Re: [PATCH] PCI: don't power-off apple thunderbolt controller on s2idle

[Bjorn Helgaas <helgaas@kernel.org>]

[+cc Lukas]

On Fri, May 07, 2021 at 12:32:20PM +0300, Konstantin Kharlamov wrote:
> On Thu, 2021-05-06 at 16:48 -0500, Bjorn Helgaas wrote:
> > [+cc Rafael, Andreas, linux-pm]
> > 
> > On Thu, May 06, 2021 at 08:38:20PM +0300, Konstantin Kharlamov wrote:
> > > On Macbook 2013 resuming from s2idle results in external monitor no
> > > longer being detected, and dmesg having errors like:
> > > 
> > >     pcieport 0000:06:00.0: can't change power state from D3hot to D0 (config
> > > space inaccessible)
> > > 
> > > and a stacktrace. The reason turned out that the hw that the quirk
> > > powers off does not get powered on back on resume.
> > 
> > quirk_apple_poweroff_thunderbolt() was added in 2014 by 1df5172c5c25
> > ("PCI: Suspend/resume quirks for Apple thunderbolt").  It claims
> > "power is automatically restored before resume," so there must be
> > something special about s2idle that prevents the power-on.
> > 
> > IIUC this change will reduce the s2idle power savings.  I would feel
> > better about this if we understood what the difference was.  
> > 
> > > Thus, add a check for s2idle to the quirk, and do nothing if the suspend
> > > mode is s2idle.
> > 
> > Obviously the *hardware* hasn't changed since 1df5172c5c25.  Is s2idle
> > something that wasn't tested back then, or is this problem connected
> > to an s2idle change since then?  Can we identify a commit that
> > introduced this problem?  That would help with backporting or stable
> > tags.
> > 
> > > Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=212767
> > 
> > Thanks for this!  Would you mind attaching the output of
> > "sudo lspci -vvv"?  If you attach any other dmesg, could you
> > use "dmesg --color=never" so the log doesn't include all the
> > escape characters?
> 
> Thank you! So, just to be clear: in lieu of Lukas Wunner's reply, do
> you still want `lspci` and `dmesg` outputs, or are you okay with the
> information Lukas provided?

Yes, please attach at least lspci output.  It helps understand the
topology and may be useful in the future.  I wish we had a similar
bugzilla with more information about the original 1df5172c5c25.

> And while on it, an unrelated question to you as a maintainer: I
> never contributed to the kernel before: in case you are okay with
> the patch, what happens now that I got R-b, should I resend a v2 of
> it with the R-b added?

You don't need to resend a patch if the only change is to add
reviewed-by, acked-by, or similar tags.

Bjorn