Re: [PATCH v2 2/2] mfd: intel_quark_i2c_gpio: enable MSI interrupt

[Andy Shevchenko <andriy.shevchenko@linux.intel.com>]

On Wed, Mar 24, 2021 at 03:10:35PM +0000, Lee Jones wrote:
> On Wed, 24 Mar 2021, Andy Shevchenko wrote:
> > On Wed, Mar 24, 2021 at 01:07:23PM +0000, Lee Jones wrote:
> > > On Wed, 24 Mar 2021, Andy Shevchenko wrote:
> > > > On Wed, Mar 24, 2021 at 11:50:33AM +0000, Lee Jones wrote:

...

> > > > It's a two way road:
> > > > a) driver states that it needs only 1 vector and it's enough to it
> > > > b) hardware must provide at least 1 vector to be served by this driver.
> > > > 
> > > > Look again into grepped output. Most of drivers that define it as an variable
> > > > may dynamically adapt to the different amount of IRQ vectors. When it's static,
> > > > usually drivers just hard code those values.
> > > > 
> > > > I'm really don't see a point to define them _in this driver_.
> > > 
> > > That's fine.  I just felt like I had to ask.
> > > 
> > > Would you consider a comment that lets people unfamiliar with the API
> > > what the values mean?
> > > 
> > > Something to the tune of:
> > > 
> > >   "This driver requests 1 (and only 1) IRQ vector"
> > 
> > 
> > Rather
> > 
> >     "This driver requests only 1 (and it's enough) IRQ vector"
> 
>  "This driver only requires 1 IRQ vector"

Thanks! v3 has been sent.

> > or something like this.
> > 
> > Should I send a patch with the comment included? If so, please suggest if it's
> > good from English grammar/style perspective.

-- 
With Best Regards,
Andy Shevchenko