Messages in this thread | | | Subject | Re: [PATCH AUTOSEL 5.7 055/388] ASoC: SOF: Do nothing when DSP PM callbacks are not set | From | Pierre-Louis Bossart <> | Date | Thu, 18 Jun 2020 08:56:22 -0500 |
| |
On 6/18/20 6:44 AM, Daniel Baluta wrote: > On 6/18/20 2:01 PM, Mark Brown wrote: >> On Wed, Jun 17, 2020 at 09:02:32PM -0400, Sasha Levin wrote: >>> From: Daniel Baluta <daniel.baluta@nxp.com> >>> >>> [ Upstream commit c26fde3b15ed41f5f452f1da727795f787833287 ] >>> >>> This provides a better separation between runtime and PM sleep >>> callbacks. >>> >>> Only do nothing if given runtime flag is set and calback is not set. >>> >>> With the current implementation, if PM sleep callback is set but runtime >>> callback is not set then at runtime resume we reload the firmware even >>> if we do not support runtime resume callback. >> This doesn't look like a bugfix, just an optimization? > > Indeed can be seen as an optimization, but it does unexpected things > which can cause trouble > > and weird behavior for people not familiar with the matter. > > For example, as explained in the commit message if you only provide > > System PM handler but not runtime PM handler, then the DSP will be resetted > > even if this is not the intention.
I think it's a bug fix for Intel legacy platforms (Baytrail, Broadwell) where runtime_pm isn't supported. However the additional fixes for system suspend/resume were only provided for 5.8, so this patch in isolation will not do much for those platforms. Put differently, even if this patch is applied to 5.7 suspend/resume would still not work for Baytrail/Broadwell. Daniel, your call if you need this for i.MX?
| |