MIME-Version: 1.0
References: <20230302074704.11371-1-johan+linaro@kernel.org>
In-Reply-To: <20230302074704.11371-1-johan+linaro@kernel.org>
From:   Doug Anderson <dianders@chromium.org>
Date:   Thu, 2 Mar 2023 10:29:29 -0800
Message-ID: <CAD=FV=WK2yxWqAro6k4N2kr-HeNZLNxi6tuq=_uL0tbSjGVEAg@mail.gmail.com>
Subject: Re: [PATCH] drm/edid: fix info leak when failing to get panel id
To:     Johan Hovold <johan+linaro@kernel.org>
Cc:     Maarten Lankhorst <maarten.lankhorst@linux.intel.com>,
        Maxime Ripard <mripard@kernel.org>,
        Thomas Zimmermann <tzimmermann@suse.de>,
        David Airlie <airlied@gmail.com>,
        Daniel Vetter <daniel@ffwll.ch>,
        Abhinav Kumar <quic_abhinavk@quicinc.com>,
        dri-devel@lists.freedesktop.org, linux-kernel@vger.kernel.org,
        stable@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Hi,

On Wed, Mar 1, 2023 at 11:49=E2=80=AFPM Johan Hovold <johan+linaro@kernel.o=
rg> wrote:
>
> Make sure to clear the transfer buffer before fetching the EDID to
> avoid leaking slab data to the logs on errors that leave the buffer
> unchanged.
>
> Fixes: 69c7717c20cc ("drm/edid: Dump the EDID when drm_edid_get_panel_id(=
) has an error")
> Cc: stable@vger.kernel.org      # 6.2
> Cc: Douglas Anderson <dianders@chromium.org>
> Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
> ---
>  drivers/gpu/drm/drm_edid.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
> index 3841aba17abd..8707fe72a028 100644
> --- a/drivers/gpu/drm/drm_edid.c
> +++ b/drivers/gpu/drm/drm_edid.c
> @@ -2797,7 +2797,7 @@ u32 drm_edid_get_panel_id(struct i2c_adapter *adapt=
er)
>          * the EDID then we'll just return 0.
>          */
>
> -       base_block =3D kmalloc(EDID_LENGTH, GFP_KERNEL);
> +       base_block =3D kzalloc(EDID_LENGTH, GFP_KERNEL);

Good catch! I'm landing this to drm-misc-fixes right away.

4d8457fe0eb9 drm/edid: fix info leak when failing to get panel id

I'm sure I copied the kmalloc() from _drm_do_get_edid(), but it looks
like in _that_ case if the read fails we never print the buffer so
that one is OK.

-Doug