Re: input: i8042 possible bug

[Dmitry Torokhov <dmitry.torokhov@gmail.com>]

On Tue, Sep 03, 2013 at 09:39:10PM +0400, Andrey Moiseev wrote:
> Here is an extract from drivers/input/serio/i8042.c:
> > 
> > static int i8042_flush(void)
> > {
> > 	unsigned long flags;
> > 	unsigned char data, str;
> > 	int i = 0;
> > 
> > 	spin_lock_irqsave(&i8042_lock, flags);
> > 
> > 	while (((str = i8042_read_status()) & I8042_STR_OBF) && (i < I8042_BUFFER_SIZE)) {
> > 		udelay(50);
> > 		data = i8042_read_data();
> > 		i++;
> > 		dbg("%02x <- i8042 (flush, %s)\n",
> > 		    data, str & I8042_STR_AUXDATA ? "aux" : "kbd");
> > 	}
> > 
> > 	spin_unlock_irqrestore(&i8042_lock, flags);
> > 
> > 	return i;
> > }
> > 
> > static int i8042_controller_check(void)
> > {
> > 	if (i8042_flush() == I8042_BUFFER_SIZE) {
> Doesn't it panic needlessly, while the buffer was initially full?

It appears so. We probably want to change it to return -1 if we still
see OBF after fetching I8042_BUFFER_SIZE characters.

Thanks.

-- 
Dmitry