操作系统_再识(Linux设备模型之input子系统详解)

来源:互联网 发布:vb 串口扫描枪 编辑:程序博客网 时间:2024/05/19 03:43

------------------------------------------

本文系本站原创,欢迎转载!

转载请注明出处:http://ericxiao.cublog.cn/

------------------------------------------

一:前言

在键盘驱动代码分析的笔记中,接触到了input子系统.键盘驱动,键盘驱动将检测到的所有按键都上报给了input子系统。Input子系统是所有I/O设备驱动的中间层,为上层提供了一个统一的界面。例如,在终端系统中,我们不需要去管有多少个键盘,多少个鼠标。它只要从input子系统中去取对应的事件(按键,鼠标移位等)就可以了。今天就对input子系统做一个详尽的分析.

下面的代码是基于linux kernel 2.6.25.分析的代码主要位于kernel2.6.25/drivers/input下面.

二:使用input子系统的例子

在内核自带的文档Documentation/input/input-programming.txt中。有一个使用input子系统的例子,并附带相应的说明。以此为例分析如下:

#include <linux/input.h>

#include <linux/module.h>

#include <linux/init.h>

 

#include <asm/irq.h>

#include <asm/io.h>

 

static void button_interrupt(int irq, void *dummy, struct pt_regs *fp)

{

       input_report_key(&button_dev, BTN_1, inb(BUTTON_PORT) & 1);

        input_sync(&button_dev);

}

 

static int __init button_init(void)

{

        if (request_irq(BUTTON_IRQ,button_interrupt, 0, "button", NULL)) {

               printk(KERN_ERR "button.c: Can't allocate irq %d/n", button_irq);

               return -EBUSY;

        }

 

        button_dev.evbit[0] = BIT(EV_KEY);

        button_dev.keybit[LONG(BTN_0)]= BIT(BTN_0);

 

       input_register_device(&button_dev);

}

 

static void __exit button_exit(void)

{

       input_unregister_device(&button_dev);

        free_irq(BUTTON_IRQ,button_interrupt);

}

 

module_init(button_init);

module_exit(button_exit);

 

这个示例module代码还是比较简单,在初始化函数里注册了一个中断处理例程。然后注册了一个input device.在中断处理程序里,将接收到的按键上报给input子系统。

文档的作者在之后的分析里又对这个module作了优化。主要是在注册中断处理的时序上。在修改过后的代码里,为input device定义了open函数,在open的时候再去注册中断处理例程。具体的信息请自行参考这篇文档。在资料缺乏的情况下,kernel自带的文档就是剖析kernel相关知识的最好资料.

文档的作者还分析了几个api函数。列举如下:

 

1):set_bit(EV_KEY, button_dev.evbit);

   set_bit(BTN_0, button_dev.keybit);

分别用来设置设备所产生的事件以及上报的按键值。Struct iput_dev中有两个成员,一个是evbit.一个是keybit.分别用表示设备所支持的动作和按键类型。

2): input_register_device(&button_dev);

用来注册一个input device.

3): input_report_key()

用于给上层上报一个按键动作

4): input_sync()

用来告诉上层,本次的事件已经完成了.

5): NBITS(x) - returns the length of a bitfield array in longs for x bits

    LONG(x)  - returns the index in the array in longsfor bit x

BIT(x)   - returns the index in a long for bitx     

这几个宏在input子系统中经常用到。上面的英文解释已经很清楚了。

 

三:input设备注册分析.

Input设备注册的接口为:input_register_device()。代码如下:

int input_register_device(struct input_dev *dev)

{

         static atomic_t input_no= ATOMIC_INIT(0);

         struct input_handler*handler;

         const char *path;

         int error;

 

         __set_bit(EV_SYN,dev->evbit);

 

         /*

          * If delay andperiod are pre-set by the driver, then autorepeating

          * is handled by thedriver itself and we don't do it in input.c.

          */

 

         init_timer(&dev->timer);

         if(!dev->rep[REP_DELAY] && !dev->rep[REP_PERIOD]) {

                  dev->timer.data = (long) dev;

                  dev->timer.function = input_repeat_key;

                  dev->rep[REP_DELAY] = 250;

                  dev->rep[REP_PERIOD] = 33;

         }

在前面的分析中曾分析过。Input_deviceevbit表示该设备所支持的事件。在这里将其EV_SYN置位,即所有设备都支持这个事件.如果dev->rep[REP_DELAY]dev->rep[REP_PERIOD]没有设值,则将其赋默认值。这主要是处理重复按键的.

