Re: [PATCH 16/18] thermal: exynos: cleanup code for enabling threshold interrupts

[Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>]

On Tuesday, May 01, 2018 01:02:39 PM Daniel Lezcano wrote:
> On Thu, Apr 26, 2018 at 01:51:31PM +0200, Bartlomiej Zolnierkiewicz wrote:
> > Cleanup code for enabling threshold interrupts in ->tmu_control
> > method implementations.
> > 
> > There should be no functional changes caused by this patch.
> > 
> > Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
> > ---
> >  drivers/thermal/samsung/exynos_tmu.c | 101 ++++++++++++-----------------------
> >  1 file changed, 34 insertions(+), 67 deletions(-)
> > 
> > diff --git a/drivers/thermal/samsung/exynos_tmu.c b/drivers/thermal/samsung/exynos_tmu.c
> > index abe0737..9639acf 100644
> > --- a/drivers/thermal/samsung/exynos_tmu.c
> > +++ b/drivers/thermal/samsung/exynos_tmu.c
> > @@ -76,9 +76,6 @@
> >  #define EXYNOS_TMU_THERM_TRIP_EN_SHIFT	12
> >  
> >  #define EXYNOS_TMU_INTEN_RISE0_SHIFT	0
> > -#define EXYNOS_TMU_INTEN_RISE1_SHIFT	4
> > -#define EXYNOS_TMU_INTEN_RISE2_SHIFT	8
> > -#define EXYNOS_TMU_INTEN_RISE3_SHIFT	12
> >  #define EXYNOS_TMU_INTEN_FALL0_SHIFT	16
> >  
> >  #define EXYNOS_EMUL_TIME	0x57F0
> > @@ -136,13 +133,6 @@
> >  #define EXYNOS7_TMU_TEMP_MASK			0x1ff
> >  #define EXYNOS7_PD_DET_EN_SHIFT			23
> >  #define EXYNOS7_TMU_INTEN_RISE0_SHIFT		0
> > -#define EXYNOS7_TMU_INTEN_RISE1_SHIFT		1
> > -#define EXYNOS7_TMU_INTEN_RISE2_SHIFT		2
> > -#define EXYNOS7_TMU_INTEN_RISE3_SHIFT		3
> > -#define EXYNOS7_TMU_INTEN_RISE4_SHIFT		4
> > -#define EXYNOS7_TMU_INTEN_RISE5_SHIFT		5
> > -#define EXYNOS7_TMU_INTEN_RISE6_SHIFT		6
> > -#define EXYNOS7_TMU_INTEN_RISE7_SHIFT		7
> >  #define EXYNOS7_EMUL_DATA_SHIFT			7
> >  #define EXYNOS7_EMUL_DATA_MASK			0x1ff
> >  
> > @@ -615,29 +605,27 @@ static void exynos4210_tmu_control(struct platform_device *pdev, bool on)
> >  {
> >  	struct exynos_tmu_data *data = platform_get_drvdata(pdev);
> >  	struct thermal_zone_device *tz = data->tzd;
> > -	unsigned int con, interrupt_en;
> > +	unsigned int con, interrupt_en = 0, i;
> >  
> >  	con = get_con_reg(data, readl(data->base + EXYNOS_TMU_REG_CONTROL));
> >  
> >  	if (on) {
> > -		con |= (1 << EXYNOS_TMU_CORE_EN_SHIFT);
> > -		interrupt_en =
> > -			(of_thermal_is_trip_valid(tz, 3)
> > -			 << EXYNOS_TMU_INTEN_RISE3_SHIFT) |
> > -			(of_thermal_is_trip_valid(tz, 2)
> > -			 << EXYNOS_TMU_INTEN_RISE2_SHIFT) |
> > -			(of_thermal_is_trip_valid(tz, 1)
> > -			 << EXYNOS_TMU_INTEN_RISE1_SHIFT) |
> > -			(of_thermal_is_trip_valid(tz, 0)
> > -			 << EXYNOS_TMU_INTEN_RISE0_SHIFT);
> > +		for (i = 0; i < data->ntrip; i++) {
> > +			if (!of_thermal_is_trip_valid(tz, i))
> > +				continue;
> > +
> > +			interrupt_en |=
> > +				(1 << (EXYNOS_TMU_INTEN_RISE0_SHIFT + i * 4));
> > +		}
> 
> As EXYNOS_TMU_INTEN_RISE0_SHIFT is equal to zero, may be you can replace this
> by BITS(i * 4) ?
> 
> Same comments for exynos5433 and exynos7 below.

Good point, I will replace it by using BIT() macro.

> I don't know which one was intended : 

The one that doesn't change the driver behavior as stated in the patch
description.

>  ((EXYNOS_TMU_INTEN_RISE0_SHIFT + i) * 4) or
>   (EXYNOS_TMU_INTEN_RISE0_SHIFT + (i * 4))
> 
> but if it is the former it is lucky it works because the macro is zero.
> 
> Is it possible to have the registers layout, that would facilitate the review.

I'm sorry but Exynos TMU documentation is not publicly available.

Best regards,
--
Bartlomiej Zolnierkiewicz
Samsung R&D Institute Poland
Samsung Electronics