Re: checkpatch induced patches...

[Joe Perches <joe@perches.com>]

On Wed, 2015-02-11 at 21:24 +0100, Pavel Machek wrote:
> On Wed 2015-02-11 12:20:25, Joe Perches wrote:
> > On Wed, 2015-02-11 at 21:02 +0100, Richard Weinberger wrote:
> > > On Wed, Feb 11, 2015 at 7:36 PM, Dan Carpenter <dan.carpenter@oracle.com> wrote:
> > > > On Wed, Feb 11, 2015 at 10:00:29AM -0800, Joe Perches wrote:
> > > >> I'm half tempted to submit some patch like this to
> > > >> make it difficult to use checkpatch on files outside
> > > >> of drivers/staging.
> > > >> o Only allow checkpatch to be used with the -f/--file
> > > >>   option for drivers/staging/
> > > >> o Add an undocumented --force command line option
> > > > Sure.  We could try that.  I once sent a patch to make -f generate a
> > > > warning about not wasting people's time, but this is also ok.
> > > >> o Make --strict the default for drivers/staging
[]
> > > FYI: We had already a heated debate on that topic.
> > > https://lkml.org/lkml/2014/7/17/415
[]
> > This is basically a patch that implements my suggestion
> > in that thread.
> > https://lkml.org/lkml/2014/7/17/427
> > 
> > I wonder if the undocumented --force option is acceptable
> > to Pavel and Kalle.

> Undocumented options are evil... You can add warning about not wasting
> people's time in --force documentation...

Yeah, I had added --force to the help text
then removed it before sending, so I suppose
adding a warning there is OK too.

Nobody reads the --help text anyway.

Dan/Andrew/Greg?  You got a preference?

Maybe some help/warning text like:

  --force                    Without --force, checkpatch will not scan files
                             using -f or --file outside of drivers/staging/...
                             Do not use this option merely to create potential
                             patches that are uncompiled or untested.