diff --git a/src/soc/rockchip/rk3399/Makefile.inc b/src/soc/rockchip/rk3399/Makefile.inc index 990edcba2c..28c633835a 100644 --- a/src/soc/rockchip/rk3399/Makefile.inc +++ b/src/soc/rockchip/rk3399/Makefile.inc @@ -49,6 +49,7 @@ romstage-y += romstage.c romstage-y += tsadc.c romstage-y += usb.c romstage-y += gpio.c +romstage-y += saradc.c romstage-y += ../common/gpio.c ################################################################################ diff --git a/src/soc/rockchip/rk3399/romstage.c b/src/soc/rockchip/rk3399/romstage.c index 3722813d6d..b65a9adc87 100644 --- a/src/soc/rockchip/rk3399/romstage.c +++ b/src/soc/rockchip/rk3399/romstage.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include #include @@ -38,6 +39,7 @@ static const uint64_t dram_size = static void init_dvs_outputs(void) { + int duty_ns; uint32_t i; write32(&rk3399_grf->iomux_pwm_0, IOMUX_PWM_0); /* GPU */ @@ -46,31 +48,41 @@ static void init_dvs_outputs(void) write32(&rk3399_pmugrf->iomux_pwm_3a, IOMUX_PWM_3_A); /* Centerlog */ /* - * Notes: + * Set up voltages for all DVS rails. * - * design_min = 0.8 - * design_max = 1.5 + * LITTLE CPU: At the speed we're running at right now and on the + * early silicon, .9V is sane. If/when we run faster, let's bump this. * - * period = 3333 # 300 kHz - * volt = 1.1 + * CENTER LOGIC: There are some claims that this should simply always + * be .9 V. There are other claims that say that we need to adjust this + * dynamically depending on the memory frequency. Until this is sorted + * out, it appears that .9 V works for the 800 MHz. * - * # Intentionally round down (higher volt) to be safe. - * int((period / (design_max - design_min)) * (design_max - volt)) + * BIG CPU / GPU: These aren't used in coreboot. Init to .9V which is + * supposed to be a good default. * - * Tested on kevin rev0 board 82 w/ all 4 PWMs: + * Details: + * design_min = 0.8 + * design_max = 1.5 + * period = 3337 # 300 kHz + * volt = 1.1 + * # Intentionally round down (higher volt) to be safe. + * int((period / (design_max - design_min)) * (design_max - volt)) * - * period = 3333, volt = 1.1: 1904 -- Worked for me! - * period = 3333, volt = 1.0: 2380 -- Bad - * period = 3333, volt = 0.9: 2856 -- Bad - * - * period = 25000, volt = 1.1: 14285 -- Bad - * period = 25000, volt = 1.0: 17857 -- Bad - * - * TODO: Almost certainly we don't need all 4 PWMs set to the same - * thing. We should experiment + * Apparently a period of 3333 is determined by EEs to be ideal for our + * board design / resistors / capacitors / regulators but due to + * clock dividers we actually get 3337. Solving, we get: + * period = 3337, volt = 1.1: 1906 + * period = 3337, volt = 1.0: 2383 + * period = 3337, volt = 0.9: 2860 */ + if (IS_ENABLED(CONFIG_BOARD_GOOGLE_KEVIN) && (board_id() <= 2)) + duty_ns = 1906; /* 1.1v */ + else + duty_ns = 2860; /* 0.9v */ + for (i = 0; i < 4; i++) - pwm_init(i, 3333, 1904); + pwm_init(i, 3337, duty_ns); } static void prepare_usb(void)