Received-SPF: pass (google.com: domain of linux-kernel+bounces-28234-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Date: Tue, 16 Jan 2024 15:09:50 -0500
From: Steven Rostedt <rostedt@goodmis.org>
To: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Cc: nathanl@linux.ibm.com, Kees Cook <keescook@chromium.org>,
 linux-kernel@vger.kernel.org
Subject: Re: [PATCH] seq_buf: make DECLARE_SEQ_BUF() usable
Message-ID: <20240116150950.5f444a04@gandalf.local.home>
In-Reply-To: <407c467e-1f0b-4549-951a-83c295b27733@wanadoo.fr>
References: <20240116-declare-seq-buf-fix-v1-1-915db4692f32@linux.ibm.com>
	<407c467e-1f0b-4549-951a-83c295b27733@wanadoo.fr>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

On Tue, 16 Jan 2024 20:40:29 +0100
Christophe JAILLET <christophe.jaillet@wanadoo.fr> wrote:

> Le 16/01/2024 =C3=A0 15:09, Nathan Lynch via B4 Relay a =C3=A9crit=C2=A0:
> > From: Nathan Lynch <nathanl@linux.ibm.com>
> >=20
> > Using the address operator on the array doesn't work:
> >=20
> > /include/linux/seq_buf.h:27:27: error: initialization of =E2=80=98char =
*=E2=80=99
> >    from incompatible pointer type =E2=80=98char (*)[128]=E2=80=99
> >    [-Werror=3Dincompatible-pointer-types]
> >     27 |                 .buffer =3D &__ ## NAME ## _buffer,       \
> >        |                           ^
> >=20
> > Apart from fixing that, we can improve DECLARE_SEQ_BUF() by using a
> > compound literal to define the buffer array without attaching a name
> > to it. This makes the macro a single statement, allowing constructs
> > such as:
> >=20
> >    static DECLARE_SEQ_BUF(my_seq_buf, MYSB_SIZE);
> >=20
> > to work as intended.
> >=20
> > Signed-off-by: Nathan Lynch <nathanl@linux.ibm.com>
> > Fixes: dcc4e5728eea ("seq_buf: Introduce DECLARE_SEQ_BUF and seq_buf_st=
r()")
> > ---
> >   include/linux/seq_buf.h | 3 +--
> >   1 file changed, 1 insertion(+), 2 deletions(-)
> >=20
> > diff --git a/include/linux/seq_buf.h b/include/linux/seq_buf.h
> > index 5fb1f12c33f9..c44f4b47b945 100644
> > --- a/include/linux/seq_buf.h
> > +++ b/include/linux/seq_buf.h
> > @@ -22,9 +22,8 @@ struct seq_buf {
> >   };
> >  =20
> >   #define DECLARE_SEQ_BUF(NAME, SIZE)			\
> > -	char __ ## NAME ## _buffer[SIZE] =3D "";		\
> >   	struct seq_buf NAME =3D {				\
> > -		.buffer =3D &__ ## NAME ## _buffer,	\
> > +		.buffer =3D (char[SIZE]) { 0 },		\
> >   		.size =3D SIZE,				\
> >   	} =20
>=20
> Hi,
>=20
> just removing the & in ".buffer =3D __ ## NAME ## _buffer, \" also works =
IIRC.
>=20
> See [1], which unfortunately has been unnoticed.
>=20
> CJ
>=20
>=20
> [1]:=20
> https://lore.kernel.org/all/2a534333-b5f6-4b1d-b4b8-a1a71f91c3ff@wanadoo.=
fr/

I guess I missed that.

But it still doesn't fix this case:

 static DECLARE_SEQ_BUF(my_seq_buf, MYSB_SIZE);

Which this patch does.


-- Steve