RE: [PATCH net-next v2 1/2] net: stmmac: Only enable enhanced addressing mode when needed

[Jose Abreu <Jose.Abreu@synopsys.com>]

From: Thierry Reding <thierry.reding@gmail.com>
Date: Sep/10/2019, 14:54:27 (UTC+00:00)

> On Tue, Sep 10, 2019 at 08:32:38AM +0000, Jose Abreu wrote:
> > From: Thierry Reding <thierry.reding@gmail.com>
> > Date: Sep/09/2019, 20:11:27 (UTC+00:00)
> > 
> > > On Mon, Sep 09, 2019 at 04:07:04PM +0000, Jose Abreu wrote:
> > > > From: Thierry Reding <thierry.reding@gmail.com>
> > > > Date: Sep/09/2019, 16:25:45 (UTC+00:00)
> > > > 
> > > > > @@ -92,6 +92,7 @@ struct stmmac_dma_cfg {
> > > > >  	int fixed_burst;
> > > > >  	int mixed_burst;
> > > > >  	bool aal;
> > > > > +	bool eame;
> > > > 
> > > > bools should not be used in struct's, please change to int.
> > > 
> > > Huh? Since when? "aal" right above it is also bool. Can you provide a
> > > specific rationale for why we shouldn't use bool in structs?
> > 
> > Please see https://lkml.org/lkml/2017/11/21/384.
> 
> The context is slightly different here. stmmac_dma_cfg exists once for
> each of these ethernet devices in the system, and I would assume that in
> the vast majority of cases there's exactly one such device in the system
> so the potential size increase is very small. On the other hand, there
> are potentially very many struct sched_dl_entity, so the size impact is
> multiplied.
> 
> Anyway, if you insist I'll rewrite this to use an unsigned int bitfield.

For new code I would rather prefer "int" but I guess it's up to David to 
decide this. I'm okay with both options as the check for this usage was 
removed in checkpatch: 
https://lkml.org/lkml/2019/1/10/975

---
Thanks,
Jose Miguel 
Abreu