>
Cc: "Peng, Yuan" <yuan.peng@intel.com>
Subject: Re: [dts] [PATCH v2] test_plans: Add rss_to_rte_flow_test_plan.rst
Date: Mon, 3 Sep 2018 02:00:17 +0000 [thread overview]
Message-ID: <86228AFD5BCD8E4EBFD2B90117B5E81E631316FE@SHSMSX103.ccr.corp.intel.com> (raw)
In-Reply-To: <1535684241-4573-1-git-send-email-yuan.peng@intel.com>
Yuan,
Few comments are inline.
Thanks,
Marvin
> -----Original Message-----
> From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of Peng Yuan
> Sent: Friday, August 31, 2018 10:57 AM
> To: dts@dpdk.org
> Cc: Peng, Yuan <yuan.peng@intel.com>
> Subject: [dts] [PATCH v2] test_plans: Add rss_to_rte_flow_test_plan.rst
>
> Add rss_to_rte_flow_test_plan.rst to test_plans
>
> Signed-off-by: Peng Yuan <yuan.peng@intel.com>
>
> diff --git a/test_plans/rss_to_rte_flow_test_plan.rst
> b/test_plans/rss_to_rte_flow_test_plan.rst
> new file mode 100644
> index 0000000..6b36d19
> --- /dev/null
> +++ b/test_plans/rss_to_rte_flow_test_plan.rst
> @@ -0,0 +1,338 @@
> +.. Copyright (c) <2018>, 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.
> +
> +====================
> +Move RSS to rte_flow
> +====================
> +Description
> +===========
> +
> + rte_flow has been actually defined to include RSS, but till now,
> + RSS is out of rte_flow. It was suggested to move existing RSS to rte_flow.
> + This can be better for users, and may save effort for CPK development.
> + RSS enabling: now, rte_flow API enabling RSS is support on igb/ixgbe/i40e.
> + RSS input set changing: now, rte flow API RSS input set is support on
> i40e
Please remove spaces in the beginning.
> +
> +Prerequisites
> +=============
> +
> +1. Hardware:
> + Fortville/82599/I350
> +
> +2. Software:
> + dpdk: http://dpdk.org/git/dpdk
> + scapy: http://www.secdev.org/projects/scapy/
> +
> +3. Bind the pf port to dpdk driver::
> +
> + ./usertools/dpdk-devbind.py -b igb_uio 05:00.0 05:00.1
> +
> +4. Configure eight packets ready sent to port 0::
> +
> + pkt1 =
> Ether(dst="00:00:00:00:01:00")/IP(src="10.0.0.1",dst="192.168.0.2")/SCTP(dpor
> t=80, sport=80)/("X"*48)
> + pkt2 =
> Ether(dst="00:00:00:00:01:00")/IP(src="10.0.0.1",dst="192.168.0.2")/UDP(dport
> =50, sport=50)/("X"*48)
> + pkt3 =
> Ether(dst="00:00:00:00:01:00")/IP(src="10.0.0.1",dst="192.168.0.3")/TCP(dport
> =50, sport=50)/("X"*48)
> + pkt4 =
> Ether(dst="00:00:00:00:01:00")/IP(src="10.0.0.1",dst="192.168.0.2")/("X"*48)
> + pkt5 =
> Ether(dst="00:00:00:00:01:00")/IPv6(src="2001::1",dst="2001::2")/SCTP(dport=8
> 0, sport=80)/("X"*48)
> + pkt6 =
> Ether(dst="00:00:00:00:01:00")/IPv6(src="2001::1",dst="2001::2")/UDP(dport=50,
> sport=50)/("X"*48)
> + pkt7 =
> Ether(dst="00:00:00:00:01:00")/IPv6(src="2001::1",dst="2001::2")/TCP(dport=50,
> sport=50)/("X"*48)
> + pkt8 =
> Ether(dst="00:00:00:00:01:00")/IPv6(src="2001::1",dst="2001::2")/("X"*48)
> +
> +Test case: enable rss
> +=====================
> +
> +1. Start the testpmd::
> +
> + ./testpmd -c 1ffff -n 4 -- -i --nb-cores=8 --rxq=4 --txq=4 --port-
> topology=chained
Yuan, please reduce the core used in this case if it is possible. Not every platform has more than 17 cpus.
