diff options
Diffstat (limited to '')
-rw-r--r-- | plat/mediatek/mt8183/plat_pm.c | 589 |
1 files changed, 589 insertions, 0 deletions
diff --git a/plat/mediatek/mt8183/plat_pm.c b/plat/mediatek/mt8183/plat_pm.c new file mode 100644 index 0000000..6094a17 --- /dev/null +++ b/plat/mediatek/mt8183/plat_pm.c @@ -0,0 +1,589 @@ +/* + * Copyright (c) 2019-2020, MediaTek Inc. All rights reserved. + * + * SPDX-License-Identifier: BSD-3-Clause + */ + +/* common headers */ +#include <arch_helpers.h> +#include <assert.h> +#include <common/debug.h> +#include <lib/mmio.h> +#include <lib/psci/psci.h> +#include <errno.h> + +/* mediatek platform specific headers */ +#include <platform_def.h> +#include <scu.h> +#include <mt_gic_v3.h> +#include <mtk_mcdi.h> +#include <mtk_plat_common.h> +#include <mtgpio.h> +#include <mtspmc.h> +#include <plat_dcm.h> +#include <plat_debug.h> +#include <plat_params.h> +#include <plat_private.h> +#include <power_tracer.h> +#include <pmic.h> +#include <spm.h> +#include <spm_suspend.h> +#include <sspm.h> +#include <rtc.h> + +/* Local power state for power domains in Run state. */ +#define MTK_LOCAL_STATE_RUN 0 +/* Local power state for retention. */ +#define MTK_LOCAL_STATE_RET 1 +/* Local power state for OFF/power-down. */ +#define MTK_LOCAL_STATE_OFF 2 + +#if PSCI_EXTENDED_STATE_ID +/* + * Macros used to parse state information from State-ID if it is using the + * recommended encoding for State-ID. + */ +#define MTK_LOCAL_PSTATE_WIDTH 4 +#define MTK_LOCAL_PSTATE_MASK ((1 << MTK_LOCAL_PSTATE_WIDTH) - 1) + +/* Macros to construct the composite power state */ + +/* Make composite power state parameter till power level 0 */ + +#define mtk_make_pwrstate_lvl0(lvl0_state, pwr_lvl, type) \ + (((lvl0_state) << PSTATE_ID_SHIFT) | ((type) << PSTATE_TYPE_SHIFT)) + +#else /* !PSCI_EXTENDED_STATE_ID */ + +#define mtk_make_pwrstate_lvl0(lvl0_state, pwr_lvl, type) \ + (((lvl0_state) << PSTATE_ID_SHIFT) | \ + ((pwr_lvl) << PSTATE_PWR_LVL_SHIFT) | \ + ((type) << PSTATE_TYPE_SHIFT)) + +#endif /* PSCI_EXTENDED_STATE_ID */ + +/* Make composite power state parameter till power level 1 */ +#define mtk_make_pwrstate_lvl1(lvl1_state, lvl0_state, pwr_lvl, type) \ + (((lvl1_state) << MTK_LOCAL_PSTATE_WIDTH) | \ + mtk_make_pwrstate_lvl0(lvl0_state, pwr_lvl, type)) + +/* Make composite power state parameter till power level 2 */ +#define mtk_make_pwrstate_lvl2( \ + lvl2_state, lvl1_state, lvl0_state, pwr_lvl, type) \ + (((lvl2_state) << (MTK_LOCAL_PSTATE_WIDTH * 2)) | \ + mtk_make_pwrstate_lvl1(lvl1_state, lvl0_state, pwr_lvl, type)) + +#define MTK_PWR_LVL0 0 +#define MTK_PWR_LVL1 1 +#define MTK_PWR_LVL2 2 + +/* Macros to read the MTK power domain state */ +#define MTK_CORE_PWR_STATE(state) (state)->pwr_domain_state[MTK_PWR_LVL0] +#define MTK_CLUSTER_PWR_STATE(state) (state)->pwr_domain_state[MTK_PWR_LVL1] +#define MTK_SYSTEM_PWR_STATE(state) ((PLAT_MAX_PWR_LVL > MTK_PWR_LVL1) ? \ + (state)->pwr_domain_state[MTK_PWR_LVL2] : 0) + +#if PSCI_EXTENDED_STATE_ID +/* + * The table storing the valid idle power states. Ensure that the + * array entries are populated in ascending order of state-id to + * enable us to use binary search during power state validation. + * The table must be terminated by a NULL entry. + */ +const unsigned