Re: [PATCH] ALSA: core: Remove trigger_tstamp_latched

[Takashi Iwai <tiwai@xxxxxxx>]

On Tue, 13 Aug 2024 14:54:42 +0200,
Zeno Endemann wrote:
> 
> Pierre-Louis Bossart wrote on 13.08.24 10:04:
> > by focusing on the trigger timestamp I think you're looking at the wrong
> > side of the problem. The timestamping is improved by using the same
> > hardware counter for the trigger AND regular timestamp during
> > playback/capture. If you look at a hardware counter during
> > playback/capture but the start position is recorded with another method,
> > would you agree that there's a systematic non-reproducible offset at
> > each run? You want the trigger and regular timestamps to be measured in
> > the same way to avoid measurement differences.
> 
> I am not sure what you are talking about. I have not seen any place in the
> code where the trigger timestamp is taken in any other more sophisticated
> way than what the default is doing, i.e. calling snd_pcm_gettime. So I do
> not see how your custom *trigger* timestamps are done "with another method".
> 
> > I will not disagree that most applications do not need precise
> > timestamping, but if you want to try to enable time-of-flight
> > measurements for presence or gesture detection you will need higher
> > sampling rates and micro-second level accuracy.
> 
> I don't know, this sounds very theoretical at best to me. However I do not
> have the desire to try to further argue and convince you otherwise.
> 
> Do you want to propose a different solution for the stop trigger timestamp
> bug? That is my main goal after all.

Ah, I guess that the discussion drifted because of misunderstanding.

This isn't about the accuracy of the audio timestamp, but rather the
timing of trigger tstamp.  The commit 2b79d7a6bf34 ("ALSA: pcm: allow
for trigger_tstamp snapshot in .trigger") allowed the trigger_tstamp
taken in the driver's trigger callback.  But, the effectiveness of
this change is dubious, because the timestamp taken in the usual code
path in PCM core is right after the trigger callback, hence the
difference should be negligible -- that's the argument.

No matter how the fix will be, could you put the Fixes tag pointing to
the culprit commit(s) at the next submission?


thanks,

Takashi