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:   Jernej =?utf-8?B?xaBrcmFiZWM=?= <jernej.skrabec@gmail.com>
To:     linux-media@vger.kernel.org, Maxime Ripard <mripard@kernel.org>,
        Paul Kocialkowski <paul.kocialkowski@bootlin.com>,
        Mauro Carvalho Chehab <mchehab@kernel.org>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Chen-Yu Tsai <wens@csie.org>,
        Samuel Holland <samuel@sholland.org>,
        Nicolas Dufresne <nicolas.dufresne@collabora.com>
Cc:     kernel@collabora.com,
        Dmitry Osipenko <dmitry.osipenko@collabora.com>,
        stable@vger.kernel.org, linux-staging@lists.linux.dev,
        linux-arm-kernel@lists.infradead.org, linux-sunxi@lists.linux.dev,
        linux-kernel@vger.kernel.org
Subject: Re: Re: [PATCH v1 3/3] media: cedrus: Fix endless loop in cedrus_h265_skip_bits()
Date:   Thu, 25 Aug 2022 23:13:52 +0200
Message-ID: <1733932.VLH7GnMWUR@kista>
In-Reply-To: <52bb86cf12450ce78d2f196a1b86b4137ec61a07.camel@collabora.com>
References: <20220818203308.439043-1-nicolas.dufresne@collabora.com> <5849126.lOV4Wx5bFT@jernej-laptop> <52bb86cf12450ce78d2f196a1b86b4137ec61a07.camel@collabora.com>
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Dne petek, 19. avgust 2022 ob 17:39:25 CEST je Nicolas Dufresne napisal(a):
> Le vendredi 19 ao=C3=BBt 2022 =C3=A0 06:16 +0200, Jernej =C5=A0krabec a =
=C3=A9crit :
> > Dne =C4=8Detrtek, 18. avgust 2022 ob 22:33:08 CEST je Nicolas Dufresne=
=20
napisal(a):
> > > From: Dmitry Osipenko <dmitry.osipenko@collabora.com>
> > >=20
> > > The busy status bit may never de-assert if number of programmed skip
> > > bits is incorrect, resulting in a kernel hang because the bit is poll=
ed
> > > endlessly in the code. Fix it by adding timeout for the bit-polling.
> > > This problem is reproducible by setting the data_bit_offset field of
> > > the HEVC slice params to a wrong value by userspace.
> > >=20
> > > Cc: stable@vger.kernel.org
> > > Reported-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
> > > Signed-off-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
> > > Signed-off-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
> >=20
> > Fixes tag would be nice.
> >=20
> > > ---
> > >=20
> > >  drivers/staging/media/sunxi/cedrus/cedrus_h265.c | 6 ++++--
> > >  1 file changed, 4 insertions(+), 2 deletions(-)
> > >=20
> > > diff --git a/drivers/staging/media/sunxi/cedrus/cedrus_h265.c
> > > b/drivers/staging/media/sunxi/cedrus/cedrus_h265.c index
> > > f703c585d91c5..f0bc118021b0a 100644
> > > --- a/drivers/staging/media/sunxi/cedrus/cedrus_h265.c
> > > +++ b/drivers/staging/media/sunxi/cedrus/cedrus_h265.c
> > > @@ -227,6 +227,7 @@ static void cedrus_h265_pred_weight_write(struct
> > > cedrus_dev *dev, static void cedrus_h265_skip_bits(struct cedrus_dev
> > > *dev,
> > > int num) {
> > >=20
> > >  	int count =3D 0;
> > >=20
> > > +	u32 reg;
> > >=20
> > >  	while (count < num) {
> > >  =09
> > >  		int tmp =3D min(num - count, 32);
> > >=20
> > > @@ -234,8 +235,9 @@ static void cedrus_h265_skip_bits(struct cedrus_d=
ev
> > > *dev, int num) cedrus_write(dev, VE_DEC_H265_TRIGGER,
> > >=20
> > >  			     VE_DEC_H265_TRIGGER_FLUSH_BITS |
> > >  			     VE_DEC_H265_TRIGGER_TYPE_N_BITS(tmp));
> > >=20
> > > -		while (cedrus_read(dev, VE_DEC_H265_STATUS) &
> > > VE_DEC_H265_STATUS_VLD_BUSY) -			udelay(1);
> > > +
> > > +		if (cedrus_wait_for(dev, VE_DEC_H265_STATUS,
> > > VE_DEC_H265_STATUS_VLD_BUSY)) +
> >=20
> > dev_err_ratelimited(dev->dev, "timed out
> >=20
> > > waiting to skip bits\n");
> >=20
> > Reporting issue is nice, but better would be to propagate error, since
> > there is no way to properly decode this slice if above code block fails.
> This mimic what was already there, mind if we do that later ? The
> propagation is doing to be a lot more intrusive.

Since backporting fixes before 6.0 isn't priority, viability for backpporti=
ng=20
isn't that important. You would only need to return 0 or -ETIMEDOUT and che=
ck=20
for error in only one location. That doesn't sound  very intrusive for me.

Best regards,
Jernej

>=20
> > Best regards,
> > Jernej
> >=20
> > >  		count +=3D tmp;
> > >  =09
> > >  	}