Re: [PATCH v2] PCI: PM: Skip devices in D0 for suspend-to-idle

[Rafael J. Wysocki]

On Mon, Jun 24, 2019 at 11:37 PM Rafael J. Wysocki <rafael@xxxxxxxxxx> wrote:
>
> On Mon, Jun 24, 2019 at 2:43 PM Jon Hunter <jonathanh@xxxxxxxxxx> wrote:
> >
> > Hi Rafael,
> >
> > On 13/06/2019 22:59, Rafael J. Wysocki wrote:
> > > From: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
> > >
> > > Commit d491f2b75237 ("PCI: PM: Avoid possible suspend-to-idle issue")
> > > attempted to avoid a problem with devices whose drivers want them to
> > > stay in D0 over suspend-to-idle and resume, but it did not go as far
> > > as it should with that.
> > >
> > > Namely, first of all, the power state of a PCI bridge with a
> > > downstream device in D0 must be D0 (based on the PCI PM spec r1.2,
> > > sec 6, table 6-1, if the bridge is not in D0, there can be no PCI
> > > transactions on its secondary bus), but that is not actively enforced
> > > during system-wide PM transitions, so use the skip_bus_pm flag
> > > introduced by commit d491f2b75237 for that.
> > >
> > > Second, the configuration of devices left in D0 (whatever the reason)
> > > during suspend-to-idle need not be changed and attempting to put them
> > > into D0 again by force is pointless, so explicitly avoid doing that.
> > >
> > > Fixes: d491f2b75237 ("PCI: PM: Avoid possible suspend-to-idle issue")
> > > Reported-by: Kai-Heng Feng <kai.heng.feng@xxxxxxxxxxxxx>
> > > Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
> > > Reviewed-by: Mika Westerberg <mika.westerberg@xxxxxxxxxxxxxxx>
> > > Tested-by: Kai-Heng Feng <kai.heng.feng@xxxxxxxxxxxxx>
> >
> > I have noticed a regression in both the mainline and -next branches on
> > one of our boards when testing suspend. The bisect is point to this
> > commit and reverting on top of mainline does fix the problem. So far I
> > have not looked at this in close detail but kernel log is showing ...
>
> Can you please collect a log like that, but with dynamic debug in
> pci-driver.c enabled?
>
> Note that reverting this commit is rather out of the question, so we
> need to get to the bottom of the failure.

I suspect that there is a problem with the pm_suspend_via_firmware()
check which returns 'false' on the affected board, but the platform
actually removes power from devices left in D0 during suspend.

I guess it would be more appropriate to check something like
pm_suspend_no_platform() which would return 'true' in the
suspend-to-idle patch w/ ACPI.