int mtk_pm_idle_states[] = { + /* State-id - 0x001 */ + mtk_make_pwrstate_lvl2(MTK_LOCAL_STATE_RUN, MTK_LOCAL_STATE_RUN, + MTK_LOCAL_STATE_RET, MTK_PWR_LVL0, PSTATE_TYPE_STANDBY), + /* State-id - 0x002 */ + mtk_make_pwrstate_lvl2(MTK_LOCAL_STATE_RUN, MTK_LOCAL_STATE_RUN, + MTK_LOCAL_STATE_OFF, MTK_PWR_LVL0, PSTATE_TYPE_POWERDOWN), + /* State-id - 0x022 */ + mtk_make_pwrstate_lvl2(MTK_LOCAL_STATE_RUN, MTK_LOCAL_STATE_OFF, + MTK_LOCAL_STATE_OFF, MTK_PWR_LVL1, PSTATE_TYPE_POWERDOWN), +#if PLAT_MAX_PWR_LVL > MTK_PWR_LVL1 + /* State-id - 0x222 */ + mtk_make_pwrstate_lvl2(MTK_LOCAL_STATE_OFF, MTK_LOCAL_STATE_OFF, + MTK_LOCAL_STATE_OFF, MTK_PWR_LVL2, PSTATE_TYPE_POWERDOWN), +#endif + 0, +}; +#endif + +#define CPU_IDX(cluster, cpu) ((cluster << 2) + cpu) +#define ON true +#define OFF false + +/* Pause MCDI when CPU hotplug */ +static bool HP_SSPM_PAUSE; +/* CPU Hotplug by SSPM */ +static bool HP_SSPM_CTRL = true; +/* Turn off cluster when CPU hotplug off */ +static bool HP_CLUSTER_OFF = true; +/* Turn off cluster when CPU MCDI off */ +static bool MCDI_C2 = true; +/* Enable MCDI */ +static bool MCDI_SSPM = true; + +static uintptr_t secure_entrypoint; + +static void mp1_L2_desel_config(void) +{ + mmio_write_64(MCUCFG_BASE + 0x2200, 0x2092c820); + + dsb(); +} + +static bool clst_single_pwr(int cluster, int cpu) +{ + uint32_t cpu_mask[2] = {0x00001e00, 0x000f0000}; + uint32_t cpu_pwr_bit[] = {9, 10, 11, 12, 16, 17, 18, 19}; + int my_idx = (cluster << 2) + cpu; + uint32_t pwr_stat = mmio_read_32(0x10006180); + + return !(pwr_stat & (cpu_mask[cluster] & ~BIT(cpu_pwr_bit[my_idx]))); +} + +static bool clst_single_on(int cluster, int cpu) +{ + uint32_t cpu_mask[2] = {0x0f, 0xf0}; + int my_idx = (cluster << 2) + cpu; + uint32_t on_stat = mcdi_avail_cpu_mask_read(); + + return !(on_stat & (cpu_mask[cluster] & ~BIT(my_idx))); +} + +static void plat_cpu_pwrdwn_common(void) +{ + /* Prevent interrupts from spuriously waking up this cpu */ + mt_gic_rdistif_save(); + mt_gic_cpuif_disable(); +} + +static void plat_cpu_pwron_common(void) +{ + /* Enable the gic cpu interface */ + mt_gic_cpuif_enable(); + mt_gic_rdistif_init(); + mt_gic_rdistif_restore(); +} + +static void plat_cluster_pwrdwn_common(uint64_t mpidr, int cluster) +{ + if (cluster > 0) + mt_gic_sync_dcm_enable(); + + /* Disable coherency */ + plat_mtk_cci_disable(); + disable_scu(mpidr); +} + +static void plat_cluster_pwron_common(uint64_t mpidr, int cluster) +{ + if (cluster > 0) { + l2c_parity_check_setup(); + circular_buffer_setup(); + mp1_L2_desel_config(); + mt_gic_sync_dcm_disable(); + } + + /* Enable coherency */ + enable_scu(mpidr); + plat_mtk_cci_enable(); + /* Enable big core dcm */ + plat_dcm_restore_cluster_on(mpidr); + /* Enable rgu dcm */ + plat_dcm_rgu_enable(); +} + +static void plat_cpu_standby(plat_local_state_t cpu_state) +{ + u_register_t scr; + + scr = read_scr_el3(); + write_scr_el3(scr | SCR_IRQ_BIT | SCR_FIQ_BIT); + + isb(); + dsb(); + wfi(); + + write_scr_el3(scr); +} + +static void mcdi_ctrl_before_hotplug_on(int cluster, int cpu) +{ + if (!HP_SSPM_CTRL && HP_SSPM_PAUSE && MCDI_SSPM) { + mcdi_pause_clr(cluster, CPU_IDX(cluster, cpu), OFF); + mcdi_pause_set(cluster, CPU_IDX(cluster, cpu), ON); + } +} + +static void mcdi_ctrl_before_hotplug_off(int