[dts][PATCH V1 1/2] test_plans/ice_dcf_switch_filter: optimize max vfs script

[dts][PATCH V1 1/2] test_plans/ice_dcf_switch_filter: optimize max vfs script

[Hongbo Li]

Change the method of obtaining the maximum number of VF that can be created on one PF

Signed-off-by: Hongbo Li <hongbox.li@intel.com>
---
 .../ice_dcf_switch_filter_test_plan.rst       | 28 ++++++++++---------
 1 file changed, 15 insertions(+), 13 deletions(-)

diff --git a/test_plans/ice_dcf_switch_filter_test_plan.rst b/test_plans/ice_dcf_switch_filter_test_plan.rst
index 63ef56e0..488683b0 100644
--- a/test_plans/ice_dcf_switch_filter_test_plan.rst
+++ b/test_plans/ice_dcf_switch_filter_test_plan.rst
@@ -2791,40 +2791,42 @@ are dropped.
 Test case: Max vfs
 ==================
 
-Description: 256 VFs can be created on a Intel® Ethernet 800 Series NIC, if 2*100G NIC, each PF
-can create 128 VFs, else if 4*25G NIC, each PF can create 64 VFs. This
-case is used to test when all VFs on a PF are used, switch filter rules can work.
+Description: This case is used to test when all VFs on a PF are used, switch filter rules can work.
 This case is designed based on 4*25G NIC.
 
-1. generate 64 VFs on PF::
+1.Query the maximum number of VF that can be created by each PF::
+
+     cat /sys/bus/pci/devices/0000\:18\:00.0/sriov_totalvfs
+
+2. generate 64 VFs on PF::
 
      echo 64 > /sys/bus/pci/devices/0000:18:00.0/sriov_numvfs
 
-2. Get the interface name of the VFs, for example::
+3. Get the interface name of the VFs, for example::
 
      ./usertools/dpdk-devbind.py -s
 
      0000:18:01.1 'Ethernet Adaptive Virtual Function 1889' if=enp24s1f1 drv=iavf unused=igb_uio
 
-3. Start the 64 VFs in the kernel, for example::
+4. Start the 64 VFs in the kernel, for example::
 
      ifconfig enp24s1f1 up
 
-4. Set VF0 as trust::
+5. Set VF0 as trust::
 
      ip link set enp24s0f0 vf 0 trust on
 
-5. bind VF0 to dpdk driver::
+6. bind VF0 to dpdk driver::
 
      ./usertools/dpdk-devbind.py -b vfio-pci 0000:18:01.0
 
-6. launch dpdk on VF0, and request DCF mode, representing VF1 to VF63::
+7. launch dpdk on VF0, and request DCF mode, representing VF1 to VF63::
 
      ./x86_64-native-linuxapp-gcc/app/dpdk-testpmd -c 0xf -n 4 \
      -a 0000:18:01.0,cap=dcf,representor=[1-63] \
      -- -i
 
-7. set a switch rule to each VF from DCF, totally 63 rules::
+8. set a switch rule to each VF from DCF, totally 63 rules::
 
      testpmd> flow create 0 ingress pattern eth / ipv4 src is 192.168.0.1 / tcp / end actions represented_port ethdev_port_id 1 / end
      testpmd> flow create 0 ingress pattern eth / ipv4 src is 192.168.0.2 / tcp / end actions represented_port ethdev_port_id 2 / end
@@ -2835,7 +2837,7 @@ This case is designed based on 4*25G NIC.
 
    check the rules exist in the list.
 
-8. send matched packets::
+9. send matched packets::
 
      sendp([Ether(dst="68:05:ca:8d:ed:a8")/IP(src="192.168.0.1")/TCP()/Raw("X"*480)], iface="ens786f0", count=1)
      sendp([Ether(dst="68:05:ca:8d:ed:a8")/IP(src="192.168.0.2")/TCP()/Raw("X"*480)], iface="ens786f0", count=1)
@@ -2862,7 +2864,7 @@ This case is designed based on 4*25G NIC.
 
    check the packets are not to any VF.
 
