* [dts] [PATCH V2 1/2] tests_plan/iavf_fdir: replace rules with less consumption TCAM rule
@ 2021-10-11 10:03 Jiale Song
2021-10-11 10:03 ` [dts] [PATCH V2 2/2] tests/iavf_fdir: create rules with less consumption of TCAM rule in suite Jiale Song
0 siblings, 1 reply; 3+ messages in thread
From: Jiale Song @ 2021-10-11 10:03 UTC (permalink / raw)
To: dts; +Cc: Jiale Song
the number of rules created is affected by the profile and TCAM. the maximum profile is 128 and
the maximum TCAM is 512. in order to test the maximum profile, need to create rules that consume
less TCAM to ensure that the profile will reach the maximum before TCAM exhausted
Signed-off-by: Jiale Song <songx.jiale@intel.com>
---
test_plans/iavf_fdir_test_plan.rst | 21 +++++++++++----------
1 file changed, 11 insertions(+), 10 deletions(-)
mode change 100644 => 100755 test_plans/iavf_fdir_test_plan.rst
diff --git a/test_plans/iavf_fdir_test_plan.rst b/test_plans/iavf_fdir_test_plan.rst
old mode 100644
new mode 100755
index 62d0ad1..0c0ce16
--- a/test_plans/iavf_fdir_test_plan.rst
+++ b/test_plans/iavf_fdir_test_plan.rst
@@ -4697,20 +4697,20 @@ Subcase 1: 128 profiles
2. create 8 rules with different patterns on each port::
- flow create 0 ingress pattern eth / ipv4 src is 192.168.0.20 dst is 192.168.0.21 / udp src is 22 dst is 23 / end actions queue index 1 / mark / end
- flow create 0 ingress pattern eth / ipv4 src is 192.168.0.20 dst is 192.168.0.21 / tcp src is 22 dst is 23 / end actions queue index 1 / mark / end
- flow create 0 ingress pattern eth / ipv4 src is 192.168.0.20 dst is 192.168.0.21 / sctp src is 22 dst is 23 / end actions queue index 1 / mark / end
- flow create 0 ingress pattern eth / ipv6 dst is CDCD:910A:2222:5498:8475:1111:3900:2020 src is 2001::2 / udp src is 22 dst is 23 / end actions queue index 1 / mark / end
- flow create 0 ingress pattern eth / ipv6 dst is CDCD:910A:2222:5498:8475:1111:3900:2020 src is 2001::2 / tcp src is 22 dst is 23 / end actions queue index 1 / mark / end
- flow create 0 ingress pattern eth / ipv6 dst is CDCD:910A:2222:5498:8475:1111:3900:2020 src is 2001::2 / sctp src is 22 dst is 23 / end actions queue index 1 / mark / end
- flow create 0 ingress pattern eth type is 0x8863 / end actions queue index 1 / mark id 1 / end
- flow create 0 ingress pattern eth / ipv4 / udp / pfcp s_field is 0 / end actions queue index 2 / end
+ flow create 0 ingress pattern eth / ipv4 / l2tpv3oip session_id is 1 / end actions queue index 1 / mark / end
+ flow create 0 ingress pattern eth / ipv6 / l2tpv3oip session_id is 2 / end actions queue index 1 / mark / end
+ flow create 0 ingress pattern eth / ipv4 / tcp / end actions queue index 2 / mark / end
+ flow create 0 ingress pattern eth / ipv6 / tcp / end actions queue index 2 / mark / end
+ flow create 0 ingress pattern eth / ipv4 / esp spi is 1 / end actions queue index 3 / mark / end
+ flow create 0 ingress pattern eth / ipv6 / esp spi is 2 / end actions queue index 3 / mark / end
+ flow create 0 ingress pattern eth / ipv4 / udp / pfcp s_field is 0 / end actions queue index 4 / mark id 1 / end
+ flow create 0 ingress pattern eth / ipv6 / udp / pfcp s_field is 1 / end actions queue index 4 / end
+
created successfully on port 0-12,
failed from rule 6 on port 13::
- testpmd> flow create 13 ingress pattern eth type is 0x8863 / end actions queue index 1 / mark id 1 / end
- eth
+ testpmd> flow create 0 ingress pattern eth / ipv4 / udp / pfcp s_field is 0 / end actions queue index 4 / mark id 1 / end
iavf_execute_vf_cmd(): No response or return failure (-5) for cmd 47
iavf_fdir_add(): fail to execute command OP_ADD_FDIR_FILTER
iavf_flow_create(): Failed to create flow
@@ -4739,6 +4739,7 @@ Subcase 1: 128 profiles
profile 0 and profile 1 are default profile for specific packet.
we design case with 2*100G card, so 110 profiles can be used for vf.
if we use 4*25G card, only 94 profiles can be used for vf.