cluster, int cpu, bool cluster_off) +{ + if (!HP_SSPM_CTRL && HP_SSPM_PAUSE && MCDI_SSPM) + mcdi_pause_set(cluster_off ? cluster : -1, + CPU_IDX(cluster, cpu), OFF); +} + +static void mcdi_ctrl_cluster_cpu_off(int cluster, int cpu, bool cluster_off) +{ + if (MCDI_SSPM) { + sspm_set_bootaddr(secure_entrypoint); + + sspm_standbywfi_irq_enable(CPU_IDX(cluster, cpu)); + + if (cluster_off) + sspm_cluster_pwr_off_notify(cluster); + else + sspm_cluster_pwr_on_notify(cluster); + } +} + +static void mcdi_ctrl_suspend(void) +{ + if (MCDI_SSPM) + mcdi_pause(); +} + +static void mcdi_ctrl_resume(void) +{ + if (MCDI_SSPM) + mcdi_unpause(); +} + +static void hotplug_ctrl_cluster_on(int cluster, int cpu) +{ + if (HP_SSPM_CTRL && MCDI_SSPM) { + mcdi_hotplug_clr(cluster, CPU_IDX(cluster, cpu), OFF); + mcdi_hotplug_set(cluster, -1, ON); + mcdi_hotplug_wait_ack(cluster, -1, ON); + } else { + /* power on cluster */ + if (!spm_get_cluster_powerstate(cluster)) + spm_poweron_cluster(cluster); + } +} + +static void hotplug_ctrl_cpu_on(int cluster, int cpu) +{ + if (HP_SSPM_CTRL && MCDI_SSPM) + mcdi_hotplug_set(cluster, CPU_IDX(cluster, cpu), ON); + else + spm_poweron_cpu(cluster, cpu); +} + +static void hotplug_ctrl_cpu_on_finish(int cluster, int cpu) +{ + spm_disable_cpu_auto_off(cluster, cpu); + + if (HP_SSPM_CTRL && MCDI_SSPM) + mcdi_hotplug_clr(cluster, CPU_IDX(cluster, cpu), ON); + else if (HP_SSPM_PAUSE && MCDI_SSPM) + mcdi_pause_clr(cluster, CPU_IDX(cluster, cpu), ON); + + mcdi_avail_cpu_mask_set(BIT(CPU_IDX(cluster, cpu))); +} + +static void hotplug_ctrl_cluster_cpu_off(int cluster, int cpu, bool cluster_off) +{ + mcdi_avail_cpu_mask_clr(BIT(CPU_IDX(cluster, cpu))); + + if (HP_SSPM_CTRL && MCDI_SSPM) { + mcdi_hotplug_set(cluster_off ? cluster : -1, + CPU_IDX(cluster, cpu), OFF); + } else { + spm_enable_cpu_auto_off(cluster, cpu); + + if (cluster_off) + spm_enable_cluster_auto_off(cluster); + + spm_set_cpu_power_off(cluster, cpu); + } +} + +static int plat_mtk_power_domain_on(unsigned long mpidr) +{ + int cpu = MPIDR_AFFLVL0_VAL(mpidr); + int cluster = MPIDR_AFFLVL1_VAL(mpidr); + int clst_pwr = spm_get_cluster_powerstate(cluster); + unsigned int i; + + mcdi_ctrl_before_hotplug_on(cluster, cpu); + hotplug_ctrl_cluster_on(cluster, cpu); + + if (clst_pwr == 0) { + /* init cpu reset arch as AARCH64 of cluster */ + for (i = 0; i < PLATFORM_MAX_CPUS_PER_CLUSTER; i++) { + mcucfg_init_archstate(cluster, i, 1); + mcucfg_set_bootaddr(cluster, i, secure_entrypoint); + } + } + + hotplug_ctrl_cpu_on(cluster, cpu); + + return PSCI_E_SUCCESS; +} + +static void plat_mtk_power_domain_off(const psci_power_state_t *state) +{ + uint64_t mpidr = read_mpidr(); + int cpu = MPIDR_AFFLVL0_VAL(mpidr); + int cluster = MPIDR_AFFLVL1_VAL(mpidr); + const plat_local_state_t *pds = state->pwr_domain_state; + bool afflvl1 = (pds[MPIDR_AFFLVL1] == MTK_LOCAL_STATE_OFF); + bool cluster_off = (HP_CLUSTER_OFF && afflvl1 && + clst_single_on(cluster, cpu)); + + plat_cpu_pwrdwn_common(); + + if (cluster_off) + plat_cluster_pwrdwn_common(mpidr, cluster); + + mcdi_ctrl_before_hotplug_off(cluster, cpu, cluster_off); + hotplug_ctrl_cluster_cpu_off(cluster, cpu, cluster_off); +} + +static void plat_mtk_power_domain_on_finish(const psci_power_state_t *state) +{ + uint64_t mpidr = read_mpidr(); + int cpu = MPIDR_AFFLVL0_VAL(mpidr); + int cluster = MPIDR_AFFLVL1_VAL(mpidr); + const plat_local_state_t *pds = state->pwr_domain_state; + bool afflvl1 = (pds[MPIDR_AFFLVL1] == MTK_LOCAL_STATE_OFF); + + if (afflvl1) + plat_cluster_pwron_common(mpidr, cluster); + + plat_cpu_pwron_common(); + + hotplug_ctrl_cpu_on_finish(cluster, cpu); +} + +static void plat_mtk_power_domain_suspend(const psci_power_state_t *state) +{ + uint64_t mpidr = read_mpidr(); + int cpu = MPIDR_AFFLVL0_VAL(mpidr); + int cluster = MPIDR_AFFLVL1_VAL(mpidr); + const plat_local_state_t *pds = state->pwr_domain_state; + bool afflvl1 = (pds[MPIDR_AFFLVL1] == MTK_LOCAL_STATE_OFF); + bool afflvl2 = (pds[MPIDR_AFFLVL2] == MTK_LOCAL_STATE_OFF); + bool cluster_off = MCDI_C2 && afflvl1 && clst_single_pwr(cluster, cpu); + + plat_cpu_pwrdwn_common(); + + plat_dcm_mcsi_a_backup(); + + if (cluster_off || afflvl2) + plat_cluster_pwrdwn_common(mpidr, cluster); + + if (afflvl2) { + spm_data_t spm_d = { .cmd = SPM_SUSPEND }; + uint32_t *d = (uint32_t *)&spm_d; + uint32_t l = sizeof(spm_d) / sizeof(uint32_t); + + mcdi_ctrl_suspend(); + + spm_set_bootaddr(secure_entrypoint); + + if (MCDI_SSPM) + sspm_ipi_send_non_blocking(IPI_ID_SUSPEND, d); + + spm_system_suspend(); + + if (MCDI_SSPM) + while (sspm_ipi_recv_non_blocking(IPI_ID_SUSPEND, d, l)) + ; + + mt_gic_distif_save(); + } else { + mcdi_ctrl_cluster_cpu_off(cluster, cpu, cluster_off); + } +} + +static void plat_mtk_power_domain_suspend_finish(const psci_power_state_t *state) +{ + uint64_t mpidr = read_mpidr(); + int cluster = MPIDR_AFFLVL1_VAL(mpidr); + const plat_local_state_t *pds = state->pwr_domain_state; + bool afflvl2 = (pds[MPIDR_AFFLVL2] == MTK_LOCAL_STATE_OFF); + + if (afflvl2) { + spm_data_t spm_d = { .cmd = SPM_RESUME }; + uint32_t *d = (uint32_t *)&spm_d; + uint32_t l = sizeof(spm_d) / sizeof(uint32_t); + + mt_gic_init(); + mt_gic_distif_restore(); + mt_gic_rdistif_restore(); + + mmio_write_32(EMI_WFIFO, 0xf); + + if (MCDI_SSPM) + sspm_ipi_send_non_blocking(IPI_ID_SUSPEND, d); + + spm_system_suspend_finish(); + + if (MCDI_SSPM) + while (sspm_ipi_recv_non_blocking(IPI_ID_SUSPEND, d, l)) + ; + + mcdi_ctrl_resume(); + } else { + plat_cpu_pwron_common(); + } + + plat_cluster_pwron_common(mpidr, cluster); + + plat_dcm_mcsi_a_restore(); +} + +#if PSCI_EXTENDED_STATE_ID + +static int plat_mtk_validate_power_state(unsigned int power_state, + psci_power_state_t *req_state) +{ + unsigned int state_id; + int i; + + assert(req_state); + + if (!MCDI_SSPM) + return PSCI_E_INVALID_PARAMS; + + /* + * Currently we are using a linear search for finding the matching + * entry in the idle power state array. This can be made a binary + * search if the number of entries justify the additional complexity. + */ + for (i = 0; !!mtk_pm_idle_states[i]; i++) { + if (power_state == mtk_pm_idle_states[i]) + break; + } + + /* Return error if entry not found in the idle state array */ + if (!mtk_pm_idle_states[i]) + return PSCI_E_INVALID_PARAMS; + + i = 0; + state_id = psci_get_pstate_id(power_state); + + /* Parse the State ID and populate the state info parameter */ + while (state_id) { + req_state->pwr_domain_state[i++] = state_id & + MTK_LOCAL_PSTATE_MASK; + state_id >>= MTK_LOCAL_PSTATE_WIDTH; + } + + return