Re: [RFC v2 2/5] PM, Add sysfs file power_off to control device power off policy

["Rafael J. Wysocki" <rjw@sisk.pl>]

On Friday, May 04, 2012, Bjorn Helgaas wrote:
> On Fri, May 4, 2012 at 2:13 AM, Huang Ying <ying.huang@intel.com> wrote:
> > From: Lan Tianyu <tianyu.lan@intel.com>
> >
> > Some devices can be powered off to save more power via some platform
> > mechanism, e.g., ACPI.  But that may not work as expected for some
> > device or platform.  So, this patch adds a sysfs file named power_off
> > under <device>/power directory to provide a mechanism for user to control
> > whether to allow the device to be power off.
> >
> > power_off => "enabled" means allowing the device to be powered off if
> > possible.
> >
> > power_off => "disabled" means the device must be power on anytime.
> >
> > Also add flag power_off_user to struct dev_pm_info to record users'
> > choice. The bus layer can use this field to determine whether to
> > power off the device.
> 
> My first thought was that writing to "power_off" would actually turn
> the power off, which isn't true.  Maybe something like
> "poweroff_allowed" would work.
> 
> I think there's only one use of this new field, in
> pci_pm_runtime_suspend().  Maybe you could pull out that hunk from
> patch 5, combine it with this one, and move it to after patch 5?

Well, please see my comment.

First, it doesn't make sense to export a sysfs file to control a feature that
the given device doesn't have.

Second, if such a file is exported _at_ _this_ _level_, the sysfs setting
should affect every situation in which power may be removed from devices, not
just the PCIe D3cold damned thing.

If this is going to be PCIe-specific, the flag should go into struct pci_dev,
and the sysfs file accordingly.

Thanks,
Rafael