 

         if (!dev->getkeycode)

                  dev->getkeycode = input_default_getkeycode;

 

         if (!dev->setkeycode)

                  dev->setkeycode = input_default_setkeycode;

 

        snprintf(dev->dev.bus_id, sizeof(dev->dev.bus_id),

                   "input%ld", (unsigned long) atomic_inc_return(&input_no) -1);

 

         error =device_add(&dev->dev);

         if (error)

                  return error;

 

         path =kobject_get_path(&dev->dev.kobj, GFP_KERNEL);

         printk(KERN_INFO"input: %s as %s/n",

                  dev->name ? dev->name : "Unspecified device", path ? path :"N/A");

         kfree(path);

 

         error =mutex_lock_interruptible(&input_mutex);

         if (error) {

                  device_del(&dev->dev);

                  return error;

         }

如果input device没有定义getkeycodesetkeycode.则将其赋默认值。还记得在键盘驱动中的分析吗?这两个操作函数就可以用来取键的扫描码和设置键的扫描码。然后调用device_add()input_dev中封装的device注册到sysfs

 

        list_add_tail(&dev->node, &input_dev_list);

 

        list_for_each_entry(handler, &input_handler_list, node)

                  input_attach_handler(dev, handler);

 

        input_wakeup_procfs_readers();

 

        mutex_unlock(&input_mutex);

 

         return 0;

}

这里就是重点了。将input device 挂到input_dev_list链表上.然后,对每一个挂在input_handler_listhandler调用input_attach_handler().在这里的情况有好比设备模型中的devicedriver的匹配。所有的input device都挂在input_dev_list链上。所有的handle都挂在input_handler_list上。

看一下这个匹配的详细过程。匹配是在input_attach_handler()中完成的。代码如下:

static int input_attach_handler(struct input_dev *dev, structinput_handler *handler)

{

         const structinput_device_id *id;

         int error;

 

         if (handler->blacklist&& input_match_device(handler->blacklist, dev))

                  return -ENODEV;

 

         id =input_match_device(handler->id_table, dev);

         if (!id)

                  return -ENODEV;

 

         error =handler->connect(handler, dev, id);

         if (error &&error != -ENODEV)

                  printk(KERN_ERR

                           "input: failed to attach handler %s to device %s, "

                           "error: %d/n",

                           handler->name, kobject_name(&dev->dev.kobj), error);

 

         return error;

}

如果handleblacklist被赋值。要先匹配blacklist中的数据跟dev->id的数据是否匹配。匹配成功过后再来匹配handle->iddev->id中的数据。如果匹配成功,则调用handler->connect().

来看一下具体的数据匹配过程,这是在input_match_device()中完成的。代码如下:

static const struct input_device_id *input_match_device(const structinput_device_id *id,

                                                                struct input_dev *dev)

{

         int i;

 

         for (; id->flags ||id->driver_info; id++) {

 

                  if (id->flags & INPUT_DEVICE_ID_MATCH_BUS)

                           if (id->bustype != dev->id.bustype)

                                    continue;

 

                  if (id->flags & INPUT_DEVICE_ID_MATCH_VENDOR)

                           if (id->vendor != dev->id.vendor)

                                    continue;

 

                  if (id->flags & INPUT_DEVICE_ID_MATCH_PRODUCT)

                           if (id->product != dev->id.product)

                                    continue;

 

                  if (id->flags & INPUT_DEVICE_ID_MATCH_VERSION)

                           if (id->version != dev->id.version)

                                    continue;

 

                  MATCH_BIT(evbit,  EV_MAX);

                  MATCH_BIT(,, KEY_MAX);

                  MATCH_BIT(relbit, REL_MAX);

                  MATCH_BIT(absbit, ABS_MAX);

                  MATCH_BIT(mscbit, MSC_MAX);

                  MATCH_BIT(ledbit, LED_MAX);

                  MATCH_BIT(sndbit, SND_MAX);

                  MATCH_BIT(ffbit,  FF_MAX);

                  MATCH_BIT(swbit,  SW_MAX);

 

                  return id;

         }

 

         return NULL;

}

MATCH_BIT宏的定义如下:

#define MATCH_BIT(bit, max) /

                  for (i = 0; i < BITS_TO_LONGS(max); i++) /

                           if ((id->bit[i] & dev->bit[i]) != id->bit[i]) /

                                    break; /

                  if (i != BITS_TO_LONGS(max)) /

                           continue;

 

由此看到。在id->flags中定义了要匹配的项。定义INPUT_DEVICE_ID_MATCH_BUS。则是要比较input deviceinput handler的总线类型。INPUT_DEVICE_ID_MATCH_VENDORINPUT_DEVICE_ID_MATCH_PRODUCTINPUT_DEVICE_ID_MATCH_VERSION分别要求设备厂商。设备号和设备版本.

