Linux 内核 ASoC 框架，在概念上将嵌入式音频系统拆分为多个可复用的组件驱动程序，包括 Codec 类驱动程序、平台类驱动程序和机器类驱动程序。在实现上，机器类驱动程序用 struct snd_soc_card 和 struct snd_soc_dai_link 结构描述，属于平台类驱动程序的 DMA 引擎驱动程序由 struct snd_soc_component_driver 结构描述，codec 类驱动程序和 I2S 等驱动程序，由 struct snd_soc_component_driver、struct snd_soc_dai_driver 和 struct snd_soc_dai_ops 等结构描述。除平台类驱动程序外的各种驱动程序都通过 component 抽象组织在一起，即这些驱动程序都作为 struct snd_soc_component_driver 注册给 Linux 内核 ASoC 框架，Linux 内核 ASoC 框架为它们各自创建 struct snd_soc_component 结构对象，并保存在 sound/soc/soc-core.c 文件中定义的全局链表 component_list 中。
一个 DMA 驱动程序的示例为 soc/pxa/pxa2xx-pcm.c：
static const struct snd_soc_component_driver pxa2xx_soc_platform = {
.pcm_construct = pxa2xx_soc_pcm_new,
.pcm_destruct = pxa2xx_soc_pcm_free,
.open = pxa2xx_soc_pcm_open,
.close = pxa2xx_soc_pcm_close,
.hw_params = pxa2xx_soc_pcm_hw_params,
.hw_free = pxa2xx_soc_pcm_hw_free,
.prepare = pxa2xx_soc_pcm_prepare,
.trigger = pxa2xx_soc_pcm_trigger,
.pointer = pxa2xx_soc_pcm_pointer,
.mmap = pxa2xx_soc_pcm_mmap,
};
static int pxa2xx_soc_platform_probe(struct platform_device *pdev)
{
return devm_snd_soc_register_component(&pdev->dev, &pxa2xx_soc_platform,
NULL, 0);
}
static struct platform_driver pxa_pcm_driver = {
.driver = {
.name = "pxa-pcm-audio",
},
.probe = pxa2xx_soc_platform_probe,
};
module_platform_driver(pxa_pcm_driver);
DMA 引擎驱动程序和 I2S 或 Codec 驱动程序一样，通过 devm_snd_soc_register_component() 函数以 struct snd_soc_component_driver 的形式注册给 Linux 内核 ASoC 框架，但比较特别的地方在于，它的 dai driver 参数为空。
机器类驱动程序定义的 struct snd_soc_dai_link 结构对象通过 struct snd_soc_dai_link_component 描述它引用的其它类型的驱动程序，如机器类驱动程序 sound/soc/pxa/e800_wm9712.c 有如下的代码片段：
SND_SOC_DAILINK_DEFS(ac97,
DAILINK_COMP_ARRAY(COMP_CPU("pxa2xx-ac97")),
DAILINK_COMP_ARRAY(COMP_CODEC("wm9712-codec", "wm9712-hifi")),
DAILINK_COMP_ARRAY(COMP_PLATFORM("pxa-pcm-audio")));
SND_SOC_DAILINK_DEFS(ac97_aux,
DAILINK_COMP_ARRAY(COMP_CPU("pxa2xx-ac97-aux")),
DAILINK_COMP_ARRAY(COMP_CODEC("wm9712-codec", "wm9712-aux")),
DAILINK_COMP_ARRAY(COMP_PLATFORM("pxa-pcm-audio")));
static struct snd_soc_dai_link e800_dai[] = {
{
.name = "AC97",
.stream_name = "AC97 HiFi",
SND_SOC_DAILINK_REG(ac97),
},
{
.name = "AC97 Aux",
.stream_name = "AC97 Aux",
SND_SOC_DAILINK_REG(ac97_aux),
},
};
static struct snd_soc_card e800 = {
.name = "Toshiba e800",
.owner = THIS_MODULE,
.dai_link = e800_dai,
.num_links = ARRAY_SIZE(e800_dai),
.dapm_widgets = e800_dapm_widgets,
