Re: [PATCH 4/4] staging: rtl8723bs: checkpatch - resolve indentation and line width

From: Greg KH
Date: Fri May 12 2017 - 09:55:14 EST


On Fri, May 12, 2017 at 06:24:34AM -0700, Matthew Giassa wrote:
> * Matthew Giassa <matthew@xxxxxxxxxx> [2017-05-12 05:57:44 -0700]:
>
> > * Greg KH <gregkh@xxxxxxxxxxxxxxxxxxx> [2017-05-12 11:30:08 +0200]:
> >
> > > On Thu, May 11, 2017 at 06:45:24PM -0700, Matthew Giassa wrote:
> > > > +#define REG_INT_MIG_8723B 0x0304 /* Interrupt Migration */
> > > > +#define REG_BCNQ_DESA_8723B 0x0308 /* TX Beacon Descriptor Address
> > > > + */
> > > > +#define REG_HQ_DESA_8723B 0x0310 /* TX High Queue Descriptor
> > > > + * Address
> > > > + */
> > >
> > > Ick, that looks worse to me now, doesn't it to you? Please leave the
> > > original as-is.
>
> Paring down CC list to reduce noise for off-topic question.
>
> Quick question: in the trivial case, such as a simple block comment, the
> style guide (process/coding-style.rst) proposes this style:
>
> /*
> * Some comments that span over several lines until column limit.
> * More comments that span over several lines until column limit.
> */
>
> Though I see a similar variant often used, which I use by default:
>
> /* Some comments that span over several lines until column limit.
> * More comments that span over several lines until column limit.
> */
>
> For cases with code plus trailing (lengthy) comment, is it preferred to
> let it go past the 80 column limit, or to use one of the following
> multi-line styles? ie:
>
> Type I:
> #define REG_BCNQ_DESA_8723B 0x0308 /* TX Beacon Descriptor
> * Address */
>
> Type II (Ugly):
> #define REG_BCNQ_DESA_8723B 0x0308 /* TX Beacon Descriptor
> * Address
> */
>

Neither, best is:
/* TX Beacon Descriptor Address */
#define REG_BCNQ_DESA_8723B 0x0308

> Finally, would it be worth proposing the addition of this minor
> exception to the style guide?

Not really, it's almost too long as it is today.

thanks,

greg k-h