如果id->flags定义的类型匹配成功。或者是id->flags没有定义,就会进入到MATCH_BIT的匹配项了.MATCH_BIT宏的定义可以看出。只有当iput deviceinput handlerid成员在evbit, keybit,… swbit项相同才会匹配成功。而且匹配的顺序是从evbit, keybitswbit.只要有一项不同,就会循环到id中的下一项进行比较.

简而言之,注册input device的过程就是为input device设置默认值,并将其挂以input_dev_list.与挂载在input_handler_list中的handler相匹配。如果匹配成功,就会调用handlerconnect函数.

 

:handler注册分析

Handler注册的接口如下所示:

int input_register_handler(struct input_handler *handler)

{

         struct input_dev *dev;

         int retval;

 

         retval =mutex_lock_interruptible(&input_mutex);

         if (retval)

                  return retval;

 

        INIT_LIST_HEAD(&handler->h_list);

 

         if (handler->fops !=NULL) {

                  if (input_table[handler->minor >> 5]) {

                           retval = -EBUSY;

                           goto out;

                  }

                  input_table[handler->minor >> 5] = handler;

         }

 

        list_add_tail(&handler->node, &input_handler_list);

 

         list_for_each_entry(dev,&input_dev_list, node)

                  input_attach_handler(dev, handler);

 

        input_wakeup_procfs_readers();

 

 out:

        mutex_unlock(&input_mutex);

         return retval;

}

handler->minor表示对应input设备节点的次设备号.handler->minor右移五位做为索引值插入到input_table[ ]..之后再来分析input_talbe[ ]的作用.

然后将handler挂到input_handler_list.然后将其与挂在input_dev_list中的input device匹配.这个过程和input device的注册有相似的地方.都是注册到各自的链表,.然后与另外一条链表的对象相匹配.

 

:handle的注册

int input_register_handle(struct input_handle *handle)

{

         struct input_handler*handler = handle->handler;

         struct input_dev *dev =handle->dev;

         int error;

 

         /*

          * We takedev->mutex here to prevent race with

          *input_release_device().

          */

         error =mutex_lock_interruptible(&dev->mutex);

         if (error)

                  return error;

        list_add_tail_rcu(&handle->d_node, &dev->h_list);

        mutex_unlock(&dev->mutex);

         synchronize_rcu();

 

         /*

          * Since we aresupposed to be called from ->connect()

          * which is mutuallyexclusive with ->disconnect()

          * we can't beracing with input_unregister_handle()

          * and so separatelock is not needed here.

          */

        list_add_tail(&handle->h_node, &handler->h_list);

 

         if (handler->start)

                  handler->start(handle);

 

         return 0;

}

在这个函数里所做的处理其实很简单.handle挂到所对应input deviceh_list链表上.还将handle挂到对应的handlerhlist链表上.如果handler定义了start函数,将调用之.

到这里,我们已经看到了input device, handlerhandle是怎么关联起来的了.以图的方式总结如下:

 

 

:event事件的处理

我们在开篇的时候曾以linux kernel文档中自带的代码作分析.提出了几个事件上报的API.这些API其实都是input_event()的封装.代码如下:

void input_event(struct input_dev *dev,

                   unsigned int type, unsigned int code, int value)

{

         unsigned long flags;

 

         //判断设备是否支持这类事件

         if(is_event_supported(type, dev->evbit, EV_MAX)) {

 

                  spin_lock_irqsave(&dev->event_lock, flags);

                  //利用键盘输入来调整随机数产生器

                  add_input_randomness(type, code, value);

                  input_handle_event(dev, type, code, value);

                  spin_unlock_irqrestore(&dev->event_lock, flags);

         }

}

首先,先判断设备产生的这个事件是否合法.如果合法,流程转入到input_handle_event().

