Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   "Fabio M. De Francesco" <fmdefrancesco@gmail.com>
To:     Eugeniu Rosca <erosca@de.adit-jv.com>
Cc:     Jaroslav Kysela <perex@perex.cz>, Takashi Iwai <tiwai@suse.com>,
        Mark Brown <broonie@kernel.org>, alsa-devel@alsa-project.org,
        linux-kernel@vger.kernel.org,
        syzbot+205eb15961852c2c5974@syzkaller.appspotmail.com,
        naveenkumar.sunkari@in.bosch.com,
        Eugeniu Rosca <erosca@de.adit-jv.com>,
        Eugeniu Rosca <roscaeugeniu@gmail.com>
Subject: Re: [PATCH] ALSA: pcm: Test for "silence" field in struct "pcm_format_data"
Date:   Tue, 14 Jun 2022 12:43:16 +0200
Message-ID: <2245197.ElGaqSPkdT@opensuse>
In-Reply-To: <20220614095851.GA4199@lxhi-065>
References: <20220409012655.9399-1-fmdefrancesco@gmail.com> <20220614095851.GA4199@lxhi-065>
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="utf-8"
Precedence: bulk

On marted=C3=AC 14 giugno 2022 11:58:51 CEST Eugeniu Rosca wrote:
> Hello Fabio, hello All,
>=20
> On Sa, Apr 09, 2022 at 03:26:55 +0200, Fabio M. De Francesco wrote:
> > Syzbot reports "KASAN: null-ptr-deref Write in
> > snd_pcm_format_set_silence".[1]
> >=20
> > It is due to missing validation of the "silence" field of struct
> > "pcm_format_data" in "pcm_formats" array.
> >=20
> > Add a test for valid "pat" and, if it is not so, return -EINVAL.
> >=20
> > [1] https://lore.kernel.org/lkml/
000000000000d188ef05dc2c7279@google.com/
> >=20
> > Reported-and-tested-by:=20
syzbot+205eb15961852c2c5974@syzkaller.appspotmail.com
> > Signed-off-by: Fabio M. De Francesco <fmdefrancesco@gmail.com>
> > ---
> >=20
> > I wasn't able to figure out the commit for the "Fixes:" tag. If this=20
patch
> > is good, can someone please help with providing this missing=20
information?
> >=20
> >  sound/core/pcm_misc.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >=20
> > diff --git a/sound/core/pcm_misc.c b/sound/core/pcm_misc.c
> > index 4866aed97aac..5588b6a1ee8b 100644
> > --- a/sound/core/pcm_misc.c
> > +++ b/sound/core/pcm_misc.c
> > @@ -433,7 +433,7 @@ int snd_pcm_format_set_silence(snd_pcm_format_t=20
format, void *data, unsigned int
> >  		return 0;
> >  	width =3D pcm_formats[(INT)format].phys; /* physical width */
> >  	pat =3D pcm_formats[(INT)format].silence;
> > -	if (! width)
> > +	if (!width || !pat)
> >  		return -EINVAL;
> >  	/* signed or 1 byte data */
> >  	if (pcm_formats[(INT)format].signd =3D=3D 1 || width <=3D 8) {
>=20
> JFYI, PVS-Studio 7.19 reports:
>=20
> sound/core/pcm_misc.c	409	warn	V560 A part of=20
conditional expression is always false: !pat.

Sorry, I assumed (wrongly!) that when we have

static const struct pcm_format_data=20
pcm_formats[(INT)SNDRV_PCM_FORMAT_LAST+1] =3D {
	[SNDRV_PCM_FORMAT_S8] =3D {
		.width =3D 8, .phys =3D 8, .le =3D -1, .signd =3D 1,
		.silence =3D {},
	},
	[snip]
	/* FIXME: the following two formats are not defined properly yet=20
*/
	[SNDRV_PCM_FORMAT_MPEG] =3D {
		.le =3D -1, .signd =3D -1,
	},
	[SNDRV_PCM_FORMAT_GSM] =3D {
		.le =3D -1, .signd =3D -1,
	},

pointer "silence", and then "pat", must be NULL.

I'd better read again how fields of global struct variables are initialized=
=20
:-(

Thanks for this finding,

=46abio

>=20
> I haven't fully validated the finding, but it appears to be legit,
> since the pointer variable (as opposed to the contents behind the
> pointer) is always non-null, hence !pat always evaluating to false.
>=20
> If the above is true, then the patch likely hasn't introduced any
> regression, but also likely hasn't fixed the original KASAN problem.
>=20
> Or are there alternative views?
>=20
> BR, Eugeniu.
>=20