[dpdk-dev] [PATCH 06/26] net/octeontx/base: probe PKI and PKO PCIe VF devices

Jerin Jacob jerin.jacob at caviumnetworks.com
Mon Sep 11 20:27:32 CEST 2017


-----Original Message-----
> Date: Tue, 5 Sep 2017 18:44:22 +0100
> From: Ferruh Yigit <ferruh.yigit at intel.com>
> To: Jerin Jacob <jerin.jacob at caviumnetworks.com>, dev at dpdk.org
> CC: santosh.shukla at caviumnetworks.com
> Subject: Re: [dpdk-dev] [PATCH 06/26] net/octeontx/base: probe PKI and PKO
>  PCIe VF devices
> User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:52.0) Gecko/20100101
>  Thunderbird/52.3.0
> 
> On 8/31/2017 3:54 PM, Jerin Jacob wrote:
> > An octeontx ethdev device consists of multiple PKO VF devices and an PKI
> > VF device. On Octeontx HW, each Rx queues are enumerated as SSOVF device
> > which is exposed as event_octeontx device, Tx queues are enumerated as
> > PKOVF device, and ingress packet configuration is accomplished through
> > PKIVF device.
> > 
> > In order to expose as an single ethdev instance, On PCIe VF probe,
> > the driver stores the information associated with the PCIe VF device and
> > later with vdev infrastructure creates ethdev device with earlier
> > probed PCIe VF device.
> 
> So, is following correct:
> 
> BGX is MAC interface,
> in ingress it consists of single PKIVF (packet input) device,
> in egress it consists of PKOVF (packet output) devices.

Yes.

> 
> PKIVF and PKOVF are physical eventdev devices.

No they are network related co processors. Not the eventdev device. The eventdev
HW is abstracted through driver/event/octeontx

> 
> First physical devices are probed, later virtual ethdev is created which
> gets/puts packets into these event devices.

Yes.

> 
> A graph in documentation can be very helpful for this.
> 
> Also patchset can create multiple ethdev ports, why is this? And how

A PKOVF PCIe VF device has 8 Tx queues which can be from any MAC interface. The
created ethdev ports maps to each physical MAC interface.

> eventdev - ethdev port mapping done for that case?

One octeontx eventdev port is mapped as one octeontx ethdev Rx queue.

> 
> Thanks,
> ferruh
> 
> > 
> > Signed-off-by: Jerin Jacob <jerin.jacob at caviumnetworks.com>
> > Co-authored-by: Santosh Shukla <santosh.shukla at caviumnetworks.com>
> 
> <...>


More information about the dev mailing list