代码如下:

static void input_handle_event(struct input_dev *dev,

                                  unsigned int type, unsigned int code, intvalue)

{

         int disposition =INPUT_IGNORE_EVENT;

 

         switch (type) {

 

         case EV_SYN:

                  switch (code) {

                  case SYN_CONFIG:

                           disposition = INPUT_PASS_TO_ALL;

                           break;

 

                  case SYN_REPORT:

                           if (!dev->sync) {

                                    dev->sync = 1;

                                    disposition = INPUT_PASS_TO_HANDLERS;

                           }

                           break;

                  }

                  break;

 

         case EV_KEY:

                  //判断按键值是否被支持

                  if (is_event_supported(code, dev->keybit, KEY_MAX) &&

                      !!test_bit(code, dev->key) != value) {

 

                           if (value != 2) {

                                    __change_bit(code, dev->key);

                                    if (value)

                                              input_start_autorepeat(dev, code);

                           }

 

                           disposition = INPUT_PASS_TO_HANDLERS;

                  }

                  break;

 

         case EV_SW:

                  if (is_event_supported(code, dev->swbit, SW_MAX) &&

                      !!test_bit(code, dev->sw) != value) {

 

                           __change_bit(code, dev->sw);

                           disposition = INPUT_PASS_TO_HANDLERS;

                  }

                  break;

 

         case EV_ABS:

                  if (is_event_supported(code, dev->absbit, ABS_MAX)) {

 

                           value = input_defuzz_abs_event(value,

                                              dev->abs[code], dev->absfuzz[code]);

 

                           if (dev->abs[code] != value) {

                                    dev->abs[code] = value;

                                    disposition = INPUT_PASS_TO_HANDLERS;

                           }

                  }

                  break;

 

         case EV_REL:

                  if (is_event_supported(code, dev->relbit, REL_MAX) && value)

                           disposition = INPUT_PASS_TO_HANDLERS;

 

                  break;

 

         case EV_MSC:

                  if (is_event_supported(code, dev->mscbit, MSC_MAX))

                           disposition = INPUT_PASS_TO_ALL;

 

                  break;

 

         case EV_LED:

                  if (is_event_supported(code, dev->ledbit, LED_MAX) &&

                      !!test_bit(code, dev->led) != value) {

 

                           __change_bit(code, dev->led);

                           disposition = INPUT_PASS_TO_ALL;

                  }

                  break;

 

         case EV_SND:

                  if (is_event_supported(code, dev->sndbit, SND_MAX)) {

 

                           if (!!test_bit(code, dev->snd) != !!value)

                                    __change_bit(code, dev->snd);

                           disposition = INPUT_PASS_TO_ALL;

                  }

                  break;

 

         case EV_REP:

                  if (code <= REP_MAX && value >= 0 && dev->rep[code] !=value) {

                           dev->rep[code] = value;

                           disposition = INPUT_PASS_TO_ALL;

                  }

                  break;

 

         case EV_FF:

                  if (value >= 0)

                           disposition = INPUT_PASS_TO_ALL;

                  break;

 

         case EV_PWR:

                  disposition = INPUT_PASS_TO_ALL;

                  break;

         }

 

         if (type != EV_SYN)

                  dev->sync = 0;

 

         if ((disposition &INPUT_PASS_TO_DEVICE) && dev->event)

                  dev->event(dev, type, code, value);

 

         if (disposition &INPUT_PASS_TO_HANDLERS)

                  input_pass_event (dev, type, code, value);

}

在这里,我们忽略掉具体事件的处理.到最后,如果该事件需要input device来完成的,就会将disposition设置成INPUT_PASS_TO_DEVICE.如果需要handler来完成的,就将dispostion设为INPUT_PASS_TO_DEVICE.如果需要两者都参与,disposition设置为INPUT_PASS_TO_ALL.

需要输入设备参与的,回调设备的event函数.如果需要handler参与的.调用input_pass_event().代码如下:

static void input_pass_event(struct input_dev *dev,

                                unsigned int type, unsigned int code, int value)

{

         struct input_handle*handle;

 

         rcu_read_lock();

 

         handle =rcu_dereference(dev->grab);

         if (handle)

                  handle->handler->event(handle, type, code, value);

         else

                  list_for_each_entry_rcu(handle, &dev->h_list, d_node)

                           if (handle->open)

                                    handle->handler->event(handle,

                                                                type, code, value);

         rcu_read_unlock();

}

如果input device被强制指定了handler,则调用该handlerevent函数.

