lkml.org 
[lkml]   [2015]   [Dec]   [1]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [PATCH v4 2/3] ASoC: cygnus: Add Cygnus audio DAI driver
From
Date


On 11/22/2015 5:44 AM, Mark Brown wrote:
> On Mon, Nov 09, 2015 at 04:17:43PM -0800, Simran Rai wrote:
>
>> +#ifdef CONFIG_PM_SLEEP
>> +static int cygnus_ssp_suspend(struct snd_soc_dai *cpu_dai)
>> +{
>> + struct cygnus_aio_port *aio = cygnus_dai_get_portinfo(cpu_dai);
>> + struct cygnus_audio *cygaud = snd_soc_dai_get_drvdata(cpu_dai);
>> +
>> + audio_ssp_out_disable(aio);
>> + audio_ssp_in_disable(aio);
>> + if (cygaud->active_ports > 0)
>> + cygaud->active_ports--;
>> +
>> + return 0;
>> +}
>> +
>> +static int cygnus_ssp_resume(struct snd_soc_dai *cpu_dai)
>> +{
>> + return 0;
>> +}
> I'm a bit confused here - why do we need to disable things on suspend
> but not reenable them on resume? I'd expect that the core would have
> quiesced any streams that need to be suspended before we get as far as
> suspending the drivers.
>
> Please also remove empty functions.
>
> Now I check I see that I'm repeating the questions I had on my previous
> review:
>
> | Blank line between functions and remove empty functions. Though I'm not
> | clear why the result doesn't undo what the suspend did...
>
> Please don't ignore review comments.

Let me investigate this further and debug why the core did not
quiescence the streams.

>
>> + parent = clk_get_parent(cygaud->audio_clk[0]);
>> + if (IS_ERR(parent)) {
>> + error = PTR_ERR(parent);
>> + goto err_get_parent;
>> + }
>> +
>> + /* Set PLL VCO Frequency (Hz) to default */
>> + error = clk_set_rate(parent, DEFAULT_VCO);
>> + if (error) {
>> + dev_err(&pdev->dev,
>> + "%s Set PLL VCO rate failed: %d\n", __func__, error);
>> + goto err_get_parent;
>> + }
> I would expect any initialisationn of clocks beyond the ones that the
> device directly interacts with to be handled within the clock API
> configuration rather than in a specific driver, this avoids the driver
> being dependent on a particular system integration.

I will have to figure out a way to propagate the frequency from leaf clocks to the parent
clock. Will send out another patch with the fix.

Thanks,
Simran



\
 
 \ /
  Last update: 2015-12-02 03:41    [W:0.050 / U:1.116 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site