All of lore.kernel.org
 help / color / mirror / Atom feed
* Re: snd_pcm_hw_params_free() immediately after snd_pcm_hw_params() ???
@ 2010-01-21 10:18 Clemens Ladisch
  2010-01-24  3:30 ` Raymond Yau
  0 siblings, 1 reply; 6+ messages in thread
From: Clemens Ladisch @ 2010-01-21 10:18 UTC (permalink / raw)
  To: Raymond Yau; +Cc: alsa-devel

Raymond Yau wrote:
> Is there any flags in substream which can indicate snd_xxx_hw_params() was
> called at first time since snd_xxx_hw_params() were called many times
> during OSS emulation (probing the supported rates , channels , .... ) ?

There is no such flag.  It is the responsibility of the driver to track
the resources it has allocated.

If a resource does not depend on the hardware parameters, it might be
a better idea to allocate/free it in the open/close callbacks.


Best regards,
Clemens

^ permalink raw reply	[flat|nested] 6+ messages in thread
* snd_pcm_hw_params_free() immediately after snd_pcm_hw_params() ???
@ 2010-01-21  4:32 Raymond Yau
  2010-01-21  6:59 ` Takashi Iwai
  0 siblings, 1 reply; 6+ messages in thread
From: Raymond Yau @ 2010-01-21  4:32 UTC (permalink / raw)
  To: alsa-devel

Refer to
http://www.alsa-project.org/~tiwai/writing-an-alsa-driver/ch05s06.html#pcm-interface-operators-hw-free-callback

hw_free callback

 static int snd_xxx_hw_free(struct snd_pcm_substream *substream);

This is called to release the resources allocated via
          *hw_params*. For example, releasing the
          buffer via snd_pcm_lib_malloc_pages() is done by
          calling the following:
snd_pcm_lib_free_pages(substream);

This function is always called before the close callback is called.
Also, the callback may be called multiple times, too.
Keep track whether the resource was already released.
Nearly all the examples in those ALSA tutorials in the web sites have the code

	
		if ((err = snd_pcm_hw_params (playback_handle, hw_params)) < 0) {
			fprintf (stderr, "cannot set parameters (%s)\n",
				 snd_strerror (err));
			exit (1);
		}
	
		snd_pcm_hw_params_free (hw_params);

This seem to violate "This function is always called before the close
callback is called."

and those hardware mixing sound cards (e.g cs46xx , au88x0, ... ) use
snd_xxx_hw_free() to deallocate the hardware mixing path in the DSP.

There are a number of audio package (e.g. gstreamer , wine , .... )
call snd_pcm_hw_params_free immediately after snd_pcm_hw_params()

Is this a bug of snd_pcm_hw_params_free() not change the state of the
PCM stream ?

or

snd_pcm_prepare() did not check the buffer is already released ?

^ permalink raw reply	[flat|nested] 6+ messages in thread
end of thread, other threads:[~2010-01-24 10:29 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-01-21 10:18 snd_pcm_hw_params_free() immediately after snd_pcm_hw_params() ??? Clemens Ladisch
2010-01-24  3:30 ` Raymond Yau
2010-01-24 10:29   ` Takashi Iwai
  -- strict thread matches above, loose matches on Subject: below --
2010-01-21  4:32 Raymond Yau
2010-01-21  6:59 ` Takashi Iwai
2010-01-21  9:12   ` Raymond Yau

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.