On Thu, Jun 26, 2025 at 08:15:05PM +0200, Rafael J. Wysocki wrote:
> From: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
>
> The PCI bus type does not expect its runtime PM callbacks,
> pci_pm_runtime_suspend() and pci_pm_runtime_resume(), to be invoked at
> any point during system-wide suspend and resume, so make it express
> that expectation by setting power.strict_midlayer for all PCI devices
> in pci_pm_prepare() and clear it in pci_pm_complete().
>
> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
Previous PM-related patches in drivers/pci/ use a subject line like:
PCI/PM: ...
Would be cool if there were hints about what
dev_pm_set_strict_midlayer() means. Maybe the comment in
get_callback() is enough, but it takes a little work to find it.
Acked-by: Bjorn Helgaas <bhelgaas@xxxxxxxxxx>
> ---
>
> v1 -> v2:
> * Set and clear the new flag in "prepare" and "complete" to allow
> pm_runtime_force_suspend() invoked from driver remove callbacks to
> work.
> * Update subject and changelog.
>
> ---
> drivers/pci/pci-driver.c | 4 ++++
> 1 file changed, 4 insertions(+)
>
> --- a/drivers/pci/pci-driver.c
> +++ b/drivers/pci/pci-driver.c
> @@ -708,6 +708,8 @@
> struct pci_dev *pci_dev = to_pci_dev(dev);
> const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
>
> + dev_pm_set_strict_midlayer(dev, true);
> +
> if (pm && pm->prepare) {
> int error = pm->prepare(dev);
> if (error < 0)
> @@ -749,6 +751,8 @@
> if (pci_dev->current_state < pre_sleep_state)
> pm_request_resume(dev);
> }
> +
> + dev_pm_set_strict_midlayer(dev, false);
> }
>
> #else /* !CONFIG_PM_SLEEP */
>
>
>