.num_dapm_widgets = ARRAY_SIZE(e800_dapm_widgets),
.dapm_routes = audio_map,
.num_dapm_routes = ARRAY_SIZE(audio_map),
};
SND_SOC_DAILINK_DEFS() 宏用于为 struct snd_soc_dai_link 方便地定义引用的 cpus、codecs 和 platforms 等 struct snd_soc_dai_link_component 数组，其中用于定义 platforms 的 DAILINK_COMP_ARRAY(COMP_PLATFORM("pxa-pcm-audio")) 引用了上面我们看到的 DMA 引擎驱动。
Linux 内核 ASoC 框架提供了一个通用的 DMA 引擎驱动程序，位于文件 sound/soc/soc-generic-dmaengine-pcm 中。这个驱动程序本身不会主动向 Linux 内核 ASoC 框架注册自己，需要使用 DMA 引擎在设备和内存之间传数据的驱动程序要在 probe 时注册它，如 sound/soc/rockchip/rockchip_pcm.c：
static const struct snd_pcm_hardware snd_rockchip_hardware = {
.info = SNDRV_PCM_INFO_MMAP |
SNDRV_PCM_INFO_MMAP_VALID |
SNDRV_PCM_INFO_PAUSE |
SNDRV_PCM_INFO_RESUME |
SNDRV_PCM_INFO_INTERLEAVED,
.period_bytes_min = 32,
.period_bytes_max = 8192,
.periods_min = 1,
.periods_max = 52,
.buffer_bytes_max = 64 * 1024,
.fifo_size = 32,
};
static const struct snd_dmaengine_pcm_config rk_dmaengine_pcm_config = {
.pcm_hardware = &snd_rockchip_hardware,
.prepare_slave_config = snd_dmaengine_pcm_prepare_slave_config,
.prealloc_buffer_size = 32 * 1024,
};
int rockchip_pcm_platform_register(struct device *dev)
{
return devm_snd_dmaengine_pcm_register(dev, &rk_dmaengine_pcm_config,
SND_DMAENGINE_PCM_FLAG_COMPAT);
}
EXPORT_SYMBOL_GPL(rockchip_pcm_platform_register);
这里的 rockchip_pcm_platform_register() 函数在 I2S 驱动程序的 probe 操作 (位于 sound/soc/rockchip/rockchip_i2s.c) 中调用：
static int rockchip_i2s_probe(struct platform_device *pdev)
{
. . . . . .
ret = devm_snd_soc_register_component(&pdev->dev,
&rockchip_i2s_component,
soc_dai, 1);
if (ret) {
dev_err(&pdev->dev, "Could not register DAI\n");
goto err_suspend;
}
ret = rockchip_pcm_platform_register(&pdev->dev);
if (ret) {
dev_err(&pdev->dev, "Could not register PCM\n");
goto err_suspend;
}
return 0;
. . . . . .
return ret;
}
rockchip_pcm_platform_register() 函数调用 devm_snd_dmaengine_pcm_register() 函数注册通用 DMA 引擎驱动程序。devm_snd_dmaengine_pcm_register() 函数定义 (位于 sound/soc/soc-devres.c) 如下：
static void devm_dmaengine_pcm_release(struct device *dev, void *res)
{
snd_dmaengine_pcm_unregister(*(struct device **)res);
}
/**
* devm_snd_dmaengine_pcm_register - resource managed dmaengine PCM registration
* @dev: The parent device for the PCM device
* @config: Platform specific PCM configuration
* @flags: Platform specific quirks
*
* Register a dmaengine based PCM device with automatic unregistration when the
* device is unregistered.
