基于uda341的OSS声卡驱动

最新推荐文章于 2021-05-16 22:58:48 发布

cocosasa

最新推荐文章于 2021-05-16 22:58:48 发布

阅读量613

点赞数 1

文章标签： audio stream iis struct buffer output

我将基于uda341的OSS声卡驱动程序全部贴在了此处，可以直接全部复制了去。不过针对不同内核版本要做一些改动。这个程序不难看懂，我也加了足够多的注释，那些注释只能代表我个人的理解，我没有参看任何资料我也不敢保证我的理解就是正确的。看懂之后可以硬着头皮编译一下，不过得有一定的心理承受能力。上百个错误出现在你面前时，但愿你还能保持以往对声卡驱动的热情。

#include <linux/module.h>
#include <linux/device.h>
#include <linux/init.h>
#include <linux/types.h>
#include <linux/fs.h>
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/delay.h>
#include <linux/sched.h>
#include <linux/poll.h>
#include <linux/interrupt.h>
#include <linux/errno.h>
#include <linux/sound.h>
#include <linux/soundcard.h>

#include <linux/pm.h>

#include <asm/uaccess.h>
#include <asm/io.h>
#include <asm/hardware.h>
#include <asm/semaphore.h>
#include <asm/dma.h>
//#include <asm/arch/cpu_s3c2410.h>
#include <asm/arch/dma.h>
#include <asm/arch/regs-gpio.h>
#include <asm/arch/regs-iis.h>
#include <asm/hardware/clock.h>
#include <asm/arch/regs-clock.h>
#include <linux/dma-mapping.h>
#include <asm/dma-mapping.h>
#include <asm/arch/hardware.h>
#include <asm/arch/map.h>
#include <asm/arch/S3C2410.h>
#define PFX "s3c2410-uda1341-superlp: "

#define MAX_DMA_CHANNELS 0

/* The S3C2410 has four internal DMA channels. */

#define MAX_S3C2410_DMA_CHANNELS S3C2410_DMA_CHANNELS

#define DMA_CH0 0
#define DMA_CH1 1
#define DMA_CH2 2
#define DMA_CH3 3

#define DMA_BUF_WR 1
#define DMA_BUF_RD 0

#define dma_wrreg(chan, reg, val) writel((val), (chan)->regs + (reg))

static struct clk *iis_clock;
static void __iomem *iis_base;

static struct s3c2410_dma_client s3c2410iis_dma_out= {
.name = "I2SSDO",
};

static struct s3c2410_dma_client s3c2410iis_dma_in = {
.name = "I2SSDI",
};

#ifdef DEBUG
#define DPRINTK printk
#else
#define DPRINTK( x... )
#endif

static void init_s3c2410_iis_bus_rx(void);
static void init_s3c2410_iis_bus_tx(void);

#define DEF_VOLUME 65 //音量预设为65%

/* UDA1341 Register bits */
#define UDA1341_ADDR 0x14 //数据手册中有这么一句The address of the UDA1341TS is 000101.
/*80~158行参看数据手册*/
#define UDA1341_REG_DATA0 (UDA1341_ADDR + 0)
#define UDA1341_REG_STATUS (UDA1341_ADDR + 2)

/* status control */
#define STAT0 (0x00)
#define STAT0_RST (1 << 6)
#define STAT0_SC_MASK (3 << 4)
#define STAT0_SC_512FS (0 << 4)
#define STAT0_SC_384FS (1 << 4)
#define STAT0_SC_256FS (2 << 4)
#define STAT0_IF_MASK (7 << 1)
#define STAT0_IF_I2S (0 << 1)
#define STAT0_IF_LSB16 (1 << 1)
#define STAT0_IF_LSB18 (2 << 1)
#define STAT0_IF_LSB20 (3 << 1)
#define STAT0_IF_MSB (4 << 1)
#define STAT0_IF_LSB16MSB (5 << 1)
#define STAT0_IF_LSB18MSB (6 << 1)
#define STAT0_IF_LSB20MSB (7 << 1)
#define STAT0_DC_FILTER (1 << 0)
#define STAT0_DC_NO_FILTER (0 << 0)

#define STAT1 (0x80)
#define STAT1_DAC_GAIN (1 << 6) /* gain of DAC */
#define STAT1_ADC_GAIN (1 << 5) /* gain of ADC */
#define STAT1_ADC_POL (1 << 4) /* polarity of ADC */
#define STAT1_DAC_POL (1 << 3) /* polarity of DAC */
#define STAT1_DBL_SPD (1 << 2) /* double speed playback */
#define STAT1_ADC_ON (1 << 1) /* ADC powered */
#define STAT1_DAC_ON (1 << 0) /* DAC powered */

/* data0 direct control */
#define DATA0 (0x00)
#define DATA0_VOLUME_MASK (0x3f)
#define DATA0_VOLUME(x) (x)

#define DATA1 (0x40)
#define DATA1_BASS(x) ((x) << 2)
#define DATA1_BASS_MASK (15 << 2)
#define DATA1_TREBLE(x) ((x))
#define DATA1_TREBLE_MASK (3)

#define DATA2 (0x80)
#define DATA2_PEAKAFTER (0x1 << 5)
#define DATA2_DEEMP_NONE (0x0 << 3)
#define DATA2_DEEMP_32KHz (0x1 << 3)
#define DATA2_DEEMP_44KHz (0x2 << 3)
#define DATA2_DEEMP_48KHz (0x3 << 3)
#define DATA2_MUTE (0x1 << 2)
#define DATA2_FILTER_FLAT (0x0 << 0)
#define DATA2_FILTER_MIN (0x1 << 0)
#define DATA2_FILTER_MAX (0x3 << 0)
/* data0 extend control */
#define EXTADDR(n) (0xc0 | (n))
#define EXTDATA(d) (0xe0 | (d))

#define EXT0 0
#define EXT0_CH1_GAIN(x) (x)
#define EXT1 1
#define EXT1_CH2_GAIN(x) (x)
#define EXT2 2
#define EXT2_MIC_GAIN_MASK (7 << 2)
#define EXT2_MIC_GAIN(x) ((x) << 2)
#define EXT2_MIXMODE_DOUBLEDIFF (0)
#define EXT2_MIXMODE_CH1 (1)
#define EXT2_MIXMODE_CH2 (2)
#define EXT2_MIXMODE_MIX (3)
#define EXT4 4
#define EXT4_AGC_ENABLE (1 << 4)
#define EXT4_INPUT_GAIN_MASK (3)
#define EXT4_INPUT_GAIN(x) ((x) & 3)
#define EXT5 5
#define EXT5_INPUT_GAIN(x) ((x) >> 2)
#define EXT6 6
#define EXT6_AGC_CONSTANT_MASK (7 << 2)
#define EXT6_AGC_CONSTANT(x) ((x) << 2)
#define EXT6_AGC_LEVEL_MASK (3)
#define EXT6_AGC_LEVEL(x) (x)

