Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 1 May 2023 18:05:59 -0700
From:   Dmitry Torokhov <dmitry.torokhov@gmail.com>
To:     Biswarup Pal <biswarupp@google.com>,
        Peter Hutterer <peter.hutterer@who-t.net>
Cc:     kernel-team@android.com, linux-input@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH] Call input_set_timestamp for events injected using uinput
Message-ID: <ZFBh948No3IAV2hf@google.com>
References: <20230427000152.1407471-1-biswarupp@google.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20230427000152.1407471-1-biswarupp@google.com>
Precedence: bulk

On Thu, Apr 27, 2023 at 12:01:51AM +0000, Biswarup Pal wrote:
> Currently, uinput doesn't use the input_set_timestamp API, so any
> event injected using uinput is not accurately timestamped in terms of
> measuring when the actual event happened. Hence, call the
> input_set_timestamp API from uinput in order to provide a more
> accurate sense of time for the event. Propagate only the timestamps
> which are a) positive, b) within a pre-defined offset (10 secs) from
> the current time, and c) not in the future.

This makes sense to me. Peter, do you see any issues?

Thanks!

> 
> Signed-off-by: Biswarup Pal <biswarupp@google.com>
> ---
>  drivers/input/misc/uinput.c | 34 ++++++++++++++++++++++++++++++++++
>  1 file changed, 34 insertions(+)
> 
> diff --git a/drivers/input/misc/uinput.c b/drivers/input/misc/uinput.c
> index f2593133e524..d98212d55108 100644
> --- a/drivers/input/misc/uinput.c
> +++ b/drivers/input/misc/uinput.c
> @@ -33,6 +33,7 @@
>  #define UINPUT_NAME		"uinput"
>  #define UINPUT_BUFFER_SIZE	16
>  #define UINPUT_NUM_REQUESTS	16
> +#define UINPUT_TIMESTAMP_ALLOWED_OFFSET_SECS 10
>  
>  enum uinput_state { UIST_NEW_DEVICE, UIST_SETUP_COMPLETE, UIST_CREATED };
>  
> @@ -569,11 +570,40 @@ static int uinput_setup_device_legacy(struct uinput_device *udev,
>  	return retval;
>  }
>  
> +/*
> + * Returns true if the given timestamp is valid (i.e., if all the following
> + * conditions are satisfied), false otherwise.
> + * 1) given timestamp is positive
> + * 2) it's within the allowed offset before the current time
> + * 3) it's not in the future
> + */
> +static bool is_valid_timestamp(const ktime_t timestamp)
> +{
> +	ktime_t zero_time;
> +	ktime_t current_time;
> +	ktime_t min_time;
> +	ktime_t offset;
> +
> +	zero_time = ktime_set(0, 0);
> +	if (ktime_compare(zero_time, timestamp) >= 0)
> +		return false;
> +
> +	current_time = ktime_get();
> +	offset = ktime_set(UINPUT_TIMESTAMP_ALLOWED_OFFSET_SECS, 0);
> +	min_time = ktime_sub(current_time, offset);
> +
> +	if (ktime_after(min_time, timestamp) || ktime_after(timestamp, current_time))
> +		return false;
> +
> +	return true;
> +}
> +
>  static ssize_t uinput_inject_events(struct uinput_device *udev,
>  				    const char __user *buffer, size_t count)
>  {
>  	struct input_event ev;
>  	size_t bytes = 0;
> +	ktime_t timestamp;
>  
>  	if (count != 0 && count < input_event_size())
>  		return -EINVAL;
> @@ -588,6 +618,10 @@ static ssize_t uinput_inject_events(struct uinput_device *udev,
>  		if (input_event_from_user(buffer + bytes, &ev))
>  			return -EFAULT;
>  
> +		timestamp = ktime_set(ev.input_event_sec, ev.input_event_usec * NSEC_PER_USEC);
> +		if (is_valid_timestamp(timestamp))
> +			input_set_timestamp(udev->dev, timestamp);
> +
>  		input_event(udev->dev, ev.type, ev.code, ev.value);
>  		bytes += input_event_size();
>  		cond_resched();
> -- 
> 2.40.1.495.gc816e09b53d-goog
> 

-- 
Dmitry