Re: Regression in v5.19-rc4 Sound Distortion

[Daniel Exner]

horsten Leemhuis wrote:
> On 12.07.22 18:35, Dmitry Osipenko wrote:
> > On 7/12/22 19:02, Daniel Exner wrote:
> >
> > > Since  v5.19-rc4 this box got some *loud* distorting sound on boot and
> > > after some time without any sound if something plays sound again from
> > > the integrated speakers in my monitor.
> > >
> > > I managed to bisect this down to:
> > >
> > > > commit 202773260023b56e868d09d13d3a417028f1ff5b
> > > > Author: Dmitry Osipenko <dmitry.osipenko@xxxxxxxxxxxxx>
> > > > Date:   Fri Jun 17 15:24:02 2022 +0300
> > > >
> > > >     PM: hibernate: Use kernel_can_power_off()
> > > Reverting that commit on top of v5.19-rc6 does indeed fix the problem here.
> > You're saying that the problem is on boot, but this patch is about
> > suspension to disk, which has nothing to do neither with the boot nor
> > with power-off. I'm afraid your bisection is incorrect.
> After quickly looking at the code this looks appropriate, but why does
> reverting that commit help then? That's a bit odd. Daniel: are you maybe
> using suspend-to-disk and forgot to mention it?

No, I am not. But it seems Dmitry was right, at some point I bisected 
wrong by declaring some point good too early because I didn't have the 
bug on boot.

Perhaps I have two effects here: some bug resulting in sound distortion 
and some bug triggering that faster.


I will redo the bisection, but I fear it will take some time.


Greetings

Daniel