[alsa-devel] [PATCH] ASoC: Factor out I2C 8 bit address 8 bit data I/O
This patch is for the AK4671 codec driver using this format.
Signed-off-by: Joonyoung Shim jy0922.shim@samsung.com --- sound/soc/soc-cache.c | 30 ++++++++++++++++++++++++++++++ 1 files changed, 30 insertions(+), 0 deletions(-)
diff --git a/sound/soc/soc-cache.c b/sound/soc/soc-cache.c index c8ceddc..404231e 100644 --- a/sound/soc/soc-cache.c +++ b/sound/soc/soc-cache.c @@ -77,6 +77,35 @@ static int snd_soc_7_9_spi_write(void *control_data, const char *data, #define snd_soc_7_9_spi_write NULL #endif
+static int snd_soc_8_8_write(struct snd_soc_codec *codec, unsigned int reg, + unsigned int value) +{ + u8 *cache = codec->reg_cache; + u8 data[2]; + + BUG_ON(codec->volatile_register); + + data[0] = reg & 0xff; + data[1] = value & 0xff; + + if (reg < codec->reg_cache_size) + cache[reg] = value; + + if (codec->hw_write(codec->control_data, data, 2) == 2) + return 0; + else + return -EIO; +} + +static unsigned int snd_soc_8_8_read(struct snd_soc_codec *codec, + unsigned int reg) +{ + u8 *cache = codec->reg_cache; + if (reg >= codec->reg_cache_size) + return -1; + return cache[reg]; +} + static int snd_soc_8_16_write(struct snd_soc_codec *codec, unsigned int reg, unsigned int value) { @@ -151,6 +180,7 @@ static struct { unsigned int (*i2c_read)(struct snd_soc_codec *, unsigned int); } io_types[] = { { 7, 9, snd_soc_7_9_write, snd_soc_7_9_spi_write, snd_soc_7_9_read }, + { 8, 8, snd_soc_8_8_write, NULL, snd_soc_8_8_read, NULL }, { 8, 16, snd_soc_8_16_write, NULL, snd_soc_8_16_read, snd_soc_8_16_read_i2c }, };
On Mon, Sep 07, 2009 at 12:04:37PM +0900, Joonyoung Shim wrote:
This patch is for the AK4671 codec driver using this format.
Signed-off-by: Joonyoung Shim jy0922.shim@samsung.com
Applied, thanks.
At Mon, 7 Sep 2009 11:15:19 +0100, Mark Brown wrote:
On Mon, Sep 07, 2009 at 12:04:37PM +0900, Joonyoung Shim wrote:
This patch is for the AK4671 codec driver using this format.
Signed-off-by: Joonyoung Shim jy0922.shim@samsung.com
Applied, thanks.
Where we are at it: Mark, could you fix the initialization of io_types[] in C99 style? For this case, C99 style initialization is suited better since some fields are optional now.
Also, I find the current unbalanced hw_read/hw_write definitions uneasy. hw_read takes the codec instance while hw_write takes i2c instance (and hw_read_t is dead). It'd be nice if you can fix these as well.
thanks,
Takashi
On Mon, Sep 07, 2009 at 12:40:02PM +0200, Takashi Iwai wrote:
Where we are at it: Mark, could you fix the initialization of io_types[] in C99 style? For this case, C99 style initialization is suited better since some fields are optional now.
Meh, I guess.
Also, I find the current unbalanced hw_read/hw_write definitions uneasy. hw_read takes the codec instance while hw_write takes i2c instance (and hw_read_t is dead). It'd be nice if you can fix these as well.
I don't think it's worth the pain until more of the drivers are using the shared cache functions - making changes like that more straightforward is one of the goals of doing this refactoring.
At Mon, 7 Sep 2009 11:50:06 +0100, Mark Brown wrote:
On Mon, Sep 07, 2009 at 12:40:02PM +0200, Takashi Iwai wrote:
Where we are at it: Mark, could you fix the initialization of io_types[] in C99 style? For this case, C99 style initialization is suited better since some fields are optional now.
Meh, I guess.
Also, I find the current unbalanced hw_read/hw_write definitions uneasy. hw_read takes the codec instance while hw_write takes i2c instance (and hw_read_t is dead). It'd be nice if you can fix these as well.
I don't think it's worth the pain until more of the drivers are using the shared cache functions - making changes like that more straightforward is one of the goals of doing this refactoring.
OK, but at least hw_read_t should be removed or fixed :)
thanks,
Takashi
participants (3)
-
Joonyoung Shim
-
Mark Brown
-
Takashi Iwai