Re: [PATCH 7/7] Input: libps2 - do not discard non-ack bytes when controlling LEDs

[Raul E Rangel]

On Thu, May 11, 2023 at 11:52:47AM -0700, Dmitry Torokhov wrote:
> Upon receiving a PS/2 command the device and controller are supposed to
> stop sending normal data (scancodes or movement packets) and instead
> immediately start delivering ACK/NAK and command response. Unfortunately
> often EC has an output buffer which may contain latched data by the time
> the EC receives a command from the host. The kernel used to ignore such
> data, but that may cause "stuck" keys if the data dropped happens to be a
> break code or a part of a break code. This occasionally happens, for
> example, on Chromebooks when the kernel tries to toggle CapsLock LED on
> a keyboard while user releases Alt+Search keyboard shortcut.
> 
> Fix this by passing the first non-ACK byte to the normal handler for a
> handful of PS/2 commands that are expected to be used during normal device
> operation (as opposed to probe/configuration time).
> 
> Signed-off-by: Dmitry Torokhov <dmitry.torokhov@xxxxxxxxx>
> ---
>  drivers/input/serio/libps2.c | 36 ++++++++++++++++++++++++++++++++----
>  1 file changed, 32 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/input/serio/libps2.c b/drivers/input/serio/libps2.c
> index 7c5fc853072a..6d78a1fe00c1 100644
> --- a/drivers/input/serio/libps2.c
> +++ b/drivers/input/serio/libps2.c
> @@ -21,7 +21,10 @@
>  
>  #define PS2_CMD_SETSCALE11	0x00e6
>  #define PS2_CMD_SETRES		0x10e8
> +#define PS2_CMD_EX_SETLEDS	0x20eb
> +#define PS2_CMD_SETLEDS		0x10ed
>  #define PS2_CMD_GETID		0x02f2
> +#define PS2_CMD_SETREP		0x10f3 /* Set repeat rate/set report rate */
>  #define PS2_CMD_RESET_BAT	0x02ff
>  
>  #define PS2_RET_BAT		0xaa
> @@ -35,6 +38,7 @@
>  #define PS2_FLAG_CMD1		BIT(2)	/* Waiting for the first byte of command response */
>  #define PS2_FLAG_WAITID		BIT(3)	/* Command executing is GET ID */
>  #define PS2_FLAG_NAK		BIT(4)	/* Last transmission was NAKed */
> +#define PS2_FLAG_PASS_NOACK	BIT(5)	/* Pass non-ACK byte to receive handler */
>  
>  static int ps2_do_sendbyte(struct ps2dev *ps2dev, u8 byte,
>  			   unsigned int timeout, unsigned int max_attempts)
> @@ -281,9 +285,28 @@ int __ps2_command(struct ps2dev *ps2dev, u8 *param, unsigned int command)
>  
>  	serio_pause_rx(ps2dev->serio);
>  
> -	/* Some mice do not ACK the "get ID" command, prepare to handle this. */
> -	ps2dev->flags = command == PS2_CMD_GETID ? PS2_FLAG_WAITID : 0;
>  	ps2dev->cmdcnt = receive;
> +
> +	switch (command) {
> +	case PS2_CMD_GETID:
> +		/*
> +		 * Some mice do not ACK the "get ID" command, prepare to
> +		 * handle this.
> +		 */
> +		ps2dev->flags = PS2_FLAG_WAITID;
> +		break;
> +
> +	case PS2_CMD_SETLEDS:
> +	case PS2_CMD_EX_SETLEDS:
> +	case PS2_CMD_SETREP:
> +		ps2dev->flags = PS2_FLAG_PASS_NOACK;
> +		break;
> +
> +	default:
> +		ps2dev->flags = 0;
> +		break;
> +	}
> +
>  	if (receive) {
>  		/* Indicate that we expect response to the command. */
>  		ps2dev->flags |= PS2_FLAG_CMD | PS2_FLAG_CMD1;
> @@ -512,14 +535,19 @@ static void ps2_handle_ack(struct ps2dev *ps2dev, u8 data)
>  		 * Do not signal errors if we get unexpected reply while
>  		 * waiting for an ACK to the initial (first) command byte:
>  		 * the device might not be quiesced yet and continue
> -		 * delivering data.
> +		 * delivering data. For certain commands (such as set leds and
> +		 * set repeat rate) that can be used during normal device
> +		 * operation, we even pass this data byte to the normal receive
> +		 * handler.
>  		 * Note that we reset PS2_FLAG_WAITID flag, so the workaround
>  		 * for mice not acknowledging the Get ID command only triggers
>  		 * on the 1st byte; if device spews data we really want to see
>  		 * a real ACK from it.
>  		 */
>  		dev_dbg(&ps2dev->serio->dev, "unexpected %#02x\n", data);
> -		ps2dev->flags &= ~PS2_FLAG_WAITID;
> +		if (ps2dev->flags & PS2_FLAG_PASS_NOACK)
> +			ps2dev->receive_handler(ps2dev, data);
> +		ps2dev->flags &= ~(PS2_FLAG_WAITID | PS2_FLAG_PASS_NOACK);
>  		return;
>  	}
>  

Great refactoring. It made this final patch super simple!

Reviewed-by: Raul E Rangel <rrangel@xxxxxxxxxxxx>