*/
int devm_snd_dmaengine_pcm_register(struct device *dev,
const struct snd_dmaengine_pcm_config *config, unsigned int flags)
{
struct device **ptr;
int ret;
ptr = devres_alloc(devm_dmaengine_pcm_release, sizeof(*ptr), GFP_KERNEL);
if (!ptr)
return -ENOMEM;
ret = snd_dmaengine_pcm_register(dev, config, flags);
if (ret == 0) {
*ptr = dev;
devres_add(dev, ptr);
} else {
devres_free(ptr);
}
return ret;
}
EXPORT_SYMBOL_GPL(devm_snd_dmaengine_pcm_register);
#endif
devm_snd_dmaengine_pcm_register() 函数和之前看到的devm_snd_soc_register_card() 与 devm_snd_soc_register_component() 函数一样，只是它封装的是 snd_dmaengine_pcm_register() 函数。snd_dmaengine_pcm_register() 函数定义 (位于 sound/soc/soc-generic-dmaengine-pcm.c) 如下：
static const struct snd_soc_component_driver dmaengine_pcm_component = {
.name = SND_DMAENGINE_PCM_DRV_NAME,
.probe_order = SND_SOC_COMP_ORDER_LATE,
.open = dmaengine_pcm_open,
.close = dmaengine_pcm_close,
.hw_params = dmaengine_pcm_hw_params,
.trigger = dmaengine_pcm_trigger,
.pointer = dmaengine_pcm_pointer,
.pcm_construct = dmaengine_pcm_new,
};
static const struct snd_soc_component_driver dmaengine_pcm_component_process = {
.name = SND_DMAENGINE_PCM_DRV_NAME,
.probe_order = SND_SOC_COMP_ORDER_LATE,
.open = dmaengine_pcm_open,
.close = dmaengine_pcm_close,
.hw_params = dmaengine_pcm_hw_params,
.trigger = dmaengine_pcm_trigger,
.pointer = dmaengine_pcm_pointer,
.copy_user = dmaengine_copy_user,
.pcm_construct = dmaengine_pcm_new,
};
static const char * const dmaengine_pcm_dma_channel_names[] = {
[SNDRV_PCM_STREAM_PLAYBACK] = "tx",
[SNDRV_PCM_STREAM_CAPTURE] = "rx",
};
static int dmaengine_pcm_request_chan_of(struct dmaengine_pcm *pcm,
struct device *dev, const struct snd_dmaengine_pcm_config *config)
{
unsigned int i;
const char *name;
struct dma_chan *chan;
if ((pcm->flags & SND_DMAENGINE_PCM_FLAG_NO_DT) || (!dev->of_node &&
!(config && config->dma_dev && config->dma_dev->of_node)))
return 0;
if (config && config->dma_dev) {
/*
* If this warning is seen, it probably means that your Linux
* device structure does not match your HW device structure.
* It would be best to refactor the Linux device structure to
* correctly match the HW structure.
*/
dev_warn(dev, "DMA channels sourced from device %s",
dev_name(config->dma_dev));
dev = config->dma_dev;
}
for_each_pcm_streams(i) {
if (pcm->flags & SND_DMAENGINE_PCM_FLAG_HALF_DUPLEX)
name = "rx-tx";
else
name = dmaengine_pcm_dma_channel_names[i];
if (config && config->chan_names[i])
name = config->chan_names[i];
chan = dma_request_chan(dev, name);
if (IS_ERR(chan)) {
/*
* Only report probe deferral errors, channels
* might not be present for devices that
* support only TX or only RX.
*/
if (PTR_ERR(chan) == -EPROBE_DEFER)
return -EPROBE_DEFER;
pcm->chan[i] = NULL;
} else {
pcm->chan[i] = chan;
}
if (pcm->flags & SND_DMAENGINE_PCM_FLAG_HALF_DUPLEX)
break;
}
if (pcm->flags & SND_DMAENGINE_PCM_FLAG_HALF_DUPLEX)
pcm->chan[1] = pcm->chan[0];
return 0;
}
static void dmaengine_pcm_release_chan(struct dmaengine_pcm *pcm)
{
unsigned int i;
for_each_pcm_streams(i) {
if (!pcm->chan[i])
continue;
dma_release_channel(pcm->chan[i]);
if (pcm->flags & SND_DMAENGINE_PCM_FLAG_HALF_DUPLEX)
break;
}
}
/**
* snd_dmaengine_pcm_register - Register a dmaengine based PCM device
* @dev: The parent device for the PCM device
* @config: Platform specific PCM configuration
* @flags: Platform specific quirks
*/
int snd_dmaengine_pcm_register(struct device *dev,
const struct snd_dmaengine_pcm_config *config, unsigned int flags)
{
const struct snd_soc_component_driver *driver;
struct dmaengine_pcm *pcm;
int ret;
pcm = kzalloc(sizeof(*pcm), GFP_KERNEL);
if (!pcm)
return -ENOMEM;
#ifdef CONFIG_DEBUG_FS
pcm->component.debugfs_prefix = "dma";
#endif
pcm->config = config;
pcm->flags = flags;
ret = dmaengine_pcm_request_chan_of(pcm, dev, config);
if (ret)
goto err_free_dma;
if (config && config->process)
driver = &dmaengine_pcm_component_process;
else
driver = &dmaengine_pcm_component;
ret = snd_soc_component_initialize(&pcm->component, driver, dev);
if (ret)
goto err_free_dma;
ret = snd_soc_add_component(&pcm->component, NULL, 0);
if (ret)
goto err_free_dma;
return 0;
err_free_dma:
dmaengine_pcm_release_chan(pcm);
kfree(pcm);
return ret;
}
EXPORT_SYMBOL_GPL(snd_dmaengine_pcm_register);
/**
* snd_dmaengine_pcm_unregister - Removes a dmaengine based PCM device
* @dev: Parent device the PCM was register with
*
* Removes a dmaengine based PCM device previously registered with
* snd_dmaengine_pcm_register.