结合handle注册的分析.我们知道.会将handle挂到input deviceh_list链表上.

如果没有为input device强制指定handler.就会遍历input device->h_list上的handle成员.如果该handle被打开,则调用与输入设备对应的handlerevent()函数.注意,只有在handle被打开的情况下才会接收到事件.

另外,输入设备的handler强制设置一般是用带EVIOCGRAB标志的ioctl来完成的.如下是发图的方示总结evnet的处理过程:

 

 

 

我们已经分析了input device,handlerhandle的注册过程以及事件的上报和处理.下面以evdev为实例做分析.来贯穿理解一下整个过程.

 

:evdev概述

 Evdev对应的设备节点一般位于/dev/input/event0 ~ /dev/input/event4.理论上可以对应32个设备节点.分别代表被handler匹配的32input device.

可以用cat /dev/input/event0.然后移动鼠标或者键盘按键就会有数据输出(两者之间只能选一.因为一个设备文件只能关能一个输入设备).还可以往这个文件里写数据,使其产生特定的事件.这个过程我们之后再详细分析.

为了分析这一过程,必须从input子系统的初始化说起.

 

:input子系统的初始化

Input子系统的初始化函数为input_init().代码如下:

static int __init input_init(void)

{

         int err;

 

         err =class_register(&input_class);

         if (err) {

                  printk(KERN_ERR "input: unable to register input_dev class/n");

                  return err;

         }

 

         err = input_proc_init();

         if (err)

                  goto fail1;

 

         err =register_chrdev(INPUT_MAJOR, "input", &input_fops);

         if (err) {

                  printk(KERN_ERR "input: unable to register char major %d",INPUT_MAJOR);

                  goto fail2;

         }

 

         return 0;

 

 fail2:        input_proc_exit();

 fail1:       class_unregister(&input_class);

         return err;

}

在这个初始化函数里,先注册了一个名为”input”的类.所有input device都属于这个类.sysfs中表现就是.所有input device所代表的目录都位于/dev/class/input下面.

然后调用input_proc_init()/proc下面建立相关的交互文件.

再后调用register_chrdev()注册了主设备号为INPUT_MAJOR(13).次设备号为0~255的字符设备.它的操作指针为input_fops.

在这里,我们看到.所有主设备号13的字符设备的操作最终都会转入到input_fops.在前面分析的/dev/input/event0~/dev/input/event4的主设备号为13.操作也不例外的落在了input_fops.

Input_fops定义如下:

static const struct file_operations input_fops = {

         .owner = THIS_MODULE,

         .open = input_open_file,

};

打开文件所对应的操作函数为input_open_file.代码如下示:

static int input_open_file(struct inode *inode, struct file *file)

{

         struct input_handler*handler = input_table[iminor(inode) >> 5];

         const structfile_operations *old_fops, *new_fops = NULL;

         int err;

 

         /* No load-on-demandhere? */

         if (!handler ||!(new_fops = fops_get(handler->fops)))

                  return -ENODEV;

 

iminor(inode)为打开文件所对应的次设备号.input_table是一个struct input_handler全局数组.在这里.它先设备结点的次设备号右移5位做为索引值到input_table中取对应项.从这里我们也可以看到.一个handle代表1<<5个设备节点(因为在input_table中取值是以次备号右移5位为索引的.即低5位相同的次备号对应的是同一个索引).在这里,终于看到了input_talbe大显身手的地方了.input_talbe[ ]中取值和input_talbe[ ]的赋值,这两个过程是相对应的.

 

input_table中找到对应的handler之后,就会检验这个handle是否存,是否带有fops文件操作集.如果没有.则返回一个设备不存在的错误.

         /*

          * That's _really_odd. Usually NULL ->open means "nothing special",

          * not "nodevice". Oh, well...

          */

         if (!new_fops->open) {

                  fops_put(new_fops);

                  return -ENODEV;

         }

         old_fops = file->f_op;

         file->f_op = new_fops;

 

         err =new_fops->open(inode, file);

 

         if (err) {

                  fops_put(file->f_op);

                  file->f_op = fops_get(old_fops);

         }

         fops_put(old_fops);

         return err;

}

然后将handler中的fops替换掉当前的fops.如果新的fops中有open()函数,则调用它.

 

:evdev的初始化

Evdev的模块初始化函数为evdev_init().代码如下:

static int __init evdev_init(void)

{

         returninput_register_handler(&evdev_handler);

}

它调用了input_register_handler注册了一个handler.