#define AUDIO_NAME "UDA1341"
#define AUDIO_NAME_VERBOSE "UDA1341 audio driver"

#define AUDIO_FMT_MASK (AFMT_S16_LE) //小端模式 16位
#define AUDIO_FMT_DEFAULT (AFMT_S16_LE)

#define AUDIO_CHANNELS_DEFAULT 2 //默认通道2
#define AUDIO_RATE_DEFAULT 44100 //默认采样频率44.1k

#define AUDIO_NBFRAGS_DEFAULT 8 //数据缓冲区默认段数
#define AUDIO_FRAGSIZE_DEFAULT 8192 //一段数据缓存默认大小

#define S_CLOCK_FREQ 384 //默认CODECLK = 384fs
#define PCM_ABS(a) (a < 0 ? -a : a) //取绝对值

typedef struct { //缓存段的管理结构体
int size; /* buffer size */
char *start; /* point to actual buffer *///虚拟地址
dma_addr_t dma_addr; //物理地址，向DMA的初始化源寄存器或目的寄存器写的应是物理地址
struct semaphore sem; /* down before touching the buffer */
int master; /* owner for buffer allocation, contain size when true */
} audio_buf_t;

typedef struct { //管理缓存链
audio_buf_t *buffers; /* pointer to audio buffer structures */
audio_buf_t *buf; /* current buffer used by read/write */
u_int buf_idx; /* index for the pointer above */
u_int fragsize; /* fragment i.e. buffer size */
u_int nbfrags; /* nbr of fragments */
dmach_t dma_ch; /* DMA channel (channel2 for audio) */
u_int dma_ok;
} audio_stream_t;

static audio_stream_t output_stream;//输出缓存
static audio_stream_t input_stream; /* input *///输入缓存
//以下宏的作用是s->buf指向下一个数据缓存段
#define NEXT_BUF(_s_,_b_) { /
(_s_)->_b_##_idx++; / //如果_b_是buf该句的意思是s->buf_idx++
(_s_)->_b_##_idx %= (_s_)->nbfrags; / //s->nbfrags为整个缓存链的缓存段数，该句为防止s->buf_idx值超出缓存链
(_s_)->_b_ = (_s_)->buffers + (_s_)->_b_##_idx; } //改变当前使用的缓存段s->buf

static u_int audio_rate; //采样频率
/*如果为2则不使用立体声，如果是1则使用立体声。
*具体使用哪个DMA通道由s->dma_ch来决定。
*/
static int audio_channels;
static int audio_fmt; //数据传输模式
static u_int audio_fragsize; //每缓存段数据长度
static u_int audio_nbfrags; //缓存段个数

static int audio_rd_refcount; //读引用计数
static int audio_wr_refcount; //写引用计数
#define audio_active (audio_rd_refcount | audio_wr_refcount) //

static int audio_dev_dsp; //存放DSP设备注册后的次设备号
static int audio_dev_mixer; //存放混频设备注册后的次设备号
static int audio_mix_modcnt;

static int uda1341_volume; //
static u8 uda_sampling; //采样频率
static int uda1341_boost; //存放低音增强值
static int mixer_igain=0x4; /* -6db*/ //混频器默认输入增益

static void uda1341_l3_address(u8 data) //地址写入
{
int i;
unsigned long flags;

local_irq_save(flags);

// write_gpio_bit(GPIO_L3MODE, 0);
s3c2410_gpio_setpin(S3C2410_GPB2,0);
// write_gpio_bit(GPIO_L3CLOCK, 1);
s3c2410_gpio_setpin(S3C2410_GPB4,1);
udelay(1);

for (i = 0; i < 8; i++) { //8为数据传输
  if (data & 0x1) { //低位在前
   s3c2410_gpio_setpin(S3C2410_GPB4,0);
   s3c2410_gpio_setpin(S3C2410_GPB3,1);
   udelay(1);
   s3c2410_gpio_setpin(S3C2410_GPB4,1);
  } else {
   s3c2410_gpio_setpin(S3C2410_GPB4,0);
   s3c2410_gpio_setpin(S3C2410_GPB3,0);
   udelay(1);
   s3c2410_gpio_setpin(S3C2410_GPB4,1);
  }
  data >>= 1;
}

s3c2410_gpio_setpin(S3C2410_GPB2,1);
s3c2410_gpio_setpin(S3C2410_GPB4,1);
local_irq_restore(flags);
}

static void uda1341_l3_data(u8 data) //数据传输函数
{
int i;
unsigned long flags;

local_irq_save(flags);
udelay(1);

for (i = 0; i < 8; i++) {
  if (data & 0x1) {
   s3c2410_gpio_setpin(S3C2410_GPB4,0);
   s3c2410_gpio_setpin(S3C2410_GPB3,1);
   udelay(1);
   s3c2410_gpio_setpin(S3C2410_GPB4,1);
  } else {
   s3c2410_gpio_setpin(S3C2410_GPB4,0);
   s3c2410_gpio_setpin(S3C2410_GPB3,0);
   udelay(1);
   s3c2410_gpio_setpin(S3C2410_GPB4,1);
  }

  data >>= 1;
}

local_irq_restore(flags);
}

static void audio_clear_buf(audio_stream_t * s)
{
DPRINTK("audio_clear_buf/n");
//停止DMA传输
if(s->dma_ok) s3c2410_dma_ctrl(s->dma_ch, S3C2410_DMAOP_FLUSH);

if (s->buffers) {
  int frag;

  for (frag = 0; frag < s->nbfrags; frag++) {
   if (!s->buffers[frag].master)
    continue;
   dma_free_coherent(NULL, //释放缓存段
    s->buffers[frag].master,
    s->buffers[frag].start,
    s->buffers[frag].dma_addr);
  }
  kfree(s->buffers); //释放内存
  s->buffers = NULL;
}