+ if card is chapman beach 100g*2, one pf port equals a general 100g*2 card,so 118 profiles can be used for vf
4. send matched packets to vf 12,
the packets are redirected to the expected queue.
--
2.17.1
^ permalink raw reply [flat|nested] 3+ messages in thread
* [dts] [PATCH V2 2/2] tests/iavf_fdir: create rules with less consumption of TCAM rule in suite
2021-10-11 10:03 [dts] [PATCH V2 1/2] tests_plan/iavf_fdir: replace rules with less consumption TCAM rule Jiale Song
@ 2021-10-11 10:03 ` Jiale Song
2021-10-25 7:57 ` Tu, Lijuan
0 siblings, 1 reply; 3+ messages in thread
From: Jiale Song @ 2021-10-11 10:03 UTC (permalink / raw)
To: dts; +Cc: Jiale Song
modify the suite according to the testplan
Signed-off-by: Jiale Song <songx.jiale@intel.com>
---
tests/TestSuite_iavf_fdir.py | 66 +++++++++++++++++++++++-------------
1 file changed, 42 insertions(+), 24 deletions(-)
diff --git a/tests/TestSuite_iavf_fdir.py b/tests/TestSuite_iavf_fdir.py
index 7f8dace..44fcb95 100755
--- a/tests/TestSuite_iavf_fdir.py
+++ b/tests/TestSuite_iavf_fdir.py
@@ -8349,6 +8349,9 @@ class TestIAVFFdir(TestCase):
create same rules on pf and vf, no conflict
"""
self.dut.kill_all()
+ self.dut.send_expect("rmmod ice", "# ", 15)
+ self.dut.send_expect("modprobe ice", "# ", 15)
+ self.setup_2pf_4vf_env()
self.session_secondary = self.dut.new_session()
self.session_third = self.dut.new_session()
@@ -8500,6 +8503,9 @@ class TestIAVFFdir(TestCase):
create same input set but different action rules on pf and vf, no conflict.
"""
self.dut.kill_all()
+ self.dut.send_expect("rmmod ice", "# ", 15)
+ self.dut.send_expect("modprobe ice", "# ", 15)
+ self.setup_2pf_4vf_env()
self.session_secondary = self.dut.new_session()
self.session_third = self.dut.new_session()
@@ -8819,6 +8825,15 @@ class TestIAVFFdir(TestCase):
nex_cnt = 94 // 8
elif len(self.dut_ports) == 2:
nex_cnt = 110 // 8
+ # check the card is chapman beach 100g*2 or not
+ pf_pci = self.dut.ports_info[0]['pci']
+ out = self.dut.send_expect('lspci -s {} -vvv |grep "Product Name"'.format(pf_pci), '#')
+ res = re.search(r"Network Adapter\s+(?P<product_name>E810-.*)", out)
+ self.verify(res, "product name not found'")
+ # if card is chapman beach 100g*2, one pf port equals a general 100g*2 card,so 118 profiles can be used for vf
+ if 'E810-2CQDA2' in res.group('product_name'):
+ nex_cnt = 118 // 8
+
else:
self.verify(False, 'The number of ports is not supported')
@@ -8828,43 +8843,46 @@ class TestIAVFFdir(TestCase):
self.config_testpmd()
for port_id in range(nex_cnt):
+ # The number of rules created is affected by the profile and TCAM. The maximum profile is 128 and the maximum
+ # TCAM is 512. In order to test the maximum profile, need to create rules that consume less TCAM to ensure
+ # that the profile will reach the maximum before TCAM exhausted
rules = [
- "flow create %d ingress pattern eth / ipv4 src is 192.168.0.20 dst is 192.168.0.21 / udp src is 22 dst is 23 / end actions queue index 1 / mark / end" % port_id,
- "flow create %d ingress pattern eth / ipv4 src is 192.168.0.20 dst is 192.168.0.21 / tcp src is 22 dst is 23 / end actions queue index 1 / mark / end" % port_id,
- "flow create %d ingress pattern eth / ipv4 src is 192.168.0.20 dst is 192.168.0.21 / sctp src is 22 dst is 23 / end actions queue index 1 / mark / end" % port_id,
- "flow create %d ingress pattern eth / ipv6 dst is CDCD:910A:2222:5498:8475:1111:3900:2020 src is 2001::2 / udp src is 22 dst is 23 / end actions queue index 1 / mark / end" % port_id,
- "flow create %d ingress pattern eth / ipv6 dst is CDCD:910A:2222:5498:8475:1111:3900:2020 src is 2001::2 / tcp src is 22 dst is 23 / end actions queue index 1 / mark / end" % port_id,
