Re: [PATCH v1 1/1] x86/platform/intel-mid: Run PWRMU command immediately

[Ingo Molnar]

* Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:

> On Thu, 2016-08-18 at 12:52 +0200, Ingo Molnar wrote:
> > * Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:
> > 
> > > 
> > > On some firmwares we have to tell how exactly we want the command to
> > > be run.
> > > The default case for now is to run it immediately.
> > > 
> > > Signed-off-by: Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>
> > > ---
> > >  arch/x86/platform/intel-mid/pwr.c | 6 +++++-
> > >  1 file changed, 5 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/arch/x86/platform/intel-mid/pwr.c
> > > b/arch/x86/platform/intel-mid/pwr.c
> > > index c901a34..0548741 100644
> > > --- a/arch/x86/platform/intel-mid/pwr.c
> > > +++ b/arch/x86/platform/intel-mid/pwr.c
> > > @@ -44,6 +44,10 @@
> > >  /* Bits in PM_CMD */
> > >  #define PM_CMD_CMD(x)		((x) << 0)
> > >  #define PM_CMD_IOC		(1 << 8)
> > > +#define PM_CMD_CM_NOP		(0 << 9)
> > > +#define PM_CMD_CM_IMMEDIATE	(1 << 9)
> > > +#define PM_CMD_CM_DELAY		(2 << 9)
> > > +#define PM_CMD_CM_TRIGGER	(3 << 9)
> > >  #define PM_CMD_D3cold		(1 << 21)
> > >  
> > >  /* List of commands */
> > > @@ -137,7 +141,7 @@ static int mid_pwr_wait(struct mid_pwr *pwr)
> > >  
> > >  static int mid_pwr_wait_for_cmd(struct mid_pwr *pwr, u8 cmd)
> > >  {
> > > -	writel(PM_CMD_CMD(cmd), pwr->regs + PM_CMD);
> > > +	writel(PM_CMD_CMD(cmd) | PM_CMD_CM_IMMEDIATE, pwr->regs +
> > > PM_CMD);
> > >  	return mid_pwr_wait(pwr);
> > >  }
> > 
> > Does this fix a bug? If yes then please also add that to the
> > changelog: what are 
> > the symptoms of the bug - how does a user notice, etc.
> 
> Unfortunately I have no firmware (I have knowledge of) to test this. On
> the board I have, i.e. Intel Edison, everything works either way. On the
> other hand the official BSP code has magic number 0x2201 to set, where
> bits [15:13] indeed has no meaning to firmware, but the rest is
> meaningful. So, I could conclude it *might* fix a bug.
> 
> [15:13] MODE_ID
> Numeric ID associated with the given mode from an OSPM perspective.
> Value not interpreted by firmware. Upon successful completion of this
> command, this value should be reflected in the PM_STS.MODE_ID field
> 
> Taking above to the consideration what would you advise me?

"This appears to be a safer approach based on the documentation." is good enough 
justification, IMHO. So if you update the changelog with this information it's 
fine to me!

Thanks,

	Ingo