linux-kernel - Re: [PATCH v1 RFC] video/hdmi: Fix HDMI_VENDOR_INFOFRAME

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 6 Feb 2023 11:58:26 +0200
From:   Ville Syrjälä <ville.syrjala@...ux.intel.com>
To:     Martin Blumenstingl <martin.blumenstingl@...glemail.com>
Cc:     dri-devel@...ts.freedesktop.org, Helge Deller <deller@....de>,
        linux-fbdev@...r.kernel.org, linux-kernel@...r.kernel.org,
        Hans Verkuil <hans.verkuil@...co.com>,
        Bartlomiej Zolnierkiewicz <b.zolnierkie@...sung.com>,
        Bernard Zhao <bernard@...o.com>
Subject: Re: [PATCH v1 RFC] video/hdmi: Fix HDMI_VENDOR_INFOFRAME_SIZE

On Mon, Jan 09, 2023 at 11:31:10PM +0100, Martin Blumenstingl wrote:
> When support for the HDMI vendor infoframe was introduced back with
> commit 7d27becb3532 ("video/hdmi: Introduce helpers for the HDMI vendor
> specific infoframe") it's payload size was either 5 or 6 bytes,
> depending on:
>   if (frame->s3d_struct >= HDMI_3D_STRUCTURE_SIDE_BY_SIDE_HALF)
> When true the size was 6 bytes, otherwise 5 bytes.
> 
> Drivers that are using hdmi_infoframe_pack() are reserving 10 bytes (4
> bytes for the header and up to 6 bytes for the infoframe payload data)
> or more (exynos_hdmi reserves 25 bytes).
> 
> Over time the frame payload length was reduced to 4 bytes. This however
> does not match the code from hdmi_hdmi_infoframe_pack() where ptr[8] and
> ptr[9] are written, which means the infoframe has to allow up to 6 bytes
> of payload data (considering that the header takes 4 bytes).
> 
> Change HDMI_VENDOR_INFOFRAME_SIZE to 6 bytes so
> hdmi_vendor_infoframe_pack_only() can properly check the passed buffer
> size and avoid an out of bounds write to ptr[8] or ptr[9].

The function should return -ENOSPC if the caller didn't
provide a big enough buffer. Are you saying there are
drivers that are passing a bogus size here?

> 
> Fixes: c5e69ab35c0d ("video/hdmi: Constify infoframe passed to the pack functions")
> Fixes: d43be2554b58 ("drivers: video: hdmi: cleanup coding style in video a bit")
> Signed-off-by: Martin Blumenstingl <martin.blumenstingl@...glemail.com>
> ---
> I'm not an expert on this topic and I'm not sure if the size still
> depends on that if condition from long time ago. So please share your
> thoughts.
> 
> 
>  include/linux/hdmi.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/include/linux/hdmi.h b/include/linux/hdmi.h
> index 2f4dcc8d060e..026c5ef5a1a5 100644
> --- a/include/linux/hdmi.h
> +++ b/include/linux/hdmi.h
> @@ -57,7 +57,7 @@ enum hdmi_infoframe_type {
>  #define HDMI_SPD_INFOFRAME_SIZE    25
>  #define HDMI_AUDIO_INFOFRAME_SIZE  10
>  #define HDMI_DRM_INFOFRAME_SIZE    26
> -#define HDMI_VENDOR_INFOFRAME_SIZE  4
> +#define HDMI_VENDOR_INFOFRAME_SIZE  6
>  
>  #define HDMI_INFOFRAME_SIZE(type)	\
>  	(HDMI_INFOFRAME_HEADER_SIZE + HDMI_ ## type ## _INFOFRAME_SIZE)
> -- 
> 2.39.0

-- 
Ville Syrjälä
Intel