Upstream: PATCH 06/21 ASoC: SOF: Intel: Add Intel specific HDA DSP HW operations #416
Description
- if (ret < 0) {
dev_err(bus->dev, "error: cannot turn on display power on i915 after resume\n");return ret;- }
- /* init hda controller and power dsp up */
- return hda_resume(sdev);
+}+int hda_dsp_runtime_resume(struct snd_sof_dev *sdev)
+{
- /* init hda controller and power dsp up */
- return hda_resume(sdev);
+}
But why can't runtime PM turn on/off the display power?
Metadata
Metadata
Assignees
Labels
No labels
Activity
jocelyn-li commentedon Dec 17, 2018
@ranj063 any help needed?
mengdonglin commentedon Jan 22, 2019
@ranj063 Do we still need this?
ranj063 commentedon Jan 22, 2019
@mengdonglin yes, this was a question from Takashi which we havent addressed. Can @libinyang take a look into this while he's making the other i915 related changes?
libinyang commentedon Jan 22, 2019
@ranj063 @mengdonglin
Which question are your talking about? I didn't find the email from Takashi for suspend/resume issue.
libinyang commentedon Jan 22, 2019
BTW: I updated a patch #569. It may be related to this topic.
ranj063 commentedon Jan 22, 2019
@libinyang the question is in the description. Basically, Takashi was asking why we power off/on i915 display only on system suspend/resume and not during runtime suspend/resume.
libinyang commentedon Jan 22, 2019
@ranj063
Actually, with Takashi's latest patch, we don't need handle i915 power any more. We can directly remove the i915 setting, please see my patch #569
We need power on/off before in suspend/resume because hdmi codec will power on i915 when suspend and the hdmi codec didn't power off i915. So we need power off in controller suspend. As hdmi codec suspend/resume sequence has been changed, we don't need power off i915 power any more in controller suspend and the same in resume.
plbossart commentedon Feb 8, 2019
closing, this doesn't seem to be applicable any longer