Alsa-Devel Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Takashi Iwai <tiwai@suse.de>
To: "Yang, Libin" <libin.yang@intel.com>
Cc: "alsa-devel@alsa-project.org" <alsa-devel@alsa-project.org>,
	"broonie@kernel.org" <broonie@kernel.org>
Subject: Re: [PATCH V2] ASoC: fix hdmi codec driver contest in S3
Date: Tue, 26 Mar 2019 08:36:50 +0100	[thread overview]
Message-ID: <s5hr2aum6i5.wl-tiwai@suse.de> (raw)
In-Reply-To: <96A12704CE18D347B625EE2D4A099D19527FB1A0@SHSMSX103.ccr.corp.intel.com>

On Tue, 26 Mar 2019 06:42:15 +0100,
Yang, Libin wrote:
(snip)
> Hi Takashi,
> Below is my draft patch, which doesn't work with pm_runtime_get_sync(). Is
> there anything wrong in my code?
(snip)
> And the dmesg is:
> [   36.087224] HDMI HDA Codec ehdaudio0D2: in hdmi_codec_resume 1890 ylb
> [   36.087230] HDMI HDA Codec ehdaudio0D2: in hdac_hdmi_runtime_resume 2122 ylb 1
> [   36.087335] HDMI HDA Codec ehdaudio0D2: in hdac_hdmi_cvt_output_widget_event 812 ylb
> [   40.097586] HDMI HDA Codec ehdaudio0D2: in hdac_hdmi_runtime_resume 2135 ylb 2
> [   40.097766] HDMI HDA Codec ehdaudio0D2: in hdac_hdmi_pin_output_widget_event 767 ylb
> [   45.108632] HDMI HDA Codec ehdaudio0D2: in hdac_hdmi_pin_mux_widget_event 857 ylb
> 
> >From the dmesg, hdac_hdmi_cvt_output_widget_event() 
> is called between "ylb 1" and "ylb 2" in runtime_resume().
> This means xxx_event() registers setting runs before display
> power is turned on.

OK, now I understood what went wrong.  It's a similar issue as we've
hit for the legacy HD-audio and figured out recently.

If my understanding is correct, the problem is the call of
pm_runtime_force_resume() in PM resume callback combined with an async
work.  pm_runtime_force_resume() sets the runtime state immediately to
RPM_ACTIVE even before finishing the task.  The code seems assuming
blindly that there is no other conflicting task while S3/S4 resume,
but this is a problem.  That's why the next pm_runtime_get_sync()
wasn't blocked but just went through; since the RPM flag was already
set to RPM_ACTIVE in pm_runtime_force_resume(), it thought as if it
were already active, while the real runtime resume code was still
processing the callback.

In the case of legacy HD-audio, we "fixed" the problem by avoiding the
trigger of async work at resume, but let it be triggered from runtime
resume.  In ASoC case, however, it's no option.

Maybe a possible solution in the sound driver side would be to move
from system suspend/resume to ASoC component suspend/resume.  The
runtime suspend/resume can be kept as is, and call
pm_runtime_force_suspend and resume from the component suspend/resume
callbacks.  Since the component callbacks are certainly processed
before DAPM events, this should assure the order.


thanks,

Takashi

  reply	other threads:[~2019-03-26  7:36 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-03-22  6:24 [PATCH V2] ASoC: fix hdmi codec driver contest in S3 libin.yang
2019-03-22  9:04 ` Takashi Iwai
2019-03-22  9:12   ` Yang, Libin
2019-03-22  9:56     ` Takashi Iwai
2019-03-22 11:46       ` Yang, Libin
2019-03-22 20:25         ` Takashi Iwai
2019-03-25  6:51           ` Yang, Libin
2019-03-25 10:08             ` Takashi Iwai
2019-03-25 14:27               ` Yang, Libin
2019-03-25 14:38                 ` Takashi Iwai
2019-03-26  5:42                   ` Yang, Libin
2019-03-26  7:36                     ` Takashi Iwai [this message]
2019-03-26  7:58                       ` Yang, Libin
2019-03-26  8:59                         ` Takashi Iwai
2019-03-26 11:02                           ` Yang, Libin
2019-03-26 11:10                             ` Takashi Iwai
2019-03-26 11:19                               ` Yang, Libin
2019-03-26 11:22                                 ` Takashi Iwai
2019-03-26 11:28                                   ` Yang, Libin
2019-03-26  8:46                       ` Yang, Libin
2019-03-26  8:53                         ` Takashi Iwai
2019-03-26 11:15                           ` Yang, Libin
2019-03-26 11:19                             ` Takashi Iwai
2019-03-26 11:21                               ` Yang, Libin
2019-03-27  6:14                               ` Yang, Libin
2019-03-28  7:10                                 ` Yang, Libin
2019-03-28  7:15                                   ` Takashi Iwai
2019-03-28  7:28                                     ` Yang, Libin
2019-03-28  7:31                                       ` 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=s5hr2aum6i5.wl-tiwai@suse.de \
    --to=tiwai@suse.de \
    --cc=alsa-devel@alsa-project.org \
    --cc=broonie@kernel.org \
    --cc=libin.yang@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