input子系统分析：3. 应用层获取数据，及驱动数据上报流程

==============================================
1. 应用程序调用open() 


以androidM Gsensor为例
hal层中acceleration.cpp中
FindDataFd()
	fd = open("/sys/class/misc/m_acc_misc/accdevnum", O_RDONLY);
	len = read(fd, buf, sizeof(buf)-1);
	buf[len] = '\0';            
	sscanf(buf, "%d\n", &num);
	sprintf(buf_s, "/dev/input/event%d", num);
	fd = open(buf_s, O_RDONLY);
	
readEvents()
	mInputReader.fill(mdata_fd)
	mInputReader.readEvent(&event)

即open("/dev/input/event%d", O_RDONLY);
-----------------------------------------
vfs
	sys_open();		// 系统调用
		struct file file->f_ops = cdev->ops;
		file->f_ops->open();
-----------------------------------------
input handler层: evdev.c
	cdev
	xxx_ops = {
		.open = xxx_open,
		.read = xxx_read,
	}

	
evdev_connect()	
	cdev_init(&evdev->cdev, &evdev_fops);


static const struct file_operations evdev_fops = {
	.owner		= THIS_MODULE,
	.read		= evdev_read,
	.write		= evdev_write,
	.poll		= evdev_poll,
	.open		= evdev_open,
};
实际上最终调用了evdev_open()

evdev_open(struct inode *inode, struct file *file)
	|
	struct evdev *evdev = container_of(inode->i_cdev, struct evdev, cdev);	// 以小博大，inode->i_cdev就是connect()中住的的cdev
	unsigned int bufsize = evdev_compute_buffer_size(evdev->handle.dev);	// 通过handle找到 input device，根据input device 获取缓冲区的大小(几个input event)，但是我们驱动中未给定缓冲区大小，系统会自动给定一个
	unsigned int