[PATCH] checkpatch: do not try to sanity test cover letters

[PATCH] checkpatch: do not try to sanity test cover letters

[Paul Gortmaker]

One possibly common workflow is this:

	git format-patch -o mypatches --cover-letter ^start end
	./scripts/checkpatch.pl mypatches/*

The problem with the above is that checkpatch.pl will try to
parse the cover-letter, and of course complain that it can not
find any unified diff within.

It is pretty safe for us to assume "0000-cover-letter.patch" is
not in fact a patch and simply skip processing of it.

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
---

[what I know about perl could be written on the back of a postage
 stamp in crayon, so please feel free to reimplement the overall
 concept of this change as you see fit.... ]

diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
index a3b9782..fc22f4b 100755
--- a/scripts/checkpatch.pl
+++ b/scripts/checkpatch.pl
@@ -6,6 +6,7 @@
 # Licensed under the terms of the GNU GPL License version 2
 
 use strict;
+use File::Basename;
 
 my $P = $0;
 $P =~ s@.*/@@g;
@@ -382,6 +383,10 @@ for my $filename (@ARGV) {
 		open($FILE, '<', "$filename") ||
 			die "$P: $filename: open failed - $!\n";
 	}
+	if (basename($filename) eq '0000-cover-letter.patch') {
+		print "Skipping cover letter $filename\n";
+		next;
+	}
 	if ($filename eq '-') {
 		$vname = 'Your patch';
 	} else {
-- 
1.7.9.1

Re: [PATCH] checkpatch: do not try to sanity test cover letters

[Joe Perches]

On Mon, 2012-02-27 at 19:29 -0500, Paul Gortmaker wrote:
> One possibly common workflow is this:
> 
> 	git format-patch -o mypatches --cover-letter ^start end
> 	./scripts/checkpatch.pl mypatches/*

I use a script for this and don't put the check
in checkpatch but put the check in bash.

Re: [PATCH] checkpatch: do not try to sanity test cover letters

[Paul Gortmaker]

On Mon, Feb 27, 2012 at 7:36 PM, Joe Perches <joe@perches.com> wrote:
> On Mon, 2012-02-27 at 19:29 -0500, Paul Gortmaker wrote:
>> One possibly common workflow is this:
>>
>>       git format-patch -o mypatches --cover-letter ^start end
>>       ./scripts/checkpatch.pl mypatches/*
>
> I use a script for this and don't put the check
> in checkpatch but put the check in bash.

Sure, and I can do the same.  But my question to you is whether
you think the above is a common workflow, and if the false positives
that it generates will decrease the number of people likely to make
using it a part of their routine?

Paul.
--

>
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

Re: [PATCH] checkpatch: do not try to sanity test cover letters

[Joe Perches]

On Mon, 2012-02-27 at 19:47 -0500, Paul Gortmaker wrote:
> On Mon, Feb 27, 2012 at 7:36 PM, Joe Perches <joe@perches.com> wrote:
> > On Mon, 2012-02-27 at 19:29 -0500, Paul Gortmaker wrote:
> >> One possibly common workflow is this:
> >>
> >>       git format-patch -o mypatches --cover-letter ^start end
> >>       ./scripts/checkpatch.pl mypatches/*
> >
> > I use a script for this and don't put the check
> > in checkpatch but put the check in bash.
> 
> Sure, and I can do the same.  But my question to you is whether
> you think the above is a common workflow, and if the false positives
> that it generates will decrease the number of people likely to make
> using it a part of their routine?

Dunno.  I generally think that tools should report
errors when the input given them is inappropriate.

cheers, Joe

Re: [PATCH] checkpatch: do not try to sanity test cover letters

[Andy Whitcroft]

On Mon, Feb 27, 2012 at 04:50:11PM -0800, Joe Perches wrote:
> On Mon, 2012-02-27 at 19:47 -0500, Paul Gortmaker wrote:
> > On Mon, Feb 27, 2012 at 7:36 PM, Joe Perches <joe@perches.com> wrote:
> > > On Mon, 2012-02-27 at 19:29 -0500, Paul Gortmaker wrote:
> > >> One possibly common workflow is this:
> > >>
> > >>       git format-patch -o mypatches --cover-letter ^start end
> > >>       ./scripts/checkpatch.pl mypatches/*
> > >
> > > I use a script for this and don't put the check
> > > in checkpatch but put the check in bash.
> > 
> > Sure, and I can do the same.  But my question to you is whether
> > you think the above is a common workflow, and if the false positives
> > that it generates will decrease the number of people likely to make
> > using it a part of their routine?
> 
> Dunno.  I generally think that tools should report
> errors when the input given them is inappropriate.

Tend to agree, the more complex we make checkpatch the more often it
will be wrong.  And it is wrong enough already.

-apw