*/
void snd_dmaengine_pcm_unregister(struct device *dev)
{
struct snd_soc_component *component;
struct dmaengine_pcm *pcm;
component = snd_soc_lookup_component(dev, SND_DMAENGINE_PCM_DRV_NAME);
if (!component)
return;
pcm = soc_component_to_pcm(component);
snd_soc_unregister_component_by_driver(dev, component->driver);
dmaengine_pcm_release_chan(pcm);
kfree(pcm);
}
EXPORT_SYMBOL_GPL(snd_dmaengine_pcm_unregister);
MODULE_LICENSE("GPL");
snd_dmaengine_pcm_register() 函数的执行过程如下：
- 动态分配一个
struct dmaengine_pcm结构对象，并初始化其config和flags等字段，struct dmaengine_pcm结构定义 (位于include/sound/dmaengine_pcm.h) 如下：
struct dmaengine_pcm {
struct dma_chan *chan[SNDRV_PCM_STREAM_LAST + 1];
const struct snd_dmaengine_pcm_config *config;
struct snd_soc_component component;
unsigned int flags;
};
这个结构有一个 struct snd_soc_component 结构成员；
- 为数据的发送和接收申请 DMA 通道，这个需要在设备树的设备节点定义中，指定发送和接收引用的 DMA 通道，像下面这样：
i2s0_8ch: i2s@fe470000 {
. . . . . .
dmas = <&dmac0 0>, <&dmac0 1>;
dma-names = "tx", "rx";
. . . . . .
};
-
根据传入的
config参数，选择struct snd_soc_component_driver，dmaengine_pcm_component_process和dmaengine_pcm_component仅有的区别是，前者多定义了一个copy_user操作； -
初始化并添加
struct snd_soc_component结构对象，有一个函数 (位于include/sound/dmaengine_pcm.h) 可以通过struct snd_soc_component结构对象获得struct dmaengine_pcm结构对象：
static inline struct dmaengine_pcm *soc_component_to_pcm(struct snd_soc_component *p)
{
return container_of(p, struct dmaengine_pcm, component);
}
通用 DMA 引擎驱动程序支持的操作由 struct snd_soc_component_driver 定义。在全局 component 链表中，相应的 struct snd_soc_component 由注册它的 dev 标识。
使用了通用 DMA 引擎驱动程序的 ASoC 机器驱动程序，示例 (位于 sound/soc/rockchip/rockchip_rt5645.c) 如下：
SND_SOC_DAILINK_DEFS(pcm,
DAILINK_COMP_ARRAY(COMP_EMPTY()),
DAILINK_COMP_ARRAY(COMP_CODEC(NULL, "rt5645-aif1")),
DAILINK_COMP_ARRAY(COMP_EMPTY()));
static struct snd_soc_dai_link rk_dailink = {
.name = "rt5645",
.stream_name = "rt5645 PCM",
.init = rk_init,
.ops = &rk_aif1_ops,
/* set rt5645 as slave */
.dai_fmt = SND_SOC_DAIFMT_I2S | SND_SOC_DAIFMT_NB_NF |
SND_SOC_DAIFMT_CBS_CFS,
SND_SOC_DAILINK_REG(pcm),
};
static struct snd_soc_card snd_soc_card_rk = {
.name = "I2S-RT5650",
.owner = THIS_MODULE,
.dai_link = &rk_dailink,
.num_links = 1,
.dapm_widgets = rk_dapm_widgets,
.num_dapm_widgets = ARRAY_SIZE(rk_dapm_widgets),
.dapm_routes = rk_audio_map,
.num_dapm_routes = ARRAY_SIZE(rk_audio_map),
.controls = rk_mc_controls,
.num_controls = ARRAY_SIZE(rk_mc_controls),
};
. . . . . .
