public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
To: Ricardo Ribalda <ribalda@chromium.org>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>,
	Hans Verkuil <hverkuil@xs4all.nl>,
	Hans Verkuil <hverkuil-cisco@xs4all.nl>,
	linux-media@vger.kernel.org, linux-kernel@vger.kernel.org,
	Hans Verkuil <hans.verkuil@cisco.com>
Subject: Re: [PATCH RESEND v2 5/7] media: uvcvideo: Fix handling on Bitmask controls
Date: Thu, 29 Dec 2022 23:25:12 +0200	[thread overview]
Message-ID: <Y64FuBVxqIv5OAgx@pendragon.ideasonboard.com> (raw)
In-Reply-To: <20220920-resend-v4l2-compliance-v2-5-b0ceb15353ac@chromium.org>

Hi Ricardo,

Thank you for the patch.

On Fri, Dec 02, 2022 at 06:21:39PM +0100, Ricardo Ribalda wrote:
> Minimum and step values for V4L2_CTRL_TYPE_BITMASK controls should be 0.
> There is no need to query the camera firmware about this and maybe get
> invalid results.
> 
> Also value should be masked to the max value advertised by the
> hardware.
> 
> Finally, handle uvc 1.5 mask controls that use MAX instead of RES to

s/uvc/UVC/

> describe the valid bits.
> 
> Fixes v4l2-compliane:
> Control ioctls (Input 0):
>                 fail: v4l2-test-controls.cpp(97): minimum must be 0 for a bitmask control
> 	test VIDIOC_QUERY_EXT_CTRL/QUERYMENU: FAIL
> 
> Signed-off-by: Ricardo Ribalda <ribalda@chromium.org>
> ---
>  drivers/media/usb/uvc/uvc_ctrl.c | 53 +++++++++++++++++++++++++++++++---------
>  1 file changed, 41 insertions(+), 12 deletions(-)
> 
> diff --git a/drivers/media/usb/uvc/uvc_ctrl.c b/drivers/media/usb/uvc/uvc_ctrl.c
> index 7153ee5aabb1..526572044e82 100644
> --- a/drivers/media/usb/uvc/uvc_ctrl.c
> +++ b/drivers/media/usb/uvc/uvc_ctrl.c
> @@ -1145,6 +1145,25 @@ static const char *uvc_map_get_name(const struct uvc_control_mapping *map)
>  	return "Unknown Control";
>  }
>  
> +static u32 uvc_get_ctrl_bitmap(struct uvc_control *ctrl,
> +			       struct uvc_control_mapping *mapping)
> +{
> +	/*
> +	 * Some controls, like CT_AE_MODE_CONTROL use GET_RES to
> +	 * represent the number of bits supported, those controls
> +	 * do not list GET_MAX as supported.

You can reflow to 80 columns.

> +	 */
> +	if (ctrl->info.flags & UVC_CTRL_FLAG_GET_MAX)
> +		return mapping->get(mapping, UVC_GET_MAX,
> +				    uvc_ctrl_data(ctrl, UVC_CTRL_DATA_MAX));
> +
> +	if (ctrl->info.flags & UVC_CTRL_FLAG_GET_RES)
> +		return mapping->get(mapping, UVC_GET_RES,
> +				    uvc_ctrl_data(ctrl, UVC_CTRL_DATA_RES));

Should we start with GET_RES to avoid any regression in case of controls
that support both GET_RES and GET_MAX ?

> +
> +	return ~0;
> +}
> +
>  static int __uvc_query_v4l2_ctrl(struct uvc_video_chain *chain,
>  	struct uvc_control *ctrl,
>  	struct uvc_control_mapping *mapping,
> @@ -1219,6 +1238,12 @@ static int __uvc_query_v4l2_ctrl(struct uvc_video_chain *chain,
>  		v4l2_ctrl->step = 0;
>  		return 0;
>  
> +	case V4L2_CTRL_TYPE_BITMASK:
> +		v4l2_ctrl->minimum = 0;
> +		v4l2_ctrl->maximum = uvc_get_ctrl_bitmap(ctrl, mapping);
> +		v4l2_ctrl->step = 0;
> +		return 0;
> +
>  	default:
>  		break;
>  	}
> @@ -1320,19 +1345,14 @@ int uvc_query_v4l2_menu(struct uvc_video_chain *chain,
>  
>  	menu_info = &mapping->menu_info[query_menu->index];
>  
> -	if (mapping->data_type == UVC_CTRL_DATA_TYPE_BITMASK &&
> -	    (ctrl->info.flags & UVC_CTRL_FLAG_GET_RES)) {
> -		s32 bitmap;
> -
> +	if (mapping->data_type == UVC_CTRL_DATA_TYPE_BITMASK) {
>  		if (!ctrl->cached) {
>  			ret = uvc_ctrl_populate_cache(chain, ctrl);
>  			if (ret < 0)
>  				goto done;
>  		}
>  
> -		bitmap = mapping->get(mapping, UVC_GET_RES,
> -				      uvc_ctrl_data(ctrl, UVC_CTRL_DATA_RES));
> -		if (!(bitmap & menu_info->value)) {
> +		if (!(uvc_get_ctrl_bitmap(ctrl, mapping) & menu_info->value)) {
>  			ret = -EINVAL;
>  			goto done;
>  		}
> @@ -1815,6 +1835,18 @@ int uvc_ctrl_set(struct uvc_fh *handle,
>  		value = xctrl->value;
>  		break;
>  
> +	case V4L2_CTRL_TYPE_BITMASK:
> +		if (!ctrl->cached) {
> +			ret = uvc_ctrl_populate_cache(chain, ctrl);
> +			if (ret < 0)
> +				return ret;
> +		}
> +
> +		xctrl->value = max(0, xctrl->value);

