ethdev: hide internal structures

[Stephen Hemminger]

On Tue, 16 Nov 2021 14:58:08 -0800
Tyler Retzlaff <roretzla at linux.microsoft.com> wrote:

> On Tue, Nov 16, 2021 at 01:25:10PM -0800, Stephen Hemminger wrote:
> > On Tue, 16 Nov 2021 11:10:18 -0800
> > Tyler Retzlaff <roretzla at linux.microsoft.com> wrote:
> >   
> > > On Tue, Nov 16, 2021 at 10:32:55AM +0000, Ananyev, Konstantin wrote:  
> > > >  
> > > > rte_eth_dev,  rte_eth_dev_data, rte_eth_rxtx_callback are internal
> > > > data structures that were used by public inline ethdev functions. 
> > > > Well behaving app should not access these data structures directly.
> > > > So, for well behaving app there should no changes in the code required.
> > > > That what I meant by 'transparent' above.
> > > > But it is still an ABI change, so yes, the app has to be re-compiled.     
> > > 
> > > so it appears the application was establishing a private context /
> > > vendor extension between the application and a pmd. the application
> > > was abusing access to the rte_eth_devices[] to get the private context
> > > from the rte_eth_dev.
> > > 
> > > is there a proper / supported way of providing this functionality
> > > through the public api?
> > >   
> > > > 
> > > > Konstantin    
> > 
> > Keep a array in application?  Portid is universally
> > available.
> > 
> > struct my_portdata *my_ports[RTE_ETH_MAXPORTS];  
> 
> i guess by this you mean maintain the storage in the application and
> then export that storage for proprietary use in the pmd. ordinarily i
> wouldn't want to have this hard-coded into the modules abi but since
> we are talking about vendor extensions it has to be managed somewhere.
> 
> anyway, i guess i have my answer.
> 
> thanks stephen, appreciate it.

Don't understand, how are application and pmd exchanging extra data?
Maybe a non-standard PMD API?