static int snd_rk_mc_probe(struct platform_device *pdev)
{
. . . . . .
rk_dailink.cpus->of_node = of_parse_phandle(np,
"rockchip,i2s-controller", 0);
if (!rk_dailink.cpus->of_node) {
dev_err(&pdev->dev,
"Property 'rockchip,i2s-controller' missing or invalid\n");
ret = -EINVAL;
goto put_codec_of_node;
}
rk_dailink.platforms->of_node = rk_dailink.cpus->of_node;
. . . . . .
}
这里为 dai link 定义的 cpus dai 数组和 platforms 数组中都只有一个空元素，但在 probe 操作中，根据设备树中设备节点的定义，查找了对应的 of_node。尽管 cpus dai 和 platforms 引用了相同的 of_node，但在 snd_soc_add_pcm_runtime() 函数中，为 CPU DAI 添加 struct snd_soc_component 的过程是，先查找对应的 struct snd_soc_dai，再从 struct snd_soc_dai 获得 struct snd_soc_component，这也就意味着，为 CPU DAI 查找 struct snd_soc_component 时，不会找到没有 struct snd_soc_dai 的通用 DMA 引擎的 struct snd_soc_component。为 platforms 添加 struct snd_soc_component 的过程，则是直接查找所有匹配的 struct snd_soc_component 并添加。
在为 PCM 创建 struct snd_soc_pcm_runtime 时，即在机器类驱动程序的 probe 操作中，如果为 snd_soc_dai_link 做了类似于这里的处理，使 CPU DAI 和 platform 指向相同的 of_node，且对应于 of_node 的设备驱动程序注册了通用 DMA 引擎驱动程序，通用 DMA 引擎的 struct snd_soc_component 将会包含在它的 component 链表中。
通用 DMA 引擎驱动程序的 struct snd_soc_component_driver 还有一个特别的地方是，指定了 probe_order 为 SND_SOC_COMP_ORDER_LATE，这使得它的 probe 和 init 操作，相对而言，执行的更晚，同时它在 snd_soc_card 中的位置也更靠后一点。如在 soc_probe_link_components() 函数 (位于 sound/soc/soc-core.c) 中：
static int soc_probe_component(struct snd_soc_card *card,
struct snd_soc_component *component)
{
struct snd_soc_dapm_context *dapm =
snd_soc_component_get_dapm(component);
struct snd_soc_dai *dai;
int probed = 0;
int ret;
if (!strcmp(component->name, "snd-soc-dummy"))
return 0;
if (component->card) {
if (component->card != card) {
dev_err(component->dev,
"Trying to bind component to card \"%s\" but is already bound to card \"%s\"\n",
card->name, component->card->name);
return -ENODEV;
}
return 0;
}
ret = snd_soc_component_module_get_when_probe(component);
if (ret < 0)
return ret;
component->card = card;
soc_set_name_prefix(card, component);
soc_init_component_debugfs(component);
snd_soc_dapm_init(dapm, card, component);
ret = snd_soc_dapm_new_controls(dapm,
component->driver->dapm_widgets,
component->driver->num_dapm_widgets);
if (ret != 0) {
dev_err(component->dev,
"Failed to create new controls %d\n", ret);
goto err_probe;
}
for_each_component_dais(component, dai) {
ret = snd_soc_dapm_new_dai_widgets(dapm, dai);
if (ret != 0) {
dev_err(component->dev,
"Failed to create DAI widgets %d\n", ret);
goto err_probe;
}
}
ret = snd_soc_component_probe(component);
if (ret < 0) {
dev_err(component->dev,
"ASoC: failed to probe component %d\n", ret);
goto err_probe;
}
WARN(dapm->idle_bias_off &&
dapm->bias_level != SND_SOC_BIAS_OFF,
"codec %s can not start from non-off bias with idle_bias_off==1\n",
component->name);
probed = 1;
/*
