Re: [PATCH] nvme-pci: Avoid to go into d3cold if device can't use npss.

[Bjorn Helgaas]

On Tue, May 25, 2021 at 09:44:26AM +0200, Christoph Hellwig wrote:
> On Thu, May 20, 2021 at 11:33:15AM +0800, Koba Ko wrote:

> > @@ -2958,6 +2959,15 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id)
> >  
> >  	dev_info(dev->ctrl.device, "pci function %s\n", dev_name(&pdev->dev));
> >  
> > +	if (pm_suspend_via_firmware() || !dev->ctrl.npss ||
> > +	    !pcie_aspm_enabled(pdev) ||
> > +	    dev->nr_host_mem_descs ||
> > +	    (dev->ctrl.quirks & NVME_QUIRK_SIMPLE_SUSPEND)) {
> 
> Before we start open coding this in even more places we really want a
> little helper function for these checks, which should be accomodated with
> the comment near the existing copy of the checks.
> 
> > +		pdev->d3cold_allowed = false;
> > +		pci_d3cold_disable(pdev);
> > +		pm_runtime_resume(&pdev->dev);
> 
> Why do we need to both set d3cold_allowed and call pci_d3cold_disable?

Ugh, this looks pretty hard to maintain.

I don't see why setting d3cold_allowed=false is useful.

pci_d3cold_disable() already sets dev->no_d3cold=true, and the only place
we look at d3cold_allowed is pci_dev_check_d3cold():

  if (dev->no_d3cold || !dev->d3cold_allowed || ...)

so we won't even look at d3cold_allowed when no_d3cold is set.

I don't know why we need both no_d3cold and d3cold_allowed in the
first place.  448bd857d48e ("PCI/PM: add PCIe runtime D3cold support")
added them, but without explanation for that.