Re: [PATCH 0/4] PCI PM refinements

[Bjorn Helgaas <helgaas@kernel.org>]

On Wed, Aug 31, 2016 at 08:15:18AM +0200, Lukas Wunner wrote:
> The following patches are part of my Thunderbolt Runtime PM series.
> I'm posting them separately to split the series into digestible chunks.
> 
> They benefit everyone, not just Thunderbolt, and Thunderbolt works fine
> without them (just not with the same degree of perfection).
> 
> Patch [1/4] affords direct-complete to devices which are not platform-
> power-manageable, yet can be suspended to D3cold with a nonstandard
> mechanism. It does so using a heuristic: If a device's current_state
> is D3cold and it is not power-manageable, it is assumed that a
> nonstandard PM mechanism is at work and the device is left asleep.
> The patch is the outcome of a discussion with Rafael and Alan on
> linux-pci@ and I would like to thank both for tirelessly responding
> to my questions.
> 
> Patch [3/4] replaces the unconditional resume after direct-complete with
> a conditional resume by determining if the platform firmware has put the
> device in reset-power-on or not.
> 
> Because the D3cold state cannot be determined from the PMCSR, patch [2/4]
> adds a hook to query the platform firmware for its opinion on the device's
> power state.
> 
> Finally, patch [4/4] changes the shutdown handling of PCI devices such
> that they are no longer woken without reason.
> 
> The result of these patches should be a speedup and decreased power
> consumption for system sleep and system shutdown (both for devices that
> are platform-power-manageable and those that aren't).
> 
> If you prefer point & click interfaces, the patches are browsable here:
> https://github.com/l1k/linux/commits/thunderbolt_runpm_v3

Rafael, do you have any comment on these?  I don't have any objections
myself, but I'd like your opinion as the PM expert :)

> Lukas Wunner (4):
>   PCI: Afford direct-complete to devices with nonstandard pm
>   PCI: Query platform firmware for device power state
>   PCI: Avoid unnecessary resume after direct-complete
>   PCI: Avoid unnecessary resume on shutdown
> 
>  drivers/pci/pci-acpi.c   | 23 +++++++++++++++++++++++
>  drivers/pci/pci-driver.c | 20 ++++++++++++++++++--
>  drivers/pci/pci.c        | 38 +++++++++++++++++++++++++++++++-------
>  drivers/pci/pci.h        |  3 +++
>  4 files changed, 75 insertions(+), 9 deletions(-)
> 
> -- 
> 2.8.1
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html