* machine specific init
* see
* snd_soc_component_set_aux()
*/
ret = snd_soc_component_init(component);
if (ret < 0)
goto err_probe;
ret = snd_soc_add_component_controls(component,
component->driver->controls,
component->driver->num_controls);
if (ret < 0)
goto err_probe;
ret = snd_soc_dapm_add_routes(dapm,
component->driver->dapm_routes,
component->driver->num_dapm_routes);
if (ret < 0) {
if (card->disable_route_checks) {
dev_info(card->dev,
"%s: disable_route_checks set, ignoring errors on add_routes\n",
__func__);
} else {
dev_err(card->dev,
"%s: snd_soc_dapm_add_routes failed: %d\n",
__func__, ret);
goto err_probe;
}
}
/* see for_each_card_components */
list_add(&component->card_list, &card->component_dev_list);
err_probe:
if (ret < 0)
soc_remove_component(component, probed);
return ret;
}
. . . . . .
static int soc_probe_link_components(struct snd_soc_card *card)
{
struct snd_soc_component *component;
struct snd_soc_pcm_runtime *rtd;
int i, ret, order;
for_each_comp_order(order) {
for_each_card_rtds(card, rtd) {
for_each_rtd_components(rtd, i, component) {
if (component->driver->probe_order != order)
continue;
ret = soc_probe_component(card, component);
if (ret < 0)
return ret;
}
}
}
return 0;
}
这里用到的 for_each_comp_order() 宏定义 (位于 include/sound/soc-component.h) 如下：
#define SND_SOC_COMP_ORDER_FIRST -2
#define SND_SOC_COMP_ORDER_EARLY -1
#define SND_SOC_COMP_ORDER_NORMAL 0
#define SND_SOC_COMP_ORDER_LATE 1
#define SND_SOC_COMP_ORDER_LAST 2
#define for_each_comp_order(order) \
for (order = SND_SOC_COMP_ORDER_FIRST; \
order <= SND_SOC_COMP_ORDER_LAST; \
order++)
Linux 内核 ASoC 通用 DMA 引擎驱动程序的操作
通用 DMA 引擎驱动程序提供的操作如下：
static const struct snd_soc_component_driver dmaengine_pcm_component = {
.name = SND_DMAENGINE_PCM_DRV_NAME,
.probe_order = SND_SOC_COMP_ORDER_LATE,
.open = dmaengine_pcm_open,
.close = dmaengine_pcm_close,
.hw_params = dmaengine_pcm_hw_params,
.trigger = dmaengine_pcm_trigger,
.pointer = dmaengine_pcm_pointer,
.pcm_construct = dmaengine_pcm_new,
};
static const struct snd_soc_component_driver dmaengine_pcm_component_process = {
.name = SND_DMAENGINE_PCM_DRV_NAME,
.probe_order = SND_SOC_COMP_ORDER_LATE,
.open = dmaengine_pcm_open,
.close = dmaengine_pcm_close,
.hw_params = dmaengine_pcm_hw_params,
.trigger = dmaengine_pcm_trigger,
.pointer = dmaengine_pcm_pointer,
.copy_user = dmaengine_copy_user,
.pcm_construct = dmaengine_pcm_new,
};
这些操作中，最早被调用的是 pcm_construct 操作，也就是 dmaengine_pcm_new() 函数。在机器类驱动程序中，调用 devm_snd_soc_register_card() 函数注册声卡，在这个函数中有如下调用过程，snd_soc_register_card() -> snd_soc_bind_card() -> soc_init_pcm_runtime() -> soc_new_pcm() -> snd_soc_pcm_component_new()，snd_soc_pcm_component_new() 函数执行各个 struct snd_soc_component_driver 的 pcm_construct 操作。snd_soc_pcm_component_new() 函数定义 (位于 sound/soc/soc-component.c) 如下：