> + testpmd> set fwd rxonly
> + testpmd> set verbose 1
> + testpmd> start
> +
> +2. Enable ipv4-udp, and set all the queues into RSS queue rule::
> +
> + testpmd> flow create 0 ingress pattern end actions rss types ipv4-udp
> end / end
> +
> +3. Send the ipv4-udp packets with different 5-tuple parameters.
> + All the packets are distributed to all the four queues.
> +
> +Test case: set valid and invalid parameter with all types
> +=========================================================
> +
> +1. Start the testpmd::
> +
> + ./testpmd -c 1ffff -n 4 -- -i --nb-cores=8 --rxq=16 --txq=16 --port-
> topology=chained
> + testpmd> set fwd rxonly
> + testpmd> set verbose 1
> + testpmd> start
> +
> +2. Set queue 0, 8 and 15 into RSS queue rule::
> +
> + testpmd> flow create 0 ingress pattern end actions rss queues 0 8 15 end
> / end
> +
> +3. Send the eight packets to port 0.
> + All the packets are distributed to queue 0/8/15.
> +
> +4. Set a second RSS queue rule::
> +
> + testpmd> flow create 0 ingress pattern end actions rss queues 3 end /
> end
> + Caught error type 2 (flow rule (handle)): Failed to create flow.
> +
> + There can't be more than one RSS queue rule.
> +
> +5. Reset the RSS queue rule::
> +
> + testpmd> flow flush 0
> + testpmd> flow create 0 ingress pattern end actions rss queues 3 end /
> end
> +
> + The rule is set successfully.
> + Send the eight packets to port 0, all of them are distributed to queue 3.
> +
> +6. Set a wrong parameter: queue ID is 16 ::
> +
> + testpmd> flow create 0 ingress pattern end actions rss queues 16 end /
> end
> + Caught error type 11 (specific action): cause: 0x7ffdb2bbe6e8, queue id >
> max number of queues
> +
> +7. Set all the queues to the rule::
> +
> + testpmd> flow create 0 ingress pattern end actions rss queues 0 1 2 3 4
> 5 6 7 8 9 10 11 12 13 14 15 end / end
> +
> + Send the eight packets to port 0, the packets may be distributed to any
> + of the queue 0-15.
> +
> +Notes: The max queue number may be different in different NIC types.
> +We can set different queue number in command line with different NIC types.
> +
> +Test case: Different packet types
> +=================================
> +
> +1. Start the testpmd::
> +
> + ./testpmd -c 1ffff -n 4 -- -i --nb-cores=8 --rxq=16 --txq=16 --port-
> topology=chained
> + testpmd> set fwd rxonly
> + testpmd> set verbose 1
> + testpmd> start
> +
> +2. Set queue 1, 8 and 15 into RSS queue rule::
> +
> + testpmd> flow create 0 ingress pattern end actions rss types udp ipv4-
> tcp ipv6-sctp ipv4-other end queues 1 8 15 end / end
> +
> +3. Send the eight packets to port 0.
> + pkt2/pkt3/pkt4/pkt5/pkt6 are distributed to queue 1/8/15.
> + Other packets are distributed to queue 0.
> +
> +4. Set a different packet type RSS queue rule::
> +
> + testpmd> flow create 0 ingress pattern end actions rss types ipv6-other
> end queues 3 end / end
> + Caught error type 2 (flow rule (handle)): Failed to create flow.
> +
> + There can't be more than one RSS queue rule.
> +
> +Test case: disable rss in command-line
> +======================================
"rss" should be capitalized.
> +
> +1. Start the testpmd::
> +
> + ./testpmd -c 1ffff -n 4 -- -i --nb-cores=8 --rxq=16 --txq=16 --disable-
> rss --port-topology=chained
> + testpmd> set fwd rxonly
> + testpmd> set verbose 1
> + testpmd> start
> +
> +2. Send the eight packets to port 0.
