Em Sat, 26 Mar 2022 00:21:09 +0100
Mauro Carvalho Chehab <mchehab@xxxxxxxxxx> escreveu:
> Em Fri, 25 Mar 2022 13:19:28 -0600
> Jonathan Corbet <corbet@xxxxxxx> escreveu:
>
> > Mauro Carvalho Chehab <mchehab@xxxxxxxxxx> writes:
> >
> > > Such output could be helpful while debugging it, but its main
> > > goal is to tell kernel_feat.py about what files were used
> > > by the script. Thie way, kernel_feat.py can add those as
> > > documentation dependencies.
> > >
> > > Signed-off-by: Mauro Carvalho Chehab <mchehab@xxxxxxxxxx>
> >
> > So I think these are worth getting into 5.18,
>
> Yeah, agreed.
>
> > but I do have one question:
> >
> > > @@ -95,6 +97,10 @@ sub parse_feat {
> > > return if ($file =~ m,($prefix)/arch-support.txt,);
> > > return if (!($file =~ m,arch-support.txt$,));
> > >
> > > + if ($enable_fname) {
> > > + printf "#define FILE %s\n", abs_path($file);
> > > + }
> > > +
> >
> > Why do you output the file names in this format? This isn't input to
> > the C preprocessor, so the #define just seems strange. What am I
> > missing here?
>
> Well, I didn't think much about that... I just ended using a way that is
> already used on get_abi.pl, and was originally imported from kernel-doc :-)
>
> It could be using whatever other tag, but I would keep those three scripts
> using a similar markup string for file names and line numbers:
>
> scripts/get_abi.pl:
> printf "#define LINENO %s%s#%s\n\n", $prefix, $file[0], $data{$what}->{line_no};
>
> scripts/kernel-doc:
> print "#define LINENO " . $lineno . "\n";
Btw, maybe we could replace them tree with a Sphinx comment, like:
get_feat.pl:
.. FILE <file_name>
kernel-doc:
.. LINE <line_number>
get_abi.pl:
.. FILE_LINE <file_name>:<line_number>
(or something similar)
Just let me know what you prefer and I can take care of the needed
changes on this patch and, if it is the case, writing the extra patches
in order to use the same model on kernel-doc and get_abi.pl.
Thanks,
Mauro
