Re: [RRC v1 2/3] thermal/drivers/exynos: Handle temperature threshold interrupts and clear corresponding IRQs

Hi Mateusz,

On Wed, 18 Jun 2025 at 17:22, Mateusz Majewski [off-list ref] wrote:

Hello :)

quoted

+#define INTSTAT_FALL2        BIT(24)
+#define INTSTAT_FALL1        BIT(20)
+#define INTSTAT_FALL0        BIT(16)
+#define INTSTAT_RISE2        BIT(8)
+#define INTSTAT_RISE1        BIT(4)
+#define INTSTAT_RISE0        BIT(0)
+
+#define INTCLEAR_FALL2       BIT(24)
+#define INTCLEAR_FALL1       BIT(20)
+#define INTCLEAR_FALL0       BIT(16)
+#define INTCLEAR_RISE2       BIT(8)
+#define INTCLEAR_RISE1       BIT(4)
+#define INTCLEAR_RISE0       BIT(0)

quoted

+     /* Map INTSTAT bits to INTCLEAR bits */
+     if (val_irq & INTSTAT_FALL2)
+             clearirq |= INTCLEAR_FALL2;
+     else if (val_irq & INTSTAT_FALL1)
+             clearirq |= INTCLEAR_FALL1;
+     else if (val_irq & INTSTAT_FALL0)
+             clearirq |= INTCLEAR_FALL0;
+     else if (val_irq & INTSTAT_RISE2)
+             clearirq |= INTCLEAR_RISE2;
+     else if (val_irq & INTSTAT_RISE1)
+             clearirq |= INTCLEAR_RISE1;
+     else if (val_irq & INTSTAT_RISE0)
+             clearirq |= INTCLEAR_RISE0;

This implies that only these 6 bits are used. Is this true for all SoCs
supported by this driver? My understanding is that Exynos 5433 in particular
uses bits 7:0 for rise interrupts and 23:16 for fall interrupts. When I tested
this patch (both alone and the whole series) on 5433 by running some CPU load,
the interrupt seemed to not fire consistently:
/sys/class/thermal/cooling_device1/cur_state would never go above 1 (which is
consistent with the interrupt firing once, not getting cleared and never firing
again; without this patch, it consistently went up to 6) and I got a quick
reboot every time.

Thanks for the feedback,

As per the user manual Exynos4412
INTSTAT and INTCLEAR have a clear mapping with bits
falling 20, 16, 12 and rising 8 4 0

whereas Exyno5422 has
INTSTAT and INTCLEAR have a clear mapping with bits
falling 24, 20, 16, and rising 8 4 0

Yes, it could differ for all the SoCs,
I don't have the user manual or TRM for these SoCs
to configure correctly.

I tried to configure this, referring to the comment in the driver
        /*
         * Clear the interrupts.  Please note that the documentation for
         * Exynos3250, Exynos4412, Exynos5250 and Exynos5260 incorrectly
         * states that INTCLEAR register has a different placing of bits
         * responsible for FALL IRQs than INTSTAT register.  Exynos5420
         * and Exynos5440 documentation is correct (Exynos4210 doesn't
         * support FALL IRQs at all).
         */

By the way, I don't see Exynos5433 and Exynos7 support
INTSTAT and INTCLEAR registers. We are using TMU_REG_INTPEND
 to read and update the same register.

        if (data->soc == SOC_ARCH_EXYNOS5260) {
                tmu_intstat = EXYNOS5260_TMU_REG_INTSTAT;
                tmu_intclear = EXYNOS5260_TMU_REG_INTCLEAR;
        } else if (data->soc == SOC_ARCH_EXYNOS7) {
                tmu_intstat = EXYNOS7_TMU_REG_INTPEND;
                tmu_intclear = EXYNOS7_TMU_REG_INTPEND;
        } else if (data->soc == SOC_ARCH_EXYNOS5433) {
                tmu_intstat = EXYNOS5433_TMU_REG_INTPEND;
                tmu_intclear = EXYNOS5433_TMU_REG_INTPEND;
        } else {
                tmu_intstat = EXYNOS_TMU_REG_INTSTAT;
                tmu_intclear = EXYNOS_TMU_REG_INTCLEAR;
        }

Thank you,
Mateusz Majewski

Thanks
Anand