From: Takashi Iwai <tiwai@suse.de>
To: Alan Young <Alan.Young@IEE.org>
Cc: alsa-devel@alsa-project.org
Subject: Re: [PATCH RESEND] pcm_rate: Do not discard slave reported delay in status result.
Date: Mon, 28 Nov 2016 20:11:14 +0100 [thread overview]
Message-ID: <s5hshqb5sj1.wl-tiwai@suse.de> (raw)
In-Reply-To: <52e39131-9141-b466-e92a-ff19ffab6d3a@IEE.org>
On Thu, 17 Nov 2016 18:12:57 +0100,
Alan Young wrote:
>
> On 17/11/16 15:24, Alan Young wrote:
> >
> > Similar to recent dshare patch.
> >
> Update with sign-off
Sorry for the late follow up, I've been drug by other things for too
long.
> >From 6f93ee59846d2c0058ef702c1fa68d723bfb14f6 Mon Sep 17 00:00:00 2001
> From: Alan Young <consult.awy@gmail.com>
> Date: Tue, 14 Jun 2016 10:15:01 +0100
> Subject: [PATCH] pcm_rate: Do not discard slave reported delay in status
> result.
>
> snd_pcm_rate_status() gets the underlying status from the slave PCM.
> This may contain a delay value that includes elements such as codec and
> other transfer delays. Use this as the base for the returned delay
> value, adjusted for any frames buffered locally (within the rate
> plugin).
>
> Also update snd_pcm_rate_delay() similarly.
>
> Signed-off-by: Alan Young <consult.awy@gmail.com>
> ---
> src/pcm/pcm_rate.c | 46 ++++++++++++++++++++++++++++++++++++++++------
> 1 file changed, 40 insertions(+), 6 deletions(-)
>
> diff --git a/src/pcm/pcm_rate.c b/src/pcm/pcm_rate.c
> index 6184def..15383ae 100644
> --- a/src/pcm/pcm_rate.c
> +++ b/src/pcm/pcm_rate.c
> @@ -559,10 +559,9 @@ snd_pcm_rate_read_areas1(snd_pcm_t *pcm,
> pcm->channels, rate);
> }
>
> -static inline void snd_pcm_rate_sync_hwptr(snd_pcm_t *pcm)
> +static inline void snd_pcm_rate_sync_hwptr0(snd_pcm_t *pcm, snd_pcm_uframes_t slave_hw_ptr)
> {
> snd_pcm_rate_t *rate = pcm->private_data;
> - snd_pcm_uframes_t slave_hw_ptr = *rate->gen.slave->hw.ptr;
>
> if (pcm->stream != SND_PCM_STREAM_PLAYBACK)
> return;
> @@ -576,6 +575,12 @@ static inline void snd_pcm_rate_sync_hwptr(snd_pcm_t *pcm)
> rate->hw_ptr %= pcm->boundary;
> }
>
> +static inline void snd_pcm_rate_sync_hwptr(snd_pcm_t *pcm)
> +{
> + snd_pcm_rate_t *rate = pcm->private_data;
> + snd_pcm_rate_sync_hwptr0(pcm, *rate->gen.slave->hw.ptr);
> +}
> +
> static int snd_pcm_rate_hwsync(snd_pcm_t *pcm)
> {
> snd_pcm_rate_t *rate = pcm->private_data;
> @@ -586,10 +591,37 @@ static int snd_pcm_rate_hwsync(snd_pcm_t *pcm)
> return 0;
> }
>
> +static snd_pcm_uframes_t snd_pcm_rate_playback_internal_delay(snd_pcm_t *pcm)
> +{
> + snd_pcm_rate_t *rate = pcm->private_data;
> +
> + if (rate->appl_ptr < rate->last_commit_ptr) {
> + return rate->appl_ptr - rate->last_commit_ptr + pcm->boundary;
> + } else {
> + return rate->appl_ptr - rate->last_commit_ptr;
> + }
> +}
> +
> static int snd_pcm_rate_delay(snd_pcm_t *pcm, snd_pcm_sframes_t *delayp)
> {
> + snd_pcm_rate_t *rate = pcm->private_data;
> + snd_pcm_sframes_t slave_delay;
> + int err;
> +
> snd_pcm_rate_hwsync(pcm);
> - *delayp = snd_pcm_mmap_hw_avail(pcm);
> +
> + err = snd_pcm_delay(rate->gen.slave, &slave_delay);
> + if (err < 0) {
> + return err;
> + }
> +
> + if (pcm->stream == SND_PCM_STREAM_PLAYBACK) {
> + *delayp = rate->ops.input_frames(rate->obj, slave_delay)
> + + snd_pcm_rate_playback_internal_delay(pcm);
> + } else {
> + *delayp = rate->ops.output_frames(rate->obj, slave_delay)
> + + snd_pcm_mmap_capture_hw_avail(pcm);
> + }
Hmm, I'm not 100% sure through a quick review whether it's the correct
calculation. Maybe it's worth to give more comments either in the
code or in the changelog.
