From: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
To: Brent Lu <brent.lu@intel.com>, alsa-devel@alsa-project.org
Cc: Cezary Rojewski <cezary.rojewski@intel.com>,
linux-kernel@vger.kernel.org, Jie Yang <yang.jie@linux.intel.com>,
Takashi Iwai <tiwai@suse.com>,
Liam Girdwood <liam.r.girdwood@linux.intel.com>,
Mark Brown <broonie@kernel.org>
Subject: Re: [PATCH] ASoC: Intel: Atom: use hardware counter to update hw_ptr
Date: Mon, 27 Jul 2020 09:09:33 -0500 [thread overview]
Message-ID: <7dec7fcd-0381-d279-4dac-77c27ea5f6fe@linux.intel.com> (raw)
In-Reply-To: <1595779727-31404-1-git-send-email-brent.lu@intel.com>
On 7/26/20 11:08 AM, Brent Lu wrote:
> The ring buffer counter runs faster than hardware counter if the
> period size in hw_param is larger than 240. Although the differce is
> not much (around 2k frames), it causes false underrun in CRAS
> sometimes because it's using 256 frames as period size in hw_param.
All the Atom firmware assumes data chunks in multiples of 1ms (typically
5, 10 or 20ms). I have never seen anyone use 256 frames, that's asking
for trouble really.
it's actually the same with Skylake and SOF in most cases.
Is this a 'real' problem or a problem detected by the Chrome ALSA
compliance tests, in the latter case that would hint at a too generic
value of min_period.
> Using the hardware counter could provide precise hw_ptr to user space
> and avoid the false underrun in CRAS.
>
> Signed-off-by: Brent Lu <brent.lu@intel.com>
> ---
> sound/soc/intel/atom/sst/sst_drv_interface.c | 15 +++------------
> 1 file changed, 3 insertions(+), 12 deletions(-)
>
> diff --git a/sound/soc/intel/atom/sst/sst_drv_interface.c b/sound/soc/intel/atom/sst/sst_drv_interface.c
> index 7624953..1949ad9 100644
> --- a/sound/soc/intel/atom/sst/sst_drv_interface.c
> +++ b/sound/soc/intel/atom/sst/sst_drv_interface.c
> @@ -485,7 +485,6 @@ static inline int sst_calc_tstamp(struct intel_sst_drv *ctx,
> struct snd_pcm_substream *substream,
> struct snd_sst_tstamp *fw_tstamp)
> {
> - size_t delay_bytes, delay_frames;
> size_t buffer_sz;
> u32 pointer_bytes, pointer_samples;
>
> @@ -493,22 +492,14 @@ static inline int sst_calc_tstamp(struct intel_sst_drv *ctx,
> fw_tstamp->ring_buffer_counter);
> dev_dbg(ctx->dev, "mrfld hardware_counter %llu in bytes\n",
> fw_tstamp->hardware_counter);
> - if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK)
> - delay_bytes = (size_t) (fw_tstamp->ring_buffer_counter -
> - fw_tstamp->hardware_counter);
> - else
> - delay_bytes = (size_t) (fw_tstamp->hardware_counter -
> - fw_tstamp->ring_buffer_counter);
> - delay_frames = bytes_to_frames(substream->runtime, delay_bytes);
> +
> buffer_sz = snd_pcm_lib_buffer_bytes(substream);
> - div_u64_rem(fw_tstamp->ring_buffer_counter, buffer_sz, &pointer_bytes);
> + div_u64_rem(fw_tstamp->hardware_counter, buffer_sz, &pointer_bytes);
> pointer_samples = bytes_to_samples(substream->runtime, pointer_bytes);
>
> - dev_dbg(ctx->dev, "pcm delay %zu in bytes\n", delay_bytes);
> -
> info->buffer_ptr = pointer_samples / substream->runtime->channels;
> + info->pcm_delay = 0;
and that seems also wrong? Why would the delay be zero?
> - info->pcm_delay = delay_frames;
> dev_dbg(ctx->dev, "buffer ptr %llu pcm_delay rep: %llu\n",
> info->buffer_ptr, info->pcm_delay);
> return 0;
>
next prev parent reply other threads:[~2020-07-27 14:10 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-07-26 16:08 [PATCH] ASoC: Intel: Atom: use hardware counter to update hw_ptr Brent Lu
2020-07-27 14:09 ` Pierre-Louis Bossart [this message]
2020-07-28 2:28 ` Lu, Brent
2020-07-28 12:06 ` Pierre-Louis Bossart
2020-07-28 17:02 ` Lu, Brent
2020-07-28 17:30 ` Pierre-Louis Bossart
2020-07-29 5:06 ` Cheng-yi Chiang
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=7dec7fcd-0381-d279-4dac-77c27ea5f6fe@linux.intel.com \
--to=pierre-louis.bossart@linux.intel.com \
--cc=alsa-devel@alsa-project.org \
--cc=brent.lu@intel.com \
--cc=broonie@kernel.org \
--cc=cezary.rojewski@intel.com \
--cc=liam.r.girdwood@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=tiwai@suse.com \
--cc=yang.jie@linux.intel.com \
/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).