Re: [PATCH v2] checkpatch: report warning with files ending with blank lines

[Joe Perches <joe@perches.com>]

On Wed, 2026-10-07 at 18:52 +0200, Christian Marangi wrote:
> On Wed, Oct 07, 2026 at 09:49:48AM -0700, Joe Perches wrote:
> On Wed, 2026-10-07 at 14:28 +0200, Christian Marangi wrote:
> > > On applying patch, git, depending on the default configuration, might
> > > complain for any file having blank lines at end of file.
> > I was not clear enough in my first reply.
> > 
> > I meant that the test should be done only once, not
> > for each line of the input.
> > 
> > After the end of the process() loop of
> > 
> 	foreach my $rawline (@rawlines) {
> 		[...]
> 	}
> > 
> > a single test of $rawlines[-1] could be done.
> 
> Yes I also was with that idea but then I quickly figure it out that we are
> not testing if the .patch have blank line at EOF. But if the single file of
> the patch have blank line at EOF.
> 
> That can only be done in the process() while scanning each line as it does
> progressively check new files in it.
> 
> (unless I'm missing something)

Dunno.  Maybe I'm missing something but maybe something like this:

diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
index f424dafce5bce..d2f99127a9a9a 100755
--- a/scripts/checkpatch.pl
+++ b/scripts/checkpatch.pl
@@ -7850,6 +7850,13 @@ sub process {
 		exit(0);
 	}
 
+	# check if the last line is a blank line
+	if (($is_patch && $rawlines[-1] =~ /^\+\s*$/) ||
+	    (!$is_patch && $rawlines[-1] =~ /^\s*$/)) {
+		WARN("LINE_SPACING",
+		     "Blank lines at EOF are not good\n");
+	}
+
 	# In mailback mode only produce a report in the negative, for
 	# things that appear to be patches.
 	if ($mailback && ($clean == 1 || !$is_patch)) {