Re: [PATCHv2 2/4] ACPI / LPSS: custom power domain for LPSS

["Rafael J. Wysocki" <rjw@rjwysocki.net>]

On Wednesday, May 21, 2014 01:05:11 PM Heikki Krogerus wrote:
> Hi Rafael,
> 
> On Tue, May 20, 2014 at 11:33:09PM +0200, Rafael J. Wysocki wrote:
> > > +#ifdef CONFIG_PM
> > 
> > It would be good to add a kerneldoc explaining what's being saved here and why.
> 
> OK.
> 
> > > +static void acpi_lpss_save_ctx(struct device *dev)
> > > +{
> > > +	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
> > > +	int i;
> > > +
> > > +	for (i = 0; i < LPSS_PRV_REG_COUNT; i++) {
> > > +		pdata->prv_reg_ctx[i] = __lpss_reg_read(pdata, i * sizeof(u32));
> > > +		dev_dbg(dev, "saving 0x%08x from LPSS reg at offset 0x%02x\n",
> > > +			pdata->prv_reg_ctx[i], (unsigned int)(i * sizeof(u32)));
> > > +	}
> > > +}
> > > +
> > > +static void acpi_lpss_restore_ctx(struct device *dev)
> > > +{
> > > +	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
> > > +	int i;
> > > +
> > > +	/* PCI spec expects 10ms delay when resuming from D3 to D0 */
> > > +	msleep(10);
> > 
> > Two questions here.
> > 
> > First, is the 10 ms sleep really necessary?  I'd expect the AML to take care of
> > such delays (this is not a PCI device formally).
> 
> Unfortunately that is not the case. There is nothing in the AML for
> this. Mika, correct me if I'm wrong.
> 
> > And because this is not a PCI device formally, why is the comment talking about
> > the PCI spec?  Why is PCI relevant in any way here?
> 
> Under the hood the devices are still PCI devices, even if they
> formally aren't. Maybe I should point that out in the comment..
> 
> We put the sleep there because without it there was no guarantee if
> the device was properly resumed by the time the drivers resume hooks
> were called. The symptom in case of a failure was simply that the
> registers could not be written, which leads into timeouts at least in
> case of the I2C and UART and making them unusable until the next
> suspend followed by resume.

OK, so the msleep() is functionally necessary.  Instead of talking about the
PCI in the comment, which will make a casual reader think "What the heck?",
please say something like "the delay is necessary for the subsequent register
writes to succeed on <example system>".

Rafael