Re: [dpdk-dev] [PATCH v2 1/4] ethdev: add Rx offload outer UDP checksum definition

["Ananyev, Konstantin" <konstantin.ananyev@intel.com>]

> -----Original Message-----
> From: Jerin Jacob [mailto:jerin.jacob@caviumnetworks.com]
> Sent: Wednesday, October 3, 2018 10:22 AM
> To: Ananyev, Konstantin <konstantin.ananyev@intel.com>
> Cc: Lu, Wenzhuo <wenzhuo.lu@intel.com>; Wu, Jingjing <jingjing.wu@intel.com>; Iremonger, Bernard <bernard.iremonger@intel.com>;
> Mcnamara, John <john.mcnamara@intel.com>; Kovacevic, Marko <marko.kovacevic@intel.com>; Thomas Monjalon
> <thomas@monjalon.net>; Yigit, Ferruh <ferruh.yigit@intel.com>; Andrew Rybchenko <arybchenko@solarflare.com>; Olivier Matz
> <olivier.matz@6wind.com>; dev@dpdk.org; shahafs@mellanox.com
> Subject: Re: [dpdk-dev] [PATCH v2 1/4] ethdev: add Rx offload outer UDP checksum definition
> 
> -----Original Message-----
> > Date: Wed, 3 Oct 2018 09:17:18 +0000
> > From: "Ananyev, Konstantin" <konstantin.ananyev@intel.com>
> > To: Jerin Jacob <jerin.jacob@caviumnetworks.com>
> > CC: "Lu, Wenzhuo" <wenzhuo.lu@intel.com>, "Wu, Jingjing"
> >  <jingjing.wu@intel.com>, "Iremonger, Bernard"
> >  <bernard.iremonger@intel.com>, "Mcnamara, John" <john.mcnamara@intel.com>,
> >  "Kovacevic, Marko" <marko.kovacevic@intel.com>, Thomas Monjalon
> >  <thomas@monjalon.net>, "Yigit, Ferruh" <ferruh.yigit@intel.com>, Andrew
> >  Rybchenko <arybchenko@solarflare.com>, Olivier Matz
> >  <olivier.matz@6wind.com>, "dev@dpdk.org" <dev@dpdk.org>,
> >  "shahafs@mellanox.com" <shahafs@mellanox.com>
> > Subject: RE: [dpdk-dev] [PATCH v2 1/4] ethdev: add Rx offload outer UDP
> >  checksum	definition
> >
> >
> > > -----Original Message-----
> > > From: Jerin Jacob [mailto:jerin.jacob@caviumnetworks.com]
> > > Sent: Wednesday, October 3, 2018 9:59 AM
> > > To: Ananyev, Konstantin <konstantin.ananyev@intel.com>
> > > Cc: Lu, Wenzhuo <wenzhuo.lu@intel.com>; Wu, Jingjing <jingjing.wu@intel.com>; Iremonger, Bernard
> <bernard.iremonger@intel.com>;
> > > Mcnamara, John <john.mcnamara@intel.com>; Kovacevic, Marko <marko.kovacevic@intel.com>; Thomas Monjalon
> > > <thomas@monjalon.net>; Yigit, Ferruh <ferruh.yigit@intel.com>; Andrew Rybchenko <arybchenko@solarflare.com>; Olivier Matz
> > > <olivier.matz@6wind.com>; dev@dpdk.org; shahafs@mellanox.com
> > > Subject: Re: [dpdk-dev] [PATCH v2 1/4] ethdev: add Rx offload outer UDP checksum definition
> > >
> > > -----Original Message-----
> > > > Date: Wed, 3 Oct 2018 08:53:22 +0000
> > > > From: "Ananyev, Konstantin" <konstantin.ananyev@intel.com>
> > > > To: Jerin Jacob <jerin.jacob@caviumnetworks.com>, "Lu, Wenzhuo"
> > > >  <wenzhuo.lu@intel.com>, "Wu, Jingjing" <jingjing.wu@intel.com>,
> > > >  "Iremonger, Bernard" <bernard.iremonger@intel.com>, "Mcnamara, John"
> > > >  <john.mcnamara@intel.com>, "Kovacevic, Marko" <marko.kovacevic@intel.com>,
> > > >  Thomas Monjalon <thomas@monjalon.net>, "Yigit, Ferruh"
> > > >  <ferruh.yigit@intel.com>, Andrew Rybchenko <arybchenko@solarflare.com>,
> > > >  Olivier Matz <olivier.matz@6wind.com>
> > > > CC: "dev@dpdk.org" <dev@dpdk.org>, "shahafs@mellanox.com"
> > > >  <shahafs@mellanox.com>
> > > > Subject: RE: [dpdk-dev] [PATCH v2 1/4] ethdev: add Rx offload outer UDP
> > > >  checksum   definition
> > > >
> > > >
> > > > Hi Jerin,
> > >
> > > Hi Konstantin,
> > >
> > > >
> > > > >
> > > > > Introduced DEV_RX_OFFLOAD_OUTER_UDP_CKSUM Rx offload flag and
> > > > > PKT_RX_EL4_CKSUM_BAD mbuf ol_flags to detect outer UDP checksum
> > > > > failure.
> > > > >
> > > > > - To use hardware Rx outer UDP checksum offload, the user needs to
> > > > > configure DEV_RX_OFFLOAD_OUTER_UDP_CKSUM offload flags in slowpath.
> > > > >
> > > > > - Driver updates the PKT_RX_EL4_CKSUM_BAD mbuf ol_flag on checksum failure
> > > > > similar to the outer L3 PKT_RX_EIP_CKSUM_BAD flag.
> > > > >
> > > > > Signed-off-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
> > > >
> > > > Looks ok to me in general.
> > > > Just wonder is there any PMD that supports all these new features?
> > >
> > > octeontx2 PMD has this feature. I am planning to push the PMD for v19.02.
> > > Before that I adding all the common code change to avoid the dependency.
> >
> > Ok, but why then ethdev/mbuf changes has to go into 18.11?
> 
> It it is a generic change then why not? What is the real concern here?

If there is no implementation for it, how we can conclude that it is really 'generic' one? :)
My main concern is that we already have several features in rte_ethdev and rte_security
that supposed to be 'generic' but right now no-one support them. 
I wouldn't to object about these particular features, they look reasonable to me.
But in general I think we need some better defined policy -
about what is allowed to propagate without real evidence (particular implementation)
and what is not.

> 
> > Do you plan your new PMD backward compatible with 18.11 LTS?
> 
> Yes.

I see.
Konstantin