[alsa-devel] [PATCH] ASoC: soc-core: add snd_soc_of_parse_sysclk_id() for dt
This patch adds snd_soc_of_parse_sysclk_id() and supports below style on dt:
sysclk-id = <XXX>;
It's an optional property, If this property is absent the defualt vaule 0 will be returned.
Signed-off-by: Xiubo Li Li.Xiubo@freescale.com --- include/sound/soc.h | 1 + sound/soc/soc-core.c | 16 ++++++++++++++++ 2 files changed, 17 insertions(+)
diff --git a/include/sound/soc.h b/include/sound/soc.h index 5a049d9..86ac72f 100644 --- a/include/sound/soc.h +++ b/include/sound/soc.h @@ -1170,6 +1170,7 @@ int snd_soc_of_parse_card_name(struct snd_soc_card *card, const char *propname); int snd_soc_of_parse_audio_routing(struct snd_soc_card *card, const char *propname); +unsigned int snd_soc_of_parse_sysclk_id(struct device_node *np); unsigned int snd_soc_of_parse_daifmt(struct device_node *np, const char *prefix); int snd_soc_of_get_dai_name(struct device_node *of_node, diff --git a/sound/soc/soc-core.c b/sound/soc/soc-core.c index be88df5..572d95d 100644 --- a/sound/soc/soc-core.c +++ b/sound/soc/soc-core.c @@ -4472,6 +4472,22 @@ int snd_soc_of_parse_audio_routing(struct snd_soc_card *card, } EXPORT_SYMBOL_GPL(snd_soc_of_parse_audio_routing);
+unsigned int snd_soc_of_parse_sysclk_id(struct device_node *np) +{ + u32 id; + + /* + * check "sysclk-id = <XX>" + * it is an optional property + * 0 will be returned if absent + */ + if (of_property_read_u32(np, "sysclk-id", &id)) + id = 0; + + return id; +} +EXPORT_SYMBOL_GPL(snd_soc_of_parse_sysclk_id); + unsigned int snd_soc_of_parse_daifmt(struct device_node *np, const char *prefix) {
On Thu, Jan 02, 2014 at 03:34:17PM +0800, Xiubo Li wrote:
This patch adds snd_soc_of_parse_sysclk_id() and supports below style on dt:
sysclk-id = <XXX>;
It's an optional property, If this property is absent the defualt vaule 0 will be returned.
The trick with this is defining what the clock should be. Drivers would have to start defining their clock IDs and then at some point we end up reinventing the common clock API and bindings. We really ought to be moving ASoC's clock handling over towards clk but there's issues with that given that the clock API is not available at all on some platforms and the common clock API isn't common. This is rather a larger bit of work than I expect you want to do right now though...
What's your use case for this - is it possible to make the driver you're looking at default to a clock that'll work for most systems? That'd sidestep the need to say which clock. Alternatively is the CODEC only likely to be used with systems supporting common clk in which case doing the sysclk configuration via the clk API might be the way forwards?
participants (2)
-
Mark Brown -
Xiubo Li