Re: [PATCH] doc: announce API change for ethdev port info

[Thomas Monjalon <thomas@monjalon.net>]

15/11/2017 09:13, Olivier MATZ:
> On Wed, Nov 15, 2017 at 12:50:55AM +0100, Thomas Monjalon wrote:
> > 16/10/2017 16:27, Ivan Boule:
> > > To help administrative tasks on ports, new per-port information need
> > > to be added into the data structure rte_eth_dev_info supplied by the
> > > dev_infos_get() function exported by a Poll Mode Driver.
> > > 
> > > See http://dpdk.org/ml/archives/dev/2017-September/074885.html for
> > > details.
> > > 
> > > Signed-off-by: Ivan Boule <ivan.boule@6wind.com>
> > > Acked-by: Olivier Matz <olivier.matz@6wind.com>
> > > ---
> > > +* librte_ether: additional fields will be added into the ``rte_eth_dev_info``
> > > +  structure in 18.02, breaking the API. These fields will contain:
> > > +
> > > +  - the set of supported link modes,
> > > +  - the set of advertised link modes,
> > > +  - the type of port connector,
> > > +  - autonegotiation enabled or not.
> > 
> > This patch is not accepted in 17.11 for following reasons:
> > - it requires at least 3 acks
> > - we are not going to break API in 18.02, except maybe for EAL devargs
> > - the link mode is redundant with the speed capabilities
> > - such fields require work (and time) for every PMD to be fully supported
> > - we should discuss more generally which infos are in the scope of this function
> > 
> > Sorry for the short notice, I was waiting for comments and to see if
> > other deprecations were sent (or not) for ethdev.
> 
> Thomas, be aware that it's very frustrating to have this kind feedback 9
> weeks after the submission of the RFC, and few days before the release.

Yes, sure I know how it can be frustrating and I am sorry about that.
I forgot to review the RFC and I forgot to comment this deprecation notice.
Then it was pending like other deprecation notices, but no new deprecation
notice appeared for 18.02, and nobody did a comment on those patches.
So I think it is better to just postpone it for 18.05.
Anyway we can comment and work on the RFC in the meantime.