Re: [PATCH v3 10/11] dtc: dt-check-style: Add warning for redundant white-spaces

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] False positives in dt-check-style whitespace checks due to unstripped strings and inline comments.
--

--- Patch [10]: [PATCH v3 10/11] dtc: dt-check-style: Add warning for redundant white-spaces ---
commit f5a3cbd31bda433ce990bca1b2244f3a3da27d87
Author: Krzysztof Kozlowski <[email protected]>

dtc: dt-check-style: Add warning for redundant white-spaces

This commit adds a warning in dt-check-style to catch excessive spaces or
tabs around '=', before '{', and before ';'. The checking is split between
relaxed and strict modes to accommodate existing in-tree usage.

> diff --git a/scripts/dtc/dt-check-style b/scripts/dtc/dt-check-style
> index 1430c675bd951..aa2072957a1a0 100755
> --- a/scripts/dtc/dt-check-style
> +++ b/scripts/dtc/dt-check-style
> @@ -348,6 +348,36 @@ def check_trailing_whitespace(ctx):
>              yield (dl.lineno, 'trailing whitespace')
>  
>  
> +def check_redundant_whitespace(ctx):
> +    """No whitespace between brackets or other code elements.
> +    See also check_value_whitespace() for more checks."""
> +    for dl in ctx.lines:
> +        if dl.linetype in (LineType.COMMENT, LineType.COMMENT_BODY,
> +                           LineType.COMMENT_END, LineType.COMMENT_START,
> +                           LineType.PREPROCESSOR):
> +            continue
> +        if re.search(r'(\s\s+|\t){', dl.raw):
> +            yield (dl.lineno, 'extra whitespace before {')
> +        if re.search(r':(\s\s+|\t)', dl.raw):
> +            yield (dl.lineno, 'extra whitespace after :')
> +        if re.search(r'\s+;', dl.raw):

[Severity: Medium]
Does using dl.raw directly here cause false positives for inline comments or
string literals?

For example, if a line has bootargs = "console=ttyS0  rdinit=/init"; or
reg = <1>; /* my comment ; */, since re.search evaluates dl.raw directly,
it looks like it will match the whitespace inside the string or the inline
comment.

Would it be safer to strip comments or strings via
_strip_strings_and_comments(dl.raw) before evaluating the regular expressions?

> +            yield (dl.lineno, 'extra whitespace before ;')
> +
> +
> +def check_redundant_whitespace_strict(ctx):
> +    """No whitespace between brackets or other code elements.
> +    See also check_value_whitespace() for more checks."""
> +    for dl in ctx.lines:
> +        if dl.linetype in (LineType.COMMENT, LineType.COMMENT_BODY,
> +                           LineType.COMMENT_END, LineType.COMMENT_START,
> +                           LineType.PREPROCESSOR):
> +            continue
> +        if re.search(r'(\s\s+|\t)=', dl.raw):
> +            yield (dl.lineno, 'extra whitespace before =')
> +        if re.search(r'=(\s\s+|\t)', dl.raw):

[Severity: Medium]
Similarly, will this strict check also trigger false positives on strings
that intentionally contain multiple spaces, such as
compatible = "vendor,device  v2";, because dl.raw retains the unstripped
string content?

> +            yield (dl.lineno, 'extra whitespace after =')

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260803-n-dts-style-checker-continued-v3-0-6c9776928cea@oss.qualcomm.com?part=10