> return 0;
> }
>
> @@ -1083,15 +1115,17 @@ static int snd_pcm_rate_status(snd_pcm_t *pcm, snd_pcm_status_t * status)
> status->state = SND_PCM_STATE_RUNNING;
> status->trigger_tstamp = rate->trigger_tstamp;
> }
> - snd_pcm_rate_sync_hwptr(pcm);
> + snd_pcm_rate_sync_hwptr0(rate, status->hw_ptr);
This can't work.
I can fix it in my side, but OTOH, this made me wonder how you tested
the patch...
> status->appl_ptr = *pcm->appl.ptr;
> status->hw_ptr = *pcm->hw.ptr;
> if (pcm->stream == SND_PCM_STREAM_PLAYBACK) {
> - status->delay = snd_pcm_mmap_playback_hw_avail(pcm);
> + status->delay = rate->ops.input_frames(rate->obj, status->delay)
> + + snd_pcm_rate_playback_internal_delay(pcm);
> status->avail = snd_pcm_mmap_playback_avail(pcm);
> status->avail_max = rate->ops.input_frames(rate->obj, status->avail_max);
> } else {
> - status->delay = snd_pcm_mmap_capture_hw_avail(pcm);
> + status->delay = rate->ops.output_frames(rate->obj, status->delay)
> + + snd_pcm_mmap_capture_hw_avail(pcm);
> status->avail = snd_pcm_mmap_capture_avail(pcm);
> status->avail_max = rate->ops.output_frames(rate->obj, status->avail_max);
Why only playback needs the special handling while the capture
doesn't? Again, some comments would be helpful for better
understanding your changes.
thanks,
Takashi
next prev parent reply other threads:[~2016-11-28 19:11 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-11-17 15:24 [PATCH] pcm_rate: Do not discard slave reported delay in status result Alan Young
2016-11-17 15:26 ` Alan Young
2016-11-17 16:47 ` Takashi Iwai
2016-11-17 16:51 ` Alan Young
2016-11-17 16:55 ` Takashi Iwai
2016-11-17 17:12 ` [PATCH RESEND] " Alan Young
2016-11-28 19:11 ` Takashi Iwai [this message]
2016-12-13 11:43 ` Alan Young
2016-12-13 11:49 ` Takashi Iwai
2016-12-13 13:05 ` Alan Young
2016-12-14 14:29 ` Takashi Iwai
2017-02-08 10:50 ` [PATCH] pcm: rate: Add capability to pass configuration node to plugins Alan Young
2017-02-08 15:36 ` Takashi Iwai
2017-02-09 15:41 ` Alan Young
2017-02-09 15:48 ` Alan Young
2017-02-13 17:20 ` [PATCH] [updated] " Alan Young
2017-02-14 7:08 ` Takashi Iwai
2017-02-14 7:30 ` Alan Young
2017-02-15 12:28 ` Alan Young
2017-02-17 16:53 ` Takashi Iwai
2017-02-17 17:44 ` Alan Young
2017-02-17 17:59 ` Takashi Iwai
2017-02-21 12:02 ` Alan Young
2017-02-21 12:39 ` Takashi Iwai
2017-02-21 14:34 ` Alan Young
2017-02-21 14:43 ` Takashi Iwai
2017-02-21 15:04 ` Alan Young
2017-02-21 15:14 ` Takashi Iwai
2017-02-21 15:24 ` Alan Young
2017-02-21 15:28 ` Takashi Iwai
2017-02-21 16:14 ` Alan Young
2017-02-21 21:30 ` Takashi Iwai
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=s5hshqb5sj1.wl-tiwai@suse.de \
--to=tiwai@suse.de \
--cc=Alan.Young@IEE.org \
--cc=alsa-devel@alsa-project.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).