> + All the packets are distributed to queue 0.
> +
> +3. Set the RSS queue rule::
> +
> + testpmd> flow create 0 ingress pattern end actions rss types ipv6-tcp
> ipv4-udp sctp ipv6-other end queues 6 7 8 end / end
> +
> + Send the eight packets to port 0.
> + pkt1/pkt2/pkt5/pkt7/pkt8 are distributed to queue 6/7/8.
> + Other packets are distributed to queue 0.
> +
> +4. Clean the rule::
> +
> + testpmd> flow flush 0
> +
> + Send the eight packets to port 0.
> + All the packets are distributed to queue 0.
> +
> +Test case: Flow director rule and RSS rule combination
> +======================================================
> +
> +1. Start the testpmd::
> +
> + ./testpmd -c 1ffff -n 4 -- -i --nb-cores=8 --rxq=16 --txq=16 --pkt-
> filter-mode=perfect
> + testpmd> set fwd rxonly
> + testpmd> set verbose 1
> + testpmd> start
> +
> +2. Set a RSS queue rule::
> +
> + testpmd> flow create 0 ingress pattern end actions rss queues 6 7 8 end
> / end
> +
> + Send pkt2 to port 0, pkt2 is distributed to queue 6.
> +
> +3. Set a flow director rule::
> +
> + testpmd> flow create 0 ingress pattern eth / ipv4 src is 192.168.0.1 dst
> is 192.168.0.2 / udp src is 50 dst is 50 / end actions queue index 1 / end
> +
> + Send pkt2 to port 0, pkt2 is distributed to queue 1.
> +
> +4. Destroy the flow director rule::
> +
> + testpmd> flow destroy 0 rule 1
> +
> + Send pkt2 to port 0, pkt2 is distributed to queue 6 again.
> + So flow director rule is priority to RSS queue rule.
> +
> +Test case: Set queue region with RSS rule API
> +=============================================
> +
> +Notes: Queue region is only supported by fortville, so this case only can
> +be implemented with fortville.
> +
> +1. Start the testpmd::
> +
> + ./testpmd -c 1ffff -n 4 -- -i --nb-cores=16 --rxq=16 --txq=16 --port-
> topology=chained
> + testpmd> set fwd rxonly
> + testpmd> set verbose 1
> + testpmd> start
> +
> +2. Set a RSS queue rule::
> +
> + testpmd> flow create 0 ingress pattern end actions rss queues 7 8 10 11
> 12 14 15 end / end
> +
> +3. Send pkt to port 0::
> +
> + pkt1 = Ether(dst="00:00:00:00:01:00",
> src="52:00:00:00:00:00")/Dot1Q(prio=1) \
> + /IP(src="10.0.0.1",dst="192.168.0.2")/TCP(dport=80, sport=80)/("X"*48)
> + pkt2 = Ether(dst="00:00:00:00:01:00",
> src="52:00:00:00:00:00")/Dot1Q(prio=2) \
> + /IP(src="10.0.0.1",dst="192.168.0.2")/TCP(dport=80, sport=80)/("X"*48)
> + pkt3 = Ether(dst="00:00:00:00:01:00",
> src="52:00:00:00:00:00")/Dot1Q(prio=3) \
> + /IP(src="10.0.0.1",dst="192.168.0.2")/TCP(dport=80, sport=80)/("X"*48)
> +
> + They are all distributed to queue 8.
> +
> +4. Set three queue regions::
> +
> + testpmd> flow create 0 ingress pattern vlan tci is 0x2000 / end actions
> rss queues 7 8 end / end
> + testpmd> flow create 0 ingress pattern vlan tci is 0x4000 / end actions
> rss queues 11 12 end / end
> + testpmd> flow create 0 ingress pattern vlan tci is 0x6000 / end actions
> rss queues 15 end / end
> +
> + Send the 3 packets to port 0. They are distributed to queue 7/11/15.
> +
> +Notes: only rss rule API only support setting UP queue region,
> +not support setting pctype queue region.
