Re: [PATCH 12/21] clk: sunxi: clk-sun6i-ar100: Demote non-conformant kernel-doc header

[Lee Jones]

On Tue, 26 Jan 2021, Maxime Ripard wrote:

> On Tue, Jan 26, 2021 at 12:45:31PM +0000, Lee Jones wrote:
> > Fixes the following W=1 kernel build warning(s):
> > 
> >  drivers/clk/sunxi/clk-sun6i-ar100.c:26: warning: Function parameter or member 'req' not described in 'sun6i_get_ar100_factors'
> > 
> > Cc: "Emilio López" <emilio@xxxxxxxxxxxxx>
> > Cc: Michael Turquette <mturquette@xxxxxxxxxxxx>
> > Cc: Stephen Boyd <sboyd@xxxxxxxxxx>
> > Cc: Maxime Ripard <mripard@xxxxxxxxxx>
> > Cc: Chen-Yu Tsai <wens@xxxxxxxx>
> > Cc: Jernej Skrabec <jernej.skrabec@xxxxxxxx>
> > Cc: Boris BREZILLON <boris.brezillon@xxxxxxxxxxxxxxxxxx>
> > Cc: linux-clk@xxxxxxxxxxxxxxx
> > Cc: linux-arm-kernel@xxxxxxxxxxxxxxxxxxx
> > Signed-off-by: Lee Jones <lee.jones@xxxxxxxxxx>
> > ---
> >  drivers/clk/sunxi/clk-sun6i-ar100.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/drivers/clk/sunxi/clk-sun6i-ar100.c b/drivers/clk/sunxi/clk-sun6i-ar100.c
> > index e1b7d0929cf7f..54babc2b4b9ee 100644
> > --- a/drivers/clk/sunxi/clk-sun6i-ar100.c
> > +++ b/drivers/clk/sunxi/clk-sun6i-ar100.c
> > @@ -16,7 +16,7 @@
> >  
> >  #include "clk-factors.h"
> >  
> > -/**
> > +/*
> >   * sun6i_get_ar100_factors - Calculates factors p, m for AR100
> >   *
> >   * AR100 rate is calculated as follows
> 
> This is the sixth patch doing the exact same thing over the files in
> that folder you sent. Please fix all the occurences at once

No.  That would make the whole clean-up process 10x harder than it
already is

Before starting this endeavour there were 18,000+ warnings spread over
100's of files and 10's of subsystems that needed addressing (only a
couple thousand left now thankfully).  Some issues vastly different,
some duplicated (much too much copy/pasting going which made things
very frustrating at times).

Anyway, in order to work though them all gracefully and in a sensible
time-frame I had to come up with a workable plan.  Each subsystem is
compiled separately and a script attempts to take out duplicate
warnings and takes me through the build-log one file at a time.  Once
all of the warnings are fixed in a source-file, it moves on to the
next file.  The method is clean and allows me to handle this
gargantuan task in bite-sized chunks.

Going though and pairing up similar changes is unsustainable for a
task like this.  It would add a lot of additional overhead and would
slow down the rate of acceptance since source files tend to have
different reviewers/maintainers - some working faster to review
patches than others, leading to excessive lag times waiting for that
one reviewer who takes weeks to review.  Having each file addressed
in a separate patch also helps revertability and bisectability.  Not
such a big problem with the documentation patches, but still.

Admittedly doing it this way *can* look a bit odd in *some* patch-sets
when they hit the MLs - particularly clock it seems, where there
hasn't even been a vague attempt to document any of the parameters in
the kernel-doc headers - however the alternative would mean nothing
would get done!

-- 
Lee Jones [李琼斯]
Senior Technical Lead - Developer Services
Linaro.org │ Open source software for Arm SoCs
Follow Linaro: Facebook | Twitter | Blog