注意到,在这里evdev_handler中定义的minorEVDEV_MINOR_BASE(64).也就是说evdev_handler所表示的设备文件范围为(13,64)à(13,64+32).

从之前的分析我们知道.匹配成功的关键在于handler中的blacklistid_talbe. Evdev_handlerid_table定义如下:

static const struct input_device_id evdev_ids[] = {

         { .driver_info = 1},     /* Matches all devices */

         {},                      /* Terminating zero entry */

};

它没有定义flags.也没有定义匹配属性值.这个handler是匹配所有input device.从前面的分析我们知道.匹配成功之后会调用handler->connect函数.

Evdev_handler,该成员函数如下所示:

 

static int evdev_connect(struct input_handler *handler, struct input_dev*dev,

                            const struct input_device_id *id)

{

         struct evdev *evdev;

         int minor;

         int error;

 

         for (minor = 0; minor< EVDEV_MINORS; minor++)

                  if (!evdev_table[minor])

                           break;

 

         if (minor ==EVDEV_MINORS) {

                  printk(KERN_ERR "evdev: no more free evdev devices/n");

                  return -ENFILE;

         }

EVDEV_MINORS定义为32.表示evdev_handler所表示的32个设备文件.evdev_talbe是一个struct evdev类型的数组.struct evdev是模块使用的封装结构.在接下来的代码中我们可以看到这个结构的使用.

这一段代码的在evdev_talbe找到为空的那一项.minor就是数组中第一项为空的序号.

 

         evdev =kzalloc(sizeof(struct evdev), GFP_KERNEL);

         if (!evdev)

                  return -ENOMEM;

 

        INIT_LIST_HEAD(&evdev->client_list);

        spin_lock_init(&evdev->client_lock);

        mutex_init(&evdev->mutex);

        init_waitqueue_head(&evdev->wait);

 

         snprintf(evdev->name,sizeof(evdev->name), "event%d", minor);

         evdev->exist = 1;

         evdev->minor = minor;

 

         evdev->handle.dev =input_get_device(dev);

         evdev->handle.name =evdev->name;

         evdev->handle.handler= handler;

         evdev->handle.private= evdev;

接下来,分配了一个evdev结构,并对这个结构进行初始化.在这里我们可以看到,这个结构封装了一个handle结构,这结构与我们之前所讨论的handler是不相同的.注意有一个字母的差别哦.我们可以把handle看成是handlerinput device的信息集合体.在这个结构里集合了匹配成功的handlerinput device

 

        strlcpy(evdev->dev.bus_id, evdev->name, sizeof(evdev->dev.bus_id));

         evdev->dev.devt =MKDEV(INPUT_MAJOR, EVDEV_MINOR_BASE + minor);

         evdev->dev.class =&input_class;

         evdev->dev.parent =&dev->dev;

         evdev->dev.release =evdev_free;

        device_initialize(&evdev->dev);

在这段代码里主要完成evdev封装的device的初始化.注意在这里,使它所属的类指向input_class.这样在/sysfs中创建的设备目录就会在/sys/class/input/下面显示.

 

         error =input_register_handle(&evdev->handle);

         if (error)

                  goto err_free_evdev;

         error =evdev_install_chrdev(evdev);

         if (error)

                  goto err_unregister_handle;

 

         error =device_add(&evdev->dev);

         if (error)

                  goto err_cleanup_evdev;

 

         return 0;

 

 err_cleanup_evdev:

         evdev_cleanup(evdev);

 err_unregister_handle:

         input_unregister_handle(&evdev->handle);

 err_free_evdev:

        put_device(&evdev->dev);

         return error;

}

注册handle,如果是成功的,那么调用evdev_install_chrdevevdev_tableminor项指向evdev.然后将evdev->device注册到sysfs.如果失败,将进行相关的错误处理.

万事俱备了,但是要接收事件,还得要等东风”.这个东风就是要打开相应的handle.这个打开过程是在文件的open()中完成的.