- "flow create %d ingress pattern eth / ipv6 dst is CDCD:910A:2222:5498:8475:1111:3900:2020 src is 2001::2 / sctp src is 22 dst is 23 / end actions queue index 1 / mark / end" % port_id,
- "flow create %d ingress pattern eth type is 0x8863 / end actions queue index 1 / mark id 1 / end" % port_id,
- "flow create %d ingress pattern eth / ipv4 / udp / pfcp s_field is 0 / end actions queue index 2 / end" % port_id]
- self.create_fdir_rule(rules, check_stats=True)
+ "flow create {} ingress pattern eth / ipv4 / l2tpv3oip session_id is 1 / end actions queue index 1 / mark / end",
+ "flow create {} ingress pattern eth / ipv6 / l2tpv3oip session_id is 2 / end actions queue index 1 / mark / end",
+ "flow create {} ingress pattern eth / ipv4 / tcp / end actions queue index 2 / mark / end",
+ "flow create {} ingress pattern eth / ipv6 / tcp / end actions queue index 2 / mark / end",
+ "flow create {} ingress pattern eth / ipv4 / esp spi is 1 / end actions queue index 3 / mark / end",
+ "flow create {} ingress pattern eth / ipv6 / esp spi is 2 / end actions queue index 3 / mark / end",
+ "flow create {} ingress pattern eth / ipv4 / udp / pfcp s_field is 0 / end actions queue index 4 / mark id 1 / end",
+ "flow create {} ingress pattern eth / ipv6 / udp / pfcp s_field is 1 / end actions queue index 4 / end"]
+ self.create_fdir_rule([rule.format(port_id) for rule in rules], check_stats=True)
rules = [
- "flow create {} ingress pattern eth / ipv4 src is 192.168.0.20 dst is 192.168.0.21 / udp src is 22 dst is 23 / end actions queue index 1 / mark / end".format(nex_cnt),
- "flow create {} ingress pattern eth / ipv4 src is 192.168.0.20 dst is 192.168.0.21 / tcp src is 22 dst is 23 / end actions queue index 1 / mark / end".format(nex_cnt),
- "flow create {} ingress pattern eth / ipv4 src is 192.168.0.20 dst is 192.168.0.21 / sctp src is 22 dst is 23 / end actions queue index 1 / mark / end".format(nex_cnt),
- "flow create {} ingress pattern eth / ipv6 dst is CDCD:910A:2222:5498:8475:1111:3900:2020 src is 2001::2 / udp src is 22 dst is 23 / end actions queue index 1 / mark / end".format(nex_cnt),
- "flow create {} ingress pattern eth / ipv6 dst is CDCD:910A:2222:5498:8475:1111:3900:2020 src is 2001::2 / tcp src is 22 dst is 23 / end actions queue index 1 / mark / end".format(nex_cnt),
- "flow create {} ingress pattern eth / ipv6 dst is CDCD:910A:2222:5498:8475:1111:3900:2020 src is 2001::2 / sctp src is 22 dst is 23 / end actions queue index 1 / mark / end".format(nex_cnt)]
- self.create_fdir_rule(rules, check_stats=True)
-
- rule = "flow create {} ingress pattern eth type is 0x8863 / end actions queue index 1 / mark id 1 / end".format(nex_cnt)
+ "flow create {} ingress pattern eth / ipv4 / l2tpv3oip session_id is 1 / end actions queue index 1 / mark / end",
+ "flow create {} ingress pattern eth / ipv6 / l2tpv3oip session_id is 2 / end actions queue index 1 / mark / end",
+ "flow create {} ingress pattern eth / ipv4 / tcp / end actions queue index 2 / mark / end",
+ "flow create {} ingress pattern eth / ipv6 / tcp / end actions queue index 2 / mark / end",
+ "flow create {} ingress pattern eth / ipv4 / esp spi is 1 / end actions queue index 3 / mark / end",
+ "flow create {} ingress pattern eth / ipv6 / esp spi is 2 / end actions queue index 3 / mark / end"]
+ self.create_fdir_rule([rule.format(nex_cnt) for rule in rules], check_stats=True)
+
+ rule = "flow create {} ingress pattern eth / ipv4 / udp / pfcp s_field is 0 / end actions queue index 4 / mark id 1 / end".format(nex_cnt)
self.create_fdir_rule(rule, check_stats=False)
