Am 07.02.19 um 18:50 schrieb Joe Perches:
On Thu, 2019-02-07 at 09:34 -0800, Matthew Wilcox wrote:
On Thu, Feb 07, 2019 at 09:31:20AM -0800, Joe Perches wrote:
It's not clear to me what you are asking checkpatch to do here.
It may be reasonable for checkpatch to invoke kernel-doc on some
portion of a patch, but I'm not sure how valuable it will be.
I was just hoping to match:
* Returns:
Or to quote it properly for regexes ...
^ +\* *Returns:
(I think ...)
I can't see that matching C or assembler.
checkpatch doesn't attempt to enforce any formatting standard
on kernel-doc comments.
There doesn't seem to be much standardization for kernel-doc
in the first place.
Just for the * return: case:
$ git grep -P -i '^\s*\*\s*returns?\s*:' -- '*.[ch]' | \
grep -P -oh -i '\*\s*returns?\s*:' | \
sort | uniq -c | sort -rn
5153 * Return:
2534 * Returns:
1077 * RETURN:
358 * RETURNS:
173 * RETURNS:
171 * returns:
153 * return:
148 * Return :
72 * Returns :
61 * Returns:
37 * Returns:
30 * returns:
27 * return:
22 * Return:
20 * Returns :
19 * Return:
15 * RETURNS:
11 * return:
6 * return :
6 * return:
5 * returns :
3 *Returns:
3 * Returns :
3 * returns:
2 *RETURNS:
2 * Returns:
2 * Returns:
2 * returns:
2 * RETURN :
2 * Return:
2 * Return:
2 * return :
2 * return:
1 * RETURNS:
1 * RETURNs:
1 * Returns:
1 * Returns:
1 * Returns:
1 * RETURN:
1 * Return:
1 * Return:
1 * return :
I think standarization is more something that scripts/kernel-doc
could or should do.
BTW: kernel-doc parser accept 'return' and 'returns':
} elsif ($newsection =~ m/^return?$/i) {
$newsection = $section_return;
Is there really a need to be standardize this? From the list above I think
there are only a few where it fails.
-- Markus --
