Re: [PATCH 2/2] net: phy: dp83826: Add support for straps reading

[Jean-Michel Hautbois]

Hello Andrew,

On 03/03/2025 18:50, Andrew Lunn wrote:
> On Mon, Mar 03, 2025 at 06:37:28PM +0100, Jean-Michel Hautbois wrote:
> > Hi Andrew,
> >
> > On 03/03/2025 18:20, Andrew Lunn wrote:
> > > On Mon, Mar 03, 2025 at 06:05:52PM +0100, Jean-Michel Hautbois wrote:
> > > > When the DP83826 is probed, read the straps, and apply the default
> > > > settings expected. The MDI-X is not yet supported, but still read the
> > > > strap.
> > >
> > > What about backwards compatibility? I expect this changes the
> > > behaviour of the device, potentially introducing regressions?  Please
> > > add an explanation of why this is safe.
> >
> > I am not certain it is safe. As far as I know that if straps are used on the
> > hardware, then it should be used, and if the behavior has to be different,
> > then userspace can change it (or any other way). Am I wrong ?
>
> First off, what does the datasheet say about these straps?
>
> Straps generally configure the hardware, without software being
> involved. It seems odd you need software to read the straps and apply
> them.
>
> Why do you need this? What is your use case. As you said, user space
> can configure all this, so why are you not doing that?

Thanks you for your remarks, indeed, ip link set + ethtool -s works fine 
to force the link mode wanted. Event the master/slave can be specified, 
so there is no need for this patch 2/2.

Thanks !
JM