s->buf_idx = 0;
s->buf = NULL;
}

static int audio_setup_buf(audio_stream_t * s)
{
int frag;
int dmasize = 0;
char *dmabuf = 0;
dma_addr_t dmaphys = 0;

if (s->buffers)
  return -EBUSY;

s->nbfrags = audio_nbfrags; //数据缓存段数
s->fragsize = audio_fragsize; //每段数据缓存的长度

s->buffers = (audio_buf_t *)
  kmalloc(sizeof(audio_buf_t) * s->nbfrags, GFP_KERNEL);
if (!s->buffers)
  goto err;
memset(s->buffers, 0, sizeof(audio_buf_t) * s->nbfrags);

for (frag = 0; frag < s->nbfrags; frag++) {
  audio_buf_t *b = &s->buffers[frag];

  if (!dmasize) {
   dmasize = (s->nbfrags - frag) * s->fragsize;
   do {
     /*建立DMA一致性映射
    *如果dmasize过大导致DMA建立不成功则dmasize -= s->fragsize，
    再次尝试建立映射。直到dmasize等于0。
    */
    dmabuf = dma_alloc_coherent(NULL, dmasize, &dmaphys, GFP_KERNEL|GFP_DMA);
    if (!dmabuf)
     dmasize -= s->fragsize;
   } while (!dmabuf && dmasize);
   if (!dmabuf)
    goto err;
   b->master = dmasize; //每次成功映射建立的缓存段大小
  }

  b->start = dmabuf;
  b->dma_addr = dmaphys;
  sema_init(&b->sem, 1); //将控制每段缓存访问的信号量初始化为1，相当于互斥信号
  DPRINTK("buf %d: start %p dma %d/n", frag, b->start, b->dma_addr);

  dmabuf += s->fragsize;
  dmaphys += s->fragsize;
  /*将成功建立的映射缓存分给各缓存段。如果没有建立足够大的缓存，当dmasize
   *等于0将导致再次DMA映射。
  */
  dmasize -= s->fragsize;
}

s->buf_idx = 0;
s->buf = &s->buffers[0];
return 0;

err:
printk(AUDIO_NAME ": unable to allocate audio memory/n ");
audio_clear_buf(s);
return -ENOMEM;
}
/*一段数据缓存耗尽时的回调函数，
在DMA中断处理函数中调用了函数s3c2410_dma_buffdone（），
在该函数中调用函数(chan->callback_fn)(chan, buf->id, buf->size, result)，
即为本函数。
*/
static void audio_dmaout_done_callback(s3c2410_dma_chan_t *ch, void *buf, int size,
            s3c2410_dma_buffresult_t result)
{
audio_buf_t *b = (audio_buf_t *) buf;
up(&b->sem);//释放信号量
wake_up(&b->sem.wait); //唤醒等待使用该段缓存的进程
}

static void audio_dmain_done_callback(s3c2410_dma_chan_t *ch, void *buf, int size,
           s3c2410_dma_buffresult_t result)
{
audio_buf_t *b = (audio_buf_t *) buf;
b->size = size;
up(&b->sem);
wake_up(&b->sem.wait);
}
/* using when write */
//同步，将当前缓存段中的数据发送出去
static int audio_sync(struct file *file)
{
audio_stream_t *s = &output_stream;
audio_buf_t *b = s->buf;

DPRINTK("audio_sync/n");

if (!s->buffers)
  return 0;

if (b->size != 0) {
  down(&b->sem);
  /*
   函数s3c2410_dma_enqueue的主要作用是申请一内存管理结构，将一段缓存
   交给该结构，将该段缓存插入DMA备用缓存链。

  在函数s3c2410_dma_enqueue函数中有如下程序段
  if (chan->flags & S3C2410_DMAF_AUTOSTART) {
   s3c2410_dma_ctrl(chan->number | DMACH_LOW_LEVEL,
      S3C2410_DMAOP_START);
  }
  而在函数__init audio_init_dma(audio_stream_t * s, char *desc) 中
  有flags = S3C2410_DMAF_AUTOSTART从而保证了函数s3c2410_dma_enqueue
   的调用即开启DMA传输。
  */
  s3c2410_dma_enqueue(s->dma_ch, (void *) b, b->dma_addr, b->size);
  b->size = 0;
  NEXT_BUF(s, buf);
}

b = s->buffers + ((s->nbfrags + s->buf_idx - 1) % s->nbfrags);
if (down_interruptible(&b->sem))
  return -EINTR;
up(&b->sem);

return 0;
}
//选择立体声时，从用户空间读取数据时使用该函数
static inline int copy_from_user_mono_stereo(char *to, const char *from, int count)
{
u_int *dst = (u_int *)to;
const char *end = from + count;

if (verify_area(VERIFY_READ, from, count))
  return -EFAULT;

if ((int)from & 0x2) { //如果起始位置非字对齐
  u_int v;
  __get_user(v, (const u_short *)from); from += 2;
  *dst++ = v | (v << 16);
}

while (from < end-2) {
  u_int v, x, y;
  __get_user(v, (const u_int *)from); from += 4;//占用了4个16位的数据空间，x和y
  x = v << 16;
  x |= x >> 16;//左右声道的16位数据相同
  y = v >> 16;
  y |= y << 16;
  *dst++ = x;
  *dst++ = y;
}

if (from < end) { //如果缓存段尾有未字对齐的数据
  u_int v;
  __get_user(v, (const u_short *)from);
  *dst = v | (v << 16);
}

return 0;
}

static ssize_t smdk2410_audio_write(struct file *file, const char *buffer,
         size_t count, loff_t * ppos)
{
const char *buffer0 = buffer; //存放数据源缓存的起始位置
audio_stream_t *s = &output_stream;
int chunksize, ret = 0;