 

:evdev设备结点的open()操作

我们知道.对主设备号为INPUT_MAJOR的设备节点进行操作,会将操作集转换成handler的操作集.evdev,这个操作集就是evdev_fops.对应的open函数如下示:

static int evdev_open(struct inode *inode, struct file *file)

{

         struct evdev *evdev;

         struct evdev_client*client;

         int i = iminor(inode) -EVDEV_MINOR_BASE;

         int error;

 

         if (i >= EVDEV_MINORS)

                  return -ENODEV;

 

         error =mutex_lock_interruptible(&evdev_table_mutex);

         if (error)

                  return error;

         evdev = evdev_table[i];

         if (evdev)

                  get_device(&evdev->dev);

        mutex_unlock(&evdev_table_mutex);

 

         if (!evdev)

                  return -ENODEV;

 

         client = kzalloc(sizeof(structevdev_client), GFP_KERNEL);

         if (!client) {

                  error = -ENOMEM;

                  goto err_put_evdev;

         }

        spin_lock_init(&client->buffer_lock);

         client->evdev = evdev;

         evdev_attach_client(evdev,client);

 

         error =evdev_open_device(evdev);

         if (error)

                  goto err_free_client;

 

         file->private_data =client;

         return 0;

 

 err_free_client:

        evdev_detach_client(evdev, client);

         kfree(client);

 err_put_evdev:

        put_device(&evdev->dev);

         return error;

}

iminor(inode) - EVDEV_MINOR_BASE就得到了在evdev_table[ ]中的序号.然后将数组中对应的evdev取出.递增devdevdevice的引用计数.

分配并初始化一个client.并将它和evdev关联起来: client->evdev指向它所表示的evdev.client挂到evdev->client_list.client赋为file的私有区.

对应handle的打开是在此evdev_open_device()中完成的.代码如下:

static int evdev_open_device(struct evdev *evdev)

{

         int retval;

 

         retval =mutex_lock_interruptible(&evdev->mutex);

         if (retval)

                  return retval;

 

         if (!evdev->exist)

                  retval = -ENODEV;

         else if(!evdev->open++) {

                  retval = input_open_device(&evdev->handle);

                  if (retval)

                           evdev->open--;

         }

 

        mutex_unlock(&evdev->mutex);

         return retval;

}

如果evdev是第一次打开,就会调用input_open_device()打开evdev对应的handle.跟踪一下这个函数:

int input_open_device(struct input_handle *handle)

{

         struct input_dev *dev =handle->dev;

         int retval;

 

         retval =mutex_lock_interruptible(&dev->mutex);

         if (retval)

                  return retval;

 

         if (dev->going_away) {

                  retval = -ENODEV;

                  goto out;

         }

 

         handle->open++;

 

         if (!dev->users++&& dev->open)

                  retval = dev->open(dev);

 

         if (retval) {

                  dev->users--;

                  if (!--handle->open) {

                           /*

                            * Make sure we are not delivering any more events

                            * through this handle

                            */

                           synchronize_rcu();

                  }

         }

 

 out:

        mutex_unlock(&dev->mutex);

         return retval;

}

在这个函数中,我们看到.递增handle的打开计数.如果是第一次打开.则调用input deviceopen()函数.

 

十一:evdev的事件处理

经过上面的分析.每当input device上报一个事件时,会将其交给和它匹配的handlerevent函数处理.evdev.这个event函数对应的代码为:

static void evdev_event(struct input_handle *handle,

                           unsigned int type, unsigned int code, int value)

{

         struct evdev *evdev =handle->private;

         struct evdev_client*client;

         struct input_event event;

 

        do_gettimeofday(&event.time);

         event.type = type;

         event.code = code;

         event.value = value;

 

         rcu_read_lock();

 

         client =rcu_dereference(evdev->grab);

         if (client)

                  evdev_pass_event(client, &event);

         else

                  list_for_each_entry_rcu(client, &evdev->client_list, node)

                           evdev_pass_event(client, &event);

 

         rcu_read_unlock();

 

        wake_up_interruptible(&evdev->wait);

}

首先构造一个struct input_event结构.并设备它的type.code,value为处理事件的相关属性.如果该设备被强制设置了handle.则调用如之对应的client.

我们在open的时候分析到.会初始化clinet并将其链入到evdev->client_list.这样,就可以通过evdev->client_list找到这个client.

对于找到的第一个client都会调用evdev_pass_event( ).代码如下:

static void evdev_pass_event(struct evdev_client *client,

                                struct input_event *event)

{

         /*

          * Interrupts aredisabled, just acquire the lock

          */

        spin_lock(&client->buffer_lock);

        client->buffer[client->head++] = *event;

         client->head &=EVDEV_BUFFER_SIZE - 1;

         spin_unlock(&client->buffer_lock);

 

        kill_fasync(&client->fasync, SIGIO, POLL_IN);

}

这里的操作很简单.就是将event保存到client->buffer.client->head就是当前的数据位置.注意这里是一个环形缓存区.写数据是从client->head.而读数据则是从client->tail中读.

