On Tue, Jul 01, 2014 at 06:42:12PM +0200, Lars-Peter Clausen wrote:
On 07/01/2014 06:25 PM, Vinod Koul wrote:
On Tue, Jul 01, 2014 at 03:49:51PM +0200, Lars-Peter Clausen wrote:
On 07/01/2014 09:47 AM, Benoit Cousson wrote:
And you should have cced me on this patch
[...] for (i = 0; i < rtd->num_codecs; i++) {
struct snd_soc_dai *codec_dai = rtd->codec_dais[i];/* check client and interface hw capabilities */snprintf(new_name, sizeof(new_name), "%s %s-%d",rtd->dai_link->stream_name, codec_dai->name, num);We may need to rethink how to do the naming for multiple CODEC DAI links here. This implementation will just keep overwriting new_name with each loop iteration and end up using the last one. Vinod may have some insights.
I am intrugued on how we do this for pcms with multi-codec support.
Looking at the patch series I dont think the soc_pcm_new() has been updated (please point out if i missed, it is late in night) Shouldnt this be updated for pcm too?
I would suggest makes sense is to keep appending the codec name for the dai-link. But yes keep a check on the size. Unless we have better ideas.
This is what the patch series does for normal PCMs
snprintf(new_name, sizeof(new_name), "%s %s-%d",
rtd->dai_link->stream_name, codec_dai->name, num);
rtd->dai_link->stream_name,(rtd->num_codecs > 1) ?"multicodec" : rtd->codec_dai->name, num);
Sounds good to me!