RE: [PATCH v3 19/27] thunderbolt: Add new Thunderbolt PCI IDs

From: Mario.Limonciello
Date: Mon Jun 05 2017 - 11:50:44 EST


> -----Original Message-----
> From: Mika Westerberg [mailto:mika.westerberg@xxxxxxxxxxxxxxx]
> Sent: Monday, June 5, 2017 10:41 AM
> To: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
> Cc: Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>; Lukas Wunner
> <lukas@xxxxxxxxx>; Andreas Noever <andreas.noever@xxxxxxxxx>; Michael
> Jamet <michael.jamet@xxxxxxxxx>; Yehezkel Bernat <yehezkel.bernat@xxxxxxxxx>;
> Amir Levy <amir.jer.levy@xxxxxxxxx>; Andy Lutomirski <luto@xxxxxxxxxx>;
> Limonciello, Mario <Mario_Limonciello@xxxxxxxx>; Dominguez, Jared
> <Jared_Dominguez@xxxxxxxx>; linux-kernel@xxxxxxxxxxxxxxx
> Subject: Re: [PATCH v3 19/27] thunderbolt: Add new Thunderbolt PCI IDs
>
> On Mon, Jun 05, 2017 at 05:20:36PM +0200, Greg Kroah-Hartman wrote:
> > On Mon, Jun 05, 2017 at 05:21:34PM +0300, Andy Shevchenko wrote:
> > > On Mon, 2017-06-05 at 16:33 +0300, Mika Westerberg wrote:
> > > > On Mon, Jun 05, 2017 at 03:55:07PM +0300, Mika Westerberg wrote:
> > > > > >
> > >
> > > > It turns out that this is not too big change after all so I think it
> > > > can
> > > > be included in the ICM patch if you want.
> > > >
> > > > Basically we would have module parameter:
> > > >
> > > > cm=auto|native|icm
> > > >
> > > > where auto is the default.
> > > >
> > > > Does that work?
> > >
> > >
> > > I remember Greg is opposite to being fan of module parameters in new
> > > code, so, please, if you are still going to do that, do it a separate
> > > patch which will be last in the series and can be discussed separately
> > > w.o. preventing everything else being applied.
> >
> > Yes, you should never need module parameters in "modern" code. As no
> > one will ever set them, or know what to do here. Make it "automatic"
> > please.
>
> Well in current series it is already automatic :-)
>
> What Lukas wants here is a way to force a certain connection manager
> implementation to be used instead.

And presumably that's mostly for debugging purposes and will lead
to additional quirks to get "automatic" to do the right thing on those
machines that's it's found to be needed or more beneficial.