 

十二:设备节点的read处理

对于evdev设备节点的read操作都会由evdev_read()完成.它的代码如下:

static ssize_t evdev_read(struct file *file, char __user *buffer,

                             size_t count, loff_t *ppos)

{

         struct evdev_client*client = file->private_data;

         struct evdev *evdev =client->evdev;

         struct input_event event;

         int retval;

 

         if (count <evdev_event_size())

                  return -EINVAL;

 

         if (client->head ==client->tail && evdev->exist &&

            (file->f_flags & O_NONBLOCK))

                  return -EAGAIN;

 

         retval =wait_event_interruptible(evdev->wait,

                  client->head != client->tail || !evdev->exist);

         if (retval)

                  return retval;

 

         if (!evdev->exist)

                  return -ENODEV;

 

         while (retval +evdev_event_size() <= count &&

               evdev_fetch_next_event(client,&event)) {

 

                  if (evdev_event_to_user(buffer + retval, &event))

                           return -EFAULT;

 

                  retval += evdev_event_size();

         }

 

         return retval;

}

首先,它判断缓存区大小是否足够.在读取数据的情况下,可能当前缓存区内没有数据可读.在这里先睡眠等待缓存区中有数据.如果在睡眠的时候,.条件满足.是不会进行睡眠状态而直接返回的.

然后根据read()提够的缓存区大小.client中的数据写入到用户空间的缓存区中.

十三:设备节点的写操作

同样.对设备节点的写操作是由evdev_write()完成的.代码如下:

 

static ssize_t evdev_write(struct file *file, const char __user *buffer,

                              size_t count, loff_t *ppos)

{

         struct evdev_client*client = file->private_data;

         struct evdev *evdev =client->evdev;

         struct input_event event;

         int retval;

 

         retval =mutex_lock_interruptible(&evdev->mutex);

         if (retval)

                  return retval;

 

         if (!evdev->exist) {

                  retval = -ENODEV;

                  goto out;

         }

 

         while (retval < count){

 

                  if (evdev_event_from_user(buffer + retval, &event)) {

                           retval = -EFAULT;

                           goto out;

                  }

 

                  input_inject_event(&evdev->handle,

                                       event.type, event.code, event.value);

                  retval += evdev_event_size();

         }

 

 out:

        mutex_unlock(&evdev->mutex);

         return retval;

}

首先取得操作设备文件所对应的evdev.

实际上,这里写入设备文件的是一个event结构的数组.我们在之前分析过,这个结构里包含了事件的type.codeevent.

将写入设备的event数组取出.然后对每一项调用event_inject_event().

这个函数的操作和input_event()差不多.就是将第一个参数handle转换为输入设备结构.然后这个设备再产生一个事件.

代码如下:

void input_inject_event(struct input_handle *handle,

                           unsigned int type, unsigned int code, int value)

{

         struct input_dev *dev =handle->dev;

         struct input_handle*grab;

         unsigned long flags;

 

         if(is_event_supported(type, dev->evbit, EV_MAX)) {

                  spin_lock_irqsave(&dev->event_lock, flags);

 

                  rcu_read_lock();

                  grab = rcu_dereference(dev->grab);

                  if (!grab || grab == handle)

                           input_handle_event(dev, type, code, value);

                  rcu_read_unlock();

 

                  spin_unlock_irqrestore(&dev->event_lock, flags);

         }

}

我们在这里也可以跟input_event()对比一下,这里设备可以产生任意事件,而不需要和设备所支持的事件类型相匹配.

由此可见.对于写操作而言.就是让与设备文件相关的输入设备产生一个特定的事件.

将上述设备文件的操作过程以图的方式表示如下:

 

 

十四:小结

在这一节点,分析了整个input子系统的架构,各个环节的流程.最后还以evdev为例.将各个流程贯穿在一起.以加深对input子系统的理解.由此也可以看出.linux设备驱动采用了分层的模式.从最下层的设备模型到设备,驱动,总线再到input子系统最后到input device.这样的分层结构使得最上层的驱动不必关心下层是怎么实现的.而下层驱动又为多种型号同样功能的驱动提供了一个统一的接口.