Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.35 as permitted sender) client-ip=23.128.96.35;
MIME-Version: 1.0
References: <20231023-imx214-v1-0-b33f1bbd1fcf@apitzsch.eu>
 <20231023-imx214-v1-2-b33f1bbd1fcf@apitzsch.eu> <CAPybu_2gya-XP2-JH8roYgyROUAeTbVBaY1ypMKyVp+ujb=t6A@mail.gmail.com>
 <8332c443fe33a74774f2375009a31e7895fcf37a.camel@apitzsch.eu>
In-Reply-To: <8332c443fe33a74774f2375009a31e7895fcf37a.camel@apitzsch.eu>
From:   Ricardo Ribalda Delgado <ribalda@kernel.org>
Date:   Sat, 28 Oct 2023 08:44:01 +0200
Message-ID: <CAPybu_3FF__5hpcSeAgc0pkxnBjbX5aGt9yif+CvYp4_JFAp-w@mail.gmail.com>
Subject: Re: [PATCH 2/4] media: i2c: imx214: Move controls init to separate function
To:     =?UTF-8?Q?Andr=C3=A9_Apitzsch?= <git@apitzsch.eu>
Cc:     Sakari Ailus <sakari.ailus@linux.intel.com>,
        Mauro Carvalho Chehab <mchehab@kernel.org>,
        linux-media@vger.kernel.org, linux-kernel@vger.kernel.org,
        ~postmarketos/upstreaming@lists.sr.ht
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Hi Andr=C3=A9

On Fri, Oct 27, 2023 at 11:23=E2=80=AFPM Andr=C3=A9 Apitzsch <git@apitzsch.=
eu> wrote:
>
> Hi Ricardo,
>
> Am Freitag, dem 27.10.2023 um 14:25 +0200 schrieb Ricardo Ribalda
> Delgado:
> > Hi Andre
> > On Mon, Oct 23, 2023 at 11:49=E2=80=AFPM Andr=C3=A9 Apitzsch <git@apitz=
sch.eu>
> > wrote:
> > >
> > > Code refinement, no functional changes.
> > >
> > > Signed-off-by: Andr=C3=A9 Apitzsch <git@apitzsch.eu>
> >
> > With Jacopos comments (don't use de_err_probe())
> > Reviewed-by: Ricardo Ribalda <ribalda@chromium.org>
> >
> > > +       ret =3D imx214_ctrls_init(imx214);
> > > +       if (ret < 0)
> > >                 goto free_ctrl;
> >
> > It seems like we can mutex_destroy a non inited mutex. Could you send
> > a follow-up patch to fix that?
> >
> Sorry, I don't get it. Could you explain what you mean. Thanks.
>

If the controls are initialized incorrectly we will jump to free_ctrl
in line 1046, which calls
mutex_destroy(&imx214->mutex);

But that mutex initialized in line 1050.

You did not introduce the bug, but since you have the hardware and are
sending the other patches it would be great if you could add a new
patch to fix it :)

Thanks!


> > Thanks!
>