int snd_soc_pcm_component_new(struct snd_soc_pcm_runtime *rtd)
{
struct snd_soc_component *component;
int ret;
int i;
for_each_rtd_components(rtd, i, component) {
if (component->driver->pcm_construct) {
ret = component->driver->pcm_construct(component, rtd);
if (ret < 0)
return soc_component_ret(component, ret);
}
}
return 0;
}
回到通用 DMA 引擎驱动程序的 pcm_construct 操作 dmaengine_pcm_new() 函数，这个函数定义 (位于 sound/soc/soc-generic-dmaengine-pcm.c) 如下：
static int dmaengine_pcm_new(struct snd_soc_component *component,
struct snd_soc_pcm_runtime *rtd)
{
struct dmaengine_pcm *pcm = soc_component_to_pcm(component);
const struct snd_dmaengine_pcm_config *config = pcm->config;
struct device *dev = component->dev;
struct snd_pcm_substream *substream;
size_t prealloc_buffer_size;
size_t max_buffer_size;
unsigned int i;
if (config && config->prealloc_buffer_size) {
prealloc_buffer_size = config->prealloc_buffer_size;
max_buffer_size = config->pcm_hardware->buffer_bytes_max;
} else {
prealloc_buffer_size = 512 * 1024;
max_buffer_size = SIZE_MAX;
}
for_each_pcm_streams(i) {
substream = rtd->pcm->streams[i].substream;
if (!substream)
continue;
if (!pcm->chan[i] && config && config->chan_names[i])
pcm->chan[i] = dma_request_slave_channel(dev,
config->chan_names[i]);
if (!pcm->chan[i] && (pcm->flags & SND_DMAENGINE_PCM_FLAG_COMPAT)) {
pcm->chan[i] = dmaengine_pcm_compat_request_channel(
component, rtd, substream);
}
if (!pcm->chan[i]) {
dev_err(component->dev,
"Missing dma channel for stream: %d\n", i);
return -EINVAL;
}
snd_pcm_set_managed_buffer(substream,
SNDRV_DMA_TYPE_DEV_IRAM,
dmaengine_dma_dev(pcm, substream),
prealloc_buffer_size,
max_buffer_size);
if (!dmaengine_pcm_can_report_residue(dev, pcm->chan[i]))
pcm->flags |= SND_DMAENGINE_PCM_FLAG_NO_RESIDUE;
if (rtd->pcm->streams[i].pcm->name[0] == '\0') {
strscpy_pad(rtd->pcm->streams[i].pcm->name,
rtd->pcm->streams[i].pcm->id,
sizeof(rtd->pcm->streams[i].pcm->name));
}
}
return 0;
}
这个函数分别为播放和录制申请 DMA 通道，并分配 DMA buffer 用于用户空间应用程序、内核 ALSA/ASoc 框架和硬件设备之间的数据交换。snd_pcm_set_managed_buffer() 函数分配 DMA 缓冲区。snd_pcm_set_managed_buffer() 函数定义 (位于 sound/core/pcm_memory.c) 如下：
static void preallocate_pages(struct snd_pcm_substream *substream,
int type, struct device *data,
size_t size, size_t max, bool managed)
{
if (snd_BUG_ON(substream->dma_buffer.dev.type))
return;
substream->dma_buffer.dev.type = type;
substream->dma_buffer.dev.dev = data;
if (size > 0 && preallocate_dma && substream->number < maximum_substreams)
preallocate_pcm_pages(substream, size);
if (substream->dma_buffer.bytes > 0)
substream->buffer_bytes_max = substream->dma_buffer.bytes;
substream->dma_max = max;
if (max > 0)
preallocate_info_init(substream);
if (managed)
substream->managed_buffer_alloc = 1;
}
. . . . . .
