Received-SPF: pass (google.com: domain of linux-kernel+bounces-146607-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Message-ID: <a582c379f1bc2bc0d16b46ece9211bac40b4b7cf.camel@perches.com>
Subject: Re: [PATCH] checkpatch: error if file terminates without a new-line
From: Joe Perches <joe@perches.com>
To: Dan Carpenter <dan.carpenter@linaro.org>, Greg KH
	 <gregkh@linuxfoundation.org>
Cc: Prasad Pandit <pj.pandit@yahoo.in>, "linux-kernel@vger.kernel.org"
	 <linux-kernel@vger.kernel.org>, Dwaipayan Ray <dwaipayanray1@gmail.com>, 
	Lukas Bulwahn <lukas.bulwahn@gmail.com>, Andy Whitcroft <apw@canonical.com>
Date: Tue, 16 Apr 2024 02:55:11 -0700
In-Reply-To: <edf23b1b-303a-46eb-b455-8c766c84ddf1@moroto.mountain>
References: <20240330033858.3272184-1-ppandit@redhat.com>
	 <1568855168.5949322.1713251087824@mail.yahoo.com>
	 <2024041627-overgrown-ripping-2348@gregkh>
	 <edf23b1b-303a-46eb-b455-8c766c84ddf1@moroto.mountain>
Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.48.4 (3.48.4-1.fc38) 
Precedence: bulk
MIME-Version: 1.0

On Tue, 2024-04-16 at 11:59 +0300, Dan Carpenter wrote:
> On Tue, Apr 16, 2024 at 10:48:27AM +0200, Greg KH wrote:
> > > diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
> > > index 9c4c4a61bc83..df34c0709410 100755
> > > --- a/scripts/checkpatch.pl
> > > +++ b/scripts/checkpatch.pl
> > > @@ -2795,6 +2795,13 @@ sub process {
> > > =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 $is_patch =3D 1;
> > > =A0=A0=A0 =A0=A0=A0 }
> > >=20
> > > +# check if patch terminates file without a new line (\n)
> > > +=A0 =A0 =A0 =A0 if ($line =3D~ /^\\ No newline at end of file$/
> > > +=A0 =A0 =A0 =A0 =A0 =A0 and $rawlines[$linenr - 2] =3D~ /^\+.*$/) {
> > > +=A0 =A0 =A0 =A0 =A0 =A0 ERROR("NOEOL_FILE",
> > > +=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 "patch terminates file without a=
 new line (\\n).");
> > > +=A0 =A0 =A0 =A0 }
> >=20
> > Why is this a problem?  files without a new line should not cause
> > problems with a compiler, right?  You don't have a justification for wh=
y
> > this change needs to be checked for anywhere.
> >=20
>=20
> I gave him such a good reason too...  It breaks `cat file.c`.  Plus, it
> looks weird in `git log -p` because it has a "No newline at the end of
> file" comment.
>=20
> regards,
> dan carpenter
>=20
> diff --git a/test.c b/test.c
> new file mode 100644
> index 000000000000..d808cac2d962
> --- /dev/null
> +++ b/test.c
> @@ -0,0 +1,12 @@
> +#include <stdio.h>
> +#include <stdbool.h>
> +#include "check_debug.h"
> +
> +void kfree(int *p);
> +
> +int *p;
> +int main(void)
> +{
> +       kfree(p);
> +       *p =3D 1;
> +}
> \ No newline at end of file
>=20
> commit f4a997924122d0094675c897a220371f0a129d90
>=20

There's an existing check for this.
I believe it at least used to work.
Fix that instead.

# check for adding lines without a newline.
		if ($line =3D~ /^\+/ && defined $lines[$linenr] && $lines[$linenr] =3D~ /=
^\\ No newline at end of file/) {
			if (WARN("MISSING_EOF_NEWLINE",
			         "adding a line without newline at end of file\n" . $herecurr) &=
&
			    $fix) {
				fix_delete_line($fixlinenr+1, "No newline at end of file");
			}
		}