linux-kernel - Re: [RFC PATCH v1 6/7] media: video: introduce face detection driver module

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]

Message-ID: <4EE4F896.3080406@gmail.com>
Date:	Sun, 11 Dec 2011 19:38:14 +0100
From:	Sylwester Nawrocki <snjw23@...il.com>
To:	Ming Lei <ming.lei@...onical.com>
CC:	linux-omap@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
	linux-kernel@...r.kernel.org, linux-media@...r.kernel.org
Subject: Re: [RFC PATCH v1 6/7] media: video: introduce face detection driver
 module

Hi Ming,

On 12/02/2011 04:02 PM, Ming Lei wrote:
> This patch introduces one driver for face detection purpose.
> 
> The driver is responsible for all v4l2 stuff, buffer management
> and other general things, and doesn't touch face detection hardware
> directly. Several interfaces are exported to low level drivers
> (such as the coming omap4 FD driver)which will communicate with
> face detection hw module.
> 
> So the driver will make driving face detection hw modules more
> easy.
> 
> TODO:
> 	- implement FD setting interfaces with v4l2 controls or
> 	ext controls
> 
> Signed-off-by: Ming Lei <ming.lei@...onical.com>
> ---
...
> +static int vidioc_g_fd_result(struct file *file, void *priv,
> +					struct v4l2_fd_result *f)
> +{
> +	struct fdif_dev *dev = video_drvdata(file);
> +	unsigned long flags;
> +	struct v4l2_fdif_result *tmp;
> +	struct v4l2_fdif_result *fr = NULL;
> +	unsigned int cnt = 0;
> +	int ret = -EINVAL;
> +
> +	spin_lock_irqsave(&dev->lock, flags);
> +	list_for_each_entry(tmp, &dev->fdif_dq.complete, list)
> +		if (tmp->index == f->buf_index) {
> +			fr = tmp;
> +			list_del(&tmp->list);
> +			break;
> +		}
> +	spin_unlock_irqrestore(&dev->lock, flags);
> +
> +	if (fr) {
> +		ret = 0;
> +		cnt = min(f->face_cnt, fr->face_cnt);
> +		if (cnt)
> +			memcpy(f->fd, fr->faces,
> +				sizeof(struct v4l2_fd_detection) * cnt);
> +		f->face_cnt = cnt;
> +		kfree(fr->faces);
> +		kfree(fr);

struct v4l2_fdif_result is allocated in HW driver at the time when FD result
is read out and it is freed in generic module here. Not sure if it is a good
practice to split memory management like this. Also IMHO it would be much
better to pre-allocate memory for FD output data, according to maximum number
of detected faces per frame. It could be more reliable than allocating memory
in interrupt context per each frame.


> +	}
> +	return ret;
> +}

--
Regards,
Sylwester
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/