void snd_pcm_set_managed_buffer(struct snd_pcm_substream *substream, int type,
struct device *data, size_t size, size_t max)
{
preallocate_pages(substream, type, data, size, max, true);
}
EXPORT_SYMBOL(snd_pcm_set_managed_buffer);
open 操作 dmaengine_pcm_open() 函数完成数据传输之前的准备工作，这个函数定义如下：
static int
dmaengine_pcm_set_runtime_hwparams(struct snd_soc_component *component,
struct snd_pcm_substream *substream)
{
struct snd_soc_pcm_runtime *rtd = asoc_substream_to_rtd(substream);
struct dmaengine_pcm *pcm = soc_component_to_pcm(component);
struct device *dma_dev = dmaengine_dma_dev(pcm, substream);
struct dma_chan *chan = pcm->chan[substream->stream];
struct snd_dmaengine_dai_dma_data *dma_data;
struct snd_pcm_hardware hw;
if (rtd->num_cpus > 1) {
dev_err(rtd->dev,
"%s doesn't support Multi CPU yet\n", __func__);
return -EINVAL;
}
if (pcm->config && pcm->config->pcm_hardware)
return snd_soc_set_runtime_hwparams(substream,
pcm->config->pcm_hardware);
dma_data = snd_soc_dai_get_dma_data(asoc_rtd_to_cpu(rtd, 0), substream);
memset(&hw, 0, sizeof(hw));
hw.info = SNDRV_PCM_INFO_MMAP | SNDRV_PCM_INFO_MMAP_VALID |
SNDRV_PCM_INFO_INTERLEAVED;
hw.periods_min = 2;
hw.periods_max = UINT_MAX;
hw.period_bytes_min = 256;
hw.period_bytes_max = dma_get_max_seg_size(dma_dev);
hw.buffer_bytes_max = SIZE_MAX;
hw.fifo_size = dma_data->fifo_size;
if (pcm->flags & SND_DMAENGINE_PCM_FLAG_NO_RESIDUE)
hw.info |= SNDRV_PCM_INFO_BATCH;
/**
* FIXME: Remove the return value check to align with the code
* before adding snd_dmaengine_pcm_refine_runtime_hwparams
* function.
*/
snd_dmaengine_pcm_refine_runtime_hwparams(substream,
dma_data,
&hw,
chan);
return snd_soc_set_runtime_hwparams(substream, &hw);
}
static int dmaengine_pcm_open(struct snd_soc_component *component,
struct snd_pcm_substream *substream)
{
struct dmaengine_pcm *pcm = soc_component_to_pcm(component);
struct dma_chan *chan = pcm->chan[substream->stream];
int ret;
ret = dmaengine_pcm_set_runtime_hwparams(component, substream);
if (ret)
return ret;
return snd_dmaengine_pcm_open(substream, chan);
}
这里调用 snd_soc_dai_get_dma_data() 函数获得类型为 struct snd_dmaengine_dai_dma_data 的 DMA 数据。这是通用 DMA 引擎驱动程序和 I2S 等 DAI 驱动程序的一个约定，即需要在 DAI 驱动程序的 probe 操作中设置 DMA 数据，类似于下面这样：
struct i2s_dev {
struct device *dev;
. . . . . .
struct snd_dmaengine_dai_dma_data capture_dma_data;
struct snd_dmaengine_dai_dma_data playback_dma_data;
. . . . . .
}
. . . . . .
static int i2s_dai_probe(struct snd_soc_dai *dai)
{
struct i2s_dev *i2s = snd_soc_dai_get_drvdata(dai);
snd_soc_dai_init_dma_data(dai,
i2s->has_playback ? &i2s->playback_dma_data : NULL,
i2s->has_capture ? &i2s->capture_dma_data : NULL);
return 0;
}
. . . . . .
static int i2s_init_dai(struct i2s_dev *i2s, struct resource *res,
struct snd_soc_dai_driver **dp)
{
. . . . . .
i2s->playback_dma_data.addr = res->start + I2S_TXFIFODATA;
i2s->playback_dma_data.addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
i2s->playback_dma_data.maxburst = 8;
. . . . . .
i2s->capture_dma_data.addr = res->start + I2S_RXFIFODATA;
i2s->capture_dma_data.addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
i2s->capture_dma_data.maxburst = 8;
. . . . . .
}
dmaengine_pcm_open() 函数主要是为流设置了运行时硬件参数。
对于其它操作，暂时不做太多说明。
Done.
标签：DMA,snd,驱动程序,Linux,component,soc,dmaengine,pcm,struct From： https://www.cnblogs.com/wolfcs/p/17658297.html