DPRINTK("audio_write : start count=%d/n", count);

switch (file->f_flags & O_ACCMODE) {
case O_WRONLY:
case O_RDWR:
  break;
default:
  return -EPERM;
}

if (!s->buffers && audio_setup_buf(s)) //建立输出数据缓存
  return -ENOMEM;

count &= ~0x03; //字对齐

while (count > 0) {
  audio_buf_t *b = s->buf;

  if (file->f_flags & O_NONBLOCK) {//如果文件为非阻塞打开
   ret = -EAGAIN;
   if (down_trylock(&b->sem)) //获取信号量
    break;
  } else {
   ret = -ERESTARTSYS;
   if (down_interruptible(&b->sem))
    break;
  }

  if (audio_channels == 2) { //如果audio_channels等于2则不使用立体声
   chunksize = s->fragsize - b->size;
   if (chunksize > count)
    chunksize = count;
   DPRINTK("write %d to %d/n", chunksize, s->buf_idx);
   if (copy_from_user(b->start + b->size, buffer, chunksize)) {
    up(&b->sem);
    return -EFAULT;
   }
   b->size += chunksize;
  } else {
   chunksize = (s->fragsize - b->size) >> 1; //计算半字数

   if (chunksize > count)
    chunksize = count;
   DPRINTK("write %d to %d/n", chunksize*2, s->buf_idx);
   if (copy_from_user_mono_stereo(b->start + b->size, //使用立体声播放
    buffer, chunksize)) {
    up(&b->sem);
    return -EFAULT;
   }

   b->size += chunksize*2; //记录该段缓存中已有的数据量
  }

  buffer += chunksize; //数据源指针前移
  count -= chunksize;
  if (b->size < s->fragsize) {
   up(&b->sem);
   break;
  }
  //本段数据缓存段已写入数据，使能DMA进行数据传输
  if((ret = s3c2410_dma_enqueue(s->dma_ch, (void *) b, b->dma_addr, b->size))) {
   printk(PFX"dma enqueue failed./n");
   return ret;
  }
  b->size = 0;
  NEXT_BUF(s, buf); //让当前使用数据段指向下一个数据段。
}

if ((buffer - buffer0))
  ret = buffer - buffer0; //计算本次数据写入的总数据量。

DPRINTK("audio_write : end count=%d/n/n", ret);

return ret;
}

static ssize_t smdk2410_audio_read(struct file *file, char *buffer,
           size_t count, loff_t * ppos)
{
const char *buffer0 = buffer;
audio_stream_t *s = &input_stream;
int chunksize, ret = 0;

DPRINTK("audio_read: count=%d/n", count);

if (ppos != &file->f_pos)
  return -ESPIPE;

if (!s->buffers) { //如果输入数据缓存没有建立，则建立并使能DMA读取数据
  int i;

  if (audio_setup_buf(s))
   return -ENOMEM;

  for (i = 0; i < s->nbfrags; i++) {
   audio_buf_t *b = s->buf;
   /*获取信号量，在本段数据读满时，DMA中断函数中调用函数
    *audio_dmain_done_callback 释放信号量
   */
   down(&b->sem);
   s3c2410_dma_enqueue(s->dma_ch, (void *) b, b->dma_addr, s->fragsize);
   NEXT_BUF(s, buf);
  }
}

while (count > 0) {
  audio_buf_t *b = s->buf;

  /* Wait for a buffer to become full */
  if (file->f_flags & O_NONBLOCK) {
   ret = -EAGAIN;
   if (down_trylock(&b->sem))
    break;
  } else {
   ret = -ERESTARTSYS;
   if (down_interruptible(&b->sem)) //如果本段缓存未读满则程序进入睡眠
    break;
  }

  chunksize = b->size;
  if (chunksize > count)
   chunksize = count;
  DPRINTK("read %d from %d/n", chunksize, s->buf_idx);
  /*如果本段数据缓存是第一次被读取，程序运行到此说明s->fragsize = b->size
  即s->fragsize - b->size等于0。因为如果count < s->fragsize则可能导致
   第二次读取本段缓存。
  */
  if (copy_to_user(buffer, b->start + s->fragsize - b->size,
   chunksize)) {
   up(&b->sem);
   return -EFAULT;
  }

  b->size -= chunksize;

  buffer += chunksize;
  count -= chunksize;
  if (b->size > 0) {
   up(&b->sem);
   break;
  }
  //本段缓存已读取则把这段缓存插入DMA传输缓存链，准备用于数据读取。
  /* Make current buffer available for DMA again */
  s3c2410_dma_enqueue(s->dma_ch, (void *) b, b->dma_addr, s->fragsize);

  NEXT_BUF(s, buf);
}

if ((buffer - buffer0)) //计算读取的数据总量。
  ret = buffer - buffer0;

// DPRINTK("audio_read: return=%d/n", ret);

return ret;
}

static unsigned int smdk2410_audio_poll(struct file *file,
          struct poll_table_struct *wait)
{
unsigned int mask = 0;
int i;

DPRINTK("audio_poll(): mode=%s/n",
  (file->f_mode & FMODE_WRITE) ? "w" : "");

if (file->f_mode & FMODE_READ) {
  if (!input_stream.buffers && audio_setup_buf(&input_stream))
   return -ENOMEM;
   /*增加一个等待队列到 poll_table 结构.
       信号量结构体中好像没有wait等待队列，
       不知道为什么有这么一句input_stream.buf->sem.wait。
       struct semaphore {
     spinlock_t  lock;
     unsigned int  count;
     struct list_head wait_list;
                };
   */
  poll_wait(file, &input_stream.buf->sem.wait, wait);
  //如果input_stream.buffers[i].sem.count大于0则该段数据缓存没被使用
  for (i = 0; i < input_stream.nbfrags; i++) {
   if (atomic_read(&input_stream.buffers[i].sem.count) > 0)
    mask |= POLLIN | POLLWRNORM;
   break;
  }
}

if (file->f_mode & FMODE_WRITE) {
  if (!output_stream.buffers && audio_setup_buf(&output_stream))
   return -ENOMEM;
  poll_wait(file, &output_stream.buf->sem.wait, wait);

  for (i = 0; i < output_stream.nbfrags; i++) {
   if (atomic_read(&output_stream.buffers[i].sem.count) > 0)
    mask |= POLLOUT | POLLWRNORM;
   break;
  }
}

DPRINTK("audio_poll() returned mask of %s/n",
  (mask & POLLOUT) ? "w" : "");

return mask;
}

static loff_t smdk2410_audio_llseek(struct file *file, loff_t offset,
int origin)
{
return -ESPIPE;
}

static int smdk2410_mixer_ioctl(struct inode *inode, struct file *file,
        unsigned int cmd, unsigned long arg)
{
int ret;
long val = 0;

switch (cmd) {
case SOUND_MIXER_INFO: //读取混频器信息
  {
   mixer_info info;
   strncpy(info.id, "UDA1341", sizeof(info.id));
   strncpy(info.name,"Philips UDA1341", sizeof(info.name));
   info.modify_counter = audio_mix_modcnt;
   return copy_to_user((void *)arg, &info, sizeof(info));
  }

case SOUND_OLD_MIXER_INFO:
  {
   _old_mixer_info info;
   strncpy(info.id, "UDA1341", sizeof(info.id));
   strncpy(info.name,"Philips UDA1341", sizeof(info.name));
   return copy_to_user((void *)arg, &info, sizeof(info));
  }

case SOUND_MIXER_READ_STEREODEVS:
  return put_user(0, (long *) arg);

case SOUND_MIXER_READ_CAPS:
  val = SOUND_CAP_EXCL_INPUT;
  return put_user(val, (long *) arg);

case SOUND_MIXER_WRITE_VOLUME: //设置音量
  ret = get_user(val, (long *) arg);
  if (ret)
   return ret;
  uda1341_volume = 63 - (((val & 0xff) + 1) * 63) / 100;//63个档位设置音量为val%
  uda1341_l3_address(UDA1341_REG_DATA0);
  uda1341_l3_data(uda1341_volume);
  break;

case SOUND_MIXER_READ_VOLUME: //读取音量
  val = ((63 - uda1341_volume) * 100) / 63;
  val |= val << 8;
  return put_user(val, (long *) arg);

case SOUND_MIXER_READ_IGAIN: //混频器增益读取
  val = ((31- mixer_igain) * 100) / 31;
  return put_user(val, (int *) arg);

case SOUND_MIXER_WRITE_IGAIN: //混频器增益设定
  ret = get_user(val, (int *) arg);
  if (ret)
   return ret;
  mixer_igain = 31 - (val * 31 / 100); //共有31个档位
  /* use mixer gain channel 1*/
  uda1341_l3_address(UDA1341_REG_DATA0);
  uda1341_l3_data(EXTADDR(EXT0));
  uda1341_l3_data(EXTDATA(EXT0_CH1_GAIN(mixer_igain)));
  break;

default:
  DPRINTK("mixer ioctl %u unknown/n", cmd);
  return -ENOSYS;
}

audio_mix_modcnt++;
return 0;
}

static int iispsr_value(int s_bit_clock, int sample_rate)
{
int i, prescaler = 0;
unsigned long tmpval;
unsigned long tmpval384;
unsigned long tmpval384min = 0xffff;
//s_bit_clock为256或384
tmpval384 = clk_get_rate(iis_clock) / s_bit_clock;
/*
   采样频率要经过了两次分频，一次预标定器分频，一次384或256分频。
   tmpval384已经过了384分频，还未经过预标定分频的tmpval384
   仍然大于或等于sample_rate。i从0到32增加的过程中sample_rate - tmpval
   的差值从大到小再到大，即tmpval384min从0xffff变小再变大。以下循环即为
   寻找第一个使这个差值最小的i。
   sample_rate为设定频率，tmpval为实际分频后的值，当然要找一个实际值与
   设定值最接近的。
*/
for (i = 0; i < 32; i++) {
  tmpval = tmpval384/(i+1);
  if (PCM_ABS((sample_rate - tmpval)) < tmpval384min) {
   tmpval384min = PCM_ABS((sample_rate - tmpval));
   prescaler = i;
  }
}

DPRINTK("prescaler = %d/n", prescaler);

return prescaler;
}

static long audio_set_dsp_speed(long val)
{
unsigned int prescaler;
/*
我认为应有如下宏的声明
#define IISPSR_A(x) (x)<<5
#define IISPSR_B(x) (x)
*/
prescaler=(IISPSR_A(iispsr_value(S_CLOCK_FREQ, val))
| IISPSR_B(iispsr_value(S_CLOCK_FREQ, val)));
writel(prescaler, iis_base + S3C2410_IISPSR);

printk(PFX "audio_set_dsp_speed:%ld prescaler:%i/n",val,prescaler);
return (audio_rate = val);
}

static int smdk2410_audio_ioctl(struct inode *inode, struct file *file,
        uint cmd, ulong arg)
{
long val;

switch (cmd) {
case SNDCTL_DSP_SETFMT: //设定数据传输模式
  get_user(val, (long *) arg);
  if (val & AUDIO_FMT_MASK) {
   audio_fmt = val;
   break;
  } else
   return -EINVAL;

case SNDCTL_DSP_CHANNELS:
case SNDCTL_DSP_STEREO: //设置是否使用立体声功能
  get_user(val, (long *) arg);
  if (cmd == SNDCTL_DSP_STEREO)
   val = val ? 2 : 1;
  if (val != 1 && val != 2)
   return -EINVAL;
  audio_channels = val;
  break;

case SOUND_PCM_READ_CHANNELS:
  put_user(audio_channels, (long *) arg);
  break;

case SNDCTL_DSP_SPEED: //设定采样频率
  get_user(val, (long *) arg);
  val = audio_set_dsp_speed(val);
  if (val < 0)
   return -EINVAL;
  put_user(val, (long *) arg);
  break;

case SOUND_PCM_READ_RATE:
  put_user(audio_rate, (long *) arg);
  break;

case SNDCTL_DSP_GETFMTS:
  put_user(AUDIO_FMT_MASK, (long *) arg);
  break;

case SNDCTL_DSP_GETBLKSIZE:
  if(file->f_mode & FMODE_WRITE)
   return put_user(audio_fragsize, (long *) arg);//写缓冲是32位的因为左右声道各16位
  else
   return put_user(audio_fragsize, (int *) arg); //读缓冲是16的因为录音只有一个通道

case SNDCTL_DSP_SETFRAGMENT: //设置一段缓存的大小
  if (file->f_mode & FMODE_WRITE) {
   if (output_stream.buffers)
    return -EBUSY;
   get_user(val, (long *) arg);
   audio_fragsize = 1 << (val & 0xFFFF); //低16位为一段数据缓存大小
   if (audio_fragsize < 16)
    audio_fragsize = 16;
   if (audio_fragsize > 16384)
    audio_fragsize = 16384;
   audio_nbfrags = (val >> 16) & 0x7FFF; //高15位为缓存段的大小
   if (audio_nbfrags < 2)
    audio_nbfrags = 2;
   if (audio_nbfrags * audio_fragsize > 128 * 1024)
    audio_nbfrags = 128 * 1024 / audio_fragsize;
   if (audio_setup_buf(&output_stream))
    return -ENOMEM;

  }
  if (file->f_mode & FMODE_READ) {
   if (input_stream.buffers)
    return -EBUSY;
   get_user(val, (int *) arg);
   audio_fragsize = 1 << (val & 0xFFFF); //全部16位为一段数据段的大小
   if (audio_fragsize < 16)
    audio_fragsize = 16;
   if (audio_fragsize > 16384)
    audio_fragsize = 16384;
   audio_nbfrags = (val >> 16) & 0x7FFF; //低15位为数据缓存的段数
   if (audio_nbfrags < 2)
    audio_nbfrags = 2;
   if (audio_nbfrags * audio_fragsize > 128 * 1024)
    audio_nbfrags = 128 * 1024 / audio_fragsize;
   if (audio_setup_buf(&input_stream))
    return -ENOMEM;

  }
  break;

case SNDCTL_DSP_SYNC: //同步
  return audio_sync(file);

case SNDCTL_DSP_GETOSPACE: //获取空余的缓存量
  {
   audio_stream_t *s = &output_stream;
   audio_buf_info *inf = (audio_buf_info *) arg;
   int err = verify_area(VERIFY_WRITE, inf, sizeof(*inf)); //检测内存inf是否可用
   int i;
   int frags = 0, bytes = 0;

   if (err)
    return err;
   for (i = 0; i < s->nbfrags; i++) {
    if (atomic_read(&s->buffers[i].sem.count) > 0) {
     if (s->buffers[i].size == 0) frags++;
     bytes += s->fragsize - s->buffers[i].size;
    }
   }
   put_user(frags, &inf->fragments);
   put_user(s->nbfrags, &inf->fragstotal);
   put_user(s->fragsize, &inf->fragsize);
   put_user(bytes, &inf->bytes);
   break;
  }

case SNDCTL_DSP_GETISPACE: //获取缓存中的数据量
  {
   audio_stream_t *s = &input_stream;
   audio_buf_info *inf = (audio_buf_info *) arg;
   int err = verify_area(VERIFY_WRITE, inf, sizeof(*inf));
   int i;
   int frags = 0, bytes = 0;

   if (!(file->f_mode & FMODE_READ))
    return -EINVAL;

   if (err)
    return err;
   for(i = 0; i < s->nbfrags; i++){
    if (atomic_read(&s->buffers[i].sem.count) > 0)
    {
     if (s->buffers[i].size == s->fragsize)
      frags++;
     bytes += s->buffers[i].size;
    }
   }
   put_user(frags, &inf->fragments);
   put_user(s->nbfrags, &inf->fragstotal);
   put_user(s->fragsize, &inf->fragsize);
   put_user(bytes, &inf->bytes);
   break;
  }
case SNDCTL_DSP_RESET: //DSP设备复位
  if (file->f_mode & FMODE_READ) {
   audio_clear_buf(&input_stream);
  }
  if (file->f_mode & FMODE_WRITE) {
   audio_clear_buf(&output_stream);
  }
  return 0;
case SNDCTL_DSP_NONBLOCK:
  file->f_flags |= O_NONBLOCK;
  return 0;
case SNDCTL_DSP_POST:
case SNDCTL_DSP_SUBDIVIDE:
case SNDCTL_DSP_GETCAPS:
case SNDCTL_DSP_GETTRIGGER:
case SNDCTL_DSP_SETTRIGGER:
case SNDCTL_DSP_GETIPTR:
case SNDCTL_DSP_GETOPTR:
case SNDCTL_DSP_MAPINBUF:
case SNDCTL_DSP_MAPOUTBUF:
case SNDCTL_DSP_SETSYNCRO:
case SNDCTL_DSP_SETDUPLEX:
  return -ENOSYS;
default:
  return smdk2410_mixer_ioctl(inode, file, cmd, arg);
}

return 0;
}

static int smdk2410_audio_open(struct inode *inode, struct file *file)
{
int cold = !audio_active;

DPRINTK("audio_open/n");
if ((file->f_flags & O_ACCMODE) == O_RDONLY) {
  if (audio_rd_refcount || audio_wr_refcount) //只读打开不能被其他进程打开
   return -EBUSY;
  audio_rd_refcount++;
} else if ((file->f_flags & O_ACCMODE) == O_WRONLY) {
  if (audio_wr_refcount) //只写打开不能被其他进程写打开
   return -EBUSY;
  audio_wr_refcount++;
} else if ((file->f_flags & O_ACCMODE) == O_RDWR) {
  if (audio_rd_refcount || audio_wr_refcount)
   return -EBUSY;
  audio_rd_refcount++;
  audio_wr_refcount++; //增加文件打开计数
} else
  return -EINVAL;

if (cold) { //如果文件没有被以任何形式打开，则初始化以下变量
  audio_rate = AUDIO_RATE_DEFAULT;
  audio_channels = AUDIO_CHANNELS_DEFAULT;
  audio_fragsize = AUDIO_FRAGSIZE_DEFAULT;
  audio_nbfrags = AUDIO_NBFRAGS_DEFAULT;
  if ((file->f_mode & FMODE_WRITE)){
   init_s3c2410_iis_bus_tx(); //将iis设为传输状态
   audio_clear_buf(&output_stream);
  }
  if ((file->f_mode & FMODE_READ)){
   init_s3c2410_iis_bus_rx(); //将iis设为接收状态
   audio_clear_buf(&input_stream);
  }
}
return 0;
}

static int smdk2410_mixer_open(struct inode *inode, struct file *file)
{
return 0;
}

static int smdk2410_audio_release(struct inode *inode, struct file *file)
{
DPRINTK("audio_release/n");

if (file->f_mode & FMODE_READ) {
  if (audio_rd_refcount == 1)
   audio_clear_buf(&input_stream);
  audio_rd_refcount = 0;
}

if(file->f_mode & FMODE_WRITE) {
  if (audio_wr_refcount == 1) {
   audio_sync(file);
   audio_clear_buf(&output_stream);
   audio_wr_refcount = 0;
  }
}

return 0;
}

static int smdk2410_mixer_release(struct inode *inode, struct file *file)
{
return 0;
}

static struct file_operations smdk2410_audio_fops = { //DSP设备的操作函数
llseek: smdk2410_audio_llseek,
write: smdk2410_audio_write,
read: smdk2410_audio_read,
poll: smdk2410_audio_poll,
ioctl: smdk2410_audio_ioctl,
open: smdk2410_audio_open,
release: smdk2410_audio_release
};

static struct file_operations smdk2410_mixer_fops = { //混频设备的操作函数
ioctl: smdk2410_mixer_ioctl,
open: smdk2410_mixer_open,
release: smdk2410_mixer_release
};

static void init_uda1341(void)
{
unsigned long flags;

uda1341_volume = 62 - ((DEF_VOLUME * 61) / 100); //初始化音量
uda1341_boost = 0;
uda_sampling = DATA2_DEEMP_NONE;
uda_sampling &= ~(DATA2_MUTE); //非静音模式

local_irq_save(flags);

s3c2410_gpio_setpin(S3C2410_GPB2,1);

s3c2410_gpio_setpin(S3C2410_GPB4,1);
local_irq_restore(flags);

uda1341_l3_address(UDA1341_REG_STATUS);
//复位，256fs不知道为什么iis设的又是384fs，右对齐模式，滤波
uda1341_l3_data(0x40 | STAT0_SC_256FS | STAT0_IF_MSB|STAT0_DC_FILTER); // reset uda1341
uda1341_l3_data(STAT1 | STAT1_ADC_ON | STAT1_DAC_ON);

uda1341_l3_address(UDA1341_REG_DATA0);
uda1341_l3_data(DATA0 |DATA0_VOLUME(0x0)); // maximum volume
uda1341_l3_data(DATA1 |DATA1_BASS(uda1341_boost)| DATA1_TREBLE(0));
uda1341_l3_data(uda_sampling); /* --;;*/
uda1341_l3_data(EXTADDR(EXT2));
uda1341_l3_data(EXTDATA(EXT2_MIC_GAIN(0x6)) | EXT2_MIXMODE_CH1);

}

static void init_s3c2410_iis_bus(void){
writel(0, iis_base + S3C2410_IISPSR);
writel(0, iis_base + S3C2410_IISCON);
writel(0, iis_base + S3C2410_IISMOD);
writel(0, iis_base + S3C2410_IISFCON);
clk_disable(iis_clock);
}
static void init_s3c2410_iis_bus_rx(void){
unsigned int iiscon, iismod, iisfcon;
char *dstr;
//Kill everything...
writel(0, iis_base + S3C2410_IISPSR);
writel(0, iis_base + S3C2410_IISCON);
writel(0, iis_base + S3C2410_IISMOD);
writel(0, iis_base + S3C2410_IISFCON);

clk_enable(iis_clock);

iiscon = iismod = iisfcon = 0;

//Setup basic stuff
iiscon |= S3C2410_IISCON_PSCEN; // Enable prescaler
iiscon |= S3C2410_IISCON_IISEN; // Enable interface

iismod |= S3C2410_IISMOD_MASTER; // Set interface to Master Mode
iismod |= S3C2410_IISMOD_LR_LLOW; // Low for left channel
iismod |= S3C2410_IISMOD_MSB; // IIS format 左对齐
iismod |= S3C2410_IISMOD_16BIT; // Serial data bit/channel is 16 bit
iismod |= S3C2410_IISMOD_384FS; // Master clock freq = 384 fs
iismod |= S3C2410_IISMOD_32FS; // 32 fs 串行时钟为32倍采样频率

iisfcon|= S3C2410_IISFCON_RXDMA; //Set RX FIFO acces mode to DMA
iisfcon|= S3C2410_IISFCON_TXDMA; //Set RX FIFO acces mode to DMA

iiscon |= S3C2410_IISCON_RXDMAEN; //Enable RX DMA service request
iiscon |= S3C2410_IISCON_TXIDLE; //Set TX channel idle
iismod |= S3C2410_IISMOD_RXMODE; //Set RX Mode
iisfcon|= S3C2410_IISFCON_RXENABLE; //Enable RX Fifo
dstr="RX";
//setup the prescaler
audio_set_dsp_speed(audio_rate);

//iiscon has to be set last - it enables the interface
writel(iismod, iis_base + S3C2410_IISMOD);
writel(iisfcon, iis_base + S3C2410_IISFCON);
writel(iiscon, iis_base + S3C2410_IISCON);
}

static void init_s3c2410_iis_bus_tx(void)
{
unsigned int iiscon, iismod, iisfcon;
char *dstr;

//Kill everything...
writel(0, iis_base + S3C2410_IISPSR);
writel(0, iis_base + S3C2410_IISCON);
writel(0, iis_base + S3C2410_IISMOD);
writel(0, iis_base + S3C2410_IISFCON);

clk_enable(iis_clock);

iiscon = iismod = iisfcon = 0;

//Setup basic stuff
iiscon |= S3C2410_IISCON_PSCEN; // Enable prescaler
iiscon |= S3C2410_IISCON_IISEN; // Enable interface

iismod |= S3C2410_IISMOD_MASTER; // Set interface to Master Mode
iismod |= S3C2410_IISMOD_LR_LLOW; // Low for left channel
iismod |= S3C2410_IISMOD_MSB; // MSB format 右对齐
iismod |= S3C2410_IISMOD_16BIT; // Serial data bit/channel is 16 bit
iismod |= S3C2410_IISMOD_256FS; // Master clock freq = 384 fs 费解
iismod |= S3C2410_IISMOD_32FS; // 32 fs //串行时钟为32倍采样频率

iisfcon|= S3C2410_IISFCON_RXDMA; //Set RX FIFO acces mode to DMA
iisfcon|= S3C2410_IISFCON_TXDMA; //Set TX FIFO acces mode to DMA

iiscon |= S3C2410_IISCON_TXDMAEN; //Enable TX DMA service request
iiscon |= S3C2410_IISCON_RXIDLE; //Set RX channel idle
iismod |= S3C2410_IISMOD_TXMODE; //Set TX Mode
iisfcon|= S3C2410_IISFCON_TXENABLE; //Enable TX Fifo
dstr="TX";

//setup the prescaler
audio_set_dsp_speed(audio_rate);

