Re: [PATCH V2] platform/x86: intel_pmc_core: Add CNP SLPS0 debug registers

From: David E. Box
Date: Thu May 31 2018 - 19:05:10 EST


On Thu, 2018-05-31 at 21:38 +0300, Andy Shevchenko wrote:
> On Fri, May 25, 2018 at 4:10 AM, David E. Box
> <david.e.box@xxxxxxxxxxxxxxx> wrote:
> > Adds debugfs access to registers in the Cannon Point PCH PMC that
> > are
> > useful for debugging #SLP_S0 signal assertion and other low power
> > related
> > activities. Device pm states are latched in these registers
> > whenever the
> > package enters C10 and can be read from slp_s0_debug_status. The pm
> > states may also be latched by writing 1 to slp_s0_dbg_latch which
> > will
> > immediately capture the current state on the next read of
> > slp_s0_debug_status. Also while in intel_pmc_core.h clean up
> > spacing.
> >
>
> Thanks for an update. My comments below.
>
> As far as I understand there is still ongoing discussion about the
> approach when and how to show data. So I'll wait a bit for a
> settlement between you, guys.
>
> > +static void pmc_core_slps0_dbg_latch(struct pmc_dev *pmcdev, bool
> > reset)
> > +{
> > + const struct pmc_reg_map *map = pmcdev->map;
> > + u32 fd;
> > +
> > + mutex_lock(&pmcdev->lock);
> > +
> > + if (!reset && !slps0_dbg_latch)
> > + goto out_unlock;
> > +
> > + fd = pmc_core_reg_read(pmcdev, map->slps0_dbg_offset);
> > + reset ? (fd &= ~CNP_PMC_LATCH_SLPS0_EVENTS) :
> > + (fd |= CNP_PMC_LATCH_SLPS0_EVENTS);
>
> I would rather use classical pattern here, i.e.
>
> if (reset)
> fd ...
> else
> fd ...
>
> > + pmc_core_reg_write(pmcdev, map->slps0_dbg_offset, fd);
> > +
> > + slps0_dbg_latch = 0;
> > +
> > +out_unlock:
> > + mutex_unlock(&pmcdev->lock);
> > +}
> > + struct pmc_dev *pmcdev = s ? s->private : &pmc;
>
> I'm not sure why do we need such condition.
>
> Simple
>
> ... pmcdev = s->private;
>
> is enough.
>
> > /* Cannonlake Power Management Controller register offsets */
> > -#define CNP_PMC_SLP_S0_RES_COUNTER_OFFSET 0x193C
> > -#define CNP_PMC_LTR_IGNORE_OFFSET 0x1B0C
> > -#define CNP_PMC_PM_CFG_OFFSET 0x1818
> > +#define CNP_PMC_SLP_S0_RES_COUNTER_OFFSET 0x193C
> > +#define CNP_PMC_LTR_IGNORE_OFFSET 0x1B0C
> > +#define CNP_PMC_PM_CFG_OFFSET 0x1818
>
> I have hard time to understand what is the difference here.
> Either do another clean up patch (white spaces vs. tabs?) or leave it
> untouched.
>
> > /* Cannonlake: PGD PFET Enable Ack Status Register(s) start */
> > -#define CNP_PMC_HOST_PPFEAR0A 0x1D90
> > +#define CNP_PMC_HOST_PPFEAR0A 0x1D90
> >
> > -#define CNP_PMC_MMIO_REG_LEN 0x2000
> > -#define CNP_PPFEAR_NUM_ENTRIES 8
> > -#define CNP_PMC_READ_DISABLE_BIT 22
> > +#define CNP_PMC_MMIO_REG_LEN 0x2000
> > +#define CNP_PPFEAR_NUM_ENTRIES 8
> > +#define CNP_PMC_READ_DISABLE_BIT 22
>
> Ditto.
>
> > +struct slps0_dbg_map {
> > + const struct pmc_bit_map *slps0_dbg_sts;
> > + int size;
> > +};
>
> Didn't pay attention to this earlier. Why do we have a separate size
> member? What does it define?

It holds the size of the pmc_bit_map array, assigned as shown here:

+static const struct slps0_dbg_map cnp_slps0_dbg_maps[] = {
+ {cnp_slps0_dbg0_map, ARRAY_SIZE(cnp_slps0_dbg0_map)},
+ {cnp_slps0_dbg1_map, ARRAY_SIZE(cnp_slps0_dbg1_map)},
+ {cnp_slps0_dbg2_map, ARRAY_SIZE(cnp_slps0_dbg2_map)},
+};


Okay on all other comments

Dave