This would prevent bit 31 from being set. It doesn't matter for the
controls currently supported by the driver, but is it needed ?

> +		xctrl->value &= uvc_get_ctrl_bitmap(ctrl, mapping);
> +		value = xctrl->value;
> +		break;
> +
>  	case V4L2_CTRL_TYPE_BOOLEAN:
>  		xctrl->value = clamp(xctrl->value, 0, 1);
>  		value = xctrl->value;
> @@ -1829,17 +1861,14 @@ int uvc_ctrl_set(struct uvc_fh *handle,
>  		 * Valid menu indices are reported by the GET_RES request for
>  		 * UVC controls that support it.
>  		 */
> -		if (mapping->data_type == UVC_CTRL_DATA_TYPE_BITMASK &&
> -		    (ctrl->info.flags & UVC_CTRL_FLAG_GET_RES)) {
> +		if (mapping->data_type == UVC_CTRL_DATA_TYPE_BITMASK) {
>  			if (!ctrl->cached) {
>  				ret = uvc_ctrl_populate_cache(chain, ctrl);
>  				if (ret < 0)
>  					return ret;
>  			}
>  
> -			step = mapping->get(mapping, UVC_GET_RES,
> -					uvc_ctrl_data(ctrl, UVC_CTRL_DATA_RES));
> -			if (!(step & value))
> +			if (!(uvc_get_ctrl_bitmap(ctrl, mapping) & value))
>  				return -EINVAL;
>  		}
>  
> 

-- 
Regards,

Laurent Pinchart

  reply	other threads:[~2022-12-29 21:25 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-12-02 17:21 [PATCH RESEND v2 0/7] Follow-up patches for uvc v4l2-compliance Ricardo Ribalda
2022-12-02 17:21 ` [PATCH RESEND v2 1/7] media: uvcvideo: uvc_ctrl_is_accessible: check for INACTIVE Ricardo Ribalda
2022-12-29 20:54   ` Laurent Pinchart
2022-12-02 17:21 ` [PATCH RESEND v2 2/7] media: uvcvideo: improve error logging in uvc_query_ctrl() Ricardo Ribalda
2022-12-29  2:59   ` Laurent Pinchart
2022-12-02 17:21 ` [PATCH RESEND v2 3/7] media: uvcvideo: Return -EACCES for Wrong state error Ricardo Ribalda
2022-12-29  3:25   ` Laurent Pinchart
2022-12-02 17:21 ` [PATCH RESEND v2 4/7] media: uvcvideo: Do not return positive errors in uvc_query_ctrl() Ricardo Ribalda
2022-12-29  3:28   ` Laurent Pinchart
2022-12-02 17:21 ` [PATCH RESEND v2 5/7] media: uvcvideo: Fix handling on Bitmask controls Ricardo Ribalda
2022-12-29 21:25   ` Laurent Pinchart [this message]
2022-12-02 17:21 ` [PATCH RESEND v2 6/7] media: uvcvideo: Implement mask for V4L2_CTRL_TYPE_MENU Ricardo Ribalda
2022-12-29  2:50   ` Laurent Pinchart
2022-12-02 17:21 ` [PATCH RESEND v2 7/7] media: uvcvideo: Use standard names for menus Ricardo Ribalda
2022-12-29 22:06   ` Laurent Pinchart
2022-12-29 22:13 ` [PATCH RESEND v2 0/7] Follow-up patches for uvc v4l2-compliance Laurent Pinchart
2023-01-03 14:37   ` Ricardo Ribalda

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=Y64FuBVxqIv5OAgx@pendragon.ideasonboard.com \
    --to=laurent.pinchart@ideasonboard.com \
    --cc=hans.verkuil@cisco.com \
    --cc=hverkuil-cisco@xs4all.nl \
    --cc=hverkuil@xs4all.nl \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=mchehab@kernel.org \
    --cc=ribalda@chromium.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox