On Tue, Feb 06, 2018 at 01:41:29AM +0000, Sasha Levin wrote:
> On Mon, Feb 05, 2018 at 11:41:54AM +0000, Mark Brown wrote:
> >On Sat, Feb 03, 2018 at 06:00:42PM +0000, Sasha Levin wrote:
> >> From: Abhijeet Kumar <abhijeet.kumar@xxxxxxxxx>
> >>
> >> [ Upstream commit d070f7c703ef26e3db613f24206823f916272fc6 ]
> >>
> >> In skylake platform, we hear a loud pop noise(0 dB) at start of
> >> audio capture power up sequence. This patch removes the pop noise
> >> from the recording by adding a delay before enabling ADC.
> >
> >> switch (event) {
> >> case SND_SOC_DAPM_POST_PMU:
> >> + msleep(125);
> >> regmap_update_bits(nau8825->regmap, NAU8825_REG_ENA_CTRL,
> >
> >This is another one of these where I'm not convinced that backporting to
> >stable isn't going to introduce regressions - it will fix problems on
> >some systems but if someone really cares about a startup pop and has
> >arranged to do something like discard the start of the audio where the
> >pop occurs then this may cut off even more and cause them to miss
> >things.
>
> It might cause regressions for users who worked around the issue, but it
> may also fix the issue for users who didn't. I think that in general
> we'd rather have mainline and stable trees have the same bugs and
> breakage - we don't want surprises when users jump around between
> kernels.
Yes, being "bug compatible" is a key feature to have :)
thanks,
greg k-h