> +
> +5. Flush the queue region::
> +
> + testpmd> flow list 0
> + ID Group Prio Attr Rule
> + 0 0 0 i- => RSS
> + 1 0 0 i- VLAN => RSS
> + 2 0 0 i- VLAN => RSS
> + 3 0 0 i- VLAN => RSS
> +
> + testpmd> flow destroy 0 rule 2
> + Flow rule #2 destroyed
> + testpmd> flow list 0
> + ID Group Prio Attr Rule
> + 0 0 0 i- => RSS
> + 1 0 0 i- VLAN => RSS
> + 3 0 0 i- VLAN => RSS
> +
> + Send the 3 packets to port 0. They are distributed to queue 0.
> + Destroy one queue region rule, all the rules become invalid.
> +
> +Test case: Set queue region with invalid parameter
> +==================================================
> +
> +1. Start the testpmd::
> +
> + ./testpmd -c 1ffff -n 4 -- -i --nb-cores=16 --rxq=16 --txq=16 --port-
> topology=chained
> + testpmd> set fwd rxonly
> + testpmd> set verbose 1
> + testpmd> start
> +
> +2. Set a RSS queue rule::
> +
> + testpmd> flow create 0 ingress pattern end actions rss queues 8 10 11 12
> 15 end / end
> +
> +3. Set invalid queue ID "9" to queue region::
> +
> + testpmd> flow create 0 ingress pattern vlan tci is 0x2000 / end actions
> rss queues 8 9 end / end
> + Caught error type 11 (specific action): cause: 0x7ffda008efe8, no valid
> queues
> +
> +4. Set discontinuous queue ID to queue region::
> +
> + testpmd> flow create 0 ingress pattern vlan tci is 0x2000 / end actions
> rss queues 8 10 end / end
> + Caught error type 11 (specific action): cause: 0x7ffda008efe8, no valid
> queues
> +
> +5. Set invalid queue number to queue region::
> +
> + testpmd> flow create 0 ingress pattern vlan tci is 0x4000 / end actions
> rss queues 10 11 12 end / end
> + i40e_flow_parse_rss_action(): The region sizes should be any of the
> following values: 1, 2, 4, 8, 16, 32, 64 as long as the total number of
> queues do not exceed the VSI allocation
> + Caught error type 2 (flow rule (handle)): Failed to create flow.
> +
> +Test case: Queue region(old API) and RSS rule combination
> +=========================================================
> +
> +Notes: Queue region is only supported by fortville, so this case only can
> +be implemented with fortville.
> +
> +1. Start the testpmd::
> +
> + ./testpmd -c 1ffff -n 4 -- -i --nb-cores=8 --rxq=16 --txq=16 --port-
> topology=chained
> + testpmd> port config all rss all
> + testpmd> set fwd rxonly
> + testpmd> set verbose 1
> + testpmd> start
> +
> +2. Set a queue region::
> +
> + testpmd> set port 0 queue-region region_id 0 queue_start_index 1
> queue_num 1
> + testpmd> set port 0 queue-region region_id 0 flowtype 31
> + testpmd> set port 0 queue-region flush on
> +
> + Send pkt2 to port 0. It is distributed to queue 1.
> +
> +3. Set a RSS queue rule::
> +
> + testpmd> flow create 0 ingress pattern end actions rss queues 6 7 end /
> end
> +
> + Send pkt2 to port 0. It is still distributed to queue 1.
> +
> +4. flush the queue region::
> +
> + testpmd> set port 0 queue-region flush off
> +
> + Send pkt2 to port 0. It is distributed to queue 7.
> + Queue region is priority to RSS queue rule.
> --
> 2.5.0
prev parent reply other threads:[~2018-09-03 2:00 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-08-31 2:57 Peng Yuan
2018-09-03 2:00 ` Liu, Yong [this message]
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=86228AFD5BCD8E4EBFD2B90117B5E81E631316FE@SHSMSX103.ccr.corp.intel.com \
--to=yong.liu@intel.com \
--cc=dts@dpdk.org \
--cc=yuan.peng@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).