PSCI_E_SUCCESS; +} + +#else /* if !PSCI_EXTENDED_STATE_ID */ + +static int plat_mtk_validate_power_state(unsigned int power_state, + psci_power_state_t *req_state) +{ + int pstate = psci_get_pstate_type(power_state); + int pwr_lvl = psci_get_pstate_pwrlvl(power_state); + int i; + + assert(req_state); + + if (pwr_lvl > PLAT_MAX_PWR_LVL) + return PSCI_E_INVALID_PARAMS; + + /* Sanity check the requested state */ + if (pstate == PSTATE_TYPE_STANDBY) { + /* + * It's possible to enter standby only on power level 0 + * Ignore any other power level. + */ + if (pwr_lvl != 0) + return PSCI_E_INVALID_PARAMS; + + req_state->pwr_domain_state[MTK_PWR_LVL0] = MTK_LOCAL_STATE_RET; + } else if (!MCDI_SSPM) { + return PSCI_E_INVALID_PARAMS; + } else { + for (i = 0; i <= pwr_lvl; i++) + req_state->pwr_domain_state[i] = MTK_LOCAL_STATE_OFF; + } + + return PSCI_E_SUCCESS; +} + +#endif /* PSCI_EXTENDED_STATE_ID */ + +/******************************************************************************* + * MTK handlers to shutdown/reboot the system + ******************************************************************************/ +static void __dead2 plat_mtk_system_off(void) +{ + INFO("MTK System Off\n"); + + rtc_power_off_sequence(); + wk_pmic_enable_sdn_delay(); + pmic_power_off(); + + wfi(); + ERROR("MTK System Off: operation not handled.\n"); + panic(); +} + +static void __dead2 plat_mtk_system_reset(void) +{ + struct bl_aux_gpio_info *gpio_reset = plat_get_mtk_gpio_reset(); + + INFO("MTK System Reset\n"); + + mt_set_gpio_out(gpio_reset->index, gpio_reset->polarity); + + wfi(); + ERROR("MTK System Reset: operation not handled.\n"); + panic(); +} + +static void plat_mtk_get_sys_suspend_power_state(psci_power_state_t *req_state) +{ + assert(PLAT_MAX_PWR_LVL >= 2); + + for (int i = MPIDR_AFFLVL0; i <= PLAT_MAX_PWR_LVL; i++) + req_state->pwr_domain_state[i] = MTK_LOCAL_STATE_OFF; +} + +/******************************************************************************* + * MTK_platform handler called when an affinity instance is about to be turned + * on. The level and mpidr determine the affinity instance. + ******************************************************************************/ +static const plat_psci_ops_t plat_plat_pm_ops = { + .cpu_standby = plat_cpu_standby, + .pwr_domain_on = plat_mtk_power_domain_on, + .pwr_domain_on_finish = plat_mtk_power_domain_on_finish, + .pwr_domain_off = plat_mtk_power_domain_off, + .pwr_domain_suspend = plat_mtk_power_domain_suspend, + .pwr_domain_suspend_finish = plat_mtk_power_domain_suspend_finish, + .system_off = plat_mtk_system_off, + .system_reset = plat_mtk_system_reset, + .validate_power_state = plat_mtk_validate_power_state, + .get_sys_suspend_power_state = plat_mtk_get_sys_suspend_power_state +}; + +int plat_setup_psci_ops(uintptr_t sec_entrypoint, + const plat_psci_ops_t **psci_ops) +{ + unsigned int i; + + *psci_ops = &plat_plat_pm_ops; + secure_entrypoint = sec_entrypoint; + + /* Init cpu reset arch as AARCH64 of cluster 0 */ + for (i = 0; i < PLATFORM_MAX_CPUS_PER_CLUSTER; i++) { + mcucfg_init_archstate(0, i, 1); + mcucfg_set_bootaddr(0, i, secure_entrypoint); + } + + if (!check_mcdi_ctl_stat()) { + HP_SSPM_CTRL = false; + MCDI_SSPM = false; + } + + return 0; +} |