test suite reviews and discussions
 help / color / mirror / Atom feed
From: "Tu, Lijuan" <lijuan.tu@intel.com>
To: "Han, YingyaX" <yingyax.han@intel.com>, "dts@dpdk.org" <dts@dpdk.org>
Subject: Re: [dts] [PATCH V1]tests/runtime_queue_number: fix the issue	ofcreating VF on eagle in two ports
Date: Mon, 5 Nov 2018 05:11:03 +0000	[thread overview]
Message-ID: <8CE3E05A3F976642AAB0F4675D0AD20E0B9B783E@SHSMSX101.ccr.corp.intel.com> (raw)
In-Reply-To: <9AC1400278453341942179604C2BF1260B8C0D3A@SHSMSX101.ccr.corp.intel.com>

ixgbe, each pf has 128 queues at most, pf and it's vfs share these queues.
如果vf个数大于8,每个vf最多16个queue,如果vf个数大于16,每个vf最多8个queue,如果vf个数大于32,每个vf最多4个queue,以此类推,pf 就用vf用完了后剩下的queue。

There can be set 128 vfs per Fortville card.
If there are 4 10G port in one fvl card, each port can be set 32 vfs.
The whole number of queue per NIC is 1536.
So if there are 4 ports, there are 384 queues per port, which are shared by pf and vfs.


根据这个把hard code 改掉,先评估下难度和时间。

> -----Original Message-----
> From: Han, YingyaX
> Sent: Monday, November 5, 2018 10:41 AM
> To: Tu, Lijuan <lijuan.tu@intel.com>; dts@dpdk.org
> Subject: RE: [dts] [PATCH V1]tests/runtime_queue_number: fix the issue
> ofcreating VF on eagle in two ports
> 
> Hi lijuan,
> 
> VF number due to FVL family NICs have different queue numbers. So  it's
> changed based on device type.
> Please refer to the test plan.
> 
> """
> Create 32 vfs on four ports fortville NIC::
> 
>      echo 32 > /sys/bus/pci/devices/0000\:05\:00.0/max_vfs
> 
> Create 64 vfs on two ports fortville NIC::
> 
>     echo 64 > /sys/bus/pci/devices/0000\:05\:00.0/max_vfs
> """
> 
> 
> 
> 
> thanks
> Yingya
> 
> -----Original Message-----
> From: Tu, Lijuan
> Sent: Monday, November 5, 2018 9:34 AM
> To: Han, YingyaX <yingyax.han@intel.com>; dts@dpdk.org
> Cc: Han, YingyaX <yingyax.han@intel.com>
> Subject: RE: [dts] [PATCH V1]tests/runtime_queue_number: fix the issue
> ofcreating VF on eagle in two ports
> 
> 
> 
> > -----Original Message-----
> > From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of han,yingya
> > Sent: Friday, November 2, 2018 3:57 PM
> > To: dts@dpdk.org
> > Cc: Han, YingyaX <yingyax.han@intel.com>
> > Subject: [dts] [PATCH V1]tests/runtime_queue_number: fix the issue
> > ofcreating VF on eagle in two ports
> >
> > FVL_eagle nic has two specifications, namely 2*10G and 4*10G.
> >
> > Signed-off-by: han,yingya <yingyax.han@intel.com>
> > ---
> >  tests/TestSuite_runtime_queue_number.py | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/tests/TestSuite_runtime_queue_number.py
> > b/tests/TestSuite_runtime_queue_number.py
> > index 12263f3..4a51fa2 100644
> > --- a/tests/TestSuite_runtime_queue_number.py
> > +++ b/tests/TestSuite_runtime_queue_number.py
> > @@ -251,12 +251,12 @@ class TestRuntime_Queue_Number(TestCase):
> >          """
> >          set max queue number when setting max VFs on 1 PF port.
> >          """
> > -        if (self.nic in ["fortville_eagle", "fortpark_TLV"]):
> > +        if (self.nic in ["fortville_eagle", "fortpark_TLV"]) and
> > len(self.dut_ports) > 2:
> [Lijuan] why require 2 pf ports, I found only 1 port used here.
> >              self.setup_env(32)
> >              # failed to set VF max queue num to 16.
> >              out = self.pmdout.start_testpmd("%s" % self.cores,
> > eal_param="-w %s,queue-num-per-vf=16 --file-prefix=test1 --socket-mem
> > 1024,1024" % self.pf_pci)
> >              self.verify("exceeds the hardware maximum 384" in out,
> > "the queue num exceeds the hardware maximum 384")
> > -        elif (self.nic in ["fortville_spirit", "fortville_spirit_single"]):
> > +        elif (self.nic in ["fortville_spirit",
> > + "fortville_spirit_single",
> > "fortville_eagle"]):
> >              self.setup_env(64)
> >              # failed to set VF max queue num to 16.
> >              out = self.pmdout.start_testpmd("%s" % self.cores,
> > eal_param="-w %s,queue-num-per-vf=16 --file-prefix=test1 --socket-mem
> > 1024,1024" % self.pf_pci)
> > --
> > 2.17.2


      reply	other threads:[~2018-11-05  5:11 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-11-02  7:57 han,yingya
2018-11-05  1:34 ` Tu, Lijuan
2018-11-05  2:40   ` Han, YingyaX
2018-11-05  5:11     ` Tu, Lijuan [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=8CE3E05A3F976642AAB0F4675D0AD20E0B9B783E@SHSMSX101.ccr.corp.intel.com \
    --to=lijuan.tu@intel.com \
    --cc=dts@dpdk.org \
    --cc=yingyax.han@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).