//iiscon has to be set last - it enables the interface
writel(iismod, iis_base + S3C2410_IISMOD);
writel(iisfcon, iis_base + S3C2410_IISFCON);
writel(iiscon, iis_base + S3C2410_IISCON);
}

static int __init audio_init_dma(audio_stream_t * s, char *desc)
{
int ret ;
s3c2410_dmasrc_t source;
int hwcfg;
unsigned long devaddr;
dmach_t channel;
int dcon;
unsigned int flags = 0;

if(s->dma_ch == DMA_CH2){
  channel = 2;
  source = S3C2410_DMASRC_MEM; //源数据为内存即为数据输出
  hwcfg = 3;
  devaddr = 0x55000010; //iis 的fifo寄存器地址此处必须是物理地址
  dcon = 1<<31;
   //该标志的设置会导致函数s3c2410_dma_enqueue调用即开启DMA操作
  flags = S3C2410_DMAF_AUTOSTART;
  //获取DMA通道申请DMA中断
  ret = s3c2410_dma_request(s->dma_ch, &s3c2410iis_dma_out, NULL);
  s3c2410_dma_devconfig(channel, source, hwcfg, devaddr); //配置初始化源，目的寄存器
  /*
  配置DMA的控制寄存器将配置的值放于chan->dcon = dcon;
     chan->xfer_unit = xferunit;放置数据传输大小单位
  */
  s3c2410_dma_config(channel, 2, dcon); //
  //设置回调函数chan->callback_fn = rtn;在DMA中断中调用
  s3c2410_dma_set_buffdone_fn(channel, audio_dmaout_done_callback);
  //设置chan->flags = flags。flags = S3C2410_DMAF_AUTOSTART;
  s3c2410_dma_setflags(channel, flags);
  s->dma_ok = 1;
  return ret;
}
else if(s->dma_ch == DMA_CH1){//我认为如以下设置不能使DMA工作起来
  ret = s3c2410_dma_request(s->dma_ch, &s3c2410iis_dma_in, NULL);
  s3c2410_dma_set_buffdone_fn(s->dma_ch, audio_dmain_done_callback);
  return ret ;
}
else
  return 1;
}

static int audio_clear_dma(audio_stream_t * s,s3c2410_dma_client_t *client)
{
s3c2410_dma_set_buffdone_fn(s->dma_ch, NULL);
s3c2410_dma_free(s->dma_ch, client);
return 0;
}

static int s3c2410iis_probe(struct device *dev) {
struct platform_device *pdev = to_platform_device(dev);
struct resource *res;

unsigned long flags;
res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
if (res == NULL) {
  printk(KERN_INFO PFX "failed to get memory region resouce/n");
  return -ENOENT;
}
/*
S3C2410_VA_IIS好像没定义我认为该段程序应该如下写
res = request_mem_region(res->start,
      res->end-res->start+1,
      pdev->name);

iis_base = ioremap(res->start, res->end - res->start + 1);
*/
iis_base = (void *)S3C2410_VA_IIS ; //?????????
if (iis_base == 0) {
  printk(KERN_INFO PFX "failed to ioremap() region/n");
  return -EINVAL;
}

iis_clock = clk_get(dev, "iis");
if (iis_clock == NULL) {
  printk(KERN_INFO PFX "failed to find clock source/n");
  return -ENOENT;
}
//在linux-2.6.30.4中找不到该函数
clk_use(iis_clock); ?????????

local_irq_save(flags);

/* GPB 4: L3CLOCK, OUTPUT */
s3c2410_gpio_cfgpin(S3C2410_GPB4, S3C2410_GPB4_OUTP);
/* GPB 3: L3DATA, OUTPUT */
s3c2410_gpio_cfgpin(S3C2410_GPB3,S3C2410_GPB3_OUTP);
/* GPB 2: L3MODE, OUTPUT */
s3c2410_gpio_cfgpin(S3C2410_GPB2,S3C2410_GPB2_OUTP);
/* GPE 3: I2SSDI */
s3c2410_gpio_cfgpin(S3C2410_GPE3,S3C2410_GPE3_I2SSDI);
/* GPE 0: I2SLRCK */
s3c2410_gpio_cfgpin(S3C2410_GPE0,S3C2410_GPE0_I2SLRCK);
/* GPE 1: I2SSCLK */
s3c2410_gpio_cfgpin(S3C2410_GPE1,S3C2410_GPE1_I2SSCLK);
/* GPE 2: CDCLK */
s3c2410_gpio_cfgpin(S3C2410_GPE2,S3C2410_GPE2_CDCLK);
/* GPE 4: I2SSDO */
s3c2410_gpio_cfgpin(S3C2410_GPE4,S3C2410_GPE4_I2SSDO);

local_irq_restore(flags);

init_s3c2410_iis_bus();

init_uda1341();

output_stream.dma_ch = DMA_CH2; //初始化输出缓存使用的DMA通道

if (audio_init_dma(&output_stream, "UDA1341 out")) {
  audio_clear_dma(&output_stream,&s3c2410iis_dma_out);
  printk( KERN_WARNING AUDIO_NAME_VERBOSE
   ": unable to get DMA channels/n" );
  return -EBUSY;
}

input_stream.dma_ch = DMA_CH1; //初始化输出缓存使用的DMA通道

if (audio_init_dma(&input_stream, "UDA1341 in")) {
  audio_clear_dma(&input_stream,&s3c2410iis_dma_in);
  printk( KERN_WARNING AUDIO_NAME_VERBOSE
   ": unable to get DMA channels/n" );
  return -EBUSY;
}
//DSP设备和混频设备注册返回其次设备号
audio_dev_dsp = register_sound_dsp(&smdk2410_audio_fops, -1);
audio_dev_mixer = register_sound_mixer(&smdk2410_mixer_fops, -1);

printk(AUDIO_NAME_VERBOSE " initialized/n");

return 0;
}

static int s3c2410iis_remove(struct device *dev) {
unregister_sound_dsp(audio_dev_dsp);
unregister_sound_mixer(audio_dev_mixer);
audio_clear_dma(&output_stream,&s3c2410iis_dma_out);
audio_clear_dma(&input_stream,&s3c2410iis_dma_in); /* input */
printk(AUDIO_NAME_VERBOSE " unloaded/n");
return 0;
}

//总线platform_bus_type表明该驱动为平台设备驱动
static struct device_driver s3c2410iis_driver = {
.name = "s3c2410-iis",
  .bus = &platform_bus_type,
  .probe = s3c2410iis_probe,
  .remove = s3c2410iis_remove,
};

static int __init s3c2410_uda1341_init(void) {
memzero(&input_stream, sizeof(audio_stream_t));
memzero(&output_stream, sizeof(audio_stream_t));
return driver_register(&s3c2410iis_driver);
}

static void __exit s3c2410_uda1341_exit(void) {
driver_unregister(&s3c2410iis_driver);
}

module_init(s3c2410_uda1341_init);
module_exit(s3c2410_uda1341_exit);

MODULE_LICENSE("GPL");
MODULE_AUTHOR("superlp<xinshengtaier@eyou.com >");
MODULE_DESCRIPTION("S3C2410 uda1341 sound driver");