On Mon, Oct 09, 2023 at 05:56:51PM -0500, Mario Limonciello wrote:
> commit 9d26d3a8f1b0 ("PCI: Put PCIe ports into D3 during suspend")
> changed pci_bridge_d3_possible() so that any vendor's PCIe ports
> from modern machines (>=2015) are allowed to be put into D3.
>
> This policy change has worked for most machines, but the behavior
> is improved with `pcie_port_pm=off` on some others.
>
> Add support for drivers to register a callback that they can optionally
> use to decide the policy on supported machines.
I would assume that drivers can decide the policy already today through
pci_d3cold_enable() / pci_d3cold_disable().
Why is this not sufficient and what's the benefit of the more complex
approach proposed here?
Thanks,
Lukas