- pkt1 = 'Ether(dst="00:11:22:33:44:55")/IP(src="192.168.0.20",dst="192.168.0.21")/UDP(sport=22, dport=23)/ Raw("x" * 80)'
+ pkt1 = 'Ether(dst="00:11:22:33:44:55")/IP(src="192.168.0.20",dst="192.168.0.21")/TCP()/Raw("x" * 80)'
out = self.send_pkts_getouput(pkts=pkt1)
- rfc.check_iavf_fdir_mark(out, pkt_num=1, check_param={"port_id": nex_cnt, "mark_id": 0, "queue": 1}, stats=True)
- pkt2 = 'Ether(dst="00:11:22:33:44:55", type=0x8863)/IP()/Raw("x" * 80)'
+ rfc.check_iavf_fdir_mark(out, pkt_num=1, check_param={"port_id": nex_cnt, "mark_id": 0, "queue": 2}, stats=True)
+ pkt2 = 'Ether(dst="00:11:22:33:44:55")/IP(src="192.168.0.20",dst="192.168.0.21")/UDP(dport=8805)/PFCP(S=0)/Raw("x" * 80)'
out = self.send_pkts_getouput(pkts=pkt2)
- rfc.check_iavf_fdir_mark(out, pkt_num=1, check_param={"port_id": nex_cnt, "mark_id": 1, "queue": 1}, stats=False)
+ rfc.check_iavf_fdir_mark(out, pkt_num=1, check_param={"port_id": nex_cnt, "mark_id": 1, "queue": 4}, stats=False)
self.dut.send_expect("flow flush {}".format(nex_cnt), "testpmd> ")
self.check_fdir_rule(port_id=(nex_cnt), stats=False)
out = self.send_pkts_getouput(pkts=pkt1)
- rfc.check_iavf_fdir_mark(out, pkt_num=1, check_param={"port_id": nex_cnt, "mark_id": 0, "queue": 1}, stats=False)
+ rfc.check_iavf_fdir_mark(out, pkt_num=1, check_param={"port_id": nex_cnt, "mark_id": 0, "queue": 2}, stats=False)
self.create_fdir_rule(rule, check_stats=True)
out = self.send_pkts_getouput(pkts=pkt2)
- rfc.check_iavf_fdir_mark(out, pkt_num=1, check_param={"port_id": nex_cnt, "mark_id": 1, "queue": 1}, stats=True)
+ rfc.check_iavf_fdir_mark(out, pkt_num=1, check_param={"port_id": nex_cnt, "mark_id": 1, "queue": 4}, stats=True)
def test_stress_port_stop_start(self):
"""
--
2.17.1
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [dts] [PATCH V2 2/2] tests/iavf_fdir: create rules with less consumption of TCAM rule in suite
2021-10-11 10:03 ` [dts] [PATCH V2 2/2] tests/iavf_fdir: create rules with less consumption of TCAM rule in suite Jiale Song
@ 2021-10-25 7:57 ` Tu, Lijuan
0 siblings, 0 replies; 3+ messages in thread
From: Tu, Lijuan @ 2021-10-25 7:57 UTC (permalink / raw)
To: Jiale, SongX, dts; +Cc: Jiale, SongX
> -----Original Message-----
> From: dts <dts-bounces@dpdk.org> On Behalf Of Jiale Song
> Sent: 2021年10月11日 18:04
> To: dts@dpdk.org
> Cc: Jiale, SongX <songx.jiale@intel.com>
> Subject: [dts] [PATCH V2 2/2] tests/iavf_fdir: create rules with less consumption
> of TCAM rule in suite
>
> modify the suite according to the testplan
>
> Signed-off-by: Jiale Song <songx.jiale@intel.com>
> ---
> tests/TestSuite_iavf_fdir.py | 66 +++++++++++++++++++++++-------------
> 1 file changed, 42 insertions(+), 24 deletions(-)
>
> diff --git a/tests/TestSuite_iavf_fdir.py b/tests/TestSuite_iavf_fdir.py index
> 7f8dace..44fcb95 100755
> --- a/tests/TestSuite_iavf_fdir.py
> +++ b/tests/TestSuite_iavf_fdir.py
> @@ -8349,6 +8349,9 @@ class TestIAVFFdir(TestCase):
> create same rules on pf and vf, no conflict
> """
> self.dut.kill_all()
> + self.dut.send_expect("rmmod ice", "# ", 15)
> + self.dut.send_expect("modprobe ice", "# ", 15)
I don't think there is necessary to rmmod ice. DTS should not do this, it will impact other devices.
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2021-10-25 7:57 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-10-11 10:03 [dts] [PATCH V2 1/2] tests_plan/iavf_fdir: replace rules with less consumption TCAM rule Jiale Song
2021-10-11 10:03 ` [dts] [PATCH V2 2/2] tests/iavf_fdir: create rules with less consumption of TCAM rule in suite Jiale Song
2021-10-25 7:57 ` Tu, Lijuan
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).