[v2,00/12] EAL and PCI ABI changes for 19.11

  Let's prepare for the ABI freeze.

The first patches are about changes that had been announced before (with
a patch from Stephen that I took as it is ready as is from my pov).

The malloc_heap structure from the memory subsystem can be hidden.
The PCI library had some forgotten deprecated APIs that are removed with
this series.

rte_logs could be hidden, but I am not that confortable about
doing it right away: I added an accessor to rte_logs.file, but I am fine
with dropping the last patch and wait for actually hiding this in the next
ABI break.

Changelog since v1:
- I went a step further, hiding rte_config after de-inlining non critical
  functions

Comments?
  

On Wed, 23 Oct 2019 20:54:12 +0200
David Marchand <david.marchand@redhat.com> wrote:

> Let's prepare for the ABI freeze.
> 
> The first patches are about changes that had been announced before (with
> a patch from Stephen that I took as it is ready as is from my pov).
> 
> The malloc_heap structure from the memory subsystem can be hidden.
> The PCI library had some forgotten deprecated APIs that are removed with
> this series.
> 
> rte_logs could be hidden, but I am not that confortable about
> doing it right away: I added an accessor to rte_logs.file, but I am fine
> with dropping the last patch and wait for actually hiding this in the next
> ABI break.

19.11 is an api/abi break so maybe do it now.
  

On Wed, Oct 23, 2019 at 11:10 PM Stephen Hemminger
<stephen@networkplumber.org> wrote:
>
> On Wed, 23 Oct 2019 20:54:12 +0200
> David Marchand <david.marchand@redhat.com> wrote:
>
> > Let's prepare for the ABI freeze.
> >
> > The first patches are about changes that had been announced before (with
> > a patch from Stephen that I took as it is ready as is from my pov).
> >
> > The malloc_heap structure from the memory subsystem can be hidden.
> > The PCI library had some forgotten deprecated APIs that are removed with
> > this series.
> >
> > rte_logs could be hidden, but I am not that confortable about
> > doing it right away: I added an accessor to rte_logs.file, but I am fine
> > with dropping the last patch and wait for actually hiding this in the next
> > ABI break.
>
> 19.11 is an api/abi break so maybe do it now.

Did you look at the 4 new patches too?

Same concern + this was not announced before either.
I went and hid more internals, I did not see an impact on really basic bench.

I would appreciate other opinions.
  

On Thu, 24 Oct 2019 09:32:10 +0200
David Marchand <david.marchand@redhat.com> wrote:

> On Wed, Oct 23, 2019 at 11:10 PM Stephen Hemminger
> <stephen@networkplumber.org> wrote:
> >
> > On Wed, 23 Oct 2019 20:54:12 +0200
> > David Marchand <david.marchand@redhat.com> wrote:
> >  
> > > Let's prepare for the ABI freeze.
> > >
> > > The first patches are about changes that had been announced before (with
> > > a patch from Stephen that I took as it is ready as is from my pov).
> > >
> > > The malloc_heap structure from the memory subsystem can be hidden.
> > > The PCI library had some forgotten deprecated APIs that are removed with
> > > this series.
> > >
> > > rte_logs could be hidden, but I am not that confortable about
> > > doing it right away: I added an accessor to rte_logs.file, but I am fine
> > > with dropping the last patch and wait for actually hiding this in the next
> > > ABI break.  
> >
> > 19.11 is an api/abi break so maybe do it now.  
> 
> Did you look at the 4 new patches too?
> 
> Same concern + this was not announced before either.
> I went and hid more internals, I did not see an impact on really basic bench.
> 
> I would appreciate other opinions.
> 
> 

These all look good. There is probably a lot more that could be
done, adding more accessors in 20.02 could help but more hiding won't happen
again until 20.11
  

On Thu, Oct 24, 2019 at 5:37 PM Stephen Hemminger
<stephen@networkplumber.org> wrote:
> > > > rte_logs could be hidden, but I am not that confortable about
> > > > doing it right away: I added an accessor to rte_logs.file, but I am fine
> > > > with dropping the last patch and wait for actually hiding this in the next
> > > > ABI break.
> > >
> > > 19.11 is an api/abi break so maybe do it now.
> >
> > I went and hid more internals, I did not see an impact on really basic bench.
> >
> > I would appreciate other opinions.
>
> These all look good. There is probably a lot more that could be
> done, adding more accessors in 20.02 could help but more hiding won't happen
> again until 20.11

Yes, I went with the low hanging fruits.
It is a long term effort in any case, when reviewing too.
  

23/10/2019 20:54, David Marchand:
> Let's prepare for the ABI freeze.
> 
> The first patches are about changes that had been announced before (with
> a patch from Stephen that I took as it is ready as is from my pov).
> 
> The malloc_heap structure from the memory subsystem can be hidden.
> The PCI library had some forgotten deprecated APIs that are removed with
> this series.
> 
> rte_logs could be hidden, but I am not that confortable about
> doing it right away: I added an accessor to rte_logs.file, but I am fine
> with dropping the last patch and wait for actually hiding this in the next
> ABI break.
> 
> Changelog since v1:
> - I went a step further, hiding rte_config after de-inlining non critical
>   functions
> 
> Comments?

Except patch 8 (hiding rte_logs),
Acked-by: Thomas Monjalon <thomas@monjalon.net>