-9. verify rules can be destroyed::
+10. verify rules can be destroyed::
 
      testpmd> flow flush 0
      testpmd> flow list 0
@@ -4467,4 +4469,4 @@ Subcase 4:  MAC_IPV4_NVGRE_MAC_IPV4_UDP + L4 MASK + single vf
      testpmd> flow list 0
 
    check the rules not exist in the list.
-   send matched packets, check port 1 can not receive the packet.
\ No newline at end of file
+   send matched packets, check port 1 can not receive the packet.
-- 
2.17.1

[dts][PATCH V1 2/2] tests/ice_dcf_switch_filter: optimize max vfs script

[Hongbo Li]

Change the method of obtaining the maximum number of VF that can be created on one PF

Signed-off-by: Hongbo Li <hongbox.li@intel.com>
---
 tests/TestSuite_ice_dcf_switch_filter.py | 17 ++++++++++++++---
 1 file changed, 14 insertions(+), 3 deletions(-)

diff --git a/tests/TestSuite_ice_dcf_switch_filter.py b/tests/TestSuite_ice_dcf_switch_filter.py
index 3560d065..4a84d084 100644
--- a/tests/TestSuite_ice_dcf_switch_filter.py
+++ b/tests/TestSuite_ice_dcf_switch_filter.py
@@ -2951,10 +2951,21 @@ class ICEDCFSwitchFilterTest(TestCase):
 
     def test_max_vfs(self):
         # get max vfs number
-        port_count = (1, 1, 2, 4, 4)[len(self.dut_ports)]
-        max_vf_number = int(256 / (port_count))
-        # set up max_vf_number vfs on 1 pf environment
         self.used_dut_port_0 = self.dut_ports[0]
+        self.domain_id_0 = self.dut.ports_info[self.used_dut_port_0]["port"].domain_id
+        self.bus_id_0 = self.dut.ports_info[self.used_dut_port_0]["port"].bus_id
+        self.devfun_id_0 = self.dut.ports_info[self.used_dut_port_0]["port"].devfun_id
+        cmd_max_vfs = (
+            "cat /sys/bus/pci/devices/"
+            + self.domain_id_0
+            + "\\:"
+            + self.bus_id_0
+            + "\\:"
+            + self.devfun_id_0
+            + "/sriov_totalvfs"
+        )
+        max_vf_number = int(self.dut.send_expect(cmd_max_vfs, "#"))
+        # set up max_vf_number vfs on 1 pf environment
         self.pf0_intf = self.dut.ports_info[self.used_dut_port_0]["intf"]
         out = self.dut.send_expect("ethtool -i %s" % self.pf0_intf, "#")
         # generate max_vf_number VFs on PF0
-- 
2.17.1

RE: [dts][PATCH V1 1/2] test_plans/ice_dcf_switch_filter: optimize max vfs script

[Tu, Lijuan]

> -----Original Message-----
> From: Hongbo Li <hongbox.li@intel.com>
> Sent: Tuesday, February 21, 2023 10:51 PM
> To: dts@dpdk.org
> Cc: Li, HongboX <hongbox.li@intel.com>
> Subject: [dts][PATCH V1 1/2] test_plans/ice_dcf_switch_filter: optimize max vfs
> script
> 
> Change the method of obtaining the maximum number of VF that can be
> created on one PF
> 
> Signed-off-by: Hongbo Li <hongbox.li@intel.com>
> ---
>  .../ice_dcf_switch_filter_test_plan.rst       | 28 ++++++++++---------
>  1 file changed, 15 insertions(+), 13 deletions(-)
> 
> diff --git a/test_plans/ice_dcf_switch_filter_test_plan.rst
> b/test_plans/ice_dcf_switch_filter_test_plan.rst
> index 63ef56e0..488683b0 100644
> --- a/test_plans/ice_dcf_switch_filter_test_plan.rst
> +++ b/test_plans/ice_dcf_switch_filter_test_plan.rst
> @@ -2791,40 +2791,42 @@ are dropped.
>  Test case: Max vfs
>  ==================
> 
> -Description: 256 VFs can be created on a Intel® Ethernet 800 Series NIC, if
> 2*100G NIC, each PF -can create 128 VFs, else if 4*25G NIC, each PF can create
> 64 VFs. This -case is used to test when all VFs on a PF are used, switch filter rules
> can work.
> +Description: This case is used to test when all VFs on a PF are used, switch filter
> rules can work.
>  This case is designed based on 4*25G NIC.
> 
> -1. generate 64 VFs on PF::
> +1.Query the maximum number of VF that can be created by each PF::
> +
> +     cat /sys/bus/pci/devices/0000\:18\:00.0/sriov_totalvfs
> +
> +2. generate 64 VFs on PF::
> 
>       echo 64 > /sys/bus/pci/devices/0000:18:00.0/sriov_numvfs

Why it get the maximum number of VF, but not use it? 
I see here 64 is still a hardcode, not the maximum number of VF.