RE: [PATCH v8 17/26] PM / devfreq: rockchip-dfi: make register stride SoC specific

[Chanwoo Choi]

> -----Original Message-----
> From: Chanwoo Choi <chanwoo@xxxxxxxxxx>
> Sent: Thursday, October 19, 2023 12:11 AM
> To: Sascha Hauer <s.hauer@xxxxxxxxxxxxxx>; linux-
> rockchip@xxxxxxxxxxxxxxxxxxx
> Cc: linux-arm-kernel@xxxxxxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx;
> linux-pm@xxxxxxxxxxxxxxx; Heiko Stuebner <heiko@xxxxxxxxx>; Kyungmin Park
> <kyungmin.park@xxxxxxxxxxx>; MyungJoo Ham <myungjoo.ham@xxxxxxxxxxx>; Will
> Deacon <will@xxxxxxxxxx>; Mark Rutland <mark.rutland@xxxxxxx>;
> kernel@xxxxxxxxxxxxxx; Michael Riesch <michael.riesch@xxxxxxxxxxxxxx>;
> Robin Murphy <robin.murphy@xxxxxxx>; Vincent Legoll
> <vincent.legoll@xxxxxxxxx>; Rob Herring <robh+dt@xxxxxxxxxx>; Krzysztof
> Kozlowski <krzysztof.kozlowski+dt@xxxxxxxxxx>; Conor Dooley
> <conor+dt@xxxxxxxxxx>; devicetree@xxxxxxxxxxxxxxx; Sebastian Reichel
> <sebastian.reichel@xxxxxxxxxxxxx>; Jonathan Cameron
> <Jonathan.Cameron@xxxxxxxxxx>
> Subject: Re: [PATCH v8 17/26] PM / devfreq: rockchip-dfi: make register
> stride SoC specific
> 
> On 23. 10. 18. 15:17, Sascha Hauer wrote:
> > The currently supported RK3399 has a stride of 20 between the channel
> > specific registers. Upcoming RK3588 has a different stride, so put the
> > stride into driver data to make it configurable.
> > While at it convert decimal 20 to hex 0x14 for consistency with RK3588
> > which has a register stride 0x4000 and we want to write that in hex as
> > well.
> >
> > Reviewed-by: Jonathan Cameron <Jonathan.Cameron@xxxxxxxxxx>
> > Reviewed-by: Sebastian Reichel <sebastian.reichel@xxxxxxxxxxxxx>
> > Signed-off-by: Sascha Hauer <s.hauer@xxxxxxxxxxxxxx>
> > ---
> >
> > Notes:
> >     Changes since v7:
> >      - Initialize ddrmon_stride for RK3568 and explain why it's not
> > needed
> >
> >  drivers/devfreq/event/rockchip-dfi.c | 13 +++++++++----
> >  1 file changed, 9 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/devfreq/event/rockchip-dfi.c
> > b/drivers/devfreq/event/rockchip-dfi.c
> > index 63977f9fc2693..a3d823ac68ace 100644
> > --- a/drivers/devfreq/event/rockchip-dfi.c
> > +++ b/drivers/devfreq/event/rockchip-dfi.c
> > @@ -113,6 +113,7 @@ struct rockchip_dfi {
> >  	int active_events;
> >  	int burst_len;
> >  	int buswidth[DMC_MAX_CHANNELS];
> > +	int ddrmon_stride;
> >  };
> >
> >  static int rockchip_dfi_enable(struct rockchip_dfi *dfi) @@ -190,13
> > +191,13 @@ static void rockchip_dfi_read_counters(struct rockchip_dfi
> *dfi, struct dmc_coun
> >  		if (!(dfi->channel_mask & BIT(i)))
> >  			continue;
> >  		res->c[i].read_access = readl_relaxed(dfi_regs +
> > -				DDRMON_CH0_RD_NUM + i * 20);
> > +				DDRMON_CH0_RD_NUM + i * dfi->ddrmon_stride);
> >  		res->c[i].write_access = readl_relaxed(dfi_regs +
> > -				DDRMON_CH0_WR_NUM + i * 20);
> > +				DDRMON_CH0_WR_NUM + i * dfi->ddrmon_stride);
> >  		res->c[i].access = readl_relaxed(dfi_regs +
> > -				DDRMON_CH0_DFI_ACCESS_NUM + i * 20);
> > +				DDRMON_CH0_DFI_ACCESS_NUM + i * dfi-
> >ddrmon_stride);
> >  		res->c[i].clock_cycles = readl_relaxed(dfi_regs +
> > -				DDRMON_CH0_COUNT_NUM + i * 20);
> > +				DDRMON_CH0_COUNT_NUM + i * dfi->ddrmon_stride);
> >  	}
> >  }
> >
> > @@ -664,6 +665,8 @@ static int rk3399_dfi_init(struct rockchip_dfi *dfi)
> >  	dfi->buswidth[0] = FIELD_GET(RK3399_PMUGRF_OS_REG2_BW_CH0, val) ==
> 0 ? 4 : 2;
> >  	dfi->buswidth[1] = FIELD_GET(RK3399_PMUGRF_OS_REG2_BW_CH1, val) ==
> 0
> > ? 4 : 2;
> >
> > +	dfi->ddrmon_stride = 0x14;
> > +
> >  	return 0;
> >  };
> >
> > @@ -690,6 +693,8 @@ static int rk3568_dfi_init(struct rockchip_dfi
> > *dfi)
> >
> >  	dfi->buswidth[0] = FIELD_GET(RK3568_PMUGRF_OS_REG2_BW_CH0, reg2) ==
> > 0 ? 4 : 2;
> >
> > +	dfi->ddrmon_stride = 0x0; /* not relevant, we only have a single
> > +channel on this SoC */
> > +
> >  	return 0;
> >  };
> >
> 
> Acked-by: Chanwoo Choi <cw00.choi@xxxxxxxxxxx>
> 
> --
> Best Regards,
> Samsung Electronics
> Chanwoo Choi


Applied it. Thanks

Best Regards,
Chanwoo Choi