On Fri, 2014-01-17 at 09:23 +0530, Vinod Koul wrote:
On Thu, Jan 16, 2014 at 05:08:01PM +0000, Liam Girdwood wrote:
On Thu, 2014-01-16 at 16:59 +0100, Takashi Iwai wrote:
At Wed, 15 Jan 2014 16:51:57 +0000, Liam Girdwood wrote:
- mutex_lock_nested(&fe->card->mutex, SND_SOC_CARD_CLASS_RUNTIME);
- if (platform->driver->compr_ops && platform->driver->compr_ops->set_params) {
ret = platform->driver->compr_ops->set_params(cstream, params);if (ret < 0)goto out;- }
- if (fe->dai_link->compr_ops && fe->dai_link->compr_ops->set_params) {
ret = fe->dai_link->compr_ops->set_params(cstream);if (ret < 0)goto out;- }
- memcpy(&fe->dpcm[fe_substream->stream].hw_params, params,
sizeof(struct snd_pcm_hw_params));Does this really work? params is snd_compr_params and hw_parmas is snd_pcm_hw_params, isn't it?
Hmmm, no one reported a failure here to me. It is complete nonsense though so I dont know why the testers didn't run into this. I'll fix it for V2.
I think that due to the fixup fn noone noticed this. The issue will be seen only by folks who dont have the fixup function.
I now know why we didn't see this on HSW. We have a fixed BE SSP configuration so we always ignore the BE hw_params. So we ignored the junk data ;)
Liam
--------------------------------------------------------------------- Intel Corporation (UK) Limited Registered No. 1134945 (England) Registered Office: Pipers Way, Swindon SN3